EFFECTS OF EWE BREED AND MANAGEMENT SYSTEM ON EFFICIENCY OF LAMB PRODUCTION: I. EWE PRODUCTIVITY'

EFFECTS OF EWE BREED AND MANAGEMENT SYSTEM ON EFFICIENCY OF LAMB PRODUCTION: I. EWE PRODUCTIVITY' D. R. Notter and F. S. McClaugherty Virginia Polytechnic Institute and State University2, Blacksburg 24061 ABSTRACT Performance of 1/2-Suffollc, 1/2-Rambouillet (Western) and 1/2-Suffolk, 1/4-Rambouillet, 1/4-Finnsheep (1/4-Finn) ewes was compared in three different lamb production systems over 3 yr. System 1 (56 ewes) involved late fall lambing over 84 d. System 2 (51 ewes) involved January and February lambing for 60 d. System 3 (47 ewes) involved March and April lambing for 45 d. Pregnancy rates for yearling ewes were lower in System 1 in yr 1 (50.7% vs 87.4% for Systems 2 and 3) but differed little among systems for older ewes or in remaining years. Average pregnancy rates for 2-yr-old and older ewes were 89.5, 94.0 and 85.7% for Systems 1, 2 and 3, respectively. When the pregnancy rate was adjusted to a 456 lambing season, means for older ewes were 78.6, 89.5 and 85.7% for System 1, 2 and 3, respectively. Ewe breeds did not differ in their pregnancy rates. Prolificacy (lambs born per ewe lambing) was higher for 1/4-Finn ewes (1.83 f.06 vs 1.55 f.07) and was higher in System 3 (1.86 f.06) than in Systems 1 (1.60 f.07) or 2 (1.63 f.05). Body weight at breeding in postyearling ewes was less in System 3 (64.3 kg) than in Systems 1 or 2 (average of 73.1 kg). Breeds did not differ in weight at 1 or 2 yr of age, but Western ewes were 2.1 f 1.1 kg heavier as 3-yr-olds. Fleece weights were higher for Western ewes (3.45 f.05 kg) than for WFinn ewes (2.69 f.06 kg); this difference increased with age from.19 f.i1 kg in yearlings to 1.12 f.os kg in 3-yr-olds. Repeatabilities of ewe performance traits were.16 f.08 for lambing date,.07 f.07 for prolificacy,.58 f.05 for body weight and.44 f.05 for fleece weight. Key Words: Sheep, Fertility, Systems, Breeds, Prolificacy, Wool Introduction The relatively short production cycle of the sheep, coupled with the variety of feed resources that can be used for sheep production, allow considerable latitude in the design of sheep production systems. Factors that influence the choice of production system include the availability and cost of various feedstuffs, availability of shelter at lambing, annual fluctuations in lamb prices, competition for labor from other farm enterprises and cash flow patterns for the entire farming operation. 'This research was supported in part by a grant from the Virginia Agricultural Foundation. 'hpt. of M. Sci. Received March 14, 1990. Accepted June 29, 1990. 13 J. Anim. Sci. 1991. 69:13-21 The most profitable production system should be expected to vary in response to these factors and may differ among farms within the same region. The purpose of this study was to compare three annual sheep production systems for the hill lands of the southeastern U.S. In addition, the value of increasing ewe prolificacy through use of Finnsheep crossbred ewes was evaluated for each system. This portion of the study specifically addresses the productivity of ewes of the different breeds in the three systems. Materials and Methods Sixtyeight 1/2-Suffolk, 112-Rambouillet (Western) ewes and 86 ewes that were 112 Suffolk, 114 Finnsheep and 1/4 Dorset or Rambouillet (1/4-Finn) were used for the project. Western ewes were purchased in

14 NOTI'ER AND McCLAUGHERTY Texas as ewe lambs in summer 1980 at approximately 7 mo of age and were representative of Western ewes imported to Virginia as replacement females. The 1/4-Finn ewes were born in spring of either 1980 (n = 48) or 1981 (n = 38) and were produced in Virginia by mating mature Finnsheep x Rambouillet ewes (Notter and Copenhaver, 1980) or Finnsheep x Dorset ewes (Cochran et al., 1984) to Suffolk rams. All 1980-born ewes were exposed to rams for 60 d beginning on August 20, 1980, and allowed to lamb in spring 1981, before entering the current project in June 1981. Ten percent of the Western ewes and 31% of the 1/ 4-Finn ewes lambed in 1981, Their lambs were weaned in May. The 1981-born ewes were given the opportunity to lamb as yearlings in early spring 1982, before entering