Effects of ewe age and season of lambing on proli cacy in US Targhee, Suffolk, and Polypay sheep

Small Ruminant Research 38 (2000) 1±7 Effects of ewe age and season of lambing on proli cacy in US Targhee, Suffolk, and Polypay sheep D.R. Notter * Department of Animal and Poultry Sciences, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061-0306, USA Received 1 June 1999; accepted 27 March 2000 Abstract Factors affecting ewe proli cacy (number of lambs born per ewe lambing) were evaluated using records on 9705 Targhee, 12,721 Suffolk, and 7231 Polypay lambings from ocks participating in the US National Sheep Improvement Program. Factors considered were ewe age and season of lambing. Proli cacy was affected by age of the ewe (P<0.001) and was highest for ewes lambing between 4 and either 7 (Polypay) or 8 (Targhee and Suffolk) years of age. Effects of lambing at younger ages were similar for Targhee and Polypay. Relative to mature ewes, proli cacy was reduced by 0.6±0.7 lambs in 1-year-old ewes, about 0.3 lambs in 2-year-old ewes and about 0.1 lambs in 3-year-old ewes. Effects of ewe age were smaller in Suffolk, with reductions relative to mature ewes of 0.47, 0.13, and 0.01 lambs in 1-, 2-, and 3-year-old ewes. Proli cacy of older (>8-yearold) ewes was consistently reduced by 0.17±0.20 lambs relative to adult ewes. Within the 1-year-old age class, age in months further affected proli cacy of Suffolk and Polypay but not Targhee ewes. For these two breeds, proli cacy of 15-month-old ewes averaged 0.21 lambs more than that observed for 12-month-old ewes. Seasonal differences in proli cacy were large for Targhee and Polypay (P<0.001) but not for Suffolk (Pˆ0.23) and the distribution of lambings among seasons differed (P<0.001) for all breeds. 22% of adult Polypay ewes, but only 1% of Targhee and Suffolk ewes lambed in summer and autumn. Proli cacy was highest in winter and spring lambings (December through May) and did not differ importantly among months within this period. Average proli cacy of 3- to 6-year-old ewes lambing in winter and spring was 1.75 for Targhee, 1.91 for Suffolk, and 2.13 for Polypay. For age groups with at least 10 observations, proli cacy in autumn (September through November) was reduced by an average of 0.14 lambs in Suffolk and 0.24 lambs in Polypay (very few Targhee lambed in autumn). Proli cacy in summer (June through August) was reduced by an average of 0.18 lambs in Targhee and 0.31 lambs in Polypay (very few Suffolks lambed in summer). Adjustment of proli cacy records for effects of ewe age is therefore necessary in genetic evaluation programs. Seasonal effects on proli cacy are probably best handled by placing ewes that lamb in different seasons in different contemporary groups. # 2000 Elsevier Science B.V. All rights reserved. Keywords: Sheep; Reproduction; Proli cacy; Breeds; Seasonality 1. Introduction Animal recording programs worldwide have concentrated their efforts at genetic evaluation upon relatively easily measured traits such as milk * Tel.: 1-540-231-5135; fax: 1-540-231-3010. E-mail address: drnotter@vt.edu (D.R. Notter) production, growth rate, and yield and quality of ber at the expense of traits involving animal reproduction and tness. The reproductive traits are dif cult to measure and are strongly in uenced by management decisions, but are also of paramount economic importance. Thus, a comprehensive animal recording program must include measures of reproductive performance. 0921-4488/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved. PII: S 0921-4488(00)00144-9

2 D.R. Notter / Small Ruminant Research 38 (2000) 1±7 The US National Sheep Improvement Program (NSIP) is a comprehensive program for animal recording and genetic evaluation designed to serve the entire US sheep industry (Wilson and Morrical, 1991). The NSIP database includes records on proli cacy, de ned as the number of lambs born per ewe lambing. A central tenet of animal recording is that records must be corrected for nongenetic effects before being used in genetic evaluation. Such correction may be achieved through the use of adjustment factors or by placing animals that are treated differently or that produce under different environmental circumstances in different contemporary groups. The purpose of this study was to evaluate effects of ewe age and season of lambing on proli cacy in three US sheep breeds: Targhee, Suffolk, and Polypay. These breeds possess the largest numbers of NSIP records and recently changed from a restricted, within- ock genetic analysis to a comprehensive, national across- ock evaluation system (Notter, 1998). The accumulated data thus now permit direct estimation of adjustment factors using records from participating farmers. 