Genetic Analysis of Somatic Cell Score and Female Fertility of Israeli Holsteins with an Individual Animal Model

GENETICS AND BREEDING Genetic Analysis of Somatic Cell Score and Female Fertility of Israeli Holsteins with an Individual Animal Model J. I. WELLER* and E. EZRA† *Institute of Animal Sciences, Agricultural Research Organization, The Volcani Center, Bet Dagan 50250, Israel †Israel Cattle Breeders’ Association, Netanya, Israel

ABSTRACT The individual animal model was used to analyze the conception index, defined as the inverse of the number of inseminations to conception, and mean lactation somatic cell scores (SCS) of Israeli Holsteins. In addition, multitrait variance components were estimated for these traits and for milk production traits. The animal model analyses included 440,558 lactation records for conception index and 224,869 SCS records collected over 10 yr. First parity heritabilities were 0.035 for conception index and 0.158 for SCS. Genetic correlations were negative between conception index and milk production traits and positive between SCS and milk production traits. Genetic correlations between first and second lactations were 0.9 for conception index and 0.7 for SCS. Phenotypic correlations were 0.08 and 0.3, respectively. Despite the low heritability for conception index, the standard deviation of the sire evaluations was 2%, and the range of evaluations was 10%. Predicted genetic trends for conception index and SCS were computed based on the current selection index of Israel. The realized genetic trend for the conception index was slightly positive, despite the predicted negative trend, and the realized genetic trend for SCS, although positive as predicted, was fourfold the predicted value. The reasons for these discrepancies are not known. ( Key words: genetic analysis, somatic cell score, female fertility, individual animal model) Abbreviation key: CI = conception index, HYS = herd-year-season, IAM = individual animal model, MSCS = mean lactation SCS. INTRODUCTION In 1991, the Israeli selection index was changed from an index based chiefly on milk production to an

Received June 7, 1996. Accepted September 9, 1996. 1997 J Dairy Sci 80:586–593

index based chiefly on protein production; the coefficient for carrier was negative (33). The breeding index coefficients are –0.274, 6.41, and 34.85 for kilograms of milk, fat, and protein, respectively. No other traits are included in the index. Routine genetic analysis of milk, fat, and protein production of the Israeli Holstein population via the individual animal model ( IAM) began in 1993 (34, 35). Routine genetic analysis of SCS and female fertility via the IAM began in 1995. Many studies have shown high genetic correlations among SCS, clinical mastitis, and bacterial infection (4, 6, 7, 9, 12, 26, 27, 37). Unlike clinical mastitis and bacterial infection, which have low heritability and are difficult to score, heritability for SCS is moderate and can be automatically scored together with fat and protein concentration. Therefore, numerous studies (4, 6, 7, 9, 12, 20, 22, 23, 24, 26, 27, 28, 37, 40) have recommended that this trait be an additional criterion for selection. Selection for an index based only on production traits should also lead to genetic changes in other traits that have genetic correlations with production traits (26, 28, 31, 32). Numerous studies have estimated genetic correlations between female fertility and milk production (8, 11, 29, 31) and between SCS and milk production (1, 2, 7, 23, 37), and most studies have found economically negative correlations. That is, selection for milk or protein production should result in increased SCS and reduced female fertility. There are only a few reports (11, 31, 40) of realized longterm genetic trends for either fertility or SCS. Analysis of female fertility is further complicated by the lack of consensus on the trait definition, although numerous definitions have been suggested, all of which are problematic (8, 10, 11, 13, 14, 15, 16, 29, 31, 36). Fertility data in Israel are unique in that all cows that do not display estrus within 60 d of the last insemination are checked for pregnancy by a veterinarian (21, 31, 36). Thus, unless the cow is culled prior to 60 d, the result of each insemination is known. Two models have been used in Israel for genetic analysis of fertility, an insemination model, in which each insemination was considered as a separate record, and a lactation model, in which the