the current project in June 1982. Seventy-four percent of the 1981-born 1/4-Finn lambed as yearlings; their lambs were weaned in May. Ewes were assigned at random within breed and birth year to one of three management systems and remained in the same system for the duration of the study. No voluntary culling of ewes occurred. Ewes in System 1 (n = 56) were exposed to rams for 84 d beginning June 1 and lambed in November, December and January. Lambs were creep fed until weaning at about 70 d of age and then were fed out to slaughter in drylot of a high-energy diet. Onehalf of the male lambs were left intact; the remainder were castrated shortly after birth. System 1 was expected to produce relatively rapid lamb gains, was not dependent on seasonal forages, would allow lambs to be marketed when lamb prices were seasonally high and could take advantage of the greater efficiency of ram lambs, but it required relatively high inputs of harvested feeds and satisfactory summer breeding of ewes. Ewes in System 2 (n = 51) were exposed to rams for 60 d beginning August 1 and lambed in January and February. All male lambs were castrated shortly after birth. Lambs were creep fed until the onset of spring grazing (about April 15), at which time they and their dams were moved to native bluegrass-white clover pasture and creep feeding was discontinued In late April or early May, ewes and lambs were moved to either alfalfa or ladino clover pastures. Ewes remained with their lambs only during the transition to legume grazing and were removed within 1 to 2 wk. Lambs rotationally grazed three.30-ha paddocks per legume. They remained on legume grazing until reaching market weight or until early fall, when the remaining lambs were moved to drylot and finished on a concentrate diet. System 2 represented an attempt to finish a high proportion of the lambs before the onset of seasonal declines in forage quality in mid- July and to provide high-quality legume grazing for remaining lambs in late summer. Ewes in System 3 (n = 47) were mated for 45 d beginning October 15 and lambed in March and April. AU male lambs were castrated shortly after birth. Lambs had access to creep feed until the onset of spring grazing, at which time they and their dams were moved to bluegrass-white clover pastures. Lambs remained with the ewes on these pastures until they reached market weight or until early September, when the remaining lambs were moved to drylot and finished on a concentrate diet. System 3 was expected to yield the highest pregnancy rates and greatest prolificacy and required lower inputs of harvested feed, labor and facilities. However, it was expected to produce slower lamb gains and required ewes and lambs to compete for summer forage and brought lambs to market at a time when lamb prices were anticipated to be seasonally low. Three lamb crops (1982 through 1984) were produced in each system. Each lamb crop included lambs born in System 1 in the previous year (e.g., the 1982 lamb crop included lambs born in fall 1981). All ewes were exposed to Suffolk rams in multiple-sire breeding pastures. In all three systems, target market weights were 55 kg for male lambs and 50 kg for female lambs, but with the constraint that estimated backfat must not exceed 7.5 mm. Details of lamb management and performance were given by Notter et al. (1991). Ewes in all systems grazed bluegrass-white clover pastures from mid-april until mid- October. Nonlactatjng ewes were wintered on stockpiled fescue with supplemental fescue hay in late winter or during periods of snow cover. Each gestating ewe in all systems received.23 kg of corn per day for approximately 4 wk before lambing; lactating ewes in all systems had ad libitum access to alfalfa hay plus approximately.68 kg of corn per day until the onset of spring grazing in mid-april. Ewes were weighed at the start of breeding and at weaning (or in late June for System 3) in each year. Breeding weights in 1982 and weaning

