| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Station d’Amélioration Génétique des Animaux, Institut National de la Recherche Agronomique, BP 27 and
Comite National Brebis Laitière and
Institut de l’Elevage, BP 18, 31326 Castanet-Tolosan cedex, France
Corresponding author:
R. Rupp; e-mail:
rupp{at}toulouse.inra.fr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The heritability estimate of lactation somatic cell score was close to 0.13 and similar for first and second parity. Heritabilities of somatic cell scores increased from first to fourth test day (from 0.07 to 0.11 in first lactation and from 0.05 to 0.13 in second lactation). Genetic correlations between somatic cell scores were high, usually more than 0.91, but lower between first test day and later test days in first lactation (0.64 to 0.88). The genetic correlations between lactation somatic cell score and milk yield, between lactation somatic cell score and fat content, and between lactation somatic cell score and protein content were 0.18, 0.04, and 0.03 in first lactation, respectively. The genetic antagonism between test day somatic cell score and milk yield measured in first lactation increased from beginning to the end of the lactation (0.05 to 0.23). This antagonism was slightly lower for somatic cell score in second lactation (from 0.09 to 0.14, and 0.08 for lactation mean). Environmental correlations in first lactation between lactation somatic cell score and milk yield, between lactation somatic cell score and fat content, and between lactation somatic cell score and protein content were -0.18, 0.13, and 0.30, respectively.
Key Words: dairy sheep • somatic cell count • production traits • genetic parameters
Abbreviation key: LSCS = lactation mean SCC
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Genetic studies of SCC in dairy sheep are more recent and are scarcer than in dairy cattle. Reported results, however, are consistent and show that the trait exhibits genetic variation in this species. Initial results based on repeatability test day models for SCS, i.e., log transformed SCC to achieve normality of distribution, indicated low heritability estimates, ranging from 0.04 (Baro et al., 1994) to 0.09 (El Saied et al., 1998). But more recent studies reported higher heritability estimates for the lactation mean SCS, from 0.11 to 0.18 (El Saied et al., 1999; Mavrogenis et al., 1999; Barillet et al., 2001; Rupp et al., 2001). These results were very similar to values from the dairy cattle literature (Mrode and Swanson, 1996; Heringstad et al., 2000). Thus, the level of genetic variability of SCC in dairy sheep appears sufficient to implement selection. Genetic correlation estimates with milk yield, however, are quite inconsistent across dairy sheep studies, ranging from antagonistic, i.e., from 0.09 to 0.20 (Mavrogenis et al., 1999; Barillet et al., 2001; Rupp et al., 2001), to favorable, i.e., from -0.15 to -0.37 (Baro et al., 1994; El Saied et al., 1998 and 1999).
Following the first genetic analyses on experimental data (Barillet et al., 1999, 2001), an extensive recording procedure for SCC has been implemented for French dairy sheep breeds as part of the milk recording system. The consistent database should allow further research and genetic evaluation of SCS.
The objective of this study was to estimate genetic parameters for SCS in first and second lactations using records from the national database. Models based on lactation means and multitrait within-lactation approaches were used. Relationships with dairy production traits were also investigated. Results should allow us to define a model to be used for genetic evaluation for SCS.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
From 1999 to 2001, the Lacaune database includes SCC records for 204,256 first and second lactations of ewes in 384 flocks, in addition to production trait information (fat and protein content, milk yield, pedigrees, etc.). To avoid bias due to selection, information gathered during parity 2 was considered only if information for parity 1 was available. Moreover, ewes had to be sired by either proven sires born after 1992 who have at least 100 daughters, or sampling sires, born from 1996 to 1999 who have at least 20 daughters. After edits, the dataset included 94,474 first lactations and 45,499 second lactations of Lacaune ewes. Selected ewes originated from 1221 and 171 sampling and proven sires, respectively. The average number of SCC records per lactation was 2.9, and most ewes (87%) had three records per lactation. For milk yield, an average of 5.1 test day records per lactation were available. Characteristics of the data set are summarized in Table 1
.
|
Monthly SCS were also considered as independent, single traits according to DIM at test-day: 25–55 (SCS1), 56–85 (SCS2), 86–115 (SCS3), and 116–145 (SCS4).
