Meta-Analysis of studies on genetic parameters of economic traits in Iranian Holstein dairy cows

Meta-analysis method was used for gathering studies done on genetic parameters estimation in Holstein cows to increase the accuracy of estimations. For this purpose, data from 96 articles on genetic parameter estimation of productive and reproductive traits on Iranian Holstein dairy cows were used. Productive traits were milk production, amount of milk fat, amount of milk protein, milk fat percentage and milk protein percentage and reproductive traits were age at first calving, calving interval, dry days, open days, pregnancy period and lactation period.

First, heritability estimates, genetic and phenotypic correlations were obtained from articles. After data preparation, the meta-analytical model with random effects using Metacor package, R 3.3.1 software and CMA ver.3 software were used for estimating weighted average of heritability and genetic and phenotypic correlations, standard errors and 95% confidence interval for productive and reproductive traits. Investigating available studies showed high heteroginty among studies. Therfore, using fixed model meta-analysis was not possible for estimating weighted average of effects. Data were reanalysid using CMA software and results of random model meta-analysis were reported as final results.

The weighted average of heritability for productive traits in dairy cows were between 0.19 and 0.27. Milk protien percentage had the highest (0.27) and milk fat had the lowest (0.19) heritability estimates amonge productive traits. The weighted average of heritability for reproductive traits was in the range of 0.03 to 0.14. Pregnancy period (0.14) had the highest and open days had the lowest (0.03) heritability estimates amonge reproductive traits. The weighted average of genetic correlations of productive and reproductive traits was in the range of -0.56 to 0.88 and the mean weight of the phenotypic correlations of traits was in the range of -0.42 to 0.83. Comparing the results of meta-analysis in present study with the results of individual studies showed that aggregating studies and analyzing by meta-analysis improves the accuracy of the results through reducing standard errors. For example, in studied articles the range of heritability estimates for milk production were between 0.047-0.41, for milk fat were between 0.05-0.56 and for milk protein were between 0-0.7. However, after using meta-analysis method, these ranges were reduced to 0.23-0.25 for milk production, 0.17-0.21 for milk fat and 0.19-0.29 for milk protein. In addition, because of aggregating studies and increasing the amount of data, standard error of estimates in meta-analysis were considerably reduced compared to standard error of estimates in individual studies, especially for productive traits.

Using methods like Meta-analysis will increase the performance of breeding programs and will improve genetic progress of economic traits of animals by gathering all available information, especially for populations with no or not accurate data.