J. Anim Sci. 2007. 85:2815-2829. doi:10.2527/jas.2006-064
© 2007 American Society of Animal Science
Breeding objectives for Targhee sheep1
R. C. Borg*,
D. R. Notter*,2,
L. A. Kuehn*,3 and
R. W. Kott
* Department of Animal and Poultry Sciences, Virginia Polytechnic Institute and State University, Blacksburg 24061; and
Department of Animal and Range Science, Montana State University, Bozeman 59717
 |
Abstract
|
|---|
Breeding objectives were developed for Targhee sheep under rangeland production conditions. Traits considered were those for which EPD were available from the US National Sheep Improvement Program and included direct and maternal effects on 120-d weaning weight (WW and MM, respectively); yearling weight (YW); yearling fleece weight, fiber diameter, and staple length; and percent lamb crop (PLC), measured as the number of lambs born per 100 ewes lambing. A bioeconomic model was used to predict the effects of a change of 1 additive SD in EPD for each trait, holding all other traits constant at their mean, on animal performance, feed requirements, feed costs, and economic returns. Resulting economic weightings were then used to derive selection indexes. Indexes were derived separately for 3 prolificacy levels (1.41, 1.55, and 1.70 lambs/ewe lambing), 2 triplet survival levels (50 and 67%), 2 lamb pricing policies (with or without discounting of prices for heavy feeder lambs), and 3 forage cost scenarios (renting pasture, purchasing hay, or reducing flock size to accommodate increased nutrient requirements for production). Increasing PLC generally had the largest impact on profitability, although an increase in WW was equally important, with low feed costs and no discounting of prices for heavy feeder lambs. Increases in PLC were recommended at all 3 prolificacy levels, but with low triplet survival the value of increasing PLC eventually declined as the mean litter size increased to approximately 2.15 lambs/ewe lambing and above. Increasing YW (independent of WW) increased ewe maintenance costs and reduced profitability. Predicted changes in breeding values for WW and YW under index selection varied with lamb pricing policy and feed costs. With low feed costs or no discounts for heavy lambs, YW increased at a modest rate in association with increasing WW, but with high feed costs or discounting of heavy lambs, genetic trends in WW were reduced by approximately 50% to constrain increases in YW. Changes in EPD for MM or fleece traits generally had smaller effects on profitability than changes in PLC, WW, and YW. Two indexes designed to address current rangeland production conditions (low forage costs and discounting of heavy feeder lambs) or more intensive and integrated production with retained ownership and value-based marketing of lambs (higher forage costs and no discounting of heavy lambs) were anticipated to meet the needs of most Targhee producers.
Key Words: breeding objective • growth • production efficiency • reproduction • selection index • sheep
|
INTRODUCTION
|
|---|
Aggregate breeding value is commonly defined as a linear function of breeding values (BV) for economically important traits. Weightings for individual BV in the aggregate represent partial regression coefficients of profit (or some other measure of economic merit) on BV for each trait (Hazel, 1943
) and may be estimated as the effect of a 1-unit change in each BV holding all other BV constant. Definition of aggregate breeding value may then be followed by development of selection strategies for genetic improvement. In many cases, nonlinear relationships exist between profit and BV, and economic weightings for BV in the aggregate depend upon the level of expression of other traits. In such situations, the definition of aggregate breeding value is expected to change over time and may differ among flocks.
The Targhee is a dual-purpose range breed developed in the 1920s for meat and wool production (Terrill, 1947). A number of potentially important nonlinear associations between profit and BV exist in Targhee sheep. Prices for wool and lamb differ with the quality of fleece and weight of lamb at market, respectively, and the distribution of litter size, which is categorically expressed, varies with ewe age, management, and lambing season. Therefore, changes in profit from a genetic change in performance may be influenced by means and phenotypic distributions for several traits.
A system for across-flock genetic evaluation of Targhee sheep was implemented in 1995 (Notter, 1998) and currently involves approximately 15 flocks and 1,600 breeding ewes. Expected progeny differences are estimated for 7 traits by the US National Sheep Improvement Program (NSIP; Bradford, 2003).
The objectives of this study were to develop breeding objectives for purebred Targhee flocks and to assess the sensitivity of these breeding objectives to changes in mean prolificacy and triplet survival rates, lamb pricing policy, and feed costs.
|
MATERIALS AND METHODS
|
|---|
Animal Care and Use Committee approval was not obtained for this study because the data were obtained from the existing NSIP database.
Traits included in the breeding objective were those for which EPD are calculated in the NSIP Targhee genetic evaluation (Table 1). This approach allowed aggregate breeding value to be predicted directly from published multitrait EPD by weighting each EPD by its economic value (Schneeberger et al., 1992). Economic weightings for EPD were determined from a bioeconomic model that predicted changes in animal performance, feed requirements, costs, and returns associated with changes in each EPD holding all others constant at their mean values. Differences in economic weightings associated with differences among flocks in ewe prolificacy, triplet lamb survival, lamb pricing policy, and costs of feed required to support increased performance were considered. Breeding objectives and associated selection indexes were developed for each combination of factors and correlated to one another and to various average breeding objectives to assess the extent of re-ranking of candidates for selection in different production conditions and the associated need for flock-specific breeding objectives.
Development of the Model
Flock Dynamics and Prediction of Performance
Most Targhee flocks are found in the northern plains of the United States including Montana and surrounding states, and the model was developed to describe production in this region. Commercial Targhee flocks rarely contain more than 1,500 ewes, but a base flock of 5,000 ewes was modeled for this study to reduce extraneous variation (Figure 1). No genotype x flock size interaction effects were included in the model; therefore, results were anticipated to be independent of flock size. Differences in flock size can have important effects on the variance of realized selection responses but should not affect the breeding objective unless changes in breeding value have different effects on costs and returns in flocks of different sizes. The model included stochastic and deterministic elements. As discussed below, stochastic procedures were used mainly to generate frequency distributions for a variety of class variables (e.g., ewe age classes, lamb survival groups, litters sizes born and reared, etc.) and to predict effects of nonrandom within-generation selection of replacement ewe lambs. Predictions of nutrient requirements and feed intake were predominantly deterministic because the parameters required to describe stochastic variation in these measurements are generally not available for rangeland production systems.
View larger version (20K):
[in this window]
[in a new window]
|
Figure 1. Diagram of a typical Targhee production system with a base ewe flock of 5,000 ewes and an average lamb crop of 1.55 lambs born per ewe lambing.
|
|
Ewes were assumed mated to sires of specified EPD to produce a single lamb crop. Costs and returns for this lamb crop were evaluated considering lambs sold and lifetime performance of ewe lambs retained for breeding. Three generations were thus considered including ewes in the base flock (generation 0), progeny of generation 0 ewes and selected sires (generation 1), and progeny of generation 1 replacement ewes mated to sires with breed-average EPD for all traits (generation 2). Profit was derived as lifetime income minus expenses for each cohort of lambs and expressed relative to the number of ewes in the base flock.
