| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Dairy Production Department, Teagasc, Moorepark Production Research Centre, Fermoy, Co. Cork, Ireland
Department of Animal Science, Faculty of Agriculture, University College Dublin, Belfield, Dublin 4, Ireland
Institute for Animal Science and Health (ID-Lelystad), P.O. Box 65, 8200 AB Lelystad, The Netherlands
Corresponding author:
Donagh P. Berry; e-mail:
dberry{at}moorepark.teagasc.ie.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: body weight • body condition score • fertility • selection index
Abbreviation key: AVGBCS = average body condition score, AVGBW = average body weight, Cumfat240 = cumulative fat yield to d 240 of lactation, Cummilk120, Cummilk180, Cummilk240 = cumulative milk yield to d 120, 180, and 240 of lactation, respectively, Cumprot240 = cumulative protein yield to d 240 of lactation, CVg = coefficient of genetic variation, DairyMIS = Dairy Management Information System, FSCO = first service to conception interval, HUK = Holstein United Kingdom, IFS = interval to first service, NS = number of services per cow, PR63 = pregnant 63 d after the start of the breeding season, PRFS = pregnant to first service
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Most estimates of genetic (co)variances between BCS, BW, and fertility traits have been derived from datasets containing small numbers of animals with repeated BCS/BW observations (Veerkamp et al., 2000; Pryce et al., 2001) or large numbers of animals with a single BCS/BW observation per animal (Moore et al., 1992; Veerkamp et al., 2001; Royal et al., 2002b). These studies indicated negative genetic correlation between BCS and fertility-related interval traits (Dechow et al., 2001; Pryce et al., 2001; Veerkamp et al., 2001; Royal et al., 2002b). Dechow et al. (2002) reported that cows that are genetically inclined to lose more BCS in early lactation will have a prolonged calving to first service interval. Many studies (Darwash et al., 1997; Royal et al., 2000) have used milk progesterone assays as indicators of the commencement of luteal activity. Using this methodology, Royal et al. (2002b) reported a significant genetic relationship between commencement of luteal activity and BCS; they suggested that each unit increase in BCS (scale 1 to 9) would bring forward the commencement of luteal activity by on average approximately 6 d. Veerkamp et al. (2001) reported that BCS in early lactation showed a stronger positive genetic correlation with first-service conception rate and 56 d nonreturn rate to first service than BCS in later lactation.
Few studies have estimated genetic correlations between BW and fertility traits. Moore et al. (1992) estimated a low but favorable genetic correlation (-0.07) between BW at calving and days open in first-lactation Holstein cows. A similar genetic correlation was observed by Veerkamp et al. (2000) on first lactation animals using milk progesterone assays as an indicator of the commencement of luteal activity. This correlation was strongest for BW at wk 15 (-0.54). Veerkamp et al. (2000) also reported a strong negative genetic correlation (-0.80) between BW change in the first 15 wk of lactation and commencement of luteal activity. However, neither of the above studies adjusted BW for differences in BCS prior to estimating the genetic correlations between BW with fertility despite BCS being reported to explain 12 to 45% of the genetic variation in BW (Veerkamp and Brotherstone, 1997; Berry et al., 2002).
The objective of this study was to estimate genetic (co)variances between BCS, BCS change, BW, BW change, milk production, and fertility traits on a large number of animals with several BCS, BW, and milk test-day observations per animal and assess the subsequent effects of alternative selection objectives on the genetic response for each trait.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Pedigree Information
Of the cows available 48% were herd book registered with Holstein United Kingdom (HUK). Four generations of ancestry on the paternal and maternal side were identified for 92 and 66% of these cows, respectively. For the remaining 52% of the cows not registered with HUK, sire and maternal grand sire were obtained from DairyMIS. For these herds the paternal ancestry and maternal grand sire ancestry was provided by HUK to the same level as for the pedigree cows. The proportion of North American Holstein-Friesian genetics for each sire/maternal grand sire present in the data set were calculated as outlined by Berry et al. (2002).
Within the edited data set, there were 818 different sires with daughters. The number of daughters per sire ranged from 1 to 415, and the average was 10.1 daughters per sire. In total 6340 of the cows with records had identified maternal grandsires. The additive genetic relationship matrix included 20,613 animals.
