| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, Centre for Reproductive Biology in Uppsala, P.O. Box 7023, SE-750 07 Uppsala, Sweden
Swedish Dairy Association, P.O. Box 7054, SE-750 07 Uppsala, Sweden
Division of Animal Physiology, School of Biosciences, University of Nottingham, Sutton Bonington Campus, LE12 5RD, United Kingdom
Roslin Institute (Edinburgh), Roslin, Midlothian, EH25 9PS, United Kingdom
# Department of Veterinary Clinical Science, Faculty of Veterinary Science, University of Liverpool, Liverpool, CH64 7TE, United Kingdom
1 Corresponding author: karl-johan.petersson{at}hgen.slu.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: luteal activity • fertility • dairy cow • heritability
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Fertility has been included in Nordic breeding programs for dairy cattle since the early 1970s (Lindhé et al., 1994). However, the genetic trend has shown a decrease in fertility of Swedish Holsteins whereas the genetic level has been relatively constant for the Swedish Red (Lindhé and Philipsson, 2001). It seems that inclusion of fertility in the breeding goal in Sweden has not been enough to withstand the effects of importation of genetic material into the Swedish Holsteins from countries that have a low, or no, weighting on fertility in their breeding objective. The traits traditionally used in the genetic evaluation for fertility (e.g., number of inseminations per service period and interval between calving and first AI) have very low heritabilities (Roxström et al., 2001; Wall et al., 2003). This may be partly a result of the large influence of management on the measurements that are used in present breeding programs for fertility. For instance, the rationale for measuring the interval from calving to first AI (CFI) is to obtain an indirect measure of interval from calving to first ovulation (CFO). However, CFI is affected by the farmer’s decision of when to start the service period, which may vary between herds and between cows within herds.
In recent studies (Darwash et al., 1997a; Royal, 1999; Veerkamp et al., 2000; Royal et al., 2002a), a measure more related to CFO has been presented, namely the interval from calving to commencement of luteal activity (C-LA). The definition of C-LA is the interval from calving until the progesterone level in milk reaches a threshold value, thereby indicating progesterone production by the corpus luteum. The occurrence of C-LA is about 4 to 5 d after first ovulation (Darwash et al., 1997a) and is thereby a direct measurement of the resumption of ovarian activity after calving. In these studies of C-LA, heritability estimates of 16 to 21% have been reported, which is considerably higher than for traditional measurements of fertility in dairy cows.
At a phenotypic level, early onset of estrus cyclicity increases the probability of an early insemination after calving, shortens the interval from calving to conception, increases conception rate, and reduces the number of services per conception (Darwash et al., 1997b). Furthermore, at a genetic level, cows with genetically longer C-LA on average have longer calving intervals and longer interval to first service (genetic correlations of 0.39 and 0.53, respectively; Royal et al., 2003). However, not only is the time until first ovulation (reflected by C-LA) important for fertility in dairy cows, but so is the subsequent progesterone pattern. Different aberrations in progesterone profiles are associated with a longer CFI, longer interval from calving to conception, and lower pregnancy rates (Royal et al., 2000a; Petersson et al., 2006a). The decreasing trend in pregnancy rates at first AI in the study from the United Kingdom was accompanied by an increase from 32 to 44% in atypical progesterone profiles, especially those with a prolonged luteal phase (Royal et al., 2000b). We have previously shown that by using the percentage of samples taken within the first 60 d after calving with luteal activity (i.e., progesterone 
3 ng/mL; PLA), we can separate not only profiles with delayed onset of ovarian activity from normal profiles (as does C-LA) but also profiles with prolonged luteal phase from normal profiles (Petersson et al., 2006a).
In studies of C-LA, progesterone samples were taken relatively frequently (2 to 3 times a week; Darwash et al., 1997a; Royal, 1999; Veerkamp et al., 2000; Royal et al., 2002a; Petersson et al., 2006b). For use in a breeding program, such frequent sampling would require an online progesterone monitoring system. However, it will probably be many years until all dairy herds are equipped with such systems. An alternative would be to use milk samples from the routine milk recording for analysis of progesterone (Darwash et al., 1999; van der Lende et al., 2004). The disadvantage is that these samples are taken relatively infrequently (once a month), and it needs to be determined how sampling frequency affects the interpretation of different progesterone profiles as well as how it affects the genetic parameters of progesterone-based measurements.