EWE PRODUCITVITY IN DIFPERENT SYSTEMS 15 TABLE 1. LEAST SQUARES MEANS AND STANDARD ERRORS FOR PREGNANCY RATE (%) BY SYSTEM, EWE BIRTH YEAR AND LAMBING YEAR System Ewe birth Lambing YW Y- 1 2 3 Average 1980 1982 50.7b f 5.3 89.0 f 6.0 85.7 f 6.1 75.1 f 3.4 1983 88.6 f 5.3 99.9 f 6.0 83.0 f 6.0 90.5 f 3.3 1984 90.1 f 5.4 91.5 f 6.0 81.4 f 6.3 87.7 f 3.4 A% 76.5 f 3.1 93.5 f 3.5 83.4 f 3.5 84.4 f 1.9 1981 1983 78.6 f 11.2 90.5 f 8.4 72.8 f 9.8 80.6 f 5.7 1984 89.7 f 10.6 90.5 f 9.0 92.8 f 10.7 91.0 f 5.8 Avg 84.2 f 7.7 90.5 f 6.2 82.8 f 7.3 85.8 f 4.1 herd av$ 79.0 f 3.6 92.5 f 3.5 83.1 f 3.8 84.9 f 2.1 Wculated as the average of 1980-born 1/4Fions. 1980-born Westerns and 1981-born 1/4-Fm. b,cmeans in a row with similar superscripts do not differ (P <.05). weights in 1983, however, were not available for ewes in System 1. Ewe condition scores (1 = extremely thin to 9 = extremely fat) were assigned at each weighing by three independent evaluators and averaged. Ewes were shorn annually in early summer and fleece weights were recorded. The analysis of ewe performance utilized least squares procedures (Harvey, 1982) and considered pregnancy rate (coded as 0 for open ewes and 1 for ewes that lambed); number of lambs born, reared (alive after 2 wk) or marketed expressed on both a per exposure and a per lambing basis; lambing date; ewe body weight and mean condition score at breeding; and fleece weight. The statistical model for these traits was as follows: Y$lm = p + Si + Gj + Bjk + SGo + SBqk + D Q + ~ Ajm + SAum + BAjkm + SBAijh + Eijklmr where Yixlm is the observation in the mrh year (A) on de Irh ewe @) of the eh breed (El) born in the j7h ewe birth year (G) and assigned to the irh system (S), p is a constant common to all observations and Euklm is residual error. Effects of breed and year of exposure were nested within ewe birth years due to partial confounding among these effects (e.g., Western ewes were born only in 1980, and 1981-born ewes produced lambs only in 1983 and 1984). Effects of system, ewe birth year, breed and their twoway interactions were tested with the mean square for random ewe effects; all other effects were tested with residual error. Ewe weights and mean condition scores at breeding and weaning were analyzed using the same model, but it considered only ewes that weaned lambs and included effects of time of weighing (breeding vs weaning) and its interactions with other fixed effects. Effects of breed of mater- nal granddam (Rambouillet or Dorset) of 1/ 4-Finn ewes were tested in preliminary analyses, found to be nonsignificant and not included in the final models. Results and Dlscusslon Average pregnancy rates did not differ among systems or ewe breeds, but significant year effects (nested within ewe birth years) and year x system interaction were observed (Table 1). For 1980-born ewes lambing in 1982, pregnancy rates were significantly less in System 1 than in Systems 2 or 3. Thus, yearling ewes exposed in June, July and August had lower pregnancy rates than ewes bred later in the year. Similar seasonal effects on reproduction in Suffolk and Suffolk crossbred ewes have been reported by Dufour (1974) and Notter and Copenhaver (1980). Somewhat surprisingly, however, the 1980-born ewes lambing in 1983 and 1984 did not differ among systems in pregnancy rate. Because no culling of ewes was practiced based on pregnancy rate, this result suggests that the fertility of 1/ 2-Suffok ewes in summer breeding improved with age, and by 2 yr of age equaled that of ewes exposed in late summer or fall. This result is consistent with data presented by Fahmy et al. (1980) and Clarke et al. (1984) suggesting that seasonality of breeding is more intense in younger ewes. For 1983 lambings, pregnancy rates of yearling ewes born in 1981 were slightly less than those of older ewes in all systems (80.6 f 5.7 vs 90.0 f 5.0%), but pregnancy rates of yearling ewes in System 1 were still higher in 1983 than in 1982 (78.6 f 11.2 vs 50.7 f 5.3%). Nugent and Notter (1990) observed that ovulation and mating rates in mature Suffolk

