Genetic parameters and evaluations from single- and multiple-trait analysis of dairy cow fertility and milk production

https://doi.org/10.1016/S0301-6226(02)00274-9Get rights and content

Abstract

Genetic parameters and breeding values for dairy cow fertility were estimated from 62 443 lactation records. Two-trait analysis of fertility and milk yield was investigated as a method to estimate fertility breeding values when culling or selection based on milk yield in early lactation determines presence or absence of fertility observations in later lactations. Fertility traits were calving interval, intervals from calving to first service, calving to conception and first to last service, conception success to first service and number of services per conception. Milk production traits were 305-day milk, fat and protein yield. For fertility traits, range of estimates of heritability (h2) was 0.012 to 0.028 and of permanent environmental variance (c2) was 0.016 to 0.032. Genetic correlations (rg) among fertility traits were generally high (>0.70). Genetic correlations of fertility with milk production traits were unfavourable (range −0.11 to 0.46). Single and two-trait analyses of fertility were compared using the same data set. The estimates of h2 and c2 were similar for two types of analyses. However, there were differences between estimated breeding values and rankings for the same trait from single versus multi-trait analyses. The range for rank correlation was 0.69–0.83 for all animals in the pedigree and 0.89–0.96 for sires with more than 25 daughters. As single-trait method is biased due to selection on milk yield, a multi-trait evaluation of fertility with milk yield is recommended.

Introduction

Impaired fertility in dairy cows is a major cause of involuntary culling in many countries. Dairy cow fertility is important both economically and ethically. Dairy cow fertility, as measured by, for example, pregnancy rates, has been shown to be declining in many countries. Economic implications of this decline in fertility are of major concern. Organic farming demands high standards of observed welfare, management and food production (Sundrum, 2001), suggesting that hormonal treatment to rectify impaired fertility is less desirable in these systems. However, regular therapeutic intervention to overcome infertility will be generally less acceptable in all future farming systems. Characters of good cow fertility can be defined as cows that show visible signs of heat at the right time after calving (days to first heat and/or first insemination) and that conceive when inseminated the first time (success of conception to first service). This definition addresses two important reproductive phenomenon: cyclicity and the ability of cows to conceive. When these criteria are met, other fertility measures such as days open and calving interval would then take their normal biologically determined values.

Many studies have shown that antagonistic phenotypic and genetic correlations of fertility traits with milk yield (e.g., Hoekstra et al., 1994; Pryce et al., 1997; Dematawewa and Berger, 1998; Kadarmideen et al., 2000a) would lead to a genetic decline in cow fertility, if selection is for milk production only. Thus, incorporation of fertility in selection decisions seems desirable. Currently, only few countries have national selection indices that include fertility traits. To support future profitability in production systems that penalise poor fertility, routine national sire or cow evaluations for fertility must be calculated and subsequently incorporated into a multi-trait national profit index. This will enable farmers to select the best animals based on a combination of production and fertility. Kadarmideen and Simm (2002) showed that, for UK circumstances, profitability from using such a ‘Production–Fertility Index’ over the next 20 years could be substantial (about 38% more) when compared to using ‘production-only’ index. These results indicate that there are strong economic reasons for breeding organisations to include fertility in the index. However, implementation hurdles arise for some countries mainly in the form of inconsistent recording of insemination events on farm: some record all, some or only the last insemination date. This non-uniform recording could be coupled with a lack of information on pregnancy diagnosis. Therefore, data from national recording schemes tend to have a mixture of poor and high quality data, which warrants stringent editing procedures to obtain sufficient data quality for genetic analysis (e.g., Kadarmideen and Coffey, 2001). The Dairy Information System (DAISY) is a computer-based recording scheme, reputed for accurate recording of insemination and health events in the UK. Genetic analyses in this study are based on the data from this DAISY scheme.

The inherent problem in fertility analysis is that these traits may have been subjected to censoring and selection based on milk yield. For example, cows that are culled for low milk production would have no calving interval regardless of their reproductive efficiency. Producers may give more opportunities to high yielding cows to conceive and may deliberately delay inseminations after calving for these cows. High yielding cows may have fertility problems resulting in a longer days to first insemination and calving interval. Differentiation between cows that had fertility problems but were genuinely inseminated at the first oestrous cycle and cows that were potentially able to conceive at the first insemination but the first insemination was deliberately delayed may not be possible. Milk yield can be included as a covariate in the analysis of fertility but that can only correct reproductive measures with respect to phenotypic differences in milk yield level. A multi-trait analysis of fertility with milk yield as an additional trait is a different approach which aims to improve accuracy of genetic evaluations for the traits involved by reducing variances of prediction error of estimated breeding values (Schaeffer, 1984) and provide breeding values for animals that are not recorded for a particular trait.

This study capitalises on one of the main advantages of multiple-trait analysis: it can provide unbiased estimates for a trait that is observed only on animals selected based on values of a correlated trait. Animals may have been highly selected for milk yield in early lactation, which can lead to biased fertility observations in the current or later lactation in the form of presence or absence of subsequent calving, and timing and frequency of inseminations. The preferential treatments of cows are also practised at genetic level such as favouring daughters of elite sires proven for their high genetic merit for daughters’ milk production. Single-trait analysis, ignoring information on selective treatment of cows with different (genetic potential for) milk yield, would lead to biased genetic parameters, which in turn, would result in inappropriate predictions based on multi-trait national selection indexes.

The main objective of this study was to estimate and compare breeding values for various fertility measures from single-trait evaluation of fertility and multiple-trait evaluation of fertility with milk yield as a correlated trait. The other objective of this study was to provide estimates of genetic parameters for fertility and milk production traits in dairy cattle.

Section snippets

Data

Data on insemination and calving events from 1986 to 1999 were obtained from DAISY for 132 herds. DAISY is used by a small group of co-operating farmers and veterinarians specifically interested in monitoring fertility and health in their herds. Although recording insemination and disease events is voluntary for DAISY farmers, recording is relatively more accurate and complete in DAISY than UK national milk recording schemes. All herds recorded fertility and health events with DAISY but

Descriptive statistics

Mean and standard deviations (S.D.; in parenthesis) of age at calving within lactations were 804 (144), 1188 (160), 1572 (168), and 1951 (189) and 2328 (203) days for lactations 1–5, respectively, with an overall mean and S.D. of 1363 (524) days. The mean (and S.D.) of days in milk (DIM) were 321 (58), 312 (55), 306 (53), 305(56) and 301 (60) days for lactations 1–5, respectively, with an overall mean and S.D. of 311 (57) days. Within-lactation phenotypic mean and S.D. of many fertility traits

Discussion

Single- and multiple-trait genetic analysis of fertility traits were investigated in this study. Analysis of fertility traits is known to be problematic as fertility observations are subject to managemental decisions and observing some fertility traits depend on observing some other fertility traits. For example, cows that do not conceive do not have value for days open or subsequent calving date (hence no value for calving interval). While no calving interval could be due to biological reason

Conclusions

Multiple-trait analysis of dairy cow fertility with milk yield was investigated here as a possible method to analyse selected or censored fertility observations, as a result of selective treatment of cows based on (phenotypic/genetic merit for) milk yield. As a part of this study, estimates of heritabilities, permanent environmental variances and genetic (and other) correlations for six fertility and three milk production traits were provided. Single-trait analysis of fertility was compared

Acknowledgements

We thank DAISY and NMR for providing data and the Milk Development Council for funding the early part of this study. Dr. Richard J. Esslemont is thanked for his help with DAISY data transfer and comments on the manuscript. We thank two anonymous reviewers of this paper for their valuable comments.

References (26)

Cited by (0)

View full text