The objective of this study was to investigate the possibility of combining progesterone sampling with routinely performed milk recording. Therefore, genetic parameters for a trait based on measures for luteal activity during the first 60 d postpartum (PLA) were estimated with different sampling intervals. Consequences for incorporating PLA in breeding programs for fertility in dairy cattle are discussed.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Milk progesterone samples were taken 3 times per week (Mondays, Wednesdays, and Fridays) from 2 to 8 d postpartum until a maximum of 24 d after the first AI (for further details, see Royal et al., 2000a). Progesterone concentration was measured in unextracted samples of whole milk using ELISA (Ridgeway Science Ltd., Alvington, UK). The intraassay coefficients of variation, calculated using a representative sample of 100 assays, for control standards at 2 and 8 ng/mL were 0.129 and 0.060, respectively. The interassay coefficients of variation for 2 and 8 ng/mL were 0.127 and 0.081, respectively. Sensitivity was 1 ng/mL, calculated using the absorption of the blank standard minus 2 standard deviations.
The occurrence of estrus was recorded. Where possible, in addition to visual routine checks by the herdsman, heat mount detectors (Kamar Inc., Steamboat Springs, CO) were used. Treatment of reproductive disorders was withheld for 80 d after calving unless required for welfare reasons. Data concerning 44 animals that received hormone treatment of reproductive disorders before insemination were removed from selected analyses, where appropriate. The incidence of dystocia, retained placenta, and uterine infection was recorded. However, only uterine infection was included as a covariate in this study, as no effects of dystocia and retained placenta on the studied fertility measures were found in preliminary analyses of the data.
Fertility Measurements
Interval from calving to C-LA was defined as the number of days from calving until the day of the first of 2 consecutive milk progesterone concentrations 3 ng/mL. Furthermore, the PLA (progesterone 3 ng/mL) was calculated for all or a subset of samples taken within 60 d after calving (Petersson et al., 2006a,b). For the definition of PLAa all samples in the current database (3 per week) was used for the calculation. For PLA based on weekly sampling (PLAw), only the first sample in each week was included. For PLA based on sampling twice a month (PLAf), only the first sample in every 2-wk period was included. For PLA based on random monthly sampling (PLAm), a sample within the first 4 wk of lactation was randomly chosen, utilizing SAS procedure SURVEYSELECT (SAS Institute, 2001). For this measure, the first sample together with the sample taken 4 wk later was used.
The progesterone profiles, constructed using the milk progesterone samples, were used to classify different ovarian patterns using definitions published by Lamming and Darwash (1998). Delayed ovulation type 1 (DOV1) was defined as prolonged anovulation postpartum with milk progesterone <3 ng/mL for 45 d after calving. Delayed ovulation type 2 (DOV2) was defined as prolonged interluteal interval with milk progesterone <3 ng/mL for 12 d between 2 luteal phases. Persistent corpus luteum type 1 (PCL1) was defined as delayed luteolysis with milk progesterone 3 ng/mL for 19 d during the first postpartum estrous cycle. Persistent corpus luteum type 2 (PCL2) was defined as delayed luteolysis with milk progesterone 3 ng/mL for 19 d during estrus cycles after the first cycle.
Statistical Analysis
The REML option of the DMU package (Jensen and Madsen, 1994) was used to fit a mixed linear animal model to the data. The following model was used for the analyses:
 |
where ymnopqrstuv is the analyzed trait; µ is the overall mean; lm is the fixed effect of lactation number (m = 1 to 9); hn is the fixed effect of herd (n = 1 to 8); yro is the fixed effect of calving year (o = 1995 to 1998); sp is the fixed effect of season (P = 1 to 4, for December to February, March to May, June to August, September to November); uiq is the fixed effect of uterine infection postpartum (q = yes or no); dr is the fixed effect of diet (r = 1 to 23); b1pchs is the fixed regression on PCH with coefficient b1; b2hett is the fixed regression on HET with coefficient b2; hysnop is the random effect of herd-year-season interaction, assumed to be normally distributed with mean zero and variance 
2hys; au is the random effect of breeding value of animals, assumed to be normally distributed with mean zero and variance A2A, where A is the numerator relationship matrix; and emnopqrstuv is the random residual term, with residuals assumed to be normally distributed with mean zero and variance 2e. The random effect of permanent environment (effect of cow over lactations) and the random effects of interaction of herd-year, herd-season, and year-season were also tested with likelihood ratio tests; but none of these effects was significant and they were omitted from further analyses. Before inclusion in the mixed linear model, C-LA was transformed (natural logarithm, lnC-LA) because this transformation was shown by Darwash et al. (1997a) to give the best model fit. In the heritability calculations, the herd-year-season variance was not included in the phenotypic variance. Variance components and breeding values were obtained from a single-trait analysis but for correlations, a bivariate analysis was applied.