2. Materials and methods Records of proli cacy (number of lambs born per ewe lambing) of Targhee, Suffolk, and Polypay ewes from ocks participating in NSIP between 1984 and 1994 were used for this study. The Targhee is a composite breed developed at the US Sheep Experiment Station in Idaho in the 1920s from crosses among Rambouillet, Columbia, and Corriedale. The Targhee is generally viewed as a dual-purpose breed for wool and meat production in improved range environments and possesses somewhat superior proli cacy and maternal ability compared to other US range breeds. Most Targhee ocks lamb once per year in the spring; 81% of the Targhee lambings used in this study occurred in March, April, or May. Records on 9705 lambings by Targhee ewes from 27 ocks were used in this study. Approximately 70% of the Targhee records came from nine ocks located in the state of Montana. The US Suffolk is a large meat breed with good proli cacy and maternal ability that is used heavily as both a terminal sire breed and a general purpose breed in improved environments. Suffolk ewes generally lamb once per year in winter or spring. Records from 12,721 lambings by Suffolk ewes from 82 ocks in 31 states from throughout the US were used for this study. The Polypay is also a composite breed developed at the US Sheep Experiment Station in the 1970s from four-way crosses among the Rambouillet, Dorset, Finnish Landrace, and Targhee (Hulet et al., 1984). The Polypay was developed as a proli c dam line with acceptable eece characteristics and reduced seasonality of breeding. It is often used in accelerated lambing systems under intensive or semi-intensive management. Records of 7231 lambings by Polypay ewes from 41 ocks in 23 states were available. Statistical analyses were performed using leastsquares procedures in the Statistical Analysis System (SAS, 1990). Factors considered in the study were ewe age and season of lambing. Analyses were carried out separately for each breed because of differences in management and location of ocks representing the various breeds. Only records resulting from naturalservice, single-sire matings were used. Ewe age in months was calculated for each lambing. Ewes lambing between 6 and 18 months of age were coded as 1-year-old ewes, ewes lambing between 19 and 30 months of age were coded as 2-year-old ewes, etc. Ewes were also combined into larger ewe age groups to compare ewes that were 1, 2, 3 through 6, or greater than 6 years old at lambing. The initial analysis of ewe age effects included effects of ock, contemporary group (nested within ock and de ned by the month and year of lambing), and ewe age using either actual age in years or the age groupings. These analyses provided estimates of average ewe age effects across ocks, years, and lambing seasons. Ewes in the 1-year-old class ranged from about 9 to 18 months of age at lambing, and were expected to differ by a considerable amount in level of maturity at lambing. Thus, a supplemental analysis was applied within the 1-year-old age class using a model that included effects of ock, contemporary group, and age in months at lambing. This analysis allowed investigation of ewe age effects on proli cacy within the youngest ewe age class. The nal set of analyses considered effects of season of lambing and of interactions between season and ewe age group. Seasonal effects were de ned by the month of lambing, or by grouping the months into four lambing seasons: winter (December through

D.R. Notter / Small Ruminant Research 38 (2000) 1±7 3 February), spring (March through May), summer (June through August), and autumn (September through November). Analyses to evaluate joint effects of season and ewe age included effects of ock, year of lambing, ewe age group, month or season of lambing, and the interaction of ewe age group with the month or season of lambing. 3. Results and discussion Proli cacy differed (P<0.001) among ewe age groups in all breeds. Least squares constants for average effects of ewe age on proli cacy across ocks, years, and seasons and expressed relative to the average of the adult (3- to 6-year-old) ewes are shown for each breed in Table 1. The pattern of ewe age constants differed somewhat among breeds. In both the Targhee and the Polypay, yearling ewes had 0.6±0.7 fewer lambs per ewe lambing than adult ewes, and 2- year-old ewes had about 0.3 fewer lambs per ewe lambing than adult ewes. For these two breeds, 3-yearold ewes had 0.07±0.11 fewer lambs per ewe lambing, suggesting that proli cacy is still increasing at 3 years of age in these breeds. The proli cacy of young Suffolk ewes was higher in relation to that of adult ewes than that observed in Targhee and Polypay. Proli cacy of 1- and 2-year-old Suffolk ewes was only 0.47 and 0.13 lambs lower, respectively, than that of adult ewes. The proli cacy of 3-year-old Suffolk ewes was essentially