BCS, BW and Milk Production Traits
A total of 7250 cows from 66 herds had recorded spring calving dates and greater than three BCS or BW records or both. Body condition score was measured on a scale of 1 (thin) to 5 (fat) with increments of 0.25 (Lowman, 1976). Body condition score and BW at d 5, 60, 120, 180, and 240 (BCS only) were estimated for each cow using a smoothing spline as outlined by Berry et al. (2002). Average BCS (AVGBCS) and average BW (AVGBW) were calculated as the mean of these estimated test day records, in which the animal had all records present; this minimized bias by ensuring all cows had the same number of BCS/BW records evenly dispersed throughout the lactation. Body condition score change and BW change were calculated as the difference between the traits of interest (where animals had an estimate for both test days). Cows with missing values for any variable were coded as such but remained in the analyses for the rest of the variables.
A total of 8541 cows from 78 herds had recorded spring calving dates and milk records. Test-day records for each cow were obtained from the Irish Dairy Recording Cooperative. Three cumulative milk yield traits to d 120, 180, and 240 (Cummilk120, Cummilk180, Cummilk240) of lactation, fat yield to d 240 of lactation (Cumfat240), and protein yield to d 240 of lactation (Cumprot240) were also derived fitting a smoothing spline for each cow (Berry et al., 2002).
Reproductive Traits
Five fertility variables similar to those used internationally (Grosshans et al., 1997; Veerkamp et al., 2001; Evans et al. 2002) were calculated: interval to first service (IFS), first service to conception interval (FSCO), pregnant to first service (PRFS), number of services per cow (NS), and pregnant 63 d after the start of the breeding season (PR63). The start of the breeding season for each herd was defined as the first service date recorded in that herd; start of breeding dates were available for both years of the study. In total 8315 cows had identified first-service records. In Ireland most farmers use AI for the first 6 wk of the breeding season and natural mating thereafter. Ninety-two percent of farmers observed cows more than twice daily for estrus during the breeding season, while 99% of farmers used tail paint or a vasectomized bull or both as an aid to estrus detection. This facilitated all services to be accurately recorded. Beginning 40 to 50 d after the start of the breeding season all herds were visited on three or four occasions, at approximately 40-d intervals, to perform pregnancy diagnosis by transrectal ultrasound imaging (Aloka 210D * II, 7.5 MH3). Cows that were inseminated at least 28 d and not observed in estrus again after insemination were scanned to confirm pregnancy. Subsequently, all cows in the study were determined to be pregnant or not by rectal palpation at least 56 d after the end of the defined breeding season.
Data Analysis
A multivariate analysis for all 26 traits simultaneously was not computationally feasible. For this reason, a series of bivariate analyses were carried out in ASREML (Gilmour et al., 2002). Herd-season groups were formed. Season was defined as month of calving. Herd-season groups with less than four cows had their records moved into an adjoining season group from the same herd to facilitate a more accurate estimate. Holstein percentage of the cows in the present study varied from 0 to 75%. Average Holstein percentage was 53%. However, the maximum Holstein percentage a cow on the study may have is 75% as the maternal granddam is assumed to have 0% Holstein genes. The latter was assumed because most maternal granddams, and their proportion of Holstein-Friesian genes were unknown, in addition the base population in Ireland prior to the mideighties was predominantly British Friesian. Least squares means for all traits were calculated using ASREML for cows with 0 and 75% Holstein genes. The number of cows with 0 and 75% Holstein genes were 426 and 1457, respectively. The output also supplied the standard error of the difference, which was used to calculate the significance of the Holstein effect at the 5% level (1.96 * standard error of the difference).
The average length of the breeding season was 15 wk across the 78 herds in both years of the study. This ranged from 9 to 25 wk for individual herds. Senatore et al. (1996) showed that PRFS was positively related to the number of ovulations prior to first service. The number of services per cow will be influenced by the length of the breeding season in each herd. For these reasons, a quadratic polynomial regression for both the number of days between calving and the start of the breeding season and between calving and the end of the breeding season were added to the model when analyzing the fertility traits.