Economic values for each trait in the breeding objective were determined as the change in profit from use of sires with EPD that were 1 additive SD better than the mean for each trait. Additive SD and genetic correlations (Tables 1
and 2) were estimated using procedures of Notter and Hough (1997) and were consistent with those reported by Notter and Hough (1997), Bromley et al. (2000), Rao and Notter (2000), and Hanford et al. (2003). In Targhee, sires with EPD in the top 2 to 3% of the population for each trait generally differ in EPD from the population mean by approximately 1 additive SD. Therefore, changes in mean performance used to estimate profit could be realized by intensive selection for each trait.
Records of 2,102 lambings from 4 Montana flocks in 2000 through 2002 were used to determine flock dynamics and phenotypic distributions of performance traits. Frequencies of ewes in each age category were similar among flocks. A constant ewe age distribution was therefore assumed (Table 3). Average prolificacy of 2-to 6-yr-old ewes varied among the 4 flocks, ranging from 1.46 to 1.70 lambs born per ewe lambing. Three baseline levels of flock prolificacy (1.41, 1.55, or 1.70 lambs per ewe) were therefore considered (Table 3).
View this table:
[in this window]
[in a new window]
|
Table 3. Predicted mean litter sizes and proportions of single, twin, and triplet litters by ewe age for different flock prolificacy levels
|
|
Base ewes were randomly placed into fertility groups (pregnant or open), assuming 70% fertility in each of two 17-d estrous cycles starting November 1 (Kott and Thomas, 1987). Overall predicted fertility was thus 91% in 34 d (Iman and Slyter, 1993; Ercanbrack and Knight, 1998). Matings were assumed to be uniformly distributed within estrous cycles. Lambing occurred 147 d after the randomly assigned mating date. Litter size was assigned at random for pregnant ewes from an underlying normal phenotypic distribution with mean of zero and SD of 1.0. Threshold values for changes in litter size from 1 to 2 and from 2 to 3 were derived from NSIP Targhee records and located at 0.25 and 2.29 SD, respectively, for 2-yr-old ewes of medium prolificacy. Frequencies of birth types for other ewe ages and prolificacy levels were determined by adjusting the mean of the underlying scale (Table 3). In the base flock, a mean shift of 0.30 SD was equivalent to a change in percentage lamb crop (PLC) of 15%, and a change of 0.18 SD corresponded to a 1 additive SD change of approximately 8.8% in EPD for PLC.
Lamb survival varied with litter size. Single and twin lambs were assumed to have mean survival rates of 90 and 80%, respectively, but survival of triplet lambs was set at either 50 or 67% to reflect differences in lambing management (Kott and Thomas, 1987; Iman and Slyter, 1996). Lamb survival was randomly determined for each lamb based on these mortality rates and was assumed to be independent of ewe age (Smith, 1977) and lamb sex (Ercanbrack and Knight, 1985; Iman and Slyter, 1996). Most preweaning lamb mortality occurs within 5 d of birth (Safford and Hoversland, 1960); thus, rearing costs of lambs that died were not considered. Ewes were assumed to raise only their surviving lambs; fostering of lambs was not considered. Overall lamb survival was close to 80%, similar to literature values of 77.5% (Safford and Hoversland, 1960), 83.4% (Shelton, 1964), and 74.9% (Hanford et al., 2003).
Body weights and fleece characteristics were assumed to be normally distributed with means and phenotypic SD obtained from NSIP Targhee records (Table 4). Adjusted 120-d weaning weights (WW) were randomly assigned and then de-adjusted for average effects of type of birth and rearing, age of ewe, and sex of lamb using multiplicative adjustment factors (Bradford, 2003) to derive unadjusted 120-d WW. Each lamb was then also randomly assigned an adjusted birth weight from a normal distribution with a mean of 5.04 kg, phenotypic SD of 0.91 kg, and phenotypic correlation with adjusted WW of 0.32. The resulting adjusted birth weights were de-adjusted for type of birth, age of ewe, and sex of lamb (Bradford, 2003) to derive phenotypic values. Resulting mean birth weights of 5.7, 4.7, and 4.1 kg for singles, twins, and triplets, respectively, were consistent with NSIP records. Preweaning ADG was calculated for each lamb by subtracting the birth weight from WW and dividing by 120 d. Lambs were assumed to be marketed in a single group in mid September at an average age of approximately 150 d. Actual market weights were calculated for each lamb by multiplying actual age at this time by ADG and adding birth weight. The resulting mean market weight was 40 kg.
Ewe lambs kept as replacements in commercial flocks are generally heavier than the average of all available ewe lambs, which affects rearing costs and revenue from lamb sales. In our model, replacement ewe lambs were identified at an average age of 150 d and came from the heaviest 54 to 63% of ewe lambs, depending on numbers of lambs weaned and replacements required to maintain flock size. To account for subjective selection, a random 5% of the selected group were removed and sold as market lambs. Yearling weights of ewe lambs kept as replacements were calculated as the sum of 120-d WW and 245-d postweaning gain (PWG). The PWG was stochastically assigned assuming a mean of 0.064 kg/d, a SD of 0.025 kg/d, and a phenotypic correlation of –0.23 with WW (Notter and Hough, 1997). Phenotypic selection of heavier-than-average ewe lambs would tend to also select against lambs from multiple births, but such selection is common in commercial flocks and correlated genetic effects of such selection would be minor relative to sire selection based on PLC EPD.
Adult BW were not recorded in NSIP, but autumn weights were available for Targhee ewes from Montana State University at 1 through 6 yr of age (R. W. Kott, unpublished). In these data, the phenotypic correlation between BW in adjacent years averaged 0.60 and was consistent across ewe ages. The regression coefficient relating adult ewe weights to yearling weight (YW) EPD was approximately 1.25 kg/kg. Ewe weights at 2, 3, and 
4 yr of age were stochastically generated with means (Table 4) based on these data and including a regression of weight at age t on weight in the previous year. Replacement ewes were bred to lamb for the first time at 2 yr of age and remained in the flock until culled for reproductive failure, health problems, age (after 6 yr), or death. Mortality of 7.5% from weaning to first lambing and between subsequent lambings was assumed (Nass, 1977; Walker et al., 1993). Ewes that were not pregnant at the end of the breeding season were assumed to be culled.