16 NO'ZTER AND McCLAUGHERTY TABLE 2. LEAST SQUARES MEANS AND STANDAFCD ERRORS FOR LAMBING DATE BY SYSTEY EWE BIRTH YEAR AND LAMBING YEAR system Ewe birth Lambing Y W Year 1 2 3 1980 1982 Jan 17b f 3 Jan22 f 3 Mar12 f 3 1983 Nov29C f 3 Jan13 f 3 Mar21 f 3 1984 DecllC f 3 Jan23 f 3 Mar 4 f 3 Avg Dec18f2 Jan20 f 2 Mar12 f 2 1981 1983 m11 f5 Jan17 f 4 Mar21 f 5 1984 Dec 2 f 5 Jan26 It 4 Mar 2 f 5 Avg Dec 7 f 4 Jan21 f 3 Mar11f4 Overall av8 Dec14f2 Jan20 f 2 Mar12 f 2 8Calculated as the average of 198C-born 1/4-Pinns, 1980-born Westems and 1981-born 1/4-Finns. b9cmeans in a column and ewe birth year with similar superscripts do not differ (P e.05). and Hampshire ewes in June were increased when the ewes cohabited with cyclic whitefaced crossbred ewes. A similar effect may have allowed the presence of cyclic, 1980-born ewes to enhance the early summer breeding performance of 1981-born ewes exposed to lamb in 1983. The 1980-born ewes, in contrast, were isolated from other ewes at their first breeding. Comparisons of pregnancy rates in different systems may be biased by differences in duration of ram exposure. From a production standpoint, we shortened the breeding season from the less favorable breeding season of System 1 (84 d) to the more favorable seasons of System 2 (60 d) and 3 (45 d). Hence, pregnancy rates may have been contingent upon differences in duration of the breeding season. In an attempt to remove this bias, data were reanalyzed with ewes in Systems 1 and 2 that lambed after d 45 of the anticipated lambing season assigned pregnancy rates of 0. These adjustments were made assuming that gestation length was 148 d (Anderson et al., 198 1). After adjustment to a constant breeding season, overall effects of the system became highly significant, but system x year interaction remained important (P <.OOOl). Relative to values shown in Table 1, overall average conception rates for Systems 1 and 2 were reduced to 59.2 f 3.6 and 88.0 f 3.5%, respectively. Adjusted differences between Systems 2 and 3 were relatively consistent across years. After adjustment to a 45-d breeding season, the pregnancy rate for yearlings exposed to lamb in 1982 was zero in System 1 (vs 85.7% in both System 2 and 3), whereas that of yearlings exposed to lamb in 1983 was 68.1 f 11.2% (vs 83.8 f 8.4% in System 2 and 72.8 f 9.8% in System 3). If yearling ewes are excluded, the overall 45-d pregnancy rate for ewes in Systems 1, 2 and 3 were 78.6 f 4.3, 89.5 f 4.1 and 85.7 f 4.6%. respectively. Differences in pregnancy rate among systems shown in Table 5Ozl thus were increased when data were restricted to a 45-d breeding season, but still this difference only approached significance (P <.lo) in older ewes. Ewe effects on pregnancy rate were significant; the repeatability of pregnancy rate was.25 *.05. Restriction of the data to a 45-d breeding season had little impact; repeatability remained at.27 f.06. These values for repeatability were higher than most values reported (Purser, 1965; Turner and Young, 1%9). When calculated separately for each system, repeatability of pregnancy rate were.10 f.09,.32 f.10 and.43 f.10 for Systems 1, 2 and 3, respectively. These results differed from those of Notter (1981), who reported that the repeatability of pregnancy rate in an accelerated lambing system was higher for April exposures (.19 f.05) than for August (.06 f.06) or November exposures (.01 f.08). Considering only the first two exposures, 92% of the ewes in System 2 and 83% of the ewes in System 3 either lambed or failed to lamb both years. This result assured that the pregnancy rate was highly repeatable for these ewes but did little to allow identification of useful variation among ewes. In contrast, in System 1, the mean pregnancy rate increased markedly from first to second exposure, so it necessarily was not repeatable for this system. Only 60% of the ewes in System 1 produced consistent pregnancy records at their first two

EWE PRODUCTIVITY IN DIFFeRENT SYSTEMS 17 TABLE 3. LEAST SQUARES MEANS AND STANDARD ERRORS FOR EWE WEIGHTS AT BREEDING (kg) BY SYSTEM EWE BIRTH YEAR. EWE BREED AND LAMBING YEAR Ewe birth Ewe Lambing system Year breed Y W 1 2 3 Avg 1980 l/bfinn 1982 61.3 f 1.9b 59.3 f 1.6 61.1 f 1.6 60.6 f 1.0 1983 69.0 f 1.3 75.3 f 1.6 65.2 f 1.6 69.8 f.9 1984 75.6 f 1.3 72.9 f 1.6 64.7 f 1.6 71.1 f.9 A% 68.6 f.9 69.2 f.9 63.7 f.9 67.2 f.5 Western 1982 62.8 f 1.7b 57.8 f 1.4 58.5 f 1.4 59.7 f.9 1983 69.0 f 1.1 74.4 f 1.4 67.5 f 1.4 70.3 f.7 1984 78.6 f 1.1 72.2 f 1.4 68.8 f 1.4 73.2 f.7 Avs 70.1 f.8 68.1 f.8 64.9 f.8 67.7 rt.5 1981 1/4-Fii 1983 