equal to that of older ewes. This result may re ect the generally high levels of management and nutrition commonly observed in purebred Suffolk ocks. It does not appear consistent with the characterization of the Suffolk as a late-maturing breed type. These results suggest that breeds which have large mature size and delayed compositional maturity in terms of fat deposition patterns may not necessarily have correspondingly delayed patterns of sexual maturation. Peak proli cacy was generally achieved between 4 and 8 years of age. Exceptions to this generalization include a somewhat sharper peak in proli cacy for the Targhee. Proli cacy of 5- and 6-year-old Targhee ewes averaged 0.06 higher than the proli cacy of 4- or 7- year-old ewes. Also, the proli cacy of Polypay ewes appeared to have already begun to decline by 8 years of age. Ewes that were more than 8-year-old at lambing had 0.17±0.20 fewer lambs per ewe lambing than the 3- to 6-year-old ewes. Thus, proli cacy did not exhibit consistent declines until after 7 years of age. Dickerson and Glimp (1975) used linear and quadratic regression to evaluate ewe age effects on proli- cacy in seven US breeds and obtained similar results to those of this study. Across all breeds, proli cacy was maximum at 5.9 years of age. Least-squares constants for ewe ages 1 through 9 years were 0.47, 0.27, 0.12, 0.01, 0.05, 0.07, 0.04, 0.03, and 0.15, respectively. However, predicted Table 1 Numbers of observations, least-square (LS) constants, and standard errors for ewe age effects on proli cacy a in Targhee, Suffolk, and Polypay ewes b Ewe age group (years) Targhee Suffolk Polypay No. LS constant No. LS constant No. LS constant 1 459 0.610.03 2304 0.470.01 2016 0.690.02 2 2784 0.300.01 3331 0.130.01 1795 0.320.02 3 2162 0.110.01 2487 0.010.01 1346 0.070.02 4 1615 0.000.02 1725 0.020.02 822 0.030.02 5 1164 0.050.02 1189 0.020.02 546 0.040.03 6 774 0.070.02 757 0.030.02 375 0.000.03 7 414 0.000.03 490 0.040.03 208 0.010.04 8 219 0.010.04 249 0.050.04 70 0.120.07 >8 c 114 0.170.05 189 0.190.04 53 0.200.08 b Constants are expressed relative to the average of the 3- to 6-year-old ewes. c Ewe age classes of >8 years were combined because of small numbers of observations.

4 D.R. Notter / Small Ruminant Research 38 (2000) 1±7 ewe age effects declined much more rapidly after 9 years that did those from the current study, suggesting that linear and quadratic regression may not be appropriate to describe ewe age effects in older ewes. Dickerson and Glimp (1975) also found that age effects in younger ewes were smaller for Suffolk than for Targhee. Least-squares constants for 1-, 2-, and 3- year-old ewes were 0.47, 0.25, and 0.10, respectively, for Suffolk and 0.62, 0.36, and 0.15, respectively, for Targhee. In other studies, Glimp (1971) reported that proli- cacy in several US breeds was maximized at 5 years and that 2-year-old ewes produced 0.19 fewer lambs than 3- to 6-year-old ewes. Hohenboken et al. (1976) reported that in Suffolk, Hampshire, and Willamette (a composite breed of 1/2 Columbia, 1/4 Dorset Horn, and 1/4 Border Cheviot), proli cacy was maximized at 4±6 years and that 2-year-old ewes produced 0.30 fewer lambs than adult ewes. Fahmy et al. (1980) found that proli cacy in the DLS (a composite breed of 1/2 Dorset, 1/4 Leicester, and 1/4 Suffolk) and Newfoundland (a native Canadian breed) was maximized at 3 years. The 1-year-old ewes produced 0.53 fewer lambs than adult ewes but 2-year-old ewes produced only 0.14 fewer lambs than adult ewes. Mean proli cacies for each ewe age group are shown in Table 2. Peak proli cacy was 1.72 for Targhee, 1.88 for Suffolk, and 2.05 for Polypay. Proli cacy of young Suffolk ewes equaled or exceeded that of young Polypay ewes, but these results are somewhat biased by the higher frequency of Polypay ewes that lambed in suboptimal lambing seasons. Table 2 Least-squares means and standard errors for proli cacy a for ewe age groups in Targhee, Suffolk, and Polypay ewes Ewe age group (years) Targhee Suffolk Polypay 1 1.150.03 1.410.01 1.380.01 2 1.450.01 1.750.01 1.750.01 3±6 1.720.01 1.880.01 2.050.01 >6 1.720.02 1.850.02 2.000.03 Effects of ewe age in months within the 1-year-old ewe age class are shown in Table 3. Ewe age constants are shown relative to the average of the 12-, 13-, and 14-month-old ewes, which were the most frequent classes for all breeds. Proli cacy varied (P<0.001) with ewe age within the 1-year-old class in Suffolk and Polypay but not Targhee. For Targhee, the proportion of the total lambings that were by 1-year-old ewes was lower (5%) than for Suffolk (18%) or Polypay (28%) and very few yearling ewes lambed outside the core ages of 12±14 months. Even within these core ages, there was no consistent trend in proli cacy. A consistent pattern of increasing