Zhenyong ZHOU,Hongbo LI,Xiangmin YAN,Jinshan ZHANG,Wei DU,Yang ZHANG

Institute of Raise Livestock,Xinjiang Academy of Animal Science,Urumqi 830000,China

Responsible editor:Yong XU Responsible proofreader:Xiaoyan WU

Xinjiang brown cattle is a new dual-purpose cattle bred by China in recent decades.It not only has strong adaptability and disease resistance as well as good appearance,but also has certain milk performance,meat perofrmance and draft performance;it mainly grows in Ili regions which is in the north of Xinjiang Tianshan Mountain and pasturing areas as well as farming-pastoral regions of Junggar Boundary Mountain; at present,the quantity of this cattle breed and its crossbred cattle is more than 100×104.With the gradual promotion of the breed of Xinjiang brown cattle,the variety breeding and popularization of Xinjiang brown cattle have been promoted by governments at all levels and scientific research institutions,and a series of relevant experiment researches have been carried out,which have made outstanding contributions to the production performance evaluation of the existing Xinjiang brown cattle population.However,with the reform of economic system and the development of market economy,the breeding work of Xinjiang brown cattle has been mainly in a few core breeding fields,and the breeding has been focused on open nucleus breeding,according to the requirements of the market to beef products and dairy products,there has been an urgent need to breed new beef and dairy strains and formulate breed planning objectives in conformity with it; what’s more,to formulate a breeding scheme must view accurate and reliable estimation of genetic parameters and breeding value as the prerequisite.Therefore,the estimation of genetic parameters and breeding value becomes the most basic task for livestock breeding work.In recent years,Du Wei[1],Wang Wenqi[2],Li Na[3],Zhang Yang[4],Zhou Zhenyong[5]et al.carried out a lot of basic research work,while the study on genetic analysis of main growth and economic characters and the estimation of genetic parameters was almost neglected.In the study,the genetic parameters of main growth and economic characters of Xinjiang brown cattle were estimated by using animal models and multiple traits derivative free restricted maximum likelihood (MTDFREML),which provided a theoretical basis for the breeding and improvement of Xinjiang brown cattle.

Materials and Methods

Materials

The archival data about the breeding of Xinjiang brown cattle in the cattle farms of autonomous region?levels (Urumchi cattle farm,Tacheng cattle farm and Ili stud farm)were collected and sorted out,and the determination data about production performance of three generations of cattle were obtained based on pedigree data(1987-2003),the production performance mainly include birth weight,body weight and measurement indexes of different stages of bull and cow,in which the body measurement indexes mainly consist of withers height(WH),body slanting length (BL),chest girth(CG),shank circumference(SC).

Data reduction

Excel was used to sort out the pedigree data and data information,and SPSS 13.0 software package was used to analyze fixed effect,moreover,the significance testing and Duncan’s multiple comparison on the differences of the fixed effect among different levels were carried out,the inapparent differences were classified as the same level,and the significant differences were divided into different levels.

Methods of statistical analysis[6]

Animal model was adopted to estimate the genetic parameters of each character.In the model,individual animal effect and random residual effect were stochastic effect,and there were different fixed effects for different characters,such as field,year,season,ages of dam,and gender and so on.

The formula of the animal model was as follows:

Thereinto,y represented the vector of observed value,X represented the structure matrix of fixed effect,β represented the vector of fixed effect,Z represented the structure matrix of stochastic effect,μ represented the vector of stochastic effect,e represented the vector of random residuals.

The assumed prerequisites of the model were:

Variance analysisBy using the variance analysis procedure of statistical analysis software SPSS 13.0,the effects of such fixed effects as field,year,season and gender on growth traits of Xinjiang brown cattle were analyzed according to the data of different economic characters.The model suitable for birth weight,6-month-age weight,one full year weight and 2-year-old weight of cattle was as follows:

The model suitable for adult cattle weight was:

Thereinto,Y represented the observed value of each character,H represented the fixed effect of field,BY represented the fixed effect of birth year,BS represented the fixed effect of birth season,A represented the fixed effect of dam age,C represented the fixed effect of calving number,X represented the fixed effect of gender,S represented sire effect,e represented random residual effect.

