| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL PRODUCTION |
,4
* Department of Animal Sciences, Washington State University, Pullman 99164-6310 and
and
USDA, ARS, Fort Keogh Livestock and Range Research Laboratory, Miles City, MT 59301
Abstract
Objectives were to 1) identify risk factors affecting the longevity of beef females, 2) evaluate the utility of measures collected early in life in predicting longevity, and 3) estimate the heritability of longevity when females were culled primarily for not being pregnant following a 45-d breeding season. Data were from 1,379 Composite Gene Combination (CGC; 
Red Angus, 
Charolais, Tarentaise) cows born from 1982 through 1999 at the USDA-ARS, Fort Keogh Livestock and Range Research Laboratory, Miles City, MT, and first calving at approximately 2 yr of age. The length of productive life was modeled using Cox regression to identify factors affecting the longevity of beef females. Age at first calving and calf birth weight did not influence longevity. Cows that experienced dystocia were at greater risk of being culled than those that calved without assistance (P < 0.01). On average, as breeding value for cow weight increased, the risk of being culled decreased (P < 0.01), whereas the risk of being culled increased with increasing maternal breeding values for preweaning gain (P < 0.05). Traits measured before 1 yr of age were not useful in predicting the subsequent longevity of cows. The heritability of functional longevity was estimated to be 0.14. Relatively low heritability and the lack of indicators of longevity expressed early in life suggest that genetic improvement of longevity will be difficult. Matching the genetic potential of cows for size and milk production to the production environment such that rebreeding performance is not compromised by concurrent lactation seems to be a consideration in retaining beef females when open cows are culled.
Key Words: Beef Cattle • Productive Life • Risk • Survival
Introduction
In beef production systems, the longevity of breeding stock has a substantial effect on economic efficiency. To be profitable, females remaining in production beyond their breakeven age must compensate for females that are culled earlier (Snelling et al., 1995
). Increasing the longevity of females reduces annual production costs associated with raising replacement heifers, increases the number of high producing mature cows, and reduces the number of cows that are culled involuntarily. However, the longevity of beef females remains a frequently overlooked aspect of genetic evaluation programs (Hunter, 1994).
The genetic evaluation of longevity is hindered by its expression late in life, by censoring, and by nonnormality of data. Progeny testing and waiting until all daughters of a sire have been culled before making selection decisions greatly increases the generation interval (Rendel and Robertson, 1950). Linear models are inadequate for a statistical analysis of longevity data due to violation of assumptions of normality (Lagakos, 1979). To consider records from cows alive at the time of evaluation as complete or to not include them in the analysis biases estimates of factors affecting longevity (Ducrocq, 1994). To overcome these statistical issues, survival analysis techniques are implemented in the genetic evaluation of longevity (Ducrocq et al., 1988).
Objectives of this research were to 1) identify risk factors affecting longevity of beef females, 2) evaluate utility of measures collected early in life in predicting longevity, and 3) estimate the heritability of longevity, when females were culled primarily as a result of their not being pregnant following a 45-d breeding season.
Materials and Methods
Data
Data were from 1,379 Composite Gene Combination (CGC; Red Angus, Charolais, Tarentaise) cows born from 1982 through 1999 at USDA-ARS, Fort Keogh Livestock and Range Research Laboratory, Miles City, MT. To alleviate differences in the length of productive life due to breed composition of the cow, data were edited to remove founding generations of this composite so that only those females with breed composition Red Angus, Charolais, and Tarentaise were included in the analyses. However, pedigree information traced the lineage of all cows back to founding Red Angus, Charolais, and Tarentaise parents. Data were edited to include only those cows that calved at approximately 2 yr of age. Additional information on the development and general performance of the CGC population is available in Newman et al. (1993a,b).