proli cacy with age was observed within the Suffolk breed, with proli cacy increasing by 0.22 from the 9±11 months age class to the 15 months age class. Proli cacy of the older, 16- to 18-month-old ewe lambs declined somewhat, and a similar trend was observed in Targhee. Both of these breeds commonly have well-de ned annual lambing seasons. The older ewe lambs therefore usually represent animals that for some reason lambed outside the normal production schedule and Table 3 Numbers of observations, least-square (LS) constants, and standard errors for effects of ewe age in months on proli cacy a within the 1-yearold age class b Ewe age group (months) Targhee Suffolk Polypay No. LS constant No. LS constant No. LS constant 9±11 c 14 0.010.12 125 0.080.05 90 0.050.05 12 110 0.110.05 511 0.050.03 448 0.090.03 13 134 0.080.04 894 0.040.02 580 0.000.02 14 150 0.030.04 564 0.090.02 405 0.090.03 15 24 0.110.09 156 0.140.04 231 0.140.04 16±18 c 27 0.080.09 54 0.040.07 262 0.250.03 b Constants are expressed relative to the average of the 12-, 13-, and 14-month-old ewes. c Classes were combined because of small numbers of observations.

D.R. Notter / Small Ruminant Research 38 (2000) 1±7 5 may not be directly comparable to those that lambed at the more typical ages of 12±15 months. Proli cacy also increased with age in Polypay, but the pattern of increase was less consistent than in Suffolk. Ewes in the youngest age class (9±11 months) were actually slightly superior to 12-month-old ewes, but within the high-frequency age classes (12±15 months), proli cacy exhibited consistent increases. Proli cacy of 16 to 18-month-old Polypay ewes was much higher than that of younger ewes. This result likely re ects use of Polypay ewes in accelerated lambing systems. The number of Polypay ewes lambing at 16, 17, and 18 months of age was higher than for other breeds (146, 61, and 55, respectively) and proli cacy was particularly high for the 17- and 18-month-old ewes (constants of 0.13, 0.36, and 0.45 at 16, 17, and 18 months, respectively). Some of the ewes lambing at 17 and 18 months of age may have been lambing for the second time in accelerated systems. Unfortunately, these ewes could not be explicitly identi ed from NSIP data les. However, for purposes of data adjustment, these young, secondparity ewes may perform differently than ewes that are of similar age but are lambing for the rst time. Seasonewe age group frequencies for each breed are shown in Table 4. The seasonal distribution of lambing dates was different for the different breeds, and, for each breed, the distribution of lambing seasons different among ewe age classes (P<0.001 from Chi-square). Lambings by Targhee and Suffolk ewes were concentrated in winter and spring, although a few Suffolk ewes (about 1%) lambed in autumn and a few Targhee ewes (about 1.5%) lambed in summer (mainly June). Within these two breeds, Targhee ewes were most likely to lamb in spring (81%) whereas Suffolk ewes were most likely to lamb in winter (69%). Even though Polypay ewes are commonly used in accelerated lambing, the highest frequencies of lambings by this breed were still in winter and spring (43% each); lambings in summer (5%) and autumn (10%) were much less frequent. Yearling ewes of all breeds were much more likely to lamb in spring (88% for Targhee, 75% for Suffolk, and 72% for Polypay) whereas the proportions of ewes in the other age classes that lambed in spring were relatively consistent at about 81% for Targhee, 21% for Suffolk, and 30% for Polypay. The high frequency of spring lambing in yearling ewes is a re ection of the greater seasonality of breeding in these animals (Notter, 1992; Notter et al., 1998). Among Polypay ewes, the proportions of lambings in autumn increased with Table 4 Numbers of observations, least squares means, and standard errors for proli cacy a by ewe age class and season b for Targhee, Suffolk, and Polypay ewes Breed Ewe age class (years) Numbers of observations a Mean prolificacy c winter In spring In summer In autumn In winter In spring In summer In autumn Targhee 1 0 405 48 0 1.190.03 0.930.08 2 485 2253 41 5 1.470.03 1.500.02 1.320.09 3±6 1050 4604 58 3 1.700.03 1.790.02 1.560.08 >6 131 610 6 0 1.720.05 1.780.03 Suffolk 1 560 1726 9 9 1.460.03 1.460.02 2 2551 742 4 34 1.760.02 1.820.03 3±6 4899 1181 9 69 1.890.02 1.920.02 1.730.11 >6 715 205 1 7 1.850.03 1.880.05 1.690.08 Polypap 1 516 1455 34 11 1.590.05 1.410.04 1.210.11 1.310.19 2 1008 562 50 175 1.840.04 1.870.05 1.670.10 1.560.06 3±6 1416 976 258 439 2.110.04 2.150.04 1.670.06 1.800.05 >6 135 89 39 68 2.050.07 2.000.08 1.730.11 1.900.08 b WinterˆDecember through February, springˆmarch through May, summerˆjune through August, and autumnˆseptember through November. c Means are shown only for classes with 10 or more observations.