Based on the results of variance analysis of each character,the significant environmental factors were taken as fixed effect and included into the genetic parameter estimation model.

The selection of stochastic effect in the modelThe stochastic effect was divided into direct additive genetic effect and maternal additive genetic effect by model(6).

Thereinto,y was observed value vector of individual character,βwas the vector of fixed effect,a was the vector of direct additive genetic effect,m was the vector of maternal additive genetic effect,Z1was the structure matrix of direct additive genetic effect,Z2was the structure matrix of maternal additive genetic effect,e was the vector of residual effect.

What’s more,the following standard was adopted for assuming and defining:

The basic theory and process were as follows:

Mixed model:

In which y was the vector of observed value,βwas the vector of fixed effect,μwas the vector of stochastic effect,e was the vector of random residual,X and Z were the structure matrixes realted toβandμ,respectively.

The assumption preconditions of the model were:

From the model,the likelihood function of y was got:

The common logarithmic function of L(y)was:

In which K was constant,p=v-1-v-1x (x′v-1x)x′v-1,c was the submatrix corresponding to the coefficient matrix of variance group of mixed model.Then lg|R|and lg|G|in the formula(12)should be evaluated based on the follows:

In which n was the dimension of observed value vector,q was the level of stochastic effect,A was the matrix of coefficient kinship among individuals.

lg|C| and y′py can be solved by Gaussian elimination method and the process was relatively complicated.According to formula (1),the following equation set of mixed model was obtained:

The coefficient matrix of formula (7) was extended as follows:

Using Gaussian elimination method,the formula (16)was carried out linear transformation,thereby upper triangular matrix was got:

By proving,it was drawn that:

By substituting(6),(7),(11) and(12)into the formula (4),the likelihood function expression which tookσ2eandσ2uas independent variables was got.Through substituting different values intoσ2eandσ2u,σ2eandσ2ucan be screened out when the lg(L)reached the maximum,which was REML estimation value of variance component.

Estimation methods of heritability,genetic correlation,environmental correlation and phenotypic correlation

The estimation results of variance and component of covariance obtained by calculating were substituted into the corresponding parameter calculation formulas,thus the estimated values of heritability and correlation coefficient of each character were got.

The formula of heritability:

The formula of genetic correlation:

The formula of environmental correlation:

The formula of phenotypic correlation:

Thereinto,h2was trait heritability,rAwas genetic correlation among different characters,rEwas environmental correlation among different characters,rPwas phenotypic correlation among different characters,σ2awas the estimated value of genetic variance component,σ2ewas the estimated value of environmental variance component,σ2pwas the estimated value of phenotypic variance component,Cov(a1,a2)was the estimated value of genetic covariance between two characters,Cov(E1,E2)was the estimated value of environmental covariance between two characters,Cov(P1,P2)was the estimated value of phenotypic covariance between two characters.

In this study,MT-DFREML software compiled based on DF-REML method by Boldman et al.was used for variance component estimation.According to Simplex method,the variance of Simplex was calculated repeatedly until it was less than 10-10.Then the heritability and genetic correlation were calculated according to the estimated results of variance component.

Results and Analyses

The basic statistics of main economic characters of Xinjiang brown cattle

As shown in Table1,the birth weight of Xinjiang brown cattle was 34.59 kg;for the 6-month-age cattle,the average weight,withers height,body slanting length,chest girth and shank circumference were 137.57 kg,105.52 cm,109.43 cm,116.68 cm and 15.01 cm,respectively; for the one full year cattle,the average weight and withers height were 266.54 kg and 112.03 cm,respectively; and for the 2-yearold cattle,the average weight,withers height,body slanting length,chest girth and shank circumference were 357.10 kg,119.90 cm,145.04 cm,162.46 cm and 17.31 cm,respectively.From the data,it can be seen that the differences on the withers height and chest girth between 5-year-old and 6-year-old Xinjiang brown cattle were not significant,indicating that the two indexes had been in a stable and equilibrium state and had reached the peak levels; however,there was still a certain increase on the indexes such as body slanting length and shank circumference,etc.,from the view of skeletal development,the growth and development of the indexes like withers height and chest girth of Xinjiang brown cattle was prior,but the tissues like body slanting length and shank circumference were still at the growth period when the cattle was 5-6 years old; from the overall development of skeleton,the preferentially complete development of Xinjiang brown cattle was height factor like body height andwidth and depth factor like chest girth,while the growth of the length factor represented by body slanting length and bone tissue growth factor like shank circumference was still at active phase and had great growth potential.