Estimates of heritabilities and genetic correlations were used for selection index calculations with CFI or lnC-LA in the breeding goal T and CFI or PLAm or both in the index I and maximizing the correlation between T and I, i.e., the accuracy (rTI). For analysis of sensitivity, the genetic correlation between PLAm and lnC-LA was changed to –0.9 and –0.8. All calculations were done assuming 50 or 100 daughters per bull.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
). The number of lactations was lower for PLAm because there had to be a corresponding sample 4 wk after the first randomly chosen sample and this was not the case for some of the lactations in the study.
|
shows the average C-LA and the different PLA measures per type of progesterone profile. The averages of all PLA measurements were generally lower for cows that had a DOV1 or DOV2 and higher for cows that had a PCL1 compared with cows with no atypical progesterone profile (Table 2). Cows with a DOV1 profile also had a later C-LA than cows with no atypical progesterone pattern. The other atypical progesterone profiles did not have a different mean of C-LA compared with cows with no atypical profile.
|
). The environmental variance increased with decreased sampling frequency.
|
. Genetic correlations between different PLA measures are not presented because they were all calculated using in part the same original samples and were consequently highly autocorrelated. The genetic correlations for lnC-LA and DOV1 with the PLA measurements were all high and negative (Table 4). Standard errors of the genetic correlations increased with decreased sampling frequency of the PLA measurements. For PLAa, PLAw, and PLAf, there was a high positive genetic correlation with PCL1 and a moderate negative correlation with MY56. Correlations with CFI were low and with high standard errors, except for the correlation with PLAm; however, this correlation was not significant.
|
) to further examine the genetic correlations between PLAa and these 2 types of atypical progesterone profiles. The regression of DOV1 on PLAa was linear (P <0.001) with decreasing incidence (lower breeding values) of DOV1 with increasing breeding values for PLAa. This associated with the strong negative genetic correlation between these measurements. The moderate positive genetic correlation between PLAa and PCL1 was associated with a more scattered plot and a flatter regression, indicating that higher breeding values for PLAa were associated with a small increased incidence (higher breeding values) of PCL1.
|
). The selection index calculations were also performed with 2 lower genetic correlations (–0.9 and –0.8) between PLAm and lnC-LA, which decreased the accuracy by almost 0.1 for each decrease in genetic correlation.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
We observed a surprisingly high heritability estimate for PLAa (29.5%) compared with heritability estimates for C-LA found here or previously reported (16 to 21%; Darwash et al., 1997a; Royal et al., 2002a; Veerkamp et al., 2000). The environmental variance of the PLA measures increased with less frequent sampling. This increase in environmental variance was probably partly a result of the decreased number of possible outcomes with decreased sampling frequency; for example, PLAm had only 3 outcomes: 0, 50, and 100%, because there were only 2 samples in the monthly sampling. The heritability estimates for lnC-LA, CFI, PFI, PCL1, and MY56 have been reported previously (Royal et al., 2002a) and the estimates from the analysis in the present study were in agreement with the earlier study. The unfavorable genetic correlations between MY56 and PLAa, PLAw, and PLAf, respectively, are in accordance with the unfavorable genetic correlation between MY56 and lnC-LA (0.36), as shown by Royal et al. (2002a).
We have previously shown in a Swedish data set that PLAa could be used to separate DOV1 and a combination of PCL1 and PCL2 profiles from normal profiles (Petersson et al., 2006a). The present study supports this, and the PLA measures (PLAa, PLAw, PLAf, and PLAm) were 33 to 38 percentage points lower for cows with a DOV1 profile and around 8 percentage points higher for cows with a PCL1 profile compared with cows with no atypical profiles. The association between these 2 types of atypical progesterone profiles and PLA was also reflected in the high negative genetic correlations between DOV1 and PLAa, PLAw, and PLAf, respectively, and the moderate to high positive genetic correlation between PCL1 and PLAa, PLAw, and PLAf, respectively. However, these high genetic correlations with PCL1 raise concerns regarding fertility disorders because it has been shown that prolonged luteal phases are associated with pyometra (Etherington et al., 1991). Thus, it appears that there is an intermediate optimum value for PLA: too high a value for PLA is unfavorable because it is associated with PCL1 but a low value could also be unfavorable because it is associated with DOV1. The genetic correlations between PLAm and DOV1 and PCL1, respectively, indicated that monthly progesterone sampling could give an indication of DOV1 profiles because of the moderate correlation with this type of profiles, but this infrequent sampling regimen cannot be used to detect PCL1 profiles. This low sampling frequency (monthly) was also insufficient to detect the negative genetic correlation that was present between the other PLA measures and MY56. We were not able to estimate genetic correlations between the PLA measures and DOV2 and PCL2 probably due to the low incidence of these 2 types of progesterone profiles.