Phenotypic values for fleece characteristics were stochastically assigned from normal distributions (Table 4). Yearling fleeces were harvested from replacement ewe lambs in February at an average age of 10 mo. Yearling fleece weights (FW) were modeled from NSIP records, which are adjusted to a constant age of 365 d, by de-adjusting for effects of age at shearing (Bradford, 2003). The 365-d adjusted FW was assumed to have a phenotypic correlation with adjusted WW of 0.34, and FW was then assumed to have phenotypic correlations with fiber diameter (FD) and staple length (SL) of 0.27 and 0.30, respectively. Phenotypic associations among WW, YW, FW, and FD were derived from NSIP Targhee records using methods of Notter and Hough (1997) and were used only to account for selection of replacement ewe lambs that were above average in WW; they were not used to predict effects of sire selection. Adult ewes were assumed shorn before lambing. Fleece weights for 4- and 5-yr-old ewes were assigned at random each year by adding 1.0 kg to the predicted yearling FW and then de-adjusting using factors from Bradford (2003) to model FW for breeding ewes of other ages. Adult fiber diameter measurements were generated from the yearling fiber diameter distribution (Table 4) by adding 0.7, 1.2, 1.6, 2.0, and 2.2 microns to yearling means for 2-, 3-, 4-, 5-, and 6-yr-old ewes, respectively (Atkins, 1990).
Dry Matter Intake and Feed Costs
The system used summer and winter range to meet forage requirements. The amount of dry matter available from winter range was the limiting resource. Flocks were assumed maintained on rangelands typical of the northern mixed-grass prairie. Dominant grasses included Idaho fescue (Festuca idahoensis), bluebunch wheatgrass (Agropyron spicatum), blue grama (Bouteloua gracilis), and western wheatgrass (Agropyron smithii). Winter range with an average TDN value of 55% (NRC, 1985; DM basis) was assumed utilized from mid August to mid May. Summer forage was available for the remaining months with an average TDN value of 65% (NRC, 1985). A grain (barley) supplement was provided for 30 d before the start of lambing, and ewes received supplemental grain and alfalfa hay in early lactation.
Energy requirements of ewes and lambs were predicted using procedures described in the Appendix and used to estimate dry matter requirements for summer and winter range and for gestation and lactation supplement. Prices for purchased feeds were 10-yr averages of regional prices (NASS, 2003) and were $0.074, $0.085, and $0.080/kg for grass hay (55% TDN), alfalfa hay (58% TDN), and barley (76% TDN), respectively (all TDN values on a DM basis). However, costs of forage DM can vary, depending on how rangeland is valued. Feed costs were therefore estimated for 3 scenarios, with additional forage requirements for increased production met by renting rangeland, purchasing hay, or reducing flock size to maintain a constant level of winter forage consumption. For renting rangeland, a ewe was assumed equal to 0.35 animal units (MontGuide Fact Sheet, 1997) at an average rental cost of $12.90/animal units/mo (NASS, 2003). Changing flock size in response to changes in performance assumed that a fixed amount of winter range was available to the flock, regardless of production level. If more DM was required due to an increase in production, fewer ewe lambs were assumed to be retained for breeding and numbers of generation 1 ewes and generation 2 lambs were reduced. In this scenario, associated reductions in alfalfa hay and grain supplement were assumed to lower feed costs, but potential revenue from unused summer range was not credited.
Income
Revenue from wool, market lambs, and cull ewes was derived using 10-yr-average prices. Prices for wool ($/kg) varied with fiber diameter at $1.77, $1.65, and $1.54/kg for fine (<23 micron), medium (23 to 25 micron), and coarse (>25 micron) fleeces, respectively, resulting in nonlinear pricing of wool relative to fiber diameter. Staple length also influences wool value; fleeces with SL below 7.35 cm are discounted (R. W. Kott, personal communication). However, few fleeces were below that threshold in these flocks, and discounts for short staples were not considered.
Two lamb pricing policies were considered. In the first, all market lambs were priced at $1.69/kg. In the second, heavy lambs were discounted such that lambs that weighed more than 43 kg were assigned a constant value of $72.67. This substantial discount for heavy lambs reflects limited opportunity to add weight and value to these lambs in the feedlot before harvest. In this production system, lambs are priced in truckload lots rather than individually; therefore, weight discounts are applied to the group as a whole based on the perceived proportion of lambs that are above some preferred weight. Records of prices and discounts are generally not available. Interviews with lamb buyers suggest that our pricing assumptions were consistent with actual pricing policies. At a mean market weight of 40 kg and a CV of 18% (Hanford et al., 2003), discounts for lambs at 1 and 2 SD above the mean would be $0.14 (8%) and $0.34/kg (20%), respectively. Options for retained ownership or alternative marketing were not considered. Cull ewe price was assumed constant at $0.66/kg. Income and costs generated after the first year of production were discounted to current values using the method of McClintock and Cunningham (1974) with a discount rate of 5.0%.
Economic Weights and Selection Indexes
Economic weights were derived from the predicted changes in profit associated with a change of 1 additive SD in sire genetic merit for each trait in Table 1, while holding sire EPD for all other traits constant. To account for variation arising from stochastic elements in the model, economic weights were derived from averages of 20 replicates of the model for the base flock and 20 replicates for each sire selection scenario. These sets of 20 replicates were repeated for each level of mean prolificacy (3) and triplet survival (2). To express weights in EPD units, economic weights were adjusted by dividing by the additive SD of the trait (Table 1), and index weightings were converted to relative weightings by expressing weights as a proportion of the index weight for WW. Breeding objectives for different scenarios were compared by calculating correlations between pairs of breeding objectives (Smith, 1983). Final indexes were determined by averaging economic weightings across scenarios, generally when correlations between breeding objectives exceeded 0.90.
|
RESULTS
|
|---|
Lamb and Ewe Performance
Comparison of predicted litter size frequencies and means and standard deviations for production traits (Tables 3 and 4) with values from NSIP records indicate that the model adequately represented current Targhee production levels. On a flock basis, predicted frequencies of single, twin, and triplet lambs differed by at most 0.035 from those observed in NSIP flocks. Predicted means and SD in Table 4 likewise differed from observed NSIP means and SD by at most 3%. Flocks with the highest prolificacy and triplet survival had, on average, the smallest predicted lamb weaning weights because more lambs were born and reared as twins and triplets. Average weaning weights of generation 1 lambs from the base flock thus ranged from 33.0 kg with high prolificacy and high triplet survival to 34.4 kg with low prolificacy and low triplet survival.
An increase of 1 additive SD in WW EPD resulted in a relatively consistent average increase of 0.90 kg in 120-d weight in generation 1 lambs, with one-half of this increase retained when replacement ewes were mated to rams of average BV to produce generation 2 lambs. If heavy lambs were not discounted, this increase in WW EPD resulted in a relatively consistent increase in average lamb value of $1.84 (2.8%) per lamb marketed in generation 1. However, if heavy lambs were discounted, an increase in WW EPD increased lamb value by an average of only $1.12 (1.75%). Changes in lamb value were larger in more prolific flocks, because fewer lamb weights reached the discounting threshold.