Milk yield per lactation was estimated using the Fleischmann method and was adjusted for milking length on a reference period of 160 d (Barillet, 1985). Lactation traits for fat and protein content were defined as the arithmetic mean of test-days’ record adjusted for DIM, following the same procedure as for LSCS.
Methods
Variance components were estimated by REML applied to multitrait sire models, using the VCE package (Neumaier and Groeneveld, 1998), which is based on a derivative-free iterative algorithm (Quasi Newton) for parameter estimation. As only 3 yr of data were available, with few dam-daughter pairs with performance records, the sire model will imply a limited loss of information compared to an animal model. A sire group effect was fitted with one level for each proven sire and one level by year of birth for sampling sires. Therefore, proven sires were treated as fixed, and only sampling sires contributed to the estimation of variance components. Underestimation of genetic variance due to highly selected rams was avoided.
Because of computation limitations, a maximum of four traits were analyzed simultaneously. To take into account the selection process from first to second lactation, which is based on milk production and somatic cell counts, all analyses including traits in second lactation also included average milk yield and LSCS in first lactation. The model for lactation measures in first or second parity was:
 | (1) |
where:
| yijklmn | = | LSCS, average milk yield, fat content, or protein content in first or second lactation;
| FYi | = | fixed effect of flock x year combination i (1143 and 680 levels for first and second lactations, respectively);
| Mj | = | fixed effect of month at lambing j within parity (5 levels);
| Ak | = | fixed effect of age at lambing k within parity (5 and 3 levels for first and second lactation, respectively);
| Ll | = | fixed effect of the number of lambs born (2 levels: 1 vs. 2 or more);
| Gm | = | fixed sire group effect (one level per sire for proven rams and one level per birth year for sampling rams);
| Sn | = | random genetic effect of sire (within sire group);
| eijklmn | = | random residual effect.
|
The four single test-day SCS traits (SCS1 to SCS4) defined according to DIM were considered as different traits in a multiple-trait test-day model approach. The model used was similar to (1) and also included the fixed effect of days in milk at sampling (15 levels with 2-day steps). Three generations of ancestors were considered to compute the pedigree file, which included 4390 individuals.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
. Mean LSCS and increase of SCS from beginning to the end of the lactation were in agreement with previous French reports (Barillet et al., 2001). However, no increase in LSCS with parity was observed, contrary to trends usually reported (Gonzalo et al., 1994; Fuertes et al., 1998). Further investigation showed that ewes culled at the end of first lactation, completed in 1999 or 2000, had much higher somatic cell counts (3.60 and 690,000 cells/ml, for LSCS and average SCC, respectively) than ewes that were allowed to start a second lactation (3.00 or 330,000 cells/ml, for LSCS and average SCC, respectively). This showed that breeders started to use SCC in culling strategies as soon as this information had been made available to them. It may explain the absence of an increase in SCC from first to second lactation, and makes it necessary to perform multitrait analyses that include milk and somatic cell count in first lactation when REML is applied to second lactation SCS data.
|
). Results were in agreement with previous French studies (Barillet et al., 2001; Rupp et al., 2001). These studies were based on data from ewes of 38 Lacaune flocks between 1993 and 1999 with 5 to 6 test-day SCC per lactation. Heritabilities of LSCS, estimated with an animal model, ranged from 0.14 to 0.15 in these studies. The value estimated presently would have been a little higher if an animal model instead of a sire model could have been used. Comparable heritabilities for LSCS (0.12) were also reported from Spanish Churra flocks (El Saied et al., 1999) and in recent dairy cattle studies (Mrode and Swanson, 1996; Boettcher et al., 1998; Rupp and Boichard, 1999).
|
and 5) were lower than the estimate for LSCS. Heritabilities of SCS increased from 0.07 for SCS1 to 0.11 for SCS4 in first lactation, and from 0.05 for SCS1 to 0.13 for SCS4 in second lactation. This increase of heritability with DIM reflected an increase of the genetic variance (Table 6). Conversely, the environmental variance was quite stable, although slightly higher at the beginning of the lactation. This result confirms similar trends observed in previous Spanish (Baro et al., 1994) and French (Barillet et al., 2001) studies. In the latter studies, heritabilities in the first 2 mo of lactation were lower (0.01 to 0.05) than the present study.