Heifers and cows were exposed to CGC bulls during a 45-d breeding season, resulting in a subsequent 60-d calving season beginning in mid-March. The breeding scheme consistently used yearling bulls that had passed a breeding soundness examination for 1 yr, with a few of these bulls being used for a second year. Cow-to-bull ratio during the breeding season varied with the population size, in response to drought, and other experimental demands on the animal resources. Across all years, the bull-to-cow ratio averaged 1:15.2, with a low of 1:7 and a high of 1:29. Pregnancy rate averaged approximately 90% and was not affected by the bull-to-cow ratio within the range used at Fort Keogh (data not shown). Thus, most sires had a limited number of daughters in the herd. The number of daughters per sire ranged from 1 to 19, with an average of 5.4 daughters with complete records and average of 1.2 uncensored daughter records. Sire effects were estimated for all sires having at least one daughter in the data (n = 237). Cows were weighed annually coincident with the weaning of their calves in the fall.
Two-year-old heifers were observed 22 h daily at the time of calving and assistance was provided to heifers not delivering the calf 1 h after being observed in Stage III of labor. Approximately 18% of heifers received assistance at parturition. Older cows were observed for difficulty in calving during daylight hours. Approximately 7% of all females experienced dystocia at some time in their life subsequent to their first calving. Arithmetic means and standard deviations for cow performance traits and her calf’s performance traits are summarized in Table 1.
|
) was used to assess relationships of the length of productive life with various measurements. The dependent variable for all analyses was length of productive life, which was defined as the number of days between first calving and last weaning date. Females were culled the first time they were diagnosed as not pregnant. Additional culling was based primarily on unsoundness and disposition. Little if any culling was based on the performance of progeny. When necessary to maintain the inventory of females exposed for breeding at approximately 360 animals, older cows were removed from the herd without respect to either their phenotype or progeny performance. In addition, from 1995 to 1997, a sample of cows was also removed from the herd to initiate a project for mapping quantitative trait loci (MacNeil and Grosz, 2002). If a cow died, was not pregnant at weaning in the fall, or was sold with an unknown pregnancy status, then she was considered to have reached the end of her productive life and her record was uncensored. Records from pregnant cows that were sold or transferred to another project and records from cows that remained in the herd in 2001 were treated as right-censored (33%). For ease of interpreting results and to fit possible nonlinear responses, continuous independent variables were coded into evenly spaced categories. Although somewhat arbitrary, the categories were designed so that a sufficient number of individuals fell into each category. Effects of independent variables are expressed as risk ratios, which represent the risk of the cow being culled, based on her level within a particular variable, with all other variables held constant. The risk ratio for the numerically lowest category within each effect was set to 1, and all other levels within the effect were expressed relative to this level. In comparing levels of effects, greater ratios indicate greater risk of being culled and conversely smaller ratios indicate a lower risk. Risk ratios for main effects of time-dependent variables should be interpreted with caution because the effect is dependent on time.
Breeding values for cow weight and 200-d gain from birth to weaning were also calculated in preliminary analyses to evaluate their relationships with longevity. Two single-trait analyses were conducted. In the first analysis, cow weights measured in the fall coincident with the weaning of calves was evaluated. Contemporary groups were formed as year-age subclasses, with cows older than 5 yr coded as 5 yr of age. There were 1,433 cows with at least one recorded weight and a total of 4,375 observations. The linear model for cow weight included fixed contemporary group effects, random direct additive effects, and uncorrelated random permanent environmental effects associated with repeated records of the cow. For the second analysis, 200-d preweaning gain (n = 4,902) was calculated as 200 times (weaning weight minus birth weight) divided by age at weaning. Contemporary groups were formed as year, sex, and age-of-dam subclasses. The linear models for 200-d preweaning gain included fixed contemporary groups, random direct and maternal additive effects, and uncorrelated random maternal permanent environmental effects of dams. Derivative-free multitrait REML (Smith and Graser, 1986; Graser et al., 1987) methods as implemented by Boldman et al. (1995) were used to predict breeding values upon convergence of estimates of the (co)variance components. Each analysis was assumed to have converged when variance of -2 log likelihoods in the simplex was less than 10-10, and analyses of the data using different starting values converged to similar estimates of the variance components.
