Comparison of genomic and classic selection method in breeding of Iranian native chickens

The indigenous stocks play an important role in villager economy in developing countries, especially in Iran. Indigenous breeds could maintain higher level of performance under Incompatible environment compared to commercial strains. In classic breeding programs for poultry, selection of birds was based on phenotypic and pedigree information. The development of genomic selection prefers new strategies in animal breeding. Genomic selection has a major limitation for implementation. This method need genotyping of large numbers of markers for implementation and the costs of genotyping for these markers is high, especially for developing countries. To overcome this problem, before applying genomic selection, different scenarios have to be compared using computer simulation. The objective of this study was comparison of results in genomic selection and classic selection method in Iranian native fowls by using ZPLAN software (Täubert 2010). ZPLAN software is a very useful tool to optimize conventional and genomic breeding programs this software allows modeling of all relevant breeding structures, while taking all relevant biological, technological and economic parameters for complex breeding programs into account (Sitzenstock et al. 2013). This software then generates results such as the annual genetic gain for the breeding objective using a pure deterministic approach (Dekkers 2007). This program has been based on the discounted gene flow-method (Hazel 1943) and selection index theory (Hazel and Lush 1942). This software calculates genetic gain, discounted return, profit and costs within and over selection paths. The results of monetary in ZPLAN are standardized to an animal unit and are given per year (Sitzenstock et al. 2013).
Material and A reference scenario and a genomic scenario were simulated on 3460 birds in both sexes. Reference scenario was simulated based on phenotypic information and genomic scenario was simulated based on combination of phenotypic and genomic data with two reference populations of 500 and 1000 birds. Genomic data was based on the information of 4000 hens and variable numbers of cocks (800 to 4000). Studied traits included egg number, body weight at 8 weeks, mean of egg weight and age at sexual maturity. In this simulation, genetic and phenotypic parameters were used (Table 1). These parameters included: phenotypic standard deviation, heritability, genetic and phenotypic correlations and traits economic coefficients. Economic weights were based on the study of Kianimanesh et al. (2001) and the other parameters were extracted from the study of Yousefi 2012. The variable costs of rearing a cock or a hen were assumed to be 440000 Rial per animal (over a period of 20 weeks). During production, daily feeding costs resulted from a feed consumption of 0.11 kg per day at a price of 12000 Rial per kg. For each cock and hen, additional costs for the animal care (20000 Rial) were assumed. The costs of selection were assumed 40000 Rial per animal. The fixed costs were not considered in conventional scenario, because it was difficult to quantify them in breeding program. But these costs must reduce from profit. The interest rate was set to %7 for discounted costs and %2 for discounted returns (Sitzenstock et al. 2013).
The generation interval for both scenarios was 14.5 months. By adding genomic information to reference scenario, accuracy of estimations increased from 0.62 for chocks and 0.64 for hens in reference scenario to 0.77 for both sex in genomic scenario. Genetic development of traits increased with increase in the number of genotyped birds and the number of birds of reference population. Genetic development for traits egg number, body weight at 8 weeks, mean of egg weight and age at sexual maturity increased from 0.22, 2.57, 0.09 and -0.23 (Table 2) respectively in reference scenario to 0.30, 3.60, 0.11 and -0.27 respectively in genomic selection with 1000 reference population. The profit of breeding program was higher than reference scenario only in genomic scenario with reference population with 1000 birds (Figure 2). The discounted return per animal unit in the conventional scenario was 20102400 Rial per year. The variable costs for conventional scenario were 1057200 Rial per animal unit. This resulted in a profit of 19044800 Rial per animal unit. The accuracy of the genomic information and the size of the reference population have a large impact on the benefits from genomic selection. Dekkers 2007 showed that with increasing of marker information increased the benefits from genomic scenarios.
The use of genomic information caused increasing genetic development in all traits. Although in this study costs were not reduced, but genomic information increased the accuracy of selection of breeding program. This study showed that genomic selection can increase the genetic improvement rate of native chickens. However, the costs of genomic scenarios were higher than conventional scenario, but genomic information increased accuracy of selection and genetic gain of breeding goals traits. To overcome this problem, use of low density SNP chips is suggested.