586

587

SOMATIC CELLS AND FEMALE FERTILITY

fertility trait was defined as the inverse of the number of inseminations to conception (21, 31, 36). This trait has been denoted previously as conception index ( CI) . Both methods have theoretical advantages and disadvantages, which have been discussed previously (31). Analysis of CI has the additional technical advantage of ready adaptation for analysis via the IAM, which is similar to the analysis model for production traits and SCS (22, 34, 35). In IAM analyses, genetic evaluations are derived for all animals that are included in the analysis. Thus, accurate estimates of genetic trends can be readily calculated. The objectives of this study were to estimate the genetic and phenotypic parameters for SCS, female fertility, and production traits; to estimate genetic and phenotypic trends for fertility and SCS via an IAM analysis; and to compare these trends to those trends predicted from the genetic and phenotypic variance matrices and by selection on the current breeding index. MATERIALS AND METHODS Determination of CI For cows that conceived, uncorrected CI was computed as 100/con, where con = number of inseminations to conception. For cows that were inseminated at least once but that were culled prior to conception, con was replaced with its expectation, E(con j) , where j = number of recorded inseminations. E(con j) was estimated using 1,196,686 first through sixth inseminations for parities 1 though 5 that occurred between January 1, 1985 and August 31, 1995 and for which conception status was known. We computed mean conception rate for each combination of insemination number and parity. For cows with j < 6, E(con j) was computed separately for each parity as follows: i = 13

E(con j) =

∑

i(p i)pc i

i = j + 1

[1]

where i = number of the insemination beginning with insemination j + 1; pi = probability the cow did not conceive prior to insemination i, given that the cow did not conceive at inseminations 1 through j; and pci = probability of conception at insemination i, computed from the mean conception rate for insemination i, computed from the data described. For insemination j + 1, pi = 1. For inseminations j + 2 to sixth, pi was computed as follows: k = i

pi =

∏

k = j + 2

( 1 – pck)

[2]

where P = multiplicative sum, k = insemination number from j + 2 to i, and the other terms are as defined previously. For k ≤6, pck was set to the appropriate mean conception rate. For i >6, pck was set equal to pc6 for the corresponding parity. The summation up to i = 13 in Equation [1] was arbitrary, but preliminary studies showed that increasing i did not significantly affect E(con j) because pi tended to 0 for high values of i. E(con j) up to insemination 5 for parities 1 through 5 are given in Table 1. The values are approximately equal to j + 3 but increased slightly as parity increased. This increase was due to a decrease in pj as parity increased (21). For cows with six or more inseminations per parity, E(con j) was arbitrarily set to E(con 5) + 1 for each parity. Previous results showed a strong seasonal effect on conception rate of Israeli Holsteins (21, 36). Because the lactation model could not correct for season of each insemination, we decided to correct CI for month of calving. To derive correction factors, uncorrected CI was computed from 518,730 lactation records from parities 1 through 5 with at least one valid insemination record. Results are given in Table 2. As expected from previous results, uncorrected CI was lowest for cows calving during June, which tended to have their first insemination during August, and highest for cows calving during November, which tended to have all inseminations during winter. Corrected conception index was then computed as follows: CI = CIu – Mij + M1–4

[3]

where CI = corrected conception index, CIu = uncorrected CI, Mij = mean of CIu for parity i and month of calving j, and M1–4 = mean of CIu for first parity cows calving during April. Thus, for first parity cows calving during April, CI = CIu, and all other combinations of parity and calving month deviated relative to this base. Computation of Mean Lactation SCS The SCS was computed as 3 + log2(cell/100), where cell = somatic cells per microliter (2, 22, 23, 24, 27, 39, 40). The mean SCS ( MSCS) was computed as the lactation mean of SCS for lactations from parities 1 through 5 with at least four valid SCS records, corrected for the effects of parity, month of test, square root of DIM, DIM, DIM2, interaction of parity and the square root of DIM, and interaction of parity and DIM. The effects of these factors were estimated with a linear model analysis of 3,593,423 Journal of Dairy Science Vol. 80, No. 3, 1997

588

WELLER AND EZRA

TABLE 1. Expected number of inseminations to conception by last recorded insemination and parity. Parity