The following linear animal model was used for the univariate and bivariate analysis of all fertility related traits:
 |
where
| Yijkpq | = | observation for trait p on animal i
|
| µp | = | overall mean for trait p,
|
| HYSj | = | fixed effect of herd by year by month of calving interaction (j = 574),
|
| lk | = | fixed effect of lactation number (k = 1, 2, 3, 4+),
|
 | = | fixed effect of a quadratic polynomial regression for the percentage of North American Holstein-Friesian genes,
|
 | = | fixed effect of a quadratic polynomial regression for the number of days between calving and the start of the breeding season,
|
 | = | fixed effect of a quadratic polynomial regression for the number of days between calving and the end of the breeding season,
|
| ai | = | random additive genetic effect,
|
| PEi | = | random permanent environmental effect for each cow, and
|
| eijkpq | = | random residual term.
|
The model applied to the nonfertility traits was the same except the quadratic regression on calving to the start of the breeding season and the quadratic regression on calving to the end of the breeding season were not included. Additive genetic, permanent environmental, and residual covariances between traits were estimated from the bivariate analysis.
Following the analyses, the genetic correlations estimated between all 26 traits were incorporated into a 26 x 26 matrix. Due to the large number of bivariate analysis carried out, some of the eigenvalues of the correlations matrix were negative and were therefore made positive, and the correlation matrix recalculated using the eigenfunctions (Hill and Thompson, 1978). The matrix was made positive definite to facilitate subsequent selection index calculations. In this new positive definite matrix, 77% of the correlations had changed by less than 0.05, and 95% of the correlations changed by less than 0.10. The standard errors, however, were not adjusted since the change in correlations were so small and are thus likely to have little effect on the standard errors of the correlations. In the present study, estimates of those genetic correlations are presented.
Adjustment of a trait (e.g., trait 1) for differences in another trait (e.g., trait 2) in the present study was achieved by including trait 2 as a covariate for trait 1 in the model and the parameters subsequently reestimated.
Selection Index Methodology
The effects of four alternative selection objectives on responses per trait were studied using selection index theory (Hazel, 1943). Traits of interest were milk yield, fat yield, protein yield, AVGBW, AVGBCS, IFS, and PR63. The four alternative selection objectives were 1) selection for increased milk production based on the yield index (shown below); 2) selection on the yield index with an economic weight of -20% on AVGBW (relative to protein yield in genetic SD terms); 3) selection on the yield index with an economic weight on AVGBCS to restrict change in PR63; 4) selection on the yield index with an economic weight on PR63 to obtain no change in PR63. All four selection objectives assumed that milk, fat, and protein yields were available from 100 half-sib daughter groups. For selection objective 2, it was assumed that daughters also had records for AVGBW, while selection objective 3 assumed measurements on all daughters for AVGBCS, as well as the production records. Selection objective 4 was derived as if records were available for PR63 on all daughters, as well as the production records.
The yield index was defined as milk yield + fat yield + protein yield with chosen economic values consistent with those currently adopted in the economic breeding index for Ireland (Veerkamp et al., 2002); €-0.076, €0.90, and €5.70 for milk yield, fat yield, and protein yield, respectively. The accuracy of the breeding value for PR63 was also calculated over different-sized half-sib daughter groups with three alternative selection indexes (a) PR63, (b) AVGBCS, and (c) PR63 + AVGBCS; the breeding objective was PR63. The genetic and phenotypic parameters used for the calculation of the optimal index weights in all index calculations were those observed in the present study. The vector of optimal index weights (b) was calculated for each of the objectives as b = P-1Ga where P-1 = the inverse of the phenotypic (co)variance matrix of the traits in the selection index, G = the genetic covariance matrix between traits in the selection goal and the selection index, and a = the vector containing the economic values for the goal traits.
The genetic change per trait from selection on each of the four objectives was calculated as 
where Rj = a vector with the genetic change per trait j, i = selection intensity (in the present study this was assumed to equal one), b' = transpose of the vector containing the index weights, Gj = the jth column of a G matrix containing the genetic covariances between the trait j and the index traits; and 
i = the standard deviation of the index used. The expected response from selection using the objectives is illustrated in the present study as genetic gain following one cycle of selection with a standardized selection intensity of one.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
and 2, respectively. The sum of IFS and FSCO equate to the number of days open (90 d), which corresponds to a calving interval of 373 d (the average gestation length was 283 d). Heritabilities for BCS and BW were larger than those reported for milk yield. Heritability estimates for the fertility traits were all less than 0.03, yet a considerable genetic variation existed for some of these traits; FSCO and PRFS both showed a coefficient of genetic variation (CVg) of greater than 10%, while IFS exhibited the lowest CVg (2.4%). The CVg for BCS, BW, and milk production were all less than 7% with the exception of Cumfat240, which was 9.1%. No permanent environmental variance existed for IFS and PR63 with the univariate analysis; hence, this component was not included in subsequent bivariate analysis for either of the two traits.
