برآورد پارامترهاي ژنتيكي اثرات مستقيم و مادري صفات رشد در گوسفند سنگسري با استفاده از روش نمونه¬گيري گيبس

Estimation of Genetic Parameters for Direct and Maternal Effects in Growth Traits of Sangsari Sheep Using Gibbs Sampling

مطالعه حاضر به منظور برآورد اجزای (كو)واریانس و پارامتر¬های ژنتیكی وزن بدن در سنین مختلف گوسفند نژاد سنگسری انجام گرفت. در این مطالعه از داده¬های گله¬های تحت پوشش ایستگاه پرورش و اصلاح نژاد سنگسری واقع در شهرستان دامغان كه طی سال¬های 1366 تا 1387 جمع آوری شده، استفاده گردید. اطلاعات مورد استفاده شامل 9707 ركورد وزن تولد، 8524 ركورد وزن از شیرگیری و 3894 ركورد وزن شش ماهگی بودند. آنالیز عوامل محیطی مؤثر بر این صفات توسط رویه GLM با استفاده از نرم افزار آماری SAS انجام شد. عوامل محیطی سال تولد، جنس بره، تیپ تولد و سن مادر هنگام زایش بر تمام صفات معنی¬دار بودند و به عنوان اثرات ثابت و سن دام هنگام وزن¬كشی به عنوان متغیر كمكی وارد مدل شدند. برآورد پارامتر¬های ژنتیكی با شش مدل مختلف حیوانی و با روش آماری بیزی، مبتنی بر نمونه¬گیری گیبس با استفاده از نرم افزار MTGSAM انجام شد. تعداد دوره¬های نمونه¬گیری گیبس 200000 دوره انتخاب شد و در هر آنالیز 20000 دوره اول به عنوان دوره¬های سوخته و همچنین فواصل نمونه¬گیری 100 انتخاب گردید. مدل¬های مختلف برازش شده و معنی¬داری مدل¬ها با شاخص معیار اطلاعات آكایكی (AIC) مورد آزمون قرار گرفت. وراثت¬پذیری مستقیم با مناسب ترین مدل برازش شده برای وزن تولد (مدل 2)، از شیرگیری (مدل 5) و شش¬ماهگی (مدل 2) به ترتیب برابر 35/0، 18/0 و 21/0 محاسبه شد. مقدار وراثت¬پذیری مادری برای وزن از شیرگیری 07/0 برآورد گردید. با توجه به بالا بودن ضریب وراثت¬پذیری وزن تولد نسبت به سایر صفات مورد مطالعه، بهینه سازی صفت به واسطه انتخاب می¬تواند در این صفت بازدهی بالاتری داشته باشد.

Introduction Small ruminants, especially native breed types, play an important role in livelihoods of a considerable part of human population in the tropics from socio-economic aspects. Therefore, integrated attempt in terms of management and genetic improvement to enhance production is of crucial importance. Knowledge of genetic variation and co-variation among traits is required for both the design of effective sheep breeding programs and the accurate prediction of genetic progress from these programs. Body weight and growth traits are one of the economically important traits in sheep production, especially in Iran where lamb sale is the main source of income for sheep breeders while other products are in secondary importance. Although mutton is the most important source of protein in Iran, meat production from the sheep does not cover the increasing consumer demand. On the other hand, increase in sheep number to increase meat production has been limited by low quality and quantity of forage range. Therefore, enhancing meat production should be achieved by selecting the animals that have maximum genetic merit as next generation parents. To design an efficient improvement program and genetic evaluation system for maximization response to selection for economically important traits, accurate estimates of the genetic parameters and the genetic relationships between the traits are necessary. Studies of various sheep breeds have shown that both direct and maternal genetic influences are of importance for lamb growth. When growth traits are included in the breeding goal, both direct and maternal genetic effects should be taken into account in order to achieve optimum genetic progress. The objective of this study was to estimate the variance components and heritability, for growth traits, by fitting six animal models in the Sangsari sheep using Gibbs sampling. Material and Method Sangsari is a fat-tailed and relatively small sized breed of sheep, native and well adapted to Semnan province. The data set and pedigree information used in this study, recorded during 1986–2008, were obtained from the breeding station of Sangsari sheep (in Damghan, Semnan Province, Iran). The data included 9707 records for birth weight (BW), 8524 records for weaning weight (WW) and 3894 records for six months weight (6MW). Records were prepared for analysis using EXCEL 97 software. During the preparation process, abnormal data were removed. The pedigree and data files were prepared using pedigree software. Firstly, the GLM procedure (SAS, 2002) was used for determining the fixed effects that had significant effect on the traits investigated (P <0.05). Interaction effects did not have significant effects on the traits and were excluded from the final models of analysis. Variance components and genetic parameters were estimated for each trait with Bayesian method based on Gibbs sampling technique using MTGSAM software by fitting six univariate animal models that exclude or include additive maternal or permanent environmental effects. Gibbs sampling is a numerical integration method and is one of several Markov chain Monte Carlo (MCMC) methods. They involve drawing samples from specified distributions; hence they are called Monte Carlo and are referred to as Markov chain because each sample depends on the previous sample. Specifically, Gibbs sampling involves generating random drawings from marginal posterior distributions through iterative sampling from the conditional posterior distributions. For each trait the most suitable model amongst all six models was determined based on Akaike 's Information Criterion (AIC). Results and Discussion The analysis of variance showed that fixed effects of birth type (single, twin), sex of kid, (male, female) age of dam (from 2 to 7 years old) and year of birth (1986–2008) were significant for weights at birth, weaning, and 6 month (P < 0.01). Direct heritability estimates for birth weight, weaning weight and weight at six months of age (based on the best model) were 0.35 (model 2), 0.18 (model 5) and 0.21 (model 2), respectively. Estimation of maternal heritability for weaning weight was 0.07. Conclusion Gibbs sampling in Bayesian statistical analysis could provide reasonable range expected estimations for parameters. The results showed that growth traits are influenced by maternal effects in the early stages of age. With increasing age of lamb, the significance of this effect is reduced, due to decrease of dependence on the dam. Whereas, the estimate of heritability of birth weight was higher than other traits, optimization of trait by selection can be more efficiency in this trait.