If both WW and YW are included in the breeding objective, YW effects arise only from differences in post-weaning growth. Thus, average weaning weights and lamb values in both generations were unchanged by selection on YW. An increase of 1 additive SD in YW EPD (independent of WW) thus resulted only in an increase in adult ewe size of 3.6 kg and an associated increase in maintenance requirements. An increase in WW EPD also resulted in an increase in adult ewe size of 2.5 kg because increased weaning weights were assumed maintained into adulthood. Prices for cull ewes ($/kg) were assumed to be unaffected by selection. Average values of cull ewes ranged from $34.32 to $34.70.
Effects of selection on maternal effects on 120-d weaning weight (MM), percent lamb crop (PLC), or wool characteristics in adult ewes were not realized until generation 2. An increase of 1 additive SD in MM resulted in an average increase in lamb weight of 0.71 kg. The associated increase in lamb value in generation 2 averaged $1.26/head (2.25%) when heavy lambs were not discounted, but only $0.78 (1.5%) with discounting of heavy lambs (Figure 2).
View larger version (11K):
[in this window]
[in a new window]
|
Figure 2. Changes in the average value of generation 2 lambs from selection for maternal milk (MM) or percent lamb crop (PLC) in flocks with different levels of prolificacy (H = high, M = medium, L = low) and triplet survival (H = high, L = low) with and without discounting of prices of heavy lambs.
|
|
An increase of 1 additive SD in PLC EPD increased numbers of lambs in generation 2 by 5.6, 4.9, and 4.5% for low, medium, and high prolificacy flocks, respectively, reflecting correlated changes in litter size distributions (Figure 3). More prolific flocks had larger increases in frequency of triplet births compared with flocks with lower prolificacy, but higher mortality rates for triplet lambs decreased the proportional rate of increase in lambs weaned. Selection to improve PLC decreased the average weight and value of generation 2 lambs, but also lowered the number of lambs that potentially received a discounted price. An increase of 1 SD in PLC EPD reduced average lamb value by an average of $0.44/head when prices for heavy lambs were discounted and $0.79/head when prices were not discounted (Figure 2). The decline in average lamb value associated with selection for PLC was smaller when ewes in the base flock were more prolific. Although average lamb values declined with selection for PLC, the total value of the generation 2 lamb crop increased by about 3.4% when prices were discounted for heavy lambs and by about 2.8% when prices were not discounted.
View larger version (7K):
[in this window]
[in a new window]
|
Figure 3. Mean numbers of lifetime lambs weaned by generation 2 ewes in the base flock and after selection for percent lamb crop (PLC) in flocks with different levels of prolificacy (H = high, M = medium, L = low) and triplet survival (H = high, L = low). Values in the columns are increases in lifetime number of lambs weaned from selection.
|
|
Effects of selection on fleece characteristics and fleece values were similar across prolificacy and triplet survival levels (Table 5). Figure 4 shows fleece values and proportions of fine (<22.09 micron), optimal (22.0 to 25.0 microns), and coarse (>25.0 microns) fleeces for yearling and adult ewes in the base flock and after selection for reduced FD. Yearling wool was finer, but total value per fleece was higher for adults, and adult ewes produced 75% of the total wool marketed. Selection to reduce FD had a greater impact on value in adult fleeces because of a larger reduction in the proportion of coarse fleeces in adults compared with yearlings. A reduction of 1 additive SD in FD EPD was associated with a 7% reduction in coarse wool and 12% increase in fine wool in adult ewes, but this selection reduced coarse wool by only 1 to 2% in yearling fleeces. Changes in fleece value from selection for SL were trivial; SL in the base flock was acceptable for current markets, and there was relatively little phenotypic variation in SL.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 4. Percentages of fleeces in different quality classes and average fleece values for adult and yearling fleeces in the base flock and after selection for reduced fiber diameter.
|
|
Feed Requirements
Increases in average DMI (kg) required per ewe bred in the base flock with selection for different traits and summed across generations are shown in Figure 5. Selection required either additional feed to support increased performance or reductions in flock size to accommodate a fixed amount of winter forage. For example, selection to increase WW increased winter forage requirements by 10.18 kg of DM per base ewe bred. If winter forage was assumed to be fixed, this requirement was met by retaining fewer ewe lambs for breeding, thereby reducing flock size by 2.15%. Selection for YW resulted in the largest increase in winter forage requirements or, if winter forage was fixed, in retention of the fewest replacement ewe lambs. In contrast, selection for fleece traits had little impact on DMI.
View larger version (11K):
[in this window]
[in a new window]
|
Figure 5. Increases in DMI from winter forage, summer forage, and gestation/lactation supplement per generation 1 ewe in generations 1 and 2 associated with different selection criteria. WW = weaning weight; MM = maternal milk; YW = yearling weight; FW = fleece weight; FD = fiber diameter; SL = staple length; PLC = percent lamb crop.
|
|
Selection for MM or PLC did not affect forage DMI until generation 2. Requirements for both summer and winter forage increased with selection for PLC. However, all lactation requirements came from summer range or lactation supplement; therefore, selection for MM resulted in little change in winter forage DMI.
Breeding Objectives and Selection Indexes
Changes in relative returns above feed costs are shown in Figure 6. Returns and resulting breeding objectives and selection indexes for scenarios involving renting pasture or reducing flock size to accommodate increased nutrient requirements were virtually identical and were henceforth combined into a single low feed-cost scenario that is contrasted to the high feed-cost scenario involving purchase of hay.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 6. Returns over feed costs across generations 1 and 2 associated with different selection criteria and with high or low feed costs and with or without discounting of prices of heavy lambs. Values are expressed as a percentage of returns over feed costs in the base flock. WW = weaning weight; MM = maternal milk; YW = yearling weight; FW = fleece weight; FD = fiber diameter; SL = staple length; PLC = percent lamb crop.
|
|
Economic weightings for each trait, expressed as the expected change in returns above feed costs per ewe bred in the base population in response to a 1 additive SD change in sire EPD for each trait, were calculated for all prolificacy, survival, forage cost, and lamb pricing scenarios, resulting in 36 different breeding objectives. However, some of these different objectives could be averaged to yield fewer unique breeding objectives (Table 6). Selection index weightings in Table 6 were scaled to represent the economic value of a change of 1 EPD unit in each trait, and further scaled to a value of 1.0 for WW EPD. Index weights in Table 6 can be directly applied to EPD, but are less biologically informative than economic weightings expressed on the additive SD scale (Figure 6).