An alternative measure that has been studied previously is the measurement C-LA50%, introduced by van der Lende et al. (2004), which is the lactation stage when 50% of the daughters of a sire had an active corpus luteum with 3- to 6-wk intervals of progesterone sampling. The basis of C-LA50% is infrequent sampling, as is our PLAm measurement, but we have applied our measurement on the cow level in contrast to C-LA50%, which was calculated on the sire level. The information obtained on cow level with PLAm could be used for management purposes; however, this has to be studied further.
Even though the interpretation of PLA is not as straightforward as C-LA, the high genetic correlation with lnC-LA makes it interesting for further analysis. The high negative correlations with lnC-LA could partly depend on the fact that the individuals with the highest breeding values for lnC-LA have the lowest breeding values for the different PLA measures.
A short interval from calving to first ovulation is generally considered desirable. In some breeding programs, CFI is used as an index trait, and used as a breeding goal trait, presumably as a proxy for CFO. Using CFI as the index trait gave an apparent high accuracy when CFI was also used as the breeding goal trait (Table 5
). However, C-LA is likely to be a more direct measure of CFO than CFI. There is only a delay of 4 to 5 d between C-LA and CFO, and both C-LA and CFO are determined by the animal’s physiology rather than by management practice. Therefore, we suggest using C-LA as the breeding goal trait in a selection index rather than CFI. With C-LA as breeding goal trait in the selection index, PLAm as index trait resulted in a much higher accuracy (0.80) compared with when CFI was the index trait (0.09).
Because PLAm is based on monthly sampling, we concluded that sampling for progesterone at the regular milk recording could be used to increase the accuracy of a breeding program toward an earlier start of cyclical ovarian activity after calving. A breeding program focused only on an earlier C-LA (by selecting on increased PLA) may, however, increase the incidence of progesterone profiles with persistent corpus luteum, because there was generally a positive genetic correlation between the PLA measures and PCL1. Further investigation of this correlation showed a nonlinear relationship between the breeding values for PCL1 and PLAa but a strongly linear relationship between DOV1 and PLAa (Figure 1). Therefore, in the current population, selection against sires with low breeding values for PLAa would also select against sires with high breeding values for DOV1 but at the same time not extremely low breeding values for PCL1. Selection for increased PLA could thus be used to decrease profiles with DOV1, but would affect PCL1 unfavorably, albeit to a much lesser extent.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication May 8, 2006. Accepted for publication August 28, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. L. Bamber, G. E. Shook, M. C. Wiltbank, J. E. P. Santos, and P. M. Fricke Genetic parameters for anovulation and pregnancy loss in dairy cattle J Dairy Sci, November 1, 2009; 92(11): 5739 - 5753. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. T. Garmo, A. D. Martin, E. Thuen, O. Havrevoll, H. Steinshamn, E. Prestlokken, A. Randby, M. Eknaes, A. Waldmann, and O. Reksen Characterization of progesterone profiles in fall-calving Norwegian Red cows J Dairy Sci, October 1, 2009; 92(10): 4919 - 4928. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Lovendahl and M. G. G. Chagunda Short communication: Genetic variation in estrus activity traits J Dairy Sci, September 1, 2009; 92(9): 4683 - 4688. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. T. Garmo, E. Ropstad, O. Havrevoll, E. Thuen, H. Steinshamn, A. Waldmann, and O. Reksen Commencement of luteal activity in three different selection lines for milk yield and fertility in Norwegian Red cows J Dairy Sci, May 1, 2009; 92(5): 2159 - 2165. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Lebeer, J. Vanderleyden, and S. C. J. De Keersmaecker Genes and Molecules of Lactobacilli Supporting Probiotic Action Microbiol. Mol. Biol. Rev., December 1, 2008; 72(4): 728 - 764. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Konig, Y. M. Chang, U. U. v. Borstel, D. Gianola, and H. Simianer Genetic and Phenotypic Relationships Among Milk Urea Nitrogen, Fertility, and Milk Yield in Holstein Cows J Dairy Sci, November 1, 2008; 91(11): 4372 - 4382. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. T. Garmo, A. O. Refsdal, K. Karlberg, E. Ropstad, A. Waldmann, J. F. Beckers, and O. Reksen Pregnancy Incidence in Norwegian Red Cows Using Nonreturn to Estrus, Rectal Palpation, Pregnancy-Associated Glycoproteins, and Progesterone J Dairy Sci, August 1, 2008; 91(8): 3025 - 3033. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Windig, B. Beerda, and R. F. Veerkamp Relationship Between Milk Progesterone Profiles and Genetic Merit for Milk Production, Milking Frequency, and Feeding Regimen in Dairy Cattle J Dairy Sci, July 1, 2008; 91(7): 2874 - 2884. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) plotted against percentage of samples with luteal activity during the first 60 d postpartum, based on all samples (PLAa). A linear trend-line (——) is shown for DOV1 and a polynomial trend-line (------) for the PCL1.