Table1 Statistics on main economic characters of Xinjiang brown cattle(mean±standard deviation)

Table2 The estimation of weight genetic parameters of Xinjiang brown cattle in different growth stages

Table3 The genetic parameters of body weight and body measurement of 6-month-age Xinjiang brown cattle

The estimation of genetic parameters of Xinjiang brown cattle’s main economic characters

As seen in Table2,the heritability of Xinjiang brown cattle’s weight character in each growth stage was above the middle level (0.34-0.45),in which the estimated value of the heritability of birth weight cattle was 0.45,this was slightly higher than 0.32 which was the estimated value of Chinese simmental cattle estimated by Wang Chunqian[7],Woldehawariat et al.estimated that the heritability value of birth weight cattle was 0.44,and that of Yanbian yellow cattle was 0.41 in 1977,which were consistent with the research results of this study.The heritability values of 6-month-age,oneyear-old and 2-year-old Xinjiang brown cattle were 0.37,0.38 and 0.34,respectively,which were basically consistent with the heritability results of 6-month-age grassland red cattle[8]and Yanbian yellow cattle[9],namely 0.36 and 0.32,respectively,but it was slightly higher than the estimated heritability value of weaning-weight Chinese simmental cattle,namely 0.11,thus it was medium heritability.Meanwhile,the above heritability values of one-year-old and 2-year-old Xinjiang brown cattle were close to that of 1.5-year-old Nanyang cattle[10]whose heritability was 0.395.In addition,the analysis results of genetic correlation and phenotypic correlation in the Table2 showed that there was extremely significant positive correlation on phenotypic correlation between one-yearold and 2-year-old cattle,and the correlation coefficient was 0.872 (P ＜0.01),moreover,the genetic correlation between the two was also significant,the correlation coefficient was 0.58,which was lower than the phenotypic correlation of the level.

From Table3,it can be seen that the heritability values of body weight,withers height,body slanting length,chest girth and shank circumference of 6-month-age Xinjiang brown cattle were 0.37,0.30,0.28,0.40 and 0.22,respectively; there was extremely significant positive correlation between body weight and each body measurement index,except the low correlation between shank circumference and withers height,shank circumference and body slanting length,the other body measurement indexes all had strong phenotypic correlation with body weight.

From the analysis results in the Table4,the heritability values of body weight,withers height,body slanting length,chest girth and shank circumference of one-year-old Xinjiang brown cattle were 0.38,0.34,0.41,0.28 and 0.31,respectively; body weight had strong positive correlation with the genetic correlation and phenotypic correlation of body measurement.In short,withers height and body slanting length were two important breeding indexes of one-year-old Xinjiang brown cattle.

Table4 The genetic parameters of body weight and body measurement of one-year-old Xinjiang brown cattle

Table5 The genetic parameters of body weight and body measurement of 2-year-old Xinjiang brown cattle

As shown in Table5,the heritability values of body weight,withers height,body slanting length,chest girth and shank circumference of oneyear-old Xinjiang brown cattle were 0.34,0.31,0.36,0.33 and 0.16,respectively; there were moderate genetic positive correlations among withers height,body slanting length and chest girth,and strong phenotypic correlations among body slanting length,chest girth and body weight.From the angles of heritability,phenotypic correlation and genetic correlation,body slanting length was an important breeding and evaluation index of 2-year-old Xinjiang brown cattle.