Last recorded insemination

1

2

3

4

5

1 2 3 4 5

3.84 5.11 6.28 7.37 8.39

3.98 5.19 6.35 7.44 8.45

3.95 5.18 6.40 7.53 8.54

4.06 5.32 6.55 7.68 8.73

4.21 5.40 6.58 7.69 8.66

valid SCS records from January 1, 1985 through September 30, 1995, which also included the effect of herd-year, as described previously (37). Records were deleted if any effects included in the model were missing or invalid or if sire of the cow was unknown. Adjustment factors are given in Table 3. All factors were additive on SCS. Factors for parity and stage of lactation were similar to those computed previously for the US and Canadian Holstein populations (39, 40), except for the effect of month of test. For the Israeli population, SCS was lowest during the early summer and highest during winter. The base value for adjustment was fourth parity SCS during December extrapolated to 0 DIM. For each SCS, the appropriate factors given in Table 3 were subtracted from the recorded value. IAM Analyses for CI and MSCS For both CI and MSCS, records were included only for cows with a valid first parity record. For the analysis of CI, only lactations with at least one valid insemination record were included. Records of parities up to 5 were included, provided that a valid record existed for all previous lactations. All known parents and grandparents of cows with records and

the paternal grandsire and granddam of sires of cows with records were included in the analysis. The CI, MSCS, and milk, fat, and protein production were analyzed with the following animal model: Yijk = HYSi + Aj + PEj + Pk + eijk

where Yijk = CI, MSCS, milk, fat, or protein record for parity k of cow j in herd-year-season i; HYSi = fixed effect of herd-year-season ( HYS) i; Aj = random portion of additive genetic merit plus the effects for unknown parent groups for cow j; PEj = random permanent environmental effect of cow j; Pk = fixed effect of parity k; and eijk = random residual associated with each record. Two seasons, beginning in April and October, were determined for each herd-year. Only animals with valid records for milk, fat, and protein and calving dates between January 1, 1985 and September 30, 1995 were included in the analyses of production traits. The numbers of animals, records, genetic groups, and HYS included in the three IAM analyses are included in Table 4. As already noted, the total number of inseminations included in the analysis was 1,196,686, and the number of lactation records included was 440,558. Thus, the mean number of inseminations per lactation was 2.7. Table 5 presents variance components for the additive genetic, permanent environment, and residual variance components as a fraction of the total phenotypic variance for the three analyses. These values are based on previous studies (5, 10, 22, 24, 27, 31, 37, 38, 40). Values for MSCS and CI were validated as described. Reliabilities were estimated as described previously (18). The genetic base for all evaluations was set to the mean of cows born during 1990. Genetic trends were computed as the regression of genetic evaluations for cows on their birth dates

TABLE 2. Mean percentage conception index by month of calving and parity. Parity

Month of calving

1

2

3

4

5

January February March April May June July August September October November December

63.9 61.8 58.9 54.7 48.4 46.5 50.9 58.0 65.2 66.5 66.9 65.7

60.8 57.5 53.7 49.9 45.3 41.1 43.2 48.4 56.0 62.2 62.9 62.8

58.9 56.2 52.8 49.2 46.4 44.7 46.0 49.0 55.6 60.6 61.8 61.1

56.2 53.0 51.5 47.9 45.8 45.0 45.5 49.2 55.6 59.3 60.0 59.0

54.3 49.3 47.5 45.4 42.4 40.7 43.5 47.6 54.2 56.6 57.5 56.9

Journal of Dairy Science Vol. 80, No. 3, 1997

[4]

589

SOMATIC CELLS AND FEMALE FERTILITY TABLE 3. Adjustment factors for SCS. Factor

Level

Effect

Month of test

January February March April May June July August September October November December 1 2 3 4

0.001 –0.018 –0.103 –0.221 –0.251 –0.279 –0.208 –0.104 0.022 –0.072 –0.039 0.000 0.197 –0.839 –0.562 0.000 –0.3653 0.0336 –3.76(10–5) –0.280 –0.064 –0.017 0.000 0.009 0.003 0.001 0.000