Five analyses were conducted using the Survival Kit (Ducrocq and Solkner, 1998) to address the specific objectives posed in this research. Model 1 considered factors affecting longevity regardless of the age at which they were expressed. Formally,
 |
where
(t)
0(t))
48 kg in increments of 3 kg
230 kg in increments of 20 kg and a seventh level coding for calves without a weaning weight
730 or >730 d of age
38 kg in increments of 12 kg
9 kg in increments of 2 kg.
Changes in time-dependent variables occurred annually. A preliminary analysis also included the fixed, time-independent effect of the cow’s direct breeding value for preweaning gain, which was found to not approach significance (P > 0.2). An additional analysis similar to Model 1 was conducted in which BVwt and BVmg were replaced with corresponding producing abilities (i.e., breeding value + permanent environmental effect) for each cow. Results from the analysis including producing abilities as independent variables were similar to those obtained using Model 1 and are not presented. Model 2 evaluated risk factors associated with phenotypes resulting from the production of each cow’s first calf. Thus,
 |
where
(t) and 0(t)
42 kg in increments of 3 kg
195 kg in increments of 15 kg and a seventh level indicated missing values
Model 3 evaluated risk factors associated with phenotypes measured early in the life of the cow. Thus,
 |
where
(t) and 0(t)
45 kg in increments of 3 kg
215 kg in increments of 15 kg
400 kg in increments of 25 kg
To handle missing phenotypes for independent variables in Model 3, we chose to assign cows the average phenotype of the herd rather than recode the missing values into separate categories. The number of observations that would have fallen into each "missing" category was very small, with a maximum number of 21 cows. Previous studies have assigned cows the average value of the herd, or an arbitrary value when information was not available for the cow (Neerhof et al., 2000; Vollema et al., 2000; Vukasinovic et al., 2001). Preliminary analysis of these data confirmed the effect of handling missing values in this way was negligible, in this investigation. Models 4 and 5 were used to derive heritability estimates for functional and true longevity, respectively, as defined by Ducrocq et al. (1988). The variance due to sires for functional longevity, longevity independent of level of production, was estimated with Model 4 as
 |
where
MGS
The random effects were grouped into a vector s, which was assumed to follow a multivariate normal distribution with a variance-covariance matrix 
Variance due to sires for true longevity, longevity that is dependent on level of production, was estimated with Model 5 as
 |
where all effects are as defined for Model 4. The independent variables BW, CD, PWG, and AGE were removed from the model because each contributes to, but does not entirely define, the phenotypic description of level of production.
Following Neerhof et al. (2000) and Vollema et al. (2000), heritability was calculated as
 |
where, var(sire) is sire variance. Yazdi et al. (2002) refer to heritability calculated in this way as effective heritability, finding little difference between this estimate and estimates from Weibull sire models with hazard increasing over time transformed to the original scale.
Results and Discussion
Life-Cycle Performance.
Model 1 was used to assess risk factors contributing to the culling of cows over their entire life cycle. Despite limited variation in age at first calving (Table 1
), its effect approached significance (P = 0.08). The trend indicated cows that were 730 d old at first calving were at less risk of being culled than cows >730 d of age at first calving. This trend seems consistent with the rationale underlying the usual recommendation that heifers be bred to calve early relative to their subsequent intended calving date (Lienard, 1975; Bogart and Taylor, 1983). Previous studies of dairy cows found that the risk of being culled increased approximately linearly with increasing age at first calving (Ducrocq, 1994; Vukasinovic et al., 2001). The increased risk associated with late calvers may be attributed to decreased fertility (Vukasinovic et al., 2001). However, both Ducrocq (1994) and Vukasinovic et al. (2001) noted that the amount of variation in longevity explained by age at first calving was minimal and subsequently removed the variable from the analysis.