6 D.R. Notter / Small Ruminant Research 38 (2000) 1±7 ewe age from less than 1% in yearlings to 10% in 2- year-old, 14% in adult ewes, and 20% in older ewes. A similar pattern was observed for summer lambings by Polypays. Summer lambings accounted for only 2% of lambings by yearlings and 3% of lambings by 2-yearold but for 8 and 12% of lambings in adult and older ewes, respectively. Means for proli cacy by ewe age group and season are shown for each breed in Table 4. Season in uenced proli cacy in Targhee and Polypay (P<0.001) but not Suffolk (Pˆ0.23). Inspection of monthly means for proli cacy in preliminary analyses did little to con rm or deny the arbitrary groupings of months into the seasons shown in Table 4. For all breeds, proli cacy, averaged across ewe age classes, was highest in March (1.63 for Targhee, 1.81 for Suffolk, and 1.93 for Polypay), and second-highest in February (Suffolk, 1.76; Polypay, 1.93) or January (Targhee, 1.62). Among subclasses with at least 10 observations (Table 4), proli cacy was highest, and not signi cantly different, for winter and spring lambings, except for yearling Polypay ewes which appeared to be more proli c in winter (1.59) than in spring (1.41). Proli cacy was lowest in summer lambs and intermediate in autumn. For adult ewes lambing in winter and spring, mean proli cacy was 1.75 for Targhee, 1.91 for Suffolk, and 2.13 for Polypay. Adjustment for season increased differences in proli cacy between Suffolk and Polypay relative to those in Table 2. The mean difference between these two breeds of 0.22 lambs per ewe lambing for adult ewes is now consistent with the prediction of Dickerson (1977) and Maijala (1984) that each 0.01 increase in Finnish Landrace breeding was associated with an increase of approximately 0.01 in proli cacy. However, differences between Polypay and Suffolk ewes at younger or older ages were less than predicted by this simple rule. The seasonewe age group interaction was not signi cant for Targhee but was signi cant for Suffolk (P<0.01) and Polypay (P<0.001). For the Suffolk, very few ewes lambed in summer or autumn (Table 4). Most interaction effects in Suffolk were associated with these seasons and were not biologically interpretable. For the Polypay, the most obvious interactive effect was the lower proli cacy of yearling ewes in spring compared to winter which was not consistent with the higher proli cacy of adult ewes in spring. A secondary source of interaction was the higher proli cacy of the older ewes compared to adult ewes in summer and autumn. It may be that older ewes that are capable of conceiving in these suboptimal seasons can also maintain higher ovulation rates. Lewis et al. (1996) have cataloged a variety of interactive effects in uencing fertility in accelerated lambing systems, and comparable interactive effects on proli cacy may exist. Thus, understanding of interactive effects on proli cacy in Polypay may require more information about the dynamics of the lambing systems involved. A number of studies have reported that proli cacy in sheep is reduced in summer and autumn lambing (e.g. Notter and Copenhaver, 1980; Fahmy and LavalleÃe, 1990; Notter and McClaugherty, 1991). However, in contrast to the current study, proli cacy in spring has often been reported to be signi cantly higher than that observed in winter. Notter and Copenhaver (1980) reported that proli cacy was 0.31 higher in Finnish LandraceRambouillet ewes, 0.16 higher in 1/4-Finnish Landrace, 3/4-Rambouillet ewes, and 0.12 higher in SuffolkRambouillet ewes in April than in January. Similarly, Notter and McClaugherty (1991) found that proli cacy in March and April averaged 0.25 higher than that observed in December, January, and February. 4. Conclusions These analyses suggest that ewe age effects on proli cacy are relatively continuous in ewes lambing between 1 and about 9 years of age (Table 1). Strati cation of ewe ages into classes is probably acceptable for purposes of adjusting proli cacy data in animal recording programs, but when suf cient data are available, separate factors for each ewe age are probably superior. Ewe age constants were similar for Targhee and Polypay, but somewhat different for Suffolk (Table 1), suggesting that breed-speci c adjustment factors may be required. Signi cant effects of ewe age (in months) within the yearling age group in Suffolk and Polypay (Table 3) suggest that proli cacy in young ewes may be particularly sensitive to differences in age and require ner distinctions among ewe age groups.