Parity

DIM0.5 DIM DIM2 Parity by DIM0.5

1 2 3 4 1 2 3 4

Parity by DIM

beginning with cows born during 1981. Phenotypic trends for CI and MSCS were estimated as the regression of first parity, corrected records on birth dates of the cows beginning in 1983 for CI and 1990 for MSCS. Cows with records prior to these dates were selected samples and do not accurately reflect the population means.

by VanRaden and Jung (30). Cow culling was scored dichotomously. Cows that were culled during first parity were scored as 1 and were scored 0 otherwise. The model included the effects of sire, HYS, sire group by year of birth, and residual. Only the sire and residual effects were considered to be random. The HYS were determined as described. Only records of cows with valid first parity records for all six traits and calving dates between January 1, 1985 and August 31, 1995 were included in the analysis. Records of cows whose sires had <10 daughters were also deleted. Relationships other than sire and daughter were not considered in this analysis. In addition, the genetic and phenotypic variancecovariance matrices between first and second parity CI and MSCS were estimated separately for each trait with the same model. Only cows with valid records for both parities and calving dates between January 1, 1985, and August 31, 1995 for first lactation were included in the analyses. The HYS was determined relative to first parity. Cows that changed herds between parities were deleted. Otherwise, edits were the same. The number of sires, groups, HYS, and records for all three analyses are given in Table 6. Heritabilities were estimated as four times the sire component of variance divided by the sum of the sire and residual variances. Genetic correlations were estimated as correlations among sire variance components. Phenotypic correlations were estimated as the correlation of the sum of the sire and residual covariances. The vector of predicted genetic trends after 10 yr of selection on the Israeli breeding index, f, was determined as follows (32):

Estimation of Genetic Parameters

f = iCb/ ( b′Pb) 0.5

Genetic and phenotypic variance-covariance matrices for milk, fat, and protein production; CI; MSCS; and cow culling during first lactation were estimated with a sire model by multivariate REML as described

[5]

where i = cumulative selection intensity, which was set equal to 4 for 10 yr of selection; C = genetic covariance matrix between the six traits in f and the

TABLE 4. The numbers of animals, records, genetic groups, and herd-year-seasons (HYS) included in the animal model analyses of conception index (CI); mean lactation SCS (MSCS); and milk, fat, and protein production. Factor

CI

MSCS

Milk, fat, and protein production

Sires Total cows Cows born since 1980 Cows with records Total records Genetic groups HYS

903 229,631 214,102 185,613 440,558 48 4842

876 217,198 209,328 120,904 224,869 48 8821

1033 301,870 285,812 221,542 500,151 48 11,782 Journal of Dairy Science Vol. 80, No. 3, 1997

590

WELLER AND EZRA TABLE 5. Variance components for animal model analyses as a fraction of the total phenotypic variance. Variance component

CI1

MSCS

Milk, fat, and protein production

Additive genetic Permanent environment Residual

0.025 0.075 0.900

0.100 0.250 0.650

0.250 0.250 0.500

1CI

= Conception index; MSCS = mean lactation SCS.

three traits in the index; b = vector of index coefficients for milk, fat, and protein production; and P = phenotypic variance matrix for the three traits in the index. RESULTS AND DISCUSSION The first parity heritabilities for milk, fat, protein production and for CI, MSCS, and culling rate, as well as the genetic and phenotypic correlations among these traits, are given in Table 7. Heritabilities were 0.035, 0.158, and 0.021 for CI, MSCS, and culling rate, respectively. Heritabilities were similar to those reported in most previous studies (1, 2, 5, 7, 8, 10, 11, 13, 14, 15, 16, 17, 19, 20, 23, 24, 25, 26, 27, 29, 31, 37, 38). Genetic correlations were negative between CI and milk production traits and positive between MSCS and milk production traits. Both of these relationships are economically unfavorable but similar to those of most previous studies (2, 7, 8, 9, 29, 37, 38). Weller ( 3 1 ) found no genetic correlation between CI and either milk or fat in a previous analysis of this population using Henderson’s method 3. Hermas et al. ( 1 1 ) found a negative genetic correlation between milk and CI, but positive correlations between CI and fat and protein. Reheja et al. ( 1 9 ) found no genetic correlations between milk production and three fertility traits. Most studies (1, 23, 25, 37) that considered later parities found positive genetic correlations between milk production and SCS for first parity but lower or even negative correlations for later parities. As expected, the genetic correlations between culling rate and milk production traits and CI were all negative. The positive genetic correlation between culling rate and MSCS was also as expected. The genetic correlation between CI and MSCS was –0.37. Thus, selection for either milk or protein production should lead to an increase in MSCS and a decrease in CI. The heritability estimates for first and second parity for CI and MSCS and the genetic and phenotypic correlations between parities are given in Table 8. Heritability estimates in these analyses were slightly lower for both traits. The genetic correlation between Journal of Dairy Science Vol. 80, No. 3, 1997