|
|
Genetic correlations among BCS and BW at the same days in milk varied from 0.37 to 0.47. Table 3 shows the genetic correlations between BCS and BCS change at different stages of lactation with milk production. All correlations were negative and ranged from -0.53 to -0.14 and were stronger than those previously reported from a subset of the data (Berry et al., 2002).
|
The genetic correlations between BCS, BW, and milk production with the five fertility traits are summarized in Tables 4, 5, and 6, respectively. Irrespective of the stage of lactation, BCS was positively correlated with PRFS and PR63 and negatively correlated with IFS and NS. Body condition score was genetically uncorrelated with FSCO. Body weight throughout lactation consistently exhibited negative genetic correlations with IFS, PRFS, and PR63, while positive genetic correlations were evident between BW with both FSCO and NS. Adjustment of BW for differences in milk yield had very little effect on the genetic correlations between BW and fertility with the exception of BW in mid- to late lactation, which became more negatively correlated with IFS.
|
|
|
Selection Indexes
The phenotypic and genetic correlations used to calculate the optimal index weights are shown in Table 7, while the expected responses from selection on the four alternative selection objectives are summarized in Table 8. Selection on the yield index alone is predicted to reduce AVGBCS but have little effect on AVGBW, suggesting that the reduction in BCS is compensated by an increase in body size. Interval to first service is expected to decrease, as is PR63 following selection on the yield index alone.
|
|
An economic weight of 45% (relative to protein yield in genetic SD terms) on BCS was required within the selection objective so as to have no effect on PR63. This selection objective reduced the expected response in protein and milk yield by 19 and 50%, respectively, over a selection objective based on the yield index alone.
To restrict the expected response in PR63 to zero—assuming only the three milk production traits and PR63 are in the index—it was necessary to apply an economic weight of 36% to PR63 in the selection objective (relative to protein yield in genetic SD terms). This implies that similar relative economic weighting in genetic standard deviation terms would have to be applied to either BCS or PR63 to restrict change in PR63 to zero. This relative weight is similar to that applied to milk yield in the index assuming a progeny group size of 100 half-sib daughters. Expected responses in both AVGBW and AVGBCS were negative using this selection index.
The inclusion of only PR63 in the selection index served as a better predictor (based on the accuracy of the selection index as a representation of the selection goal) of PR63 in the selection objective than the inclusion of only AVGBCS in the index, when the number of daughters with records was greater than 18 daughters per sire (Figure 1). However, a combined index of PR63 and BCS consistently produced a more accurate estimate of the breeding value for PR63 up to 100 daughters per sire than an index with only PR63.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Variance Components
Heritability estimates from the present study are very similar to most other international studies, which looked at BCS (Veerkamp and Brotherstone, 1997; Gallo et al., 2001; Pryce et al., 2001), BCS change (Pryce et al., 2001; Dechow et al., 2002), BW (Veerkamp and Brotherstone, 1997; Veerkamp et al., 2000; Søndergaard et al., 2002), milk yield (Veerkamp and Brotherstone, 1997; Pryce et al., 1998; Dechow et al., 2001), and fertility traits (Grosshans et al., 1997; Pryce et al., 1998). However, Royal et al. (2002a) reported large heritability estimates for some endocrine fertility parameters and stressed their potential for inclusion in a selection index to improve fertility. The lack of a permanent environmental effect for IFS and PR63 may be due to the seasonal calving system adopted in Ireland (i.e., animals with short IFS, which conceive to that service in 1 yr, will have long IFS in the following year to maintain the seasonal calving pattern, and cows, which do not become pregnant 63 d after the start of the breeding season, are more likely to be culled). The substantial CVg for all the fertility traits with the possible exception of IFS were in very close agreement with the CVg estimated for comparable fertility traits within a similar seasonal calving system adopted in New Zealand (Grosshans et al., 1997). This confirms that direct selection for fertility may prove beneficial (Philipsson, 1981). Similar to most other studies (Hoekstra et al., 1994; Roxström et al., 2001; Evans et al., 2002) higher genetic merit for milk production was associated with a longer FSCO, a greater NS, poorer PRFS, and finally lower PR63. Genetic correlations between milk production and IFS were all close to zero (-0.09 to 0.04) although the standard errors were large (0.09 to 0.11). These correlations are weaker than those reported by previous authors (Hoekstra et al., 1994; Roxström et al., 2001) between IFS and milk yield (0.27 to 0.44).