Although the birth weight of a cow’s calf did not influence her risk of being culled, cows experiencing dystocia were at a 58% greater risk of being culled (Figure 1) than cohorts that did not experience dystocia (P < 0.01). Thus, in these data, birth weight per se was not a source of increased risk, except to the extent that excessive birth weights contribute to increased incidence of dystocia. The importance of birth weight as an indicator of calving difficulty has been supported by high genetic correlations reported between birth weight and dystocia (Bellows et al., 1971; Gregory et al., 1995). Furthermore, birth weight has been reported to be the single most important factor leading to the incidence of dystocia (Bellows et al., 1971; Colburn et al., 1997).
|
).
|
and 4, respectively. As the BV for cow weight—observed at weaning—increased, the risk of being culled decreased, whereas the risk of being culled increased with increasing BV for maternal preweaning gain. One possible explanation for these trends is that cows with lower milk production (i.e., lower maternal BV for preweaning gain) may maintain and accumulate energy reserves during lactation and therefore be heavier at weaning than contemporaries with greater milk production (Davis et al., 1983). Conversely, cows producing more milk may have expended more of their body energy reserves and therefore be lighter and not in adequate condition to rebreed (Short and Adams, 1988). Likewise, Vollema et al. (2000) found that moderately high producing dairy cows were at a greater risk of being culled than low producing cows, possibly due to increased health and fertility problems. In contrast, a strong positive relationship between milk production and longevity has been observed in dairy cattle when culling was based, at least in part, on performance level (Ducrocq et al., 1988; Ducrocq, 1994; Vukasinovic et al., 2001).
|
|
Early-in-Life Indicators of Future Longevity.
Patterns of significance and direction of effects observed in Model 2 were similar to those observed for the corresponding source of variation in Model 1. This result was anticipated due to the part-whole relationship between sources of variation in the two models. Cows experiencing dystocia as first-calf heifers were at 25% greater risk of being culled than cohorts not experiencing dystocia (P < 0.05). The risk ratio associated with dystocia in first-calf heifers is less than half the corresponding risk ratio over all parities. This may result from closer observation of heifers at calving time and earlier intervention in difficult deliveries, relative to older cows. Neither calf birth weight nor calf gain from birth to weaning had a significant effect on subsequent longevity of the dam.
Birth year of the female affected her risk of being culled (P < 0.01, Figure 5). Culling policies are usually more lenient when a herd is expanding in size, and cows within an expanding herd are at a lower risk of being culled in comparison to cows in a herd with decreasing size (Ducrocq, 1994; Vollema et al., 2000). This phenomenon was observed in the early years of this study when few animals of the breed composition Red Angus, Charolais, and Tarentaise had been produced. During the earlier part of the herd’s development, fewer than 50 first-calf heifers calved each year, whereas approximately 100 first-calf heifers calved annually in later years. Across the birth years 1984 to 1998, there was a tendency for the risk of a female being culled to increase. This increase reflects a decrease in the length of productive life as the desired population size was attained, and increased culling did not compromise other research objectives. The risk ratio associated with females born in 1999 is artificially inflated due to the lack of uncensored observations from this cohort.
|
).
Heritability.
Estimated sire variances for functional and true longevity were, respectively, 0.03738, with a support interval of 0.03672 to 0.03820, and 0.02813, with a support interval of 0.02746 to 0.02880. Corresponding heritability estimates were 0.14 and 0.11. In comparison, heritability estimates for functional longevity using data from dairy cows ranged from 0.02 to 0.2 (Vollema and Groen, 1998; Neerhof et al., 2000). Stayability, defined as the probability that a cow weans five calves given that she weans one, may be viewed as an alternative phenotype for genetic improvement of the length of productive life. Similar to results of this study, heritability estimates for stayability ranged from 0.11 to 0.14 in two herds of Angus cattle evaluated by Snelling et al. (1995).