parities was higher, and the phenotypic correlation was lower, for CI than for MSCS. These results are similar to those of some previous studies that also found high genetic correlations between parities for both CI (13, 15) and MSCS (1, 20). However, Raheja et al. ( 1 9 ) found no genetic correlations among first, second, and third parities for three fertility traits, and Da et al. ( 5 ) estimated the genetic correlation between first and second parity for MSCS as only 0.55. Because the genetic correlation for both traits was >0.7 in the current study, the evaluation model used, which assumes a common genetic effect over all parities, should be a reasonable approximation. Correlations among IAM sire evaluations for milk, fat, and protein production and for CI and MSCS are given in Table 9. Only 351 sires with reliabilities >0.5 for all three analyses are included. Correlations among the sire evaluations were generally similar to the genetic correlations for first lactation that are given in Table 7. All correlations shown in the two tables were in the same direction, except for the correlation between MSCS and fat production, which was very close to 0 in both cases. Coffey et al. ( 4 ) found small negative correlations between sire evaluations for MSCS and both milk and fat production. Standard deviations and ranges for PTA of the 351 sires with reliabilities >0.5 for milk, fat, and protein production and for CI and MSCS are given in Table 10. Values for MSCS were similar to those of most previous studies (2, 22, 40). Despite the low heritability for CI, the standard deviation of the PTA was

TABLE 6. The number of sires, groups, herd-year-seasons (HYS), and records for the variance component analyses. Factor

Six traits, first parity

CI,1 Parities 1 and 2

MSCS, Parities 1 and 2

Sires Groups HYS Records

264 7 2363 59,147

448 6 4329 122,008

241 7 6096 63,575

1CI

= Conception index; MSCS = mean lactation SCS.

591

SOMATIC CELLS AND FEMALE FERTILITY TABLE 7. The first parity heritabilities and the genetic and phenotypic correlations for milk, fat, and protein production as well as conception index (CI), mean lactation SCS (MSCS), and culling rate.1

Milk Fat Protein CI MSCS Culling

Milk

Fat

Protein

CI

MSCS

Culling

0.251 0.597 0.873 0.037 –0.067 –0.144

0.445 0.354 0.663 0.023 –0.045 –0.100

0.743 0.608 0.238 0.023 –0.029 –0.133

–0.286 –0.301 –0.419 0.035 –0.024 –0.344

0.116 0.054 0.172 –0.366 0.158 0.080

–0.510 –0.270 –0.400 –0.469 0.203 0.020

1Genetic correlations are above the diagonal, phenotypic correlations are below the diagonal, and heritabilities are on the diagonal.

2%, and the range was 10%, which was similar to results of previous studies (31). Thus, there are economically important differences between sires. Annual genetic and phenotypic means for first lactation by birth year are plotted for CI in Figure 1 and for MSCS in Figure 2. Annual genetic means for milk, fat, and protein are plotted in Figure 3. Functions were nearly monotonic for all traits except CI, and positive genetic trends are clearly evident for all traits except CI. Predicted and realized genetic trends after 10 yr of selection on the current breeding index and the phenotypic trends are given in Table 11. The realized genetic trends were computed as the annual genetic trends times 10. Realized genetic gains were greater than the predicted gain for milk production and less than the predicted gains for fat and protein production. This result was expected, considering that selection until 1990 was chiefly for milk. The realized genetic trends for both MSCS and CI are not at all similar to the predicted values by either selection on