D.R. Notter / Small Ruminant Research 38 (2000) 1±7 7 Seasonal effects on proli cacy were substantial, but differences within the main winter and spring lambing seasons were minor. Animals lambing in different seasons would normally be placed in different contemporary groups for purposes of genetic evaluation, and this approach is preferred to attempting to adjust proli cacy data for seasonal effects because these seasonal effects can easily be confounded with other management variables. Although signi cant ewe age classseason interaction was observed in Suffolk and Polypay (Table 4), available data provide little justi cation for season-speci c ewe age adjustment factors. Acknowledgements Partial nancial support for this study came from the US National Sheep Improvement Program. References Dickerson, G.E., 1977. Crossbreeding evaluation of Finnsheep and some U.S. breeds for market lamb production. North Central Regional Pub. No. 246, University of Nebraska, Lincoln. Dickerson, G.E., Glimp, H.A., 1975. Breed and age effects on lamb production of ewes. J. Anim. Sci. 40, 397±408. Fahmy, M.H., LavalleÃe, D., 1990. Productivity of Polypay, Dorset and PolypayDorset ewes under two accelerated lambing systems. Small Rumin. Res. 3, 269±281. Fahmy, M.H., MacIntyre, T.M., Chancey, H.W.R., 1980. Date of lambing and reproductive performance of Newfoundland and DLS breeds of sheep raised under extensive management in Nova Scotia. J. Anim. Sci. 51, 1078±1086. Glimp, H.A., 1971. Effects of breed and mating season on reproductive performance of sheep. J. Anim. Sci. 32, 1176± 1182. Hohenboken, W., Corum, K., Bogart, R., 1976. Genetic, environmental and interaction effects in sheep. I. Reproduction and lamb production per ewe. J. Anim. Sci. 42, 299±306. Hulet, C.V., Ercanbrack, S.K., Knight, A.D., 1984. Development of the Polypay breed of sheep. J. Anim. Sci. 58, 15±24. Lewis, R.M., Notter, D.R., Hogue, D.E., Magee, B.H., 1996. Ewe fertility in the STAR accelerated lambing system. J. Anim. Sci. 74, 1511±1522. Maijala, K.J., 1984. Review of experiences about the use of Finnsheep in improving fertility. In: Hofmeyr, J.H., Meyer, E.H.H. (Eds.), Proceedings of the 2nd World Congress on Sheep Beef Cattle Breed. South African Stud Book and Livestock Improvement Association, Bloemfontein, pp. 519±531. Notter, D.R., 1992. Genetic improvement of out-of-season breeding through selection. In: Morrical, D.G. (Ed.), Proceedings of the Out of Season Breeding Symposium. Iowa State University, Ames, pp. 55±81. Notter, D.R., 1998. The U.S. National Sheep Improvement Program: across- ock genetic evaluations and new trait development. J. Anim. Sci. 76, 2324±2330. Notter, D.R., Copenhaver, J.S., 1980. Performance of Finnish Landrace crossbred ewes under accelerated lambing. I. Fertility, proli cacy and ewe productivity. J. Anim. Sci. 51, 1033±1042. Notter, D.R., McClaugherty, F.C., 1991. Effects of ewe breed and management system on ef ciency of lamb production. I. Ewe productivity. J. Anim. Sci. 69, 13±21. Notter, D.R., Al-Shorepy, S.A., Vincent, J.N., McQuown, E.C., 1998. Selection to improve fertility in fall lambing. Proc. 6th World Congr. Genet. Appl. Livest. Prod. 27, 43±46. SAS, 1990. SAS Procedures Guide, Version 6, 3rd Edition. SAS Institute, Cary, NC. Wilson, D.E., Morrical, D.G., 1991. The National Sheep Improvement Program: a review. J. Anim. Sci. 69, 3872±3881.