Implications
The relatively low heritability and the lack of indicators of longevity expressed early in life suggest that the genetic improvement of longevity may continue to be difficult due to prolonged generation intervals and relatively little response per unit of selection applied. Matching the genetic potential of cows to the production environment, such that rebreeding performance is not compromised by concurrent lactation, seems to be a consideration in retaining beef females when open cows are culled. Calving difficulty seems to be an important risk factor contributing to the early culling of beef females, and management may seek to reduce its frequency or mitigate its effects.
Footnotes
1 This research was conducted under a cooperative agreement between USDA-ARS and the Montana Agric. Exp. Stn. USDA, ARS, Northern Plains Area, is an equal opportunity/affirmative action employer and all agency services are available without discrimination. 
2 Mention of a proprietary product does not constitute a guarantee or warranty of the product by USDA, Montana Agric. Exp. Stn. or the authors and does not imply its approval to the exclusion of other products that may also be suitable.
3 Present address: Department of Animal and Poultry Sciences, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061-0306.
4 Correspondence: 243 Fort Keogh Rd. (phone: 406-232-8213; fax: 406-232-8209; e-mail: mike{at}larrl.ars.usda.gov).
Received for publication August 26, 2003. Accepted for publication October 28, 2003.
Literature Cited
Cox, D. R. and D. Oakes. 1984. Analysis of Survival Data. Chapman and Hall, London.
Bellows, R. A., R. E. Short, D. C. Anderson, B. W. Knapp, and O. F. Pahnish. 1971. Cause and effect relationships associated with calving difficulty and calf birth weight. J. Anim. Sci. 33:407–415.
Bogart, R., and R. E. Taylor. 1983. Scientific Farm Animal Production, 2nd ed. Burgess, Minneapolis, MN.
Boldman, K. G., L. A. Kriese, L. D. VanVleck, C. P. VanTassell, and S. D. Kachman. 1995. Page 114 in a Manual for Use of MTDFREML: A Set of Programs to Obtain Estimates of Variances and Covariances (draft). USDA, ARS, Washington, DC.
Colburn, D. J., G. H. Deutscher, M. K. Nielsen, and D. C. Adams. 1997. Effects of sire, dam traits, calf traits, and environment on dystocia and subsequent reproduction of two-year-old heifers. J. Anim. Sci. 75:1452–1460.
Davis, M. E., J. J. Rutledge, L. V. Cundiff, and E. R. Hauser. 1983. Life cycle efficiency of beef production: II. Relationship of cow efficiency ratios to traits of the dam and progeny weaned. J. Anim. Sci. 57:852–866.
Ducrocq, V. 1994. Statistical analysis of length of productive life for dairy cows of the Normande breed. J. Dairy Sci. 77:855–866.[Abstract]
Ducrocq, V., and H. Solkner. 1998. The Survival Kit—V3.0: A Package for Large Analysis of Survival Data. Proc. 6th World Cong. Genet. Appl. Livest. Prod., Armidale, Australia 27:447–448.
Ducrocq, V., R. L. Quaas, E. J. Pollak, and G. Casella. 1988. Length of productive life of dairy cows. 1. Justification of a Weibull mode. J. Dairy Sci. 71:3061–3070.
Graser, H.-U., S. P. Smith, and B. Tier. 1987. A derivative free approach for estimating variance components in animal models by restricted maximum likelihood. J. Anim. Sci. 64:1362–1370.
Gregory, K. E., L. V. Cundiff, and R. M. Koch. 1995. Genetic and phenotypic (co)variances for production traits of female populations of purebred and composite beef cattle. J. Anim. Sci. 73:2235–2242.[Abstract]
Hunter, C. 1994. More Years Mean More Money: Longevity Pays. Texas Longhorn Reference Library. Available: http://www.longhornshowcase.com/Library/LeanBeef/longevity.shtml. Accessed April 3, 2003.
Lagakos, S. W. 1979. General right censoring and its impact on the analysis of survival data. Biometrics 35:139–156.[Medline]
Lienard, G. 1975. Economic aspects of early calving in suckling herds. Pages 177–190 in the Early Calving of Heifers and Its Impact on Beef Production. J. C. Taylor, ed. Commission of the European Communities, Brussels.