the breeding index or by direct selection on milk production. The predicted negative trend for CI is economically significant, despite the low heritability for this trait, and this trend has been used previously to justify selection for fertility (11). The realized genetic trend for CI was slightly positive, despite the predicted negative trend, and the realized genetic trend for SCS, although positive, as predicted, was fourfold the predicted value. In a previous analysis of this population, Weller ( 3 1 ) found a positive genetic trend of nearly 1%/yr for CI, based on a sire model without inclusion of a parity effect in the model. However, selection during the period considered was for the previous index based chiefly on milk quantity. Furthermore, recent studies have shown that, in an analysis conducted across multiple parities, estimates of genetic trends can be severely biased if a parity effect was not included in the model of analysis ( 3 ) . Zhang et al. ( 4 0 ) found no apparent genetic trend for MSCS in the Canadian

TABLE 8. First and second parity heritabilities for conception index ( C I ) and mean lactation SCS (MSCS) and the genetic and phenotypic correlations between parities.

Heritability, first parity Heritability, second parity Genetic correlation Phenotypic correlation

CI

MSCS

0.017 0.028 0.891 0.083

0.133 0.132 0.737 0.329

TABLE 9. Correlations among genetic evaluations of 351 sires with reliabilities >0.5 for milk, fat, and protein production; conception index (CI); and mean lactation SCS (MSCS).

Milk Fat Protein CI

Fat

Protein

CI

MSCS

0.359

0.747 0.602

–0.011 –0.020 –0.070

0.145 –0.038 0.184 –0.142

Figure 1. Annual genetic ( ⁄) and phenotypic ( π) means for first parity by birth year for conception index. Journal of Dairy Science Vol. 80, No. 3, 1997

592

WELLER AND EZRA TABLE 10. Standard deviations and ranges for PTA of 351 sires with reliabilities >0.5 for milk, fat, and protein production; conception index (CI); and mean lactation SCS (MSCS).

Figure 2. Annual genetic ( ⁄) and phenotypic ( π) means for first parity by birth year for mean lactation SCS.

Holstein population with an IAM analysis including a parity effect, but only 6 yr of data were analyzed. The discrepancies between the realized and predicted genetic trends for MSCS and CI cannot be explained by random variation among cows. The standard error for the 10-yr genetic trend for CI was 0.017. The standard errors of all other genetic trends were <1% of the trends. However, the regression standard errors do not reflect random variation among sires. About 50 sires are progeny tested annually in the Israeli Holstein population, and only 4 to 5 of these sires are returned to general reproductive service (32). Thus, random variation among sires for CI

Trait

SD

Minimum

Maximum

Milk, kg Fat, kg Protein, kg CI, % MSCS

286.2 9.98 7.30 2.01 0.188

–900.4 –32.5 –21.6 –4.69 –0.323

719.3 30.7 20.4 5.38 0.783

and MSCS could be a significant factor. The positive trend for CI may also be due to positive selection for fertility at the farm level, as indicated by the negative genetic and phenotypic correlation between cow culling and CI. Other possible explanations could be selection on criteria other than the recommended breeding index, for example, conformation traits, which are not included in the index. CONCLUSIONS First parity heritabilities were 0.035 for CI and 0.158 for MSCS. Genetic correlations were negative between CI and milk production traits and positive between MSCS and milk production traits. Genetic correlations between first and second parity were 0.9 for CI and 0.7 for MSCS. Phenotypic correlations were 0.08 and 0.3, respectively. Despite the low heritability for CI, the SD of the sire PTA was 2%, and the range of evaluations was 10%. The realized genetic trend for CI was slightly positive, despite the predicted negative trend, and the realized genetic trend for SCS, although positive as predicted, was fourfold the predicted value. The reasons for these discrepancies are not known.