Genetic Correlations with BCS
The negative genetic correlations between BCS with milk, fat, and protein yield (-0.51 to -0.14) agrees with previous studies (Pryce et al., 2001; Veerkamp et al., 2001). There was a tendency for BCS measured in early lactation to give the weakest correlations with milk production, which is consistent with results from Veerkamp et al. (2001) using random regression models. Based on these genetic correlations, a cow with a superior genetic merit for Cummilk240 of 1000-kg milk, will have a 0.25 BCS unit lower average BCS than a cow with a genetic merit of 0 kg for milk. Genetic correlations between BCS change in early lactation and milk production (-0.45 to -0.27) agree with those from Dechow et al. (2002); the discrepancy in the signs of the correlations may be explained by the different trait definitions of the BCS loss traits. Therefore, if selection for higher milk production alone continues, the genetic merit for BCS will reduce continuously, and management practices may have to be altered to compensate for the deleterious genetic effects.
A biological interpretation of the negative genetic correlation between BCS and milk production is the apparent relationship between BCS with energy balance and tissue mobilization (Pryce and Løvendahl, 1999). Because body tissue may be used in part to fuel milk production, a moderate to strong antagonistic genetic correlation between BCS and milk production is therefore expected.
The favorable genetic correlations between BCS and fertility observed in the present study agree with previous phenotypic (Domecq et al., 1997) and genetic studies (Dechow et al., 2001; Pryce et al., 2001; Veerkamp et al., 2001). Adjusting these relationships for phenotypic milk yield had no effect on the direction of the correlations between BCS with IFS, PRFS, NS, and PR63; correlations involving BCS with FSCO were all close to zero. Thus, regardless of yield, cows with low BCS will exhibit poorer fertility, suggesting that genes associated with body tissue mobilization may have pleiotrophic effects or be closely linked to genes controlling fertility in animals. Royal et al. (2002b) suggested that the pathways in which these correlations express themselves could be through either 1) hormones controlling intermediary metabolism having a direct effect on ovarian function, or 2) reproductive hormones, which regulate ovarian function having a direct effect on intermediary metabolism. Thus, low BCS resulting from selective breeding for higher milk yield may alter circulating hormonal levels of growth hormone and corticosteroids, while simultaneously reducing circulating levels of insulin and insulin-like growth factor-1, both of which may have deleterious effects on subsequent fertility. Similarly, low BCS as a consequence of selection for higher yield may alter the level of circulating reproductive hormones such as gonadotrophins resulting in possible implications for subsequent fertility (Royal et al., 2002b). However, evidence is still not clear, and several alternative pathways are possible.
Genetic correlations, involving BCS change and fertility, expressed large standard errors and therefore make interpretation of the results difficult. Nevertheless, the genetic correlation between change in BCS between d 5 and 60 and NS was similar to that reported by Dechow et al. (2002) on similar traits in second lactation animals. Based on the genetic parameters reported in the present study between BCS and fertility, an increase in genetic merit for average BCS of 1.00 BCS unit will reduce the IFS by 3 d, increase PRFS by 9 percentage units, reduce the NS by 0.32, and increase the PR63 by 11 percentage units.
Genetic Correlations with BW
The near zero genetic correlations between BW and milk production signify that selection for milk yield has a negligible genetic effect on the BW of an animal. However, when BW was adjusted for differences in BCS, all genetic correlations between BW and milk production became positive; correlations between AVGBW and the milk production traits ranged from 0.15 to 0.39. These correlations agree more closely with correlations estimated between measures of size and milk production; Brotherstone (1994) estimated genetic correlations between stature and milk production of between 0.16 and 0.25. Therefore, although selection for increased milk production may increase cow size, the reduction in BCS associated with such an increase in milk production will ultimately result in a negligible genetic effect of milk production on BW.