57.0 f 1.8 65.7 f 1.5 58.1 f 1.8 60.3 f 1.0 1984 74.2 f 1.8 70.9 f 1.4 59.9 f. 1.6 68.3 f.9 Avg 65.6 f 1.3 68.3 f 1.0 59.0 f 1.2 64.3 f.7 Overall avpa 68 68.1 f.6 68.5 f.5 62.5 f.6 66.4 f.3 acalculated as the average of 198Ckborn 1/4-Finns, 1986bom Westem and 1981-born l/bfinns. %ese subclasses were missing in the actual data set. Means are estimates from the least squares analysis. exposures. These results suggest that repeatability estimates for zero-one measures of reproduction must be interpreted carefully when based on small numbers of exposures or when mean pregnancy rates vary across years. For ewes that lambed, the analysis of lambing date (Table 2) tended to support results of the pregnancy rate analysis. Lambing date was significantly affected by system, year within ewe birth year and system x year interaction. In System 1, 1980-born ewes lambed later as yearlings (average date of January 17) than as older ewes (average date of December 5). Thus, mean date of cunception for yearling 1980-born ewes that lambed in System 1 was about August 17, whereas in later years about half of the same ewes had conceived by late June. In contrast, 1981-born ewes lambing in 1983 had an average lambing date that was only 12 f 6 d later than that of contemporary 1980-born ewes. Differences in lambing date among years in System 2 and 3 were smaller than those observed in System 1 and showed no consistent association with ewe age. Repeatability of lambing date was.27 f.07. Repeatability of lambing date was the same for System 1 and 2 (.31 f.11) but was lower for System 3 (.20 f.13). Thus, the repeatability of lambing date was higher in summer breedings, when the duration of exposure was increased. This result was in part expected because ewes exposed for 45 d were with rams for only the equivalent of about 2.8 estrous cycles, whereas ewes exposed for 84 d were with rams for the equivalent of about 5.25 estrous cycles. Variation in lambing date due to random variation in stage of the estrous cycle at ram introduction is not expected to be repeatable. It would become a progressively smaller fraction of the total variation in lambing date as the length of the breeding season increased. When data were adjusted to a constant, 45-d breeding season, the overall repeatability of lambing date was reduced to.16 f.08; ewe effects remained as an important source of variation (P s:.05). After adjustment to a constant duration of breeding, the repeatability of lambing date was similar across systems (.17 f.19,.21 f.ll and.20 f.13 for Systems 1, 2 and 3 respectively). Number of lambs born, reared (alive at 2 wk) and marketed per ewe lambing were influenced (P <.05) by breed of ewe. For 1980-born ewes, 1/4-Finn ewes produced, raised and marketed an average of 1.83 f.06, 1.66 f.06 and 135 f.07 lambs, respectively, per ewe lambing. Comparable values for Western ewes were 1.55 f.05, 1.30 *.05 and 1.16 f.06 lambs, respectively. The difference in prolificacy between 1980-born 1/4-Finn and Western ewes was reasonably consistent across systems (.20 f.13,.28 f.12 and.37 f.13 lambs for Systems 1, 2 and 3, respectively) and averaged.28 f.08 lambs per lambing. This difference is consistent with the prediction of Dickerson (1977) that each 1% increase in Finn ancestry relative to most traditional U.S. breeds is expected to increase prolificacy by.01 lambs. Progeny of Western ewes also had higher average postnatal death losses

~ 18 N O m AND McCLAUGHERTY TABLE 4. LEAST SQUARES MEANS AND STANDARD ERRORS FOR EWE CONDITION SCORES AT BREEDING BY SYSTEM, EWE BIRTH YEAR, EWE BREED AND LAMBING YEAR' Ewe birth Ewe Lambing system Year breed Y W 1 2 3 Avg 1980 1/4-Fh 1982 6.6 f.3' 5.1 f.3 6.8 f.3 6.2 f.2 1983 6.6 f.3 7.4 f.3 4.9 f.3 6.3 f.2 1984 6.6 f.3 5.6 f.3 4.8 f.3 5.7 f.2 Ava 6.6 f.2 6.0 f.2 5.5 f.2 6.0 f.1 Western 1982 6.1 f.3' 4.8 f.3 6.2 f.3 5.7 f.2 1983 6.3 f.2 7.3 f.3 5.1 f.3 6.2 f.2 1984 6.1 f.2 6.3 f.3 5.0 f.3 5.8 f.2 Avg 6.2 f.1 6.1 f.2 5.4 f.2 5.9 f.1 1981 1/4-Fh 1983 5.1 f.4 7.0 f.3 5.8 f.4 6.0 f.2 1984 6.3 f.4 6.5 f.3 5.3 f.4 6.0 f.2 Avg 5.7 f.3 6.8 f.2 5.6 f.3 6.0 f.2 Over& avgb 6.2 *.I 6.3 f.1 5.5 f.i 6.0 f.1 Yondition scores on a scale of 1 to 9 with 9 fattest. bcalculated as the average