MacNeil, M. D., and M. D. Grosz. 2002. Genome-wide scans for QTL affecting carcass traits in Hereford x composite double backcross populations. J. Anim. Sci. 80:2316–2324.
Neerhof, H. J., P. Madsen, V. P. Ducrocq, A. R. Vollema, J. Jensen, and I. R. Korsgaard. 2000. Relationships between mastitis and functional longevity in Danish Black and White dairy cattle estimated using survival analysis. J. Dairy Sci. 83:1064–1071.[Abstract]
Newman, S., M. D. MacNeil, W. L. Reynolds, B. W. Knapp, and J. J. Urick. 1993a. Fixed effects in the formation of a composite line of beef cattle: I. Experimental design and reproductive performance. J. Anim. Sci. 71:2026–2032.[Abstract]
Newman, S., M. D. MacNeil, W. L. Reynolds, B. W. Knapp, and J. J. Urick. 1993b. Fixed effects in the formation of a composite line of beef cattle: II. Pre- and postweaning growth and carcass composition. J. Anim. Sci. 71:2033–2039.[Abstract]
Rendel, J. M., and A. Robertson. 1950. Estimation of genetic gain in milk yield by selection in a closed herd of dairy cattle. J. Genet. 50:1–9.
Short, R. E., and D. C. Adams. 1988. Nutritional and hormonal interrelationships in beef cattle reproduction. Can. J. Anim. Sci. 68:29–39.
Smith, S. P., and H.-U. Graser. 1986. Estimating variance components in a class of mixed models by restricted maximum likelihood. J. Dairy Sci. 69:1156–1165.
Snelling, W. M., B. L. Golden, and R. M. Bourdon. 1995. Within herd genetic analyses of stayability of beef females. J. Anim. Sci. 73:993–1001.[Abstract]
Vollema, A. R., and A. F. Groen. 1998. A comparison of breeding value predictors for longevity using a linear model and survival analysis. J. Dairy Sci. 81:3315–3320.[Abstract]
Vollema, A. R., S. Van Der Beek, A. G. F. Harbers, and G. De Jong. 2000. Genetic evaluation for longevity of Dutch dairy bulls. J. Dairy Sci. 83:2629–2639.[Abstract]
Vukasinovic, N., J. Moll, and L. Casanova. 2001. Implementation of a routing genetic evaluation for longevity based on survival analysis techniques in dairy cattle populations in Switzerland. J. Dairy Sci. 84:2073–2080.[Abstract]
Yazdi, M. H., P. M. Visscher, V. Ducrocq, and R. Thompson. 2002. Heritability, reliability of genetic evaluations and response to selection in proportional hazard models. J. Dairy Sci. 85:1563–1577.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  G. L. Bennett Experimental selection for calving ease and postnatal growth in seven cattle populations. I. Changes in estimated breeding values J Anim Sci, September 1, 2008; 86(9): 2093 - 2102. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. E. Ricklefs and C. D. Cadena Heritability of Longevity in Captive Populations of Nondomesticated Mammals and Birds J. Gerontol. A Biol. Sci. Med. Sci., May 1, 2008; 63(5): 435 - 446. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. MacNeil and T. B. Mott Genetic analysis of gain from birth to weaning, milk production, and udder conformation in Line 1 Hereford cattle J Anim Sci, July 1, 2006; 84(7): 1639 - 1645. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. MacNeil Genetic evaluation of the ratio of calf weaning weight to cow weight J Anim Sci, April 1, 2005; 83(4): 794 - 802. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Tarres, J. Casellas, and J. Piedrafita Genetic and environmental factors influencing mortality up to weaning of Bruna dels Pirineus beef calves in mountain areas. A survival analysis J Anim Sci, March 1, 2005; 83(3): 543 - 551. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