TABLE 11. The predicted and realized genetic trends and the phenotypic trends for first lactation after 10 yr of selection on the current breeding index.1 Genetic trends

Figure 3. Annual genetic means for milk ( ⁄) , fat ( π) , and protein ( ÿ) by birth year. Journal of Dairy Science Vol. 80, No. 3, 1997

Trait

Predicted

Realized

Phenotypic trends

Milk, kg Fat, kg Protein, kg CI,2 % MSCS Culling rate, %3

718.7 37.0 31.1 –5.50 0.13 –3.4

1039.5 22.3 26.8 0.40 0.53 . . .

1372.2 41.1 32.3 –2.39 0.28 . . .

1The breeding index = –0.274(kg milk) + 6.41(kg fat) + 34.85(kg protein). 2CI = Conception index; MSCS = mean lactation SCS. 3Genetic evaluations and realized trends were not computed for culling rate.

SOMATIC CELLS AND FEMALE FERTILITY

ACKNOWLEDGMENTS This research was supported by a grant from the Israel Milk Marketing Board. REFERENCES 1 Banos, G., and G. E. Shook. 1990. Genotype by environment interaction and genetic correlations among parities for somatic cell count and milk yield. J. Dairy Sci. 73:2563. 2 Boettcher, P. J., L. B. Hansen, P. M. VanRaden, and C. A. Ernst. 1992. Genetic evaluation of Holstein bulls for somatic cells in milk of daughters. J. Dairy Sci. 75:1127. 3 Boichard, D., B. Bonaiti, A. Barbat, and S. Mattalia. 1995. Three methods to validate the estimation of genetic trend for dairy cattle. J. Dairy Sci. 78:431. 4 Coffey, E. M., W. E. Vinson, and R. E. Pearson. 1986. Potential of somatic cell concentration in milk as a sire selection criterion to reduce mastitis in dairy cattle. J. Dairy Sci. 69:2163. 5 Da, Y., M. Grossman, I. Misztal, and G. R. Wiggans. 1992. Estimation of genetic parameters for somatic cell score in Holsteins. J. Dairy Sci. 75:2265. 6 Emanuelson, U. 1988. Recording of production diseases in cattle and possibilities for genetic improvements: a review. Livest. Prod. Sci. 20:89. 7 Emanuelson, U., B. Danell, and J. Philipsson. 1988. Genetic parameters for clinical mastitis, somatic cell counts, and milk production estimated by multiple-trait restricted maximum likelihood. J. Dairy Sci. 71:467. 8 Hansen, L. B., A. E. Freeman, and P. J. Berger. 1983. Yield and fertility relationships in dairy cattle. J. Dairy Sci. 66:293. 9 Harmon, R. J. 1994. Physiology of mastitis and factors affecting somatic cell counts. J. Dairy Sci. 77:2103. 10 Hayes, J. F., R. I. Cue, and H. G. Monardes. 1992. Estimates of repeatability of reproductive measures of Canadian Holsteins. J. Dairy Sci. 75:1701. 11 Hermas, S. A., C. M. Young, and J. W. Rust. 1987. Genetic relationships and additive genetic variation of productive and reproductive traits in Guernsey dairy cattle. J. Dairy Sci. 70: 1252. 12 Heuven, H.C.M. 1987. Diagnostic and genetic analysis of mastitis field data. Ph.D. Diss., Univ. Wisconsin, Madison. 13 Jansen, J. 1985. Genetic aspects of fertility in dairy cattle based on analysis of A. I. data—a review with emphasis on areas for further research. Livest. Prod. Sci. 12:1. 14 Jansen, J. 1986. Direct and maternal genetic parameters of fertility traits in Friesian cattle. Livest. Prod. Sci. 15:153. 15 Jansen, J., J. Van der Werf, and W. de Boer. 1987. Genetic relationships between fertility traits for dairy cows in different parities. Livest. Prod Sci.17:337. 16 Janson, L. 1980. Studies on fertility traits in Swedish dairy cattle. II. Genetic parameters. Acta Agric. Scand. 30:427. 17 Miglior, F., E. B. Burnside, and J.C.M. Dekkers. 1996. Nonadditive genetic effects and inbreeding depression for somatic cell counts of Holstein cattle. J. Dairy Sci. 78:1168. 18 Misztal, I., T. J. Lawlor, T. H. Short, and G. R. Wiggans. 1991. Continuous genetic evaluation of Holstein for type. J. Dairy Sci. 74:2001.