of 198Gborn 1/4-Finns, 198Gborn Westerns and 1981-born 1/4-Finns. CThese subclasses were missing in the actual data set. Means are estimates from the least squares analysis. (Notter et al., 1991), such that the advantage of the 1/4-Finn ewe in lamb production was increased to.36 f.08 lambs reared and to.39 f.09 lambs marketed per ewe lambing. The number of lambs born per ewe lambing differed among systems; mean prolificacy (averaged across breeds and ewe birth years) was 1.60 f.07, 1.63 f.05 and 1.86 f.w lambs per ewe lambing for Systems 1,2 and 3, respectively. These results reflect the expected reduction in prolificacy between fall-bred and spring-bred ewes (Whiteman et al., 1972; Notter and Copenhaver, 1980; Dzakuma et al., 1982a). Ewe effects on prolificacy were not significant; repeatability was.07 f.07, which was somewhat lower than values reported for Finnsheep crossbred ewes by Notter (1981;.17 f.03), Cochran et al., (1984,.14 f.03), Dzakuma et al. (1982b;.14 f.07) and Hanrahan and Quirke (1985;.09 f.03 and.11 f.05). The between-ewe variance component was small and negative for number of lambs weaned and marketed per ewe lambing. When number of lambs born was expressed on a per ewe exposed basis, the advantage of 1/4-Finn ewes over Western ewes remained significant but was reduced to.20 f.11 lambs produced,.26 f.10 lambs raised and.29 f.09 lambs marketed. Effects of system were not significant for number of lambs born per ewe exposed (1.29, 1.51 and 1.54 [It.08] for Systems 1, 2 and 3, respectively) but were significant for number of lambs raised (1.08, 1.37 and 1.39 [f.08], respectively) and approached significance (P <.lo) for number marketed (1.00, 1.29 and 1.23 [f.os], respectively). Number of lambs produced per ewe exposed was affected by year (P <.01) and system x year interaction (P <.01 for number born; P <.10 for numbers raised and marketed). These effects followed those on conception rate and suggest that the poorer overall productivity of System 1 was attributable to the lower conception rate of yearling ewes. Repeatabilities for number of lambs born, raised and marketed per ewe exposed were.20 f.06,.14 f.05 and.08 f.w and reflect the high repeatability of conception rate in these data. Ewe body weights and condition scores at breeding were affected (P c.05) by system, year, system x year interaction and random ewe effects (Tables 3 and 4). Body weight was also affected (P c.05) by breed x year interaction. Differences among systems were consistent for older ewes (1980-born ewes lambing in 1983 and 1984 and 1981-born ewes lambing in 1984). For these ewes, ewes in System 3 were lightest (64.3 kg) and poorest in condition (mean score of 5.1) at breeding. Older ewes in Systems 1 and 2 differed little in weight at breeding (73.4 and 72.8 kg for System 1 and 2 ewes, respectively) or condition score (6.4 and 6.6, respectively). System 2 ewes tended to be heavier and fatter in 1983 but lighter and thinner in 1984. System 3 ewes were bred in October and November and had lactated throughout the

EWE PRODUCTIVITY IN DEFERENT SYSTEMS 19 TABLE 5. LEAST SQUARES MEANS AND S TANDM ERRORS FOR EWE FLEECE WEIGHTS (kg) BY SYSTEM, EWE BIRTH YEAR, EWE BREED AND YEAR Ewe birth Ewe System Year breed Year 1 2 3 Avg 1980 1/4-Pinn 1981 1.84 f.13 1.85 f.14 1.83 f.13 1.84 f.08 1982 2.93 f.13 3.17 f.15 2.98 f.14 3.03 f.08 1983 3.02 f.13 3.16 f.14 2.79 f.14 2.99 f.08 1984 3.09 f.13 2.91 f.14 2.70 f.14 2.90 f.08 A% 2.72 f.10 2.77 f.ll 2.58 f.10 2.69 f.06 Western 1981 2.02 f.10 2.02 f.12 2.05 f.12 2.03 f.07 1982 3.60 f.ll 3.71 f.12 3.70 f.13 3.67 f.07 1983 4.14 f.10 4.58 f.12 4.11 f.13 4.28 f.07 1984 4.01 f.10 4.01 f.12 3.49 f.13 3.84 f.07 Avg 3.44 f.08 3.57 f.09 3.34 f.09 3.45 f.05 1981 l/z-j?inn 1982 1.97 f.16 2.15 f.i5 1.87 f.16 2.00 f.09 1983 3.10 f.16 3.35 f.13 3.04 f.15 3.16 f.os 1984 3.09 f.16 2.97 f.13 2.73 f.15 2.93 f.os Avg 2.72 f.09 2.82 f.os 2.55 f.09 2.70 f.05 Overall avp 2.96 f.05 3.05 f.05 2.82 f.05 2.94 f.03 ~ %alculated as the average of 1980-born 1/4-Finns, 1980-born Westerns and 1981-born 1/4-Fi~s. summer months prior to placing of their lambs in drylot in early September. The stress of a long lactation during summer reduced breeding weights and