Selection indexes for each of the 36 prolificacy, triplet survival, forage cost, and lamb price combinations were correlated with each other and with average selection indexes from Table 6. Correlations of 0.90 or larger suggest that selection on the different indexes would yield similar results (Smith, 1983). Correlations among pairs of indexes ranged from 0.429 to 0.999; the overall index had correlations with individual indexes that ranged from 0.850 to 0.989. Correlations of individual indexes with average indexes derived for each lamb pricing policy ranged from 0.928 to 0.987 and those with average indexes derived for each feed cost scenario ranged from 0.942 to 1.000. Selection indexes developed for each lamb price x forage cost combination were very highly correlated with their component indexes (0.986 to 1.000).
|
DISCUSSION
|
|---|
Percent lamb crop (PLC) EPD, defined as the number of lambs born per 100 ewes lambing, was the most important factor influencing profitability in all scenarios, which in large part explains the strong positive intercorrelations among breeding objectives. Relative to other traits in the breeding objective, effects of PLC were larger when prices for heavy feeder lambs were discounted. Increases in prolificacy in this situation increased both the number of lambs and total pounds of lamb marketed, but with fewer heavy lambs and, therefore, a higher average price per kilogram of lamb marketed. Somewhat greater emphasis on PLC was likewise observed when forage costs were high.
Increasing PLC also increases the frequency of triplet (and potentially quadruplet) births. However, because of the low predicted triplet survival rates (67 or 50%) and lower frequencies of triplet births in young ewes, proportions of ewes rearing triplets remained low following selection for PLC, at less than 1% with low triplet survival rates and at most 3.2%, even in high-prolificacy flocks with high triplet survival.
Many producers in range environments express a clear desire to avoid triplet births, but our results suggest that such a strategy will limit opportunities to increase profit. The impact on number of lambs weaned from increasing PLC was largest at the intermediate level of flock prolificacy, corresponding to an average litter size in adult ewes of 1.55 lambs per lambing (Figure 3). This result suggests that the relative positions of thresholds between single and twin, and twin and triplet births in ewes of different ages (Table 3) were most favorable for improvement of PLC at this level of prolificacy. Further increases in PLC led to diminishing returns in numbers of lambs marketed. With lamb survival rates of 90% for singles and 80% for twins, replacement of a single litter with a twin litter yields a net increase of 0.7 lambs. At a triplet survival rate of 67%, replacing a twin litter with a triplet litter resulted in a net gain of 0.4 lambs, but if triplet survival is only 50%, this replacement was associated with a net loss of 0.1 lambs/litter.
Fostering lambs from triplet litters to ewes nursing singles is common in some flocks, but was not considered here. Artificial rearing of surplus lambs from multiple births is not normally practiced in this production environment. Fostering is effective if overall lamb survival is improved, but little information on the survival of fostered lambs is available in the literature. Most lamb losses occur near the time of birth; therefore, fostering would not necessarily greatly increase lamb survival, and the effectiveness of fostering procedures likely vary considerably among flocks. In terms of our results, effective fostering would be approximately equivalent to an increase in triplet survival rates. Our breeding objectives were very highly correlated across different levels of ewe prolificacy and triplet lamb survival; therefore, fostering of triplet lambs is unlikely to affect our results.
Some decrease in emphasis on PLC relative to other traits was observed at the highest prolificacy level, but did not meaningfully change rankings of candidates for selection. Long-term increases in PLC were not explicitly considered, but effects of additional incremental increases of 1 additive SD were modeled for low forage costs, discounting of heavy lambs, and high or low triplet survival to assess changes in average weight of lamb weaned per ewe bred (Figure 7). With low triplet survival, the average weight of lamb weaned increased at a decreasing rate with increasing prolificacy. At the highest prolificacy level, the whole-flock frequency of triplet births was 38.8%. With low triplet survival, the frequency of ewes raising triplets remained below 5%, but overall lamb mortality exceeded 35%, a level that would be unacceptable to most producers. The average weight of lamb weaned in Figure 7 was projected (by second-degree polynomial regression) to peak at a whole-flock prolificacy level of 2.47 lambs born per ewe lambing, but using a prolificacy level of 1.55 lambs/ewe lambing (the medium prolificacy level in Table 3) as a base, 90% of the potential increase in total weight of lamb weaned was achieved at a prolificacy of approximately 2.15 lambs/ewe lambing. At this prolificacy level the whole-flock frequency of triplet births was 24%, but only 3.1% of ewes raised triplet litters and overall lamb mortality was 26%. This level of prolificacy slightly exceeds the whole-flock prolificacy level of 2.06 predicted for spring-lambing NSIP Polypay flocks with the age distribution shown in Table 3 (D. R. Notter, unpublished data). The Polypay was developed from Finnsheep crosses in order to create an optimally prolific breed for range production (Hulet et al., 1984). Wang and Dickerson (1991) and Amer et al. (1999) likewise reported that increased death losses caused the economic value for prolificacy to become negative when mean prolificacy exceeded approximately 2.1 lambs per ewe lambing.
View larger version (12K):
[in this window]
[in a new window]
|
Figure 7. Effect of flock prolificacy on average weight of lamb weaned with high or low triplet survival. Dashed and solid lines are second-degree polynomial prediction equations for high and low triplet survival, respectively.
|
|
With high triplet survival, the total weight of lamb weaned continued to increase in a nearly linear manner with increasing prolificacy. This result supports the contention of Wang and Dickerson (1991) and Amer et al. (1999) that maternal breeds would benefit from improvements in lamb survival, but there is little evidence of useful levels of genetic variation in survival (Safari et al., 2005). Thus, changes in management will likely be more effective than selection in improving triplet survival but may not be practical in extensive rangeland production systems.
These results suggest that periodic adjustments of the breeding objective should be adequate to maintain near-optimal indexes. Rates of genetic change in PLC with relatively intensive selection are unlikely to exceed 2%/yr (Bradford, 1985). Thus, a change in PLC of 10% could be approximately accommodated by recalculating breeding objectives every 5 to 10 yr (i.e., after 2 to 4 generations of selection). This result is consistent with the assertion of Brascamp et al. (1985) that most nonlinear profit relationships can be accommodated by a linear approximation because of the slow change in mean performance from selection.
Modification of indexes to reduce emphasis on PLC and limit increases in triplet births is straightforward, since PLC was considered to be genetically independent of other traits in the index (Rao and Notter, 2000). This independence allows emphasis on PLC to be changed by changing its coefficient in the index while maintaining the same relative weightings for remaining EPD. A coefficient of zero for PLC is expected to result in no change in the distribution of litter sizes, which would be preferred by some breeders. However, for the overall average index in Table 6, the anticipated change in returns above feed cost if PLC was held constant would be only $1.16/generation, 39% of that expected with optimal attention to increasing PLC. This simple approach to adjust index weightings is permissible only for traits that are genetically independent of all other traits in the index.