|
|
|
) was high (0.93), as observed previously (Rupp et al., 2001), and in agreement with dairy cattle data (Mrode and Swanson, 1996; Boettcher et al., 1998). This finding suggested that LSCS in first and second lactation were essentially the same trait genetically. Similarly, genetic correlations between single traits SCS1 to SCS4 and LSCS (Tables 4 and 5) and among single traits were high (0.64 to 0.99), especially for SCS2 to SCS4. However, SCS1 exhibited a lower correlation with SCS2 to SCS4, which could be due to the influence of both the suckling period (for 1 mo following lambing) and a biological effect of starting lactation for the first time in primiparous ewes. Environmental correlations (Tables 4 and 5) were highest between successive test day SCS and decreased with increasing interval between measures, i.e., from 0.55 to 0.36, reflecting a repeatability of monthly SCS of approximately 0.4 (Barillet et al., 2001).
Relationships between SCS and Milk Production Traits
Genetic parameters for production traits (Table 3) were in accordance with the ovine literature. Heritabilities for milk yield were about 0.30, and heritabilities for fat and protein content were between 0.40 and 0.60 (Barillet and Boichard, 1987; Barillet and Boichard, 1994; Sanna et al., 1997; Barillet et al., 2001).
Genetic correlations between LSCS and fat and protein content were close to zero (Table 3). As expected, the genetic correlation (Table 3) between milk yield in first lactation and LSCS in second lactation (0.08) was slightly lower than the genetic correlation between milk yield in first lactation and LSCS in first lactation (0.18). These values reflected a moderate genetic antagonism between the two traits and were in the range of values previously reported on a smaller French dataset (Rupp et al., 2001). Indeed, Rupp et al. (2001) found genetic correlations between LSCS and milk yield ranging from 0.09 to 0.19. Such a genetic antagonism is also reported from dairy cattle data (Mrode and Swanson, 1996; Boichard and Rupp, 1997; Rupp and Boichard, 1999). El Saied et al. (1998and 1999), however, found negative and favorable genetic correlations (-0.15 to -0.23) between SCC and milk yield in the ovine Churra breed. This discrepancy could be due to both the structure of data and the model used (Barillet et al., 2001).
The unfavorable genetic relationship between SCS and milk yield increased during first lactation, from 0.05 at first test day to 0.23 at fourth test day (Table 7). The very low correlation between SCS at first test day and milk yield may be related to the low heritability of SCS1. Nevertheless, results did not confirm the favorable correlation at first test day (-0.48) observed previously (Barillet et al., 2001), but standard errors were substantial in some cases. In the present study, all first and second lactation results showed genetic antagonism between milk yield and SCS, indicating that the udder health may be deteriorating under selection for production traits. As reported by others (Mrode and Swanson, 1996; Barillet et al., 2001), environmental correlation between milk yield and SCS are of opposite sign (e.g., negative), reflecting the reduction in the production ability of the mammary gland when an infection occurs.
|
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication April 8, 2002. Accepted for publication November 12, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Rupp, D. Bergonier, S. Dion, M. C. Hygonenq, M. R. Aurel, C. Robert-Granie, and G. Foucras Response to somatic cell count-based selection for mastitis resistance in a divergent selection experiment in sheep J Dairy Sci, March 1, 2009; 92(3): 1203 - 1219. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Riggio, R. Finocchiaro, J. B. C. H. M. van Kaam, B. Portolano, and H. Bovenhuis Parameters for Milk Somatic Cell Score and Relationships with Production Traits in Primiparous Dairy Sheep J Dairy Sci, April 1, 2007; 90(4): 1998 - 2003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Gutierrez-Gil, M. F. El-Zarei, Y. Bayon, L. Alvarez, L. F. de la Fuente, F. S. Primitivo, and J. J. Arranz Short Communication: Detection of Quantitative Trait Loci Influencing Somatic Cell Score in Spanish Churra Sheep J Dairy Sci, January 1, 2007; 90(1): 422 - 426. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Legarra and E. Ugarte Genetic Parameters of Udder Traits, Somatic Cell Score, and Milk Yield in Latxa Sheep J Dairy Sci, June 1, 2005; 88(6): 2238 - 2245. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