condition scores. However, the mean condition score of 5.1 for these ewes was adequate for breeding and suggested that ewes had not experienced an unreasonable nutritional stress. System 1 ewes entered breeding in early June after lactating for only about 70 d and had access to highquality spring forage immediately before breeding. Their relatively high prebreeding body weights and condition scores thus are not surprising. System 2 ewes ended lactation in mid-may and thus had considerable time to recover body weight before beginning breeding in August. Differences among systems in prebreeding body weights and condition scores of yearling ewes did not correspond to those observed in older ewes. For 1980-born ewes, differences in prebreeding body weight were modest. However, weights and condition scores for 1981-born ewes were highest in System 2. This variation among ewe birth years in prebreeding weight of yearling ewes is not readily explainable and would be contingent upon the management system used to prepare replacement females for entry into the various systems. The ewe breed x year interaction was significant for prebreeding weight (Table 3) but not condition score. Among 1980-born ewes, Western ewes were somewhat lighter (.9 f 1.3 kg) than 1/4-Finn ewes in 1982, but they were slightly heavier in 1983 (by.5 f 1.1 kg) and considerably heavier in 1984 (2.1 f 1.1 kg). Thus, Western ewes continued to grow for a longer time than 1/4-Finn ewes. This pattern is consistent with the slower maturing rate of the Rambouillet relative to the Finnsheep (Notter et al., 1984). Random ewe effects on prebreeding weight also were significant. Repeatability was.58 f.05 for weight and.30 f.04 for condition score. For ewes that weaned lambs, ewe weight changes during the first 60 to 80 d of lactation differed (P <.05) among systems. Ewes in Systems 1 and 2 lost an average of 5.2 and 3.0 kg, respectively, during early lactation, whereas ewes in System 3 gained an average of 2.3 kg. Thus, ewes in Systems 1 and 2 were in early lactation before the onset of spring grazing and lost the most weight during this period, but they were able to compensate for this weight loss after weaning and were heaviest at their subsequent breeding. In contrast, ewes in System 3 gained weight during early lactation on spring pastures but had lower gains in the 4 mo before breeding than ewes in the other systems and weighed less at the start of breeding. Ewe weight losses in early lactation also were influenced by ewe age, with younger ewes experiencing smaller weight losses. Yearling ewes in Systems 1 and 2 lost an average

20 NOTTER AND McCLAUGHERTY 5.0 and 2.5 kg, respectively, and those in System 3 gained an average of 2.5 kg. In contrast, older ewes lost an average of 6.0 and 4.1 kg in Systems 1 and 2, respectively, and gained an average of 2.2 kg in System 3. Ewe condition scores at weaning were considerably less than those observed at breeding. Scores averaged 3.0, 3.8 and 3.7 for ewes in flocks 1, 2 and 3, respectively. No consistent differences in weaning condition scores were observed across ewe ages or between breeds. Fleece weight was affected by system (P <.05), breed (P <.001) and year (P <.001); by system x year and breed x year interaction (P <.001); and by random ewe effects (P <.001). Repeatability of fl- weight was.44 f.05, which was somewhat smaller than the value of.61 f.08 reported by Cochran et al. (1984) for Dorset and Finnsheep crossbred ewes at the same location, but it is within the range of values reported by Turner and Young (1969) for non-merino breeds. Differences among systems in mean fleece weight (Table 5) were small for yearling and 2-yr-old ewes but were larger for 3- and 4-yr-old ewes and favored ewes in Systems 1 (3.41 f.07 kg) and 2 (3.43 f.06) relative to those in System 3 (3.09 f.07 kg). This result is consistent with the longer lactations, lighter fall weights and lower fall conditions scores of flock 3 ewes. Western ewes had substantially higher fleece weights than did 1/4-Finn ewes (3.45 f.05 vs 2.69 f.06 kg for 1980-born ewes). This difference was considerably larger than the 10% reduction in mean fleece weight reported by Ercanbrack and Knight (1985) and Lewis and Burfening (1988) when 1/4-Finn ewes were compared to purebred Rambouillet, Targhee and Columbia ewes. Differences between ewe breed groups became larger with age, increasing from.19 f.ll kg in yearling ewes to.64 f.l 1 kg in 2-yr-old ewes and to 1.12 f.08 kg in older ewes. The 1/4-Finn ewes appeared to reach their mature fleece weights by 2 yr of age, whereas fleece weights in Western ewes continued to increase until ewes were 3 yr old. lmpllcatlons Late-fall lambing resulted in lower pregnancy rates in yearling ewes, but not in older ewes. holificacy also was lower (by about 1/4 of a lamb) in fall lambing. Ewes with 1/4-Finnsheep breeding produced about.25 more lambs/lambing and about.30 more lambs/exposure from these ewes were marketable than lambs from Western ewes. The 1/4- Finnsheep ewes were only slightly smaller than Western ewes, but they produced substantially lighter fleeces. Use of 1/4-Finnsheep ewes increased the overall rate of lamb production. Llterature Cited Anderson, G. B., G. E. Bradford and P. T. Cupps. 1981. Length of gestation in ewes carrying lambs of two different breeds. Theriogenology 16119. Clarke, L. S., D. R. Notter and J. S. Copenhaver. 1984. Effects of breed group and season on conception rate in the ewe. J. Anim. Sci. 59(SuppI. 1):155 (Abstr.). Cochran, K. P., D. R. Notter and F. S. McClaugherty, 1984. A comparison of Dorset and Fish Landrace crossbred ewes. J. Anim. Sci. 59:329. Dickerson, G. E. 1977. Crossbreeding evaluation of Finnsheep and some U.S. Breeds for market lamb production. North Central Reg. hbl. No. 246. University of Nebraska, Lincoln. Dufour, J. J. 1974. The duration of the breeding season of four breeds of sheep. Can. J. Anim. Sci. 54389. Dzalruma, I. M., D. J. Stritzke and J. V. Whiteman. 1982a. Fertility and prolificacy of ewes under two cycles of accelerated lambing. J. Anim. Sci. 54213. Dzahuna. J. M., J. V. Whiteman and R. W. McNew. 1982b. Repeatability of lambing rate. J. Aoim. Sci. 54540. Ercanbrack, S. K. and A. D. Knight. 1985. Lifetime (seven years) production of 1/4 and 1/2 Fish Landrace ewes from Rambouillet, Targhee and Columbia dams under range conditions. J. Anim. Sci. 61:66. Fahmy, M. H., T. M. Machtyre and H.W.R. Chancey. 1980. Date of lambing and reproductive performance of NewfoundIand "DU" breeds of sheep raised under extensive management in Nova Scotia. J. Anim. Sci. 5 1: 1078. Hanrahan, J. P. and J. F. Quirke. 1985. Contribution of variation in ovulation rate and embryo survival to within breed variation in litter size. In: R. B. Laud and D. W. Robinson (Ed.) Genetics of Reproduction in Sheep. pp 193-201. Buttenvorths, London. Harvey, W. R. 1982. Least-squares and maximum likelihood computer program. Ohio State Univ., Columbus (hfimeo.). Lewis, R. D. and P. J. Burfenjng. 1988. Comparison of Finnish Landrace crossbred ewes with Columbia, Rambouillet and Targhee ewes on western range. J. Anim. Sci. 661059. Notter, D. R. 1981. Repeatability of conception rate and litter size for ewes in an accelerated lambing system. J. Anim. Sci. 53643. Notter, D. R. and J. S. Copenhaver. 1980. Performance of Finnish Landrace crossbred ewes under accelerated lambing. I. Fertility, prolificacy and ewe productivity. J. Anim. Sci. 51:1033. Notter, D. R., C. L. Fenell and R. A. Field. 1984. Effects of breed and intake level on growth and feed efficiency in ram lambs. J. Anim. Sci. 58560. Notter, D. R., R. F. Kelly and F. S. McClaugherty. 1991. Effects of ewe breed and management system on

EWE PRODUCTIVITY IN DIFFERENT SYSTEMS 21 efficiency of lamb production: II. Lamb growth, in hill sheep. Anim. Prod. 7:75. survival andcarcass characteristics. J. Anim. Sci. 69: Tumer. H. N. and S.S.Y. Young. 1969. Quantitative 22. Genetics in Sheep Breeding. Cornell Univ. Press, Nugenf R. A.. IU and D. R. Notter. 1990. Effects of Ithaca, NY. cohabitation with white-faced ewes on estrous activity Whiteman, I. V., W. A. Zollinger, F. A. Thrift and M. B. of Hampshire and Suffolk ewes exposed to rams in Gould. 1972. Postpartum mating performance of ewes June. J. Anim. Sci. 68:1513. involved in a twiceyearly lambing program. J. Anim. Purser, A. P. 1965. Repeatability aad heritability 01 fertility Sci. 35:836.