The genetic correlation between WW and YW is large and positive (0.74), but independent effects of the 2 weights on profitability were antagonistic (Figure 6). Selection for YW independent of WW increased ewe size and maintenance requirements, with no compensating effects on lamb value, and coefficients for YW EPD in breeding objectives were always negative. Genetic differences in YW that are independent of WW are small, but may indicate potentially important differences in postweaning gain, feed efficiency, carcass leanness at a fixed weight, or carcass weight at a fixed composition that could be advantageous with value-based marketing of feeder lambs or retained ownership of animals to harvest. However, no premiums for feeder lambs with greater postweaning growth potential could be documented in this production system.
Scenarios in Table 6 differ in the relative importance assigned to increasing WW vs. limiting YW to control ewe maintenance costs. Predicted changes in component traits associated with index selection based on EPD depend on accuracies of the EPD for each trait, and, therefore, on the information that is available to evaluate candidates for selection (Roughsedge et al., 2005). For progeny-tested rams with daughters in production (i.e., sires with at least modest accuracies for EPD for all traits), increases in WW and YW were predicted to occur if forage costs are low and heavy lambs are not discounted, but if forage is expensive and heavy lambs are discounted, little change was predicted in WW in order to avoid increases in ewe maintenance costs.
The impact of increasing ewe size on fitness under extensive grazing conditions remains subject to debate. If the range is heterogeneous and selective grazing is favored, larger ewes may be less capable of harvesting required quantities of preferred forages. NRC (1985) suggests that maintenance requirements may be increased by 50 to 70% under extensive grazing conditions. Such an increase was not modeled, but would almost certainly result in additional negative emphasis on YW EPD.
Summer grazing on public lands with a fixed cost per ewe regardless of production level was likewise not explicitly considered. This system would presumably further reduce feed costs relative to the low-cost scenarios shown in Figure 6 and Table 6 and allow somewhat greater weights on WW, MM, and PLC in the breeding objective, as well as less negative emphasis on YW. However, basing breeding objectives on the assumption of long-term, low-cost access to public grazing is likely questionable.
Commercial producers who purchase Targhee rams to produce both feeder lambs and replacement ewes should likely place negative emphasis on the YW EPD unless they can realize a premium price for lambs with superior feedlot performance or carcass merit, or unless summer feed costs are very low and independent of ewe maintenance requirements. Seedstock breeders, with longer planning horizons, must also consider future opportunities for their customers for value-based marketing or retained ownership of feeder lambs. Since 1970, average harvest weights of lambs in the United States have increased by approximately 0.50 kg/yr to a current mean of 63 kg (LMIC, 2005). Thus, while producers may not receive premiums for lambs with greater growth potential, failure to produce lambs that can meet changing weight targets may result in future discounts, and some positive emphasis on lamb postweaning growth potential may therefore be desired to ensure future competitiveness. Prediction of the long-term impact of changing YW is thus difficult, but breeding systems involving use of terminal sire breeds may be preferred to placing greater emphasis on postweaning growth in the ewe flock.
Effects of ewe milk production (MM) and fleece traits on breeding objectives were smaller than those associated with PLC, WW, and YW (Figure 6). The MM EPD was less important than WW EPD, but its importance varied in a similar manner across production scenarios. Thus, improvement in MM was less advantageous with discounting of heavy lambs or high winter forage costs. Wool contributed only about 10% of total receipts. Despite the low additional costs associated with improvement in FW and, particularly, FD, emphasis on these traits in the breeding objective remained low, supporting the greater emphasis on meat production observed in most US flocks. Staple length EPD had little economic impact at current mean SL. Nonlinear associations between wool quality and fleece value, such as those discussed by Fogarty and Gilmour (1993) for Australian Corriedale and Polworth ewes, thus had little impact in US Targhees.
The decision to use customized indexes such as those in Table 6 requires careful consideration (Fogarty and Gilmour, 1993; Koots and Gibson, 1998; Amer, 2000). Commercial producers generally wish to choose rams on criteria that are appropriate to their production situation, and the development of breeding objectives that are unique to individual flock circumstances is attractive to these producers. Seedstock producers, however, must consider the needs of potential customers as a group and develop correspondingly more general breeding objectives.
Traits included in the breeding objective in this study were those available to Targhee breeders from NSIP. The choice of traits that should appear in a breeding objective is inevitably somewhat arbitrary. In general, traits should be included if they influence costs, returns, or both, exhibit additive genetic variance that cannot be accounted for by other traits in the breeding objective, and can be measured in a cost-effective manner or predicted from other measured traits. Golden et al. (2000) have argued persuasively that EPD should be calculated and included in the breeding objective for all economically relevant traits, regardless of whether these traits can be directly measured. Ponzoni (1986) likewise recommended that reduction in feed intake be explicitly incorporated into the breeding objective and concluded that its omission reduced genetic gain in economic units by 30%. However, changes in feed intake beyond those predicted from changes in production level can be anticipated only if some of the traits in the selection criteria have genetic relationships with feed intake that differ from those predicted from changes in anticipated nutrient requirements. Thus, if 2 traits, A and B, have the same predicted relationship with nutrient requirements, but different relationships with feed intake, appropriate weightings of the 2 traits in a selection index can lead to changes in feed intake relative to nutrient requirements, and, presumably, to changes in efficiency of feed use. However, if the relationship of these traits to feed intake is simply a reflection of anticipated changes in nutrient requirements to support production, changes in feed intake independent of production are not anticipated. Genetic parameters to support relationships among commonly measured production traits and feed intake are, in some cases, available for growing animals under feedlot conditions, but are generally not available for extensively managed grazing animals. Feed intake was therefore not explicitly considered as a component of the breeding objective.
This study further assumed relative consistency of the production environment, even though periodic drought conditions occur and may limit feed availability and the carrying capacity of the range. The development of breeding objectives that allow for consistent genetic improvement, but can also optimally accommodate environmental fluctuations among years, is an important area, but is beyond the scope of the current study. Genotype x environment interactions involving specific plant communities, seasonal variation in forage availability and quality (e.g., between summer and winter range) or other manifestations of the physical or climatic environment may exist. Genetic relationships between production traits and the capability of the animal to harvest the additional forage necessary to support increased production may be particularly important. Unfortunately, these relationships are generally poorly documented and understood and could not be modeled with confidence. Our results thus mainly reflect effects of increasing productive capacity on predicted nutrient requirements without addressing the capacity of the animal to harvest those nutrients. Indeed, one may argue that measures of fitness in a range environment (e.g., ewe and ram fertility, lamb survival, ewe longevity) have not been adequately addressed in the current study. Limited consideration of such traits reflects the generally low accuracy of recording open ewes and the timing and causes of lamb and ewe deaths or disposal in these flocks, as well as the low heritabilities of these traits. However, efforts to improve recording of fitness traits, assess associations between fitness traits and reported EPD (Borg et al., 2005), and investigate the potential impact of negative phenotypic associations between production and fitness traits are under way.