Discussions

The factors affecting the accuracy of genetic parameter estimation

Animal model was used to estimate the genetic parameters of each character,and it was necessary to include the main environmental influencing factors into the genetic parameter estimation model as fixed effect,however,the growth and development of animal is a continuous process,and there is heritable variation on continuous age or time; because the growth performance of calf is a concentrated expression of genetic potential of an individual itself and the maternal ability of its mother,if the maternal effect is neglected,it will not only can not reflect the maternal heredity and perpetual environmental effect,but also can not well describe the individual heredity and perpetual environmental effect,which will cause the overestimation of individual additive genetic variance-covariance,thereby causing the overestimation of direct heritability[6].There are different structures,feeding and management,and parameter estimation methods of drove along with the time,thus the results of genetic parameter estimation are of course different[8].In practical use,the environmental factors are different in different growth stages; some studies have shown that based on the variance analysis of the factors affecting each growth trait,the weight of Chinese simmental cattle in different growth stages was affected by different environmental factors,so the main factors of significant environmental factors can be included into the genetic parameter estimation model as fixed effect,so as to ensure that the obtained genetic parameters can truly reflect its hereditary basis.Therefore,to improve the accuracy of genetic parameter estimation,it is still necessary to combine with the external conditions of Xinjiang brown cattle themselves to explore the production influencing factors,moreover,the objective evaluation must be given.

The application of genetic parameter in breeding

Heritability,genetic correlation and phenotypic correlation are three common genetic parameter indexes in breeding selection,domestic breeding workers have already carried out genetic parameter estimation on many animals for variety breeding[6-19].In recent years,there have been many studies on the heritability of growth performance and economic character of cattle,the results have basically showed that the heritability of growth character was moderate,indicating that growth character had great additive genetic effects,and the rapid genetic progress could be got by system breeding.Currently,seed selection is based on the correlations among each character,because the degree of correlation plays an important role in the selection.Jin Haiguo et al.[8]suggested that the individual with great birth weight was very important in the breeding,because the birth weight had a close relationship with the development degree of each part of the body.In this study,the genetic relevancy of one-year-old and 2-year-old cattle was high,and the correlation coefficient was 0.58,thus one year old can be viewed as the breeding index,at the same time,the breeding indexes of withers height and body slanting length etc.which were closely related to body weight can be determined according the genetic correlation of the character.

Conclusions

(1)From the results of the study,body weight and body measurement of Xinjiang brown cattle in each growth stage were moderate heritability,which was basically in line with the methods and results of the predecessors.Family selection and population selection can be used to speed up the pace of breeding.

(2)According to the estimation of genetic parameters,such as heritability and genetic correlation,withers height and body slanting length were preliminarily determined as the important breeding indexes of Xinjiang brown cattle’s body type traits,moreover,the breeding of Xinjiang brown cattle can be carried out in one year old.

(3)For Xinjiang brown cattle,the indexes like body weight and body measurement had larger standard deviation and greater heritable variation and diversity,which laid the foundation for speeding up breeding progress.

[1]WANG WQ (王文奇),LIU YF (劉艷豐),ZHAO YJ(趙蕓君),et al.Effect of natural grazing and supplementary feeding on production performance of Xinjiang brown cattle (放牧加補飼對新疆褐牛生產(chǎn)性能的影響) [J].Xinjiang Agricultural Sciences (新疆農(nóng)業(yè)科學),2010,47(1):200-203.

[2]DU W(杜瑋),ZHANG Y(張楊),LIN HK(藺宏凱),et al.Study on body size of Xinjiang brown cattle under the herd condition(新疆褐牛放牧條件下體尺指標的研究)[J].Xinjiang Agricultural Sciences(新 疆 農(nóng) 業(yè) 科 學),2010,47(2):401-405.

[3]LI N(李娜),ZHANG Y(張楊),ZHOU ZY(周振勇),et al.Compare of high-grade beef of Xinjiang brown calf by sensory evaluation(新疆褐牛高檔牛肉感官評定比較)[J].Meat Industry (肉類工業(yè)),2012,370(2):4-7.

[4]ZHANG Y (張楊),LI HB (李紅波),ZHANG JS(張金山),et al.Investigation on population resources of Xinjiang brown cattle (新疆褐牛種群資源調查研究)[J].China Cattle Science (中國牛業(yè)科學),2012,38(1):24-28.