593

19 Raheja, K. L., E. B. Burnside, and L. R. Schaeffer. 1989. Relationships between fertility and production in Holstein dairy cattle in different lactations. J. Dairy Sci. 72:2670. 20 Rogers, G. W., G. L. Hargrove, and J. B. Cooper. 1995. Correlations among somatic cell scores of milk within and across lactations and linear type traits of Jerseys. J. Dairy Sci. 78:914. 21 Ron, M., R. Bar-Anan, and G. R. Wiggans. 1984. Factors affecting conception rate of Israeli Holstein cattle. J. Dairy Sci. 67: 854. 22 Schutz, M. M. 1994. Genetic evaluation of somatic cell scores for United States dairy cattle. J. Dairy Sci. 77:2113. 23 Schutz, M. M., L. B. Hansen, G. R. Steuernagel, J. K. Reneau, and A. L. Kuck. 1990. Genetic parameters for somatic cells, protein, and fat in milk of Holsteins. J. Dairy Sci. 73:494. 24 Schutz, M. M., P. M. VanRaden, and G. R. Wiggans. 1994. Genetic variation in lactation means of somatic cell scores for six breeds of dairy cattle. J. Dairy. Sci. 77:284. 25 Seykora, A. J., and B. T. McDaniel. 1986. Genetic statistics and relationships of teat and udder traits, somatic cell counts, and milk production. J. Dairy Sci. 69:2395. 26 Shook, G. E. 1989. Selection for disease resistance. J. Dairy Sci. 72:1349. 27 Shook, G. E., and M. M. Schutz. 1994. Selection on somatic cell score to improve resistance to mastitis in the United States. J. Dairy Sci. 77:648. 28 Strandberg, E., and G. E. Shook. 1989. Genetic and economic responses to breeding programs that consider mastitis. J. Dairy Sci. 72:2136. 29 VanRaden, P. M., A. E. Freeman, and P. J. Berger. 1987. Relationships among production and reproduction in nulliparous and first lactation Holstein cattle. J. Dairy Sci. 70(Suppl. 1):231.(Abstr.) 30 VanRaden, P. M., and Y. C. Jung. 1988. A general purpose approximation to restricted maximum likelihood: the tilde-hat approach. J. Dairy Sci. 71:187. 31 Weller, J. I. 1989. Genetic analysis of fertility traits in the Israeli dairy cattle population. J. Dairy Sci. 72:2644. 32 Weller, J. I. 1994. Economic Aspects of Animal Breeding. Chapman & Hall, London, England. 33 Weller, J. I., and E. Ezra. 1991. Derivation of the March 1991 breeding index. Meshek Habakar Vehahalav 233:9. 34 Weller, J. I., C. Israel, and E. Ezra. 1993. The individual animal model. Heker Umas 15:5. 35 Weller, J. I., C. Israel, and E. Ezra. 1994. A simple procedure for obtaining approximate interim cow solutions from an animal model. J. Dairy Sci. 77:1126. 36 Weller, J. I., and M. Ron. 1992. Genetic analysis of fertility traits in Israeli Holsteins by linear and threshold models. J. Dairy Sci. 75:2541. 37 Weller, J. I., A. Saran, and Y. Zeliger. 1992. Genetic and environmental relationships among somatic cell count bacterial infection and clinical mastitis. J. Dairy Sci. 75:2532. 38 Welper, R. D., and A. E. Freeman. 1992. Genetic parameters for yield traits of Holsteins, including lactose and somatic cell score. J. Dairy Sci. 75:1342. 39 Wiggans, G. R., and G. E. Shook. 1987. A lactation measure of somatic cell count. J. Dairy Sci. 70:2666. 40 Zhang, W. C., J.C.M. Dekkers, G. Banos, and E. B. Burnside. 1994. Adjustment factors and genetic evaluations for somatic cell score and relationships with other traits of Canadian Holsteins. J. Dairy Sci. 77:659.

Journal of Dairy Science Vol. 80, No. 3, 1997