In conclusion, selection indexes derived in this study should meet the needs of most Targhee producers. Some Targhee producers sell lambs that are not discounted for heavy weights, but most flocks do not have that option and should therefore focus on increasing prolificacy and limiting increases in ewe size and maintenance requirements. Differences among flocks in cost and availability of additional forage to support increased production will also affect the breeding goal. High forage costs again favor greater emphasis on limiting adult ewe size. Two alternative indexes that together should lead to near-optimal selection outcomes would be those for low forage costs and discounting of heavy feeder lambs to accommodate existing rangeland production conditions or for higher forage costs without discounting of heavy lambs to accommodate possible future opportunities for retained ownership and value-based marketing of lambs.
|
APPENDIX: SIMULATION OF ENERGY REQUIREMENTS
|
|---|
Energy requirements were estimated from ARC (1980) and NRC (1985) recommendations. Maintenance energy accounts for over 55% of total energy requirements for adult ewes and NE for maintenance was calculated as:
 | [1] |
where W is ewe weight (ARC, 1980). Additional NE requirements for gestation were considered only in the last 60 d of gestation. A single lamb with a 5.45-kg birth weight was assumed to require 253 MJ of NE over the last 60 d of gestation, with an adjustment of 52.11 MJ of NE per kg of fetus used to account for heavier or lighter litters.
Mean lactation curves for ewes nursing singles, twins, and triplets were derived from Ramsey et al. (1994, 1998; Figure 8). Milk production by ewes rearing twins and triplets was increased to 160 and 200%, respectively, of that of ewes of the same age rearing singles. Gross energy of milk (kJ/kg) was calculated as [2,035 + 34.45 x (g of fat/kg)] (ARC, 1980), which gave an estimated energy content of 5.18 MJ/kg at an assumed average of 9.13% milk fat (Ramsey et al., 1998). Efficiency of utilization of ME for milk production was assumed to be 0.60, and average production was 1.0, 1.6, and 2.0 kg/d for ewes nursing singles, twins, and triplets, respectively. A baseline value of 1,900 MJ of ME during lactation was derived for an adult ewe nursing a single lamb and was adjusted assuming that each additional kilogram of milk required an additional 7.9 MJ of ME (ARC, 1980). Energy requirements for fleece growth (MJ/yr) were spread equally over the year and calculated from the ADG in fleece weight (DFG, g) as [(DFG x 0.022 – 0.0138) x 365] (derived from White et al., 1979 and ARC, 1980).
Energy for maintenance and gain in lambs was calculated as averages for ewe and wether lambs and based on predicted BW, growth rates, and energy densities (ED) of gain in 5 stages of lactation: 0 to 30, 31 to 60, 61 to 90, 91 to 120 d, and after 120 d. Energy for maintenance and growth of lambs in the first 30 d of life was assumed to be derived only from milk, and milk production was predicted to be adequate to meet maintenance requirements of suckling lambs through 90 d of age. Digestibility of gross energy from milk at maintenance was 95%, and efficiency of ME use from milk for maintenance was 85%. In the first 30 d of lactation, NE for maintenance for milk-fed lambs was calculated as
 | [2] |
(ARC, 1980). After 90 d, milk alone did not meet maintenance requirements, and forage contributed to meeting maintenance requirements with NE for maintenance from equation [1]. During stages 2 and 3 of lactation (31 to 90 d), NE for maintenance of lambs was calculated as the average of equations [1] and [2]. Average digestibility of summer forage was assumed to be 65%, with ME requirements for maintenance of NE/0.85 assuming 3.08 MJ of ME/kg of TDN. Milk NE not required for maintenance was used to meet NE requirements for gain using a digestibility of milk gross energy of 93% and efficiency of milk ME for gain of 70%. Requirements for gain not met by milk were achieved from forage assuming 3.08 MJ of ME/kg of TDN and efficiency of ME use for gain of 0.422.
Energy densities of gains by twin-born lambs in the base ewe flock were predicted from ARC (1980), but multiplied by 0.925 to account for the larger frame size of US lambs (Table 7). There were few surviving triplet lambs; therefore, ED was assumed to be the same for triplets and twins. Single lambs were assumed to have ED values proportional to their increased BW relative to twin lambs of the same age.
View this table:
[in this window]
[in a new window]
|
Table 7. Predicted means for weight and energy density values for twin lambs in the base flock at different stages of growth
|
|
|
Footnotes
|
|---|
1 This research was supported in part by the US National Sheep Improvement Program (NSIP). The authors would also like to thank participating Targhee breeders and NSIP for access to data for the study. 
3 Current address: USDA, ARS, Roman L. Hruska US Meat Animal Research Center, Clay Center, NE 68933.
2 Corresponding author: drnotter{at}vt.edu
Received for publication February 3, 2007.
Accepted for publication June 28, 2007.
|
LITERATURE CITED
|
|---|
Amer, P. C. 2000. Trait economic weights for genetic improvement with SIL. Proc. N. Z. Anim. Prod. 60:189–191.
Amer, P. C., J. C. McEwan, K. G. Dodds, and G. H. Davis. 1999. Economic values for ewe prolificacy and lamb survival in New Zealand sheep. Livest. Prod. Sci. 58:75–90.[CrossRef]
ARC. 1980. Pages 114–119 in The Nutritional Requirements for Ruminant Livestock. Agric. Res. Counc., Commonw. Agric. Bureaux, Slough, UK.
Atkins, K. D. 1990. Incorporating parameters for lifetime productivity into breeding objectives for sheep. Proc. 4th World Congr. Genet. Appl. Livest. Prod., Edinburgh, UK 15:17–26.
Borg, R. C., D. R. Notter, R. W. Kott, and L. A. Kuehn. 2005. Genetic factors influencing body weights and condition scores in adult Targhee ewes. J. Anim. Sci. 83(Suppl. 1):102. (Abstr.)
Bradford, G. E. 1985. Selection for litter size. Pages 3–18 in Genetics of Reproduction in Sheep. R. B. Land and D. W. Robinson, ed. Butterworths, London, UK.
Bradford, G. E. (ed.). 2003. Breeding and selection. Pages 1–80 in SID Sheep Production Handbook, 2002 ed. American Sheep Industry Inc., Centennial, CO.
Brascamp, E. W., C. Smith, and D. R. Guy. 1985. Derivation of economic weights from profit equations. Anim. Prod. 40:175–180.
Bromley, C. M., G. D. Snowder, and L. D. Van Vleck. 2000. Genetic parameters among growth, prolificacy, and wool traits of Columbia, Polypay, Rambouillet, and Targhee sheep. J. Anim. Sci. 78:846–858.[Abstract/Free Full Text]
Ercanbrack, S. K., and A. D. Knight. 1985. Ten year linear trends in reproduction and wool production among inbred and noninbred lines of Rambouillet, Targhee and Columbia sheep. J. Anim. Sci. 71:341–354.