Genetic correlations, estimated between BW and the fertility traits, indicated that although genetically heavier cows are served sooner, they require more services and have a longer interval from first service to conception, which may be due in part to their inferior PRFS. All these factors combine to give lower PR63. However, the genetic correlations involving PRFS, NS, and PR63 did possess large standard errors although the direction of the correlations is consistent across the different BW test-day records. The correlations also agree with a long-term selection experiment conducted in Minnesota, where cows selected for high body size tended to require more services per conception than cows selected for low body size (Hansen et al., 1999). The negative correlations between BW and IFS are consistent with those reported by Veerkamp et al. (2000) between BW and commencement of luteal activity before and after adjusting for genetic merit in milk production. Following the adjustment of BW for differences in milk yield, the genetic correlations between adjusted BW and IFS were stronger for BW in midlactation than for BW in early lactation, which also agrees with Veerkamp et al. (2000). Moore et al. (1992) reported a slightly negative genetic correlation (-0.07) between BW at calving and days open in Holsteins, which has been shown to be a similar trait to IFS (Jansen, 1985; Grosshans et al., 1997).
Although a considerable proportion of the variation in BW is due to BCS (Veerkamp and Brotherstone, 1997; Berry et al., 2002), the difference in signs of the genetic correlations between BW with PRFS, NS, and PR63 to the genetic correlations observed between BCS with PRFS, NS, and PR63 indicate that factors other than BCS also contribute to the genetic variation in BW. This was substantiated when BW was adjusted for differences in BCS, and the genetic correlations between BW and fertility were reestimated. The signs of the correlations remained the same although the strength of the correlations increased (with the exception of the correlations between BW and IFS, which became slightly weaker).
Selection Indexes
It is important to bear in mind that the expected responses from selection are influenced by the parameters used in calculating the optimal index weights. The standard errors of some of the genetic correlations estimated were large; therefore, care should be taken in interpreting the magnitude in selection responses. Lindhé and Philipsson (1998) reported large effects on expected genetic gains per trait when genetic correlations were assumed between the traits in the index as opposed to zero genetic correlations. In agreement with previous studies, selection for milk yield alone is likely to reduce BCS (Pryce et al., 2001; Veerkamp et al., 2001; Pryce et al., 2002) and pregnancy rates (Grosshans et al., 1997). It is therefore imperative that selection indexes incorporate other nonproduction traits to minimize the antagonistic effect on correlated traits that may prove to have economical or ethical implications in the future.
Selection for reduced BW as a means of reducing maintenance costs has been adopted in the breeding objectives of some countries (Visscher et al., 1994; Montgomerie, 2002). In a national progeny-testing program the number of BW records available is likely to be low. However, the genetic correlations between AVGBW and BW at different stages of lactation were all greater than 0.95; this indicates that one BW measurement may suffice for inclusion in a national selection objective. The heritability for individual BW test-day records were lower than for AVGBW, which implies that a higher weight on BW might be required to achieve the same genetic gain, when only one BW record is available nationally. Similar conclusions are applicable to BCS measured only once in a national progeny-testing program.
In the present study, applying a negative weight to BW in the selection objective reduced the favorable trend in IFS over selection on the yield index alone, while BCS was also predicted to decline further. The larger decline in BCS from selecting for reduced BW is due to the moderate genetic correlations between BCS and BW found in the present study. Because body tissue reserves act as a biological buffer making up the deficit in energy required for milk synthesis, a continued reduction in genetic merit for BCS may in time become the major limiting factor in response to selection for increased milk production. Nevertheless, genetic correlations estimated in the present study between production and BCS were less than an absolute value of 1, suggesting that simultaneous selection for increased milk production while increasing or maintaining BCS at its current level is possible.
The ability to select for increased milk production without a corresponding deterioration in BCS was highlighted when an economic weighting of 45% relative to protein yield was applied to BCS. Although a reduction in BCS of -0.06 BCS units from selection on the production index appears small, the inclusion of BCS in the selection objective with a positive economic value emphasizes the role that BCS mobilization plays in achieving high milk yield. Substituting the genetic parameters for AVGBCS with those for BCS on d 5 of lactation had very little effect on the economic weighting necessary to achieve similar results; the economic weighting applied to BCS relative to protein yield was increased from 45 to 47%.