Ercanbrack, S. K., and A. D. Knight. 1998. Responses to various selection protocols for lamb production in Rambouillet, Targhee, Columbia, and Polypay sheep. J. Anim. Sci. 76:1311–1325.[Abstract/Free Full Text]
Fogarty, N. M., and A. R. Gilmour. 1993. Sensitivity of breeding objectives to prices and genetic parameters in Australian Corriedale and Polworth dual-purpose sheep. Aust. J. Exp. Agric. 33:259–268.[CrossRef]
Golden, B. L., D. J. Garrick, S. Newman, and R. M. Enns. 2000. A framework for the next generation of EPDs. Pages 2–13 in Proc. Beef Improv. Fed. Res. Symp. Annu. Meet., Wichita, KS.
Hanford, K. J., L. D. Van Vleck, and G. D. Snowder. 2003. Estimates of genetic parameters and genetic change for reproduction, weight, and wool characteristics of Targhee sheep. J. Anim. Sci. 81:630–640.[Abstract/Free Full Text]
Hazel, L. N. 1943. The genetic basis for constructing selection indexes. Genetics 28:476–490.[Free Full Text]
Hulet, C. V., S. K. Ercanbrack, and A. D. Knight. 1984. Development of the Polypay breed of sheep. J. Anim. Sci. 58:15–24.[Abstract/Free Full Text]
Iman, N. Y., and A. L. Slyter. 1993. Production of yearling Targhee or Finn-Dorset-Targhee ewes managed as a farm or range flock. J. Anim. Sci. 71:3206–3210.[Abstract]
Iman, N. Y., and A. L. Slyter. 1996. Lifetime lamb and wool production of Targhee or Finn-Dorset-Targhee ewes managed as farm or range flock: I. Average annual ewe performance. J. Anim. Sci. 74:1757–1764.[Abstract]
Koots, K. R., and J. P. Gibson. 1998. Effects of production and marketing circumstances on economic values for beef production traits. Can. J. Anim. Sci. 78:47–55.
Kott, R. W., and W. M. Thomas. 1987. Montana sheep IRM activities—Phase I. SID Res. Digest Winter:35–39.
LMIC. 2005. Livestock Marketing Information Center. USDA. http://www.lmic.info/ Accessed Dec. 1, 2005.
McClintock, A. E., and E. P. Cunningham. 1974. Selection in dual purpose cattle populations: Defining the breeding objective. Anim. Prod. 18:237–247.
MontGuide Fact Sheet. 1997. Determining forage production and stocking rates: A clipping procedure for rangelands. MT#199704 AG. http://www.montant.edu./wwwpb/pubs/mt9704.html Accessed Nov. 12, 2003.
NASS. 2003. National Agricultural Statistics Service: Montana. USDA. http://www.nass.usda.gov/mt/ Accessed Sep. 1, 2003.
Nass, R. D. 1977. Mortality associated with sheep operations in Idaho. J. Range Man. 30:253–258.
Notter, D. R. 1998. The U.S. National Sheep Improvement Program: Across-flock genetic evaluations and new trait development. J. Anim. Sci. 76:2324–2330.[Abstract/Free Full Text]
Notter, D. R., and J. D. Hough. 1997. Genetic parameter estimates for growth and fleece characteristics in Targhee sheep. J. Anim. Sci. 75:1729–1737.[Abstract/Free Full Text]
NRC. 1985. Nutrient Requirements of Sheep, 5th ed. Natl. Acad. Press, Washington, DC.
Ponzoni, R. W. 1986. A profit equation for the definition of the breeding objective of Australian Merino sheep. J. Anim. Breed. Genet. 103:342–357.
Ramsey, W. S., P. G. Hatfield, and J. D. Wallace. 1998. Relationships among ewe milk production and ewe and lamb forage intake in Suffolk and Targhee ewes nursing single or twin lambs. J. Anim. Sci. 76:1247–1253.[Abstract/Free Full Text]
Ramsey, W. S., P. G. Hatfield, J. D. Wallace, and G. M. Southward. 1994. Relationships among ewe milk production and ewe and lamb forage intake in Targhee ewes nursing single or twin lambs. J. Anim. Sci. 72:811–816.[Abstract]
Rao, S., and D. R. Notter. 2000. Genetic analysis of litter size in Targhee, Suffolk, and Polypay sheep. J. Anim. Sci. 78:2113–2120.[Abstract/Free Full Text]
Roughsedge, T., P. R. Amer, R. Thompson, and G. Simm. 2005. Development of a maternal breeding goal and tools to select for this goal in UK beef production. Anim. Sci. 81:221–232.[CrossRef]
Safari, E., N. M. Fogarty, and A. R. Gilmour. 2005. A review of genetic parameter estimates for wool, growth, meat and reproduction traits in sheep. Livest. Prod. Sci. 92:271–289.[CrossRef]
Safford, J. W., and A. S. Hoversland. 1960. A study of lamb mortality in a western range flock. I. Autopsy findings on 1051 lambs. J. Anim. Sci. 19:265–273.[Medline]
Schneeberger, M., S. A. Barwick, G. H. Crow, and K. Hammond. 1992. Economic indices using breeding values predicted by BLUP. J. Anim. Breed. Genet. 109:180–187.
Shelton, M. 1964. Relation of birth weight to death losses and to certain production characters of fall-born lambs. J. Anim. Sci. 23:355–359.[Abstract/Free Full Text]
Smith, C. 1983. Effects of changes in economic weights on the efficiency of index selection. J. Anim. Sci. 56:1057–1064.[Abstract/Free Full Text]
Smith, G. M. 1977. Factors affecting birth weight, dystocia, and preweaning survival in sheep. J. Anim. Sci. 44:745–753.[Abstract/Free Full Text]
Terrill, C. E. 1947. Breed crosses used in the development of Targhee sheep. J. Anim. Sci. 6:83–92.[Abstract/Free Full Text]
Walker, J. W., H. A. Glimp, S. L. Kronberg, and T. R. Kellom. 1993. When less may mean more: Studies on range versus shed lambing in a cold desert shrub ecosystem. Prof. Anim. Sci. 9:153–158.
Wang, C. T., and G. E. Dickerson. 1991. Simulation of life-cycle efficiency of lamb and wool production for genetic levels of component traits and alternative management options. J. Anim. Sci. 69:4324–4337.[Abstract]
White, D. H., B. N. Nagorcka, and H. A. Birrell. 1979. Predicting wool growth of sheep under field conditions. Pages 139–161 in Physiological & Environmental Limitations to Wool Growth. J. L. Black and P. J. Reis, ed. Univ. New England, Armidale, Australia.
Top
Abstract
INTRODUCTION