The slight negative effect that the first two selection objectives had on PR63 may indicate the need for the inclusion of a fertility-related trait in the selection objective. PR63 would be a possible fertility trait for inclusion because it measures the ability of an animal to conceive early in the breeding season—a criteria of great importance in seasonal calving systems. In the present study, PR63 had one of the largest heritabilities of the fertility traits investigated and also had a large CVg (8.5%). However, PR63 is not a routinely available measure for the Irish dairy cow population. Nevertheless, the current economic index in Ireland includes survivability (Veerkamp et al., 2002), which is likely to be correlated with PR63, since animals not pregnant have a greater chance of being culled; infertility accounted for 50 and 46% of the total culling reasons on DairyMIS farms for 1999 and 2000, respectively. The final selection objective illustrates that it is possible to restrict the expected change in PR63 to zero by including PR63 in the selection objective with an economic weight of 36% relative to protein yield. This objective was achieved at the expense of a slight reduction (4%) in the expected response in protein yield over selection for increased production alone and a continued decline in both BCS and BW. Selection objective 4 is more efficient than selection objective 3, when records are available on 100 daughters per sire, since the expected response in milk production is greater in the former. However, with smaller progeny group sizes BCS is a better indicator of PR63 than PR63 itself (Figure 1
).
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication October 17, 2002. Accepted for publication December 19, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. K. Toshniwal, C. D. Dechow, B. G. Cassell, J. A. D. R. N. Appuhamy, and G. A. Varga Heritability of Electronically Recorded Daily Body Weight and Correlations with Yield, Dry Matter Intake, and Body Condition Score J Dairy Sci, August 1, 2008; 91(8): 3201 - 3210. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Macdonald, J. W. Penno, J. A. S. Lancaster, and J. R. Roche Effect of Stocking Rate on Pasture Production, Milk Production, and Reproduction of Dairy Cows in Pasture-Based Systems J Dairy Sci, May 1, 2008; 91(5): 2151 - 2163. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. M. Chagas, J. J. Bass, D. Blache, C. R. Burke, J. K. Kay, D. R. Lindsay, M. C. Lucy, G. B. Martin, S. Meier, F. M. Rhodes, et al. Invited Review: New Perspectives on the Roles of Nutrition and Metabolic Priorities in the Subfertility of High-Producing Dairy Cows J Dairy Sci, September 1, 2007; 90(9): 4022 - 4032. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Macdonald, L. R. McNaughton, G. A. Verkerk, J. W. Penno, L. J. Burton, D. P. Berry, P. J. S. Gore, J. A. S. Lancaster, and C. W. Holmes A Comparison of Three Strains of Holstein-Friesian Cows Grazed on Pasture: Growth, Development, and Puberty J Dairy Sci, August 1, 2007; 90(8): 3993 - 4003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. P. Berry, J. M. Lee, K. A. Macdonald, K. Stafford, L. Matthews, and J. R. Roche Associations Among Body Condition Score, Body Weight, Somatic Cell Count, and Clinical Mastitis in Seasonally Calving Dairy Cattle J Dairy Sci, February 1, 2007; 90(2): 637 - 648. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Roche, K. A. Macdonald, C. R. Burke, J. M. Lee, and D. P. Berry Associations Among Body Condition Score, Body Weight, and Reproductive Performance in Seasonal-Calving Dairy Cattle J Dairy Sci, January 1, 2007; 90(1): 376 - 391. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Brotherstone, M. P. Coffey, and G. Banos Genetic Parameters of Growth in Dairy Cattle and Associations Between Growth and Health Traits J Dairy Sci, January 1, 2007; 90(1): 444 - 450. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. T. Kuhn, J. L. Hutchison, and G. R. Wiggans Characterization of Holstein Heifer Fertility in the United States J Dairy Sci, December 1, 2006; 89(12): 4907 - 4920. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Karacaoren, F. Jaffrezic, and H. N. Kadarmideen Genetic Parameters for Functional Traits in Dairy Cattle from Daily Random Regression Models J Dairy Sci, February 1, 2006; 89(2): 791 - 798. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. G. Mitchell, G. W. Rogers, C. D. Dechow, J. E. Vallimont, J. B. Cooper, U. Sander-Nielsen, and J. S. Clay Milk Urea Nitrogen Concentration: Heritability and Genetic Correlations with Reproductive Performance and Disease J Dairy Sci, December 1, 2005; 88(12): 4434 - 4440. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Wall, I. M. S. White, M. P. Coffey, |
), body condition score (
), or PR63+body condition score (x).