[5]ZHOU ZY (周振勇),ZHANG Y (張楊),LIN HK (藺宏凱),et al.Weight model based on regression of principle component analysis of Xinjiang brown cattle(基于主成分逐步回歸法的新疆褐牛體重預測模型研究)[J].China Cattle Science(中國牛業(yè)科學),2012,38(1):1-4.

[6]WANG CQ(汪春乾).Genetic evaluation and studies on growth traits of Chinese simmental(中國西門塔爾牛生長性狀的遺傳評估和研究)[D].Hefei:Anhui Agricultural University(合肥:安徽農(nóng)業(yè)大學),2002.

[7]WANG CQ (汪春乾),ZU SZ (許尚忠),CHEN HQ (陳宏權),et al.Estimating parameters growth traits of Chinese simmental(中國西門塔爾牛生長性狀遺傳參數(shù)估計)[J].Journal of Anhui Agricultural University (安徽農(nóng)業(yè)大學學報),2004,31(4):417-420.

[8]JIN HG(金海國),YU MZ(俞美子),et al.Estimation of genetic parameters for economic traits in red steppe(草原紅牛主要經(jīng)濟性狀的遺傳參數(shù)估計)[J].Journal of Agricultural University of Hebei(河北農(nóng)業(yè)大學學報),2002,25(3):74-77.

[9]JIN CZ(金昌洙).Study on the bull selection technology of Yanbian yellow cattle in the early stage using genetic parameter(應用遺傳參數(shù)早期選擇延邊黃牛種公牛技術的研究)[J].Journal of Agricultural Science Yanbian University (延邊農(nóng)學院學報),1989(2):125-128.

[10]YANG CD (楊春德),WANG GL (王冠立).A preliminarily estimation on genetic parameters of main characters of Nanyang cattle (南陽牛主要性狀遺傳參數(shù)的初步估測) [J].Chinese Journal of Animal Science (中國畜牧雜志),1987(1):46-47.

[11]ZHOU GL (周國利),JIN HG (金海國),WU YH (吳玉厚).Genetic analysis on main economic traits in Luxi yellow cattle(魯西黃牛主要經(jīng)濟性狀的數(shù)量遺傳學分析)[J].Chinese Journal of Animal Science (中國畜牧雜志),2006,42(5):11-13.

[12]LIU SQ (劉召乾).Research on the establishment of Luxi cattle comprehensive selection index(魯西黃牛綜合選擇指數(shù)制定研究)[D].Beijing:Chinese Academy of Agricultural Sciences(北京:中國農(nóng)業(yè)科學院),2008.

[13]LIU JY (劉建勇).Study on breeding scheme (strain)of new meat breed for BMY cattle(BMY 肉牛新品種(系)優(yōu)化育種規(guī)劃的研究)[D].Yangling:Northwest A&F University (楊陵:西北農(nóng)林科技大學),2007.

[14]XU TS(徐鐵山).Genetic analysis of the main economical traits in Pekin ducks(北京鴨主要經(jīng)濟性狀的遺傳分析)[D].Yangling:Northwest A & F University(楊陵:西北農(nóng)林科技大學),2004.

[15]WANG CY(汪聰勇).Design on breeding management system and analysis of breeding effect and study on the microsatellite marker for Xianan cattle(夏南牛育種管理系統(tǒng)開發(fā)及育種效果分析和微衛(wèi)星標記分析) [D].Zhengzhou:Henan Agricultural University(鄭州:河南農(nóng)業(yè)大學),2008.

[16]LIU JY (劉建勇),ZAN LS (昝林森),HUANG BZ (黃必志).Estimated BMY beef cattle breeding traits＇ inheritance parameter (BMY 肉牛育種性狀的遺傳參數(shù)估計)[J].Acta Agriculturae Boreali-occidentalis Sinica(西北農(nóng)業(yè)學報),2007,16(6):47-50.

[17]YANG M (楊敏).Genetic parameter estimation and genetic evaluation of Pinan cows(皮南牛遺傳參數(shù)估計與遺傳評定)[D].Hefei:Anhui Agricultural University(合肥:安徽農(nóng)業(yè)大學),2009.

[18]FAN DY (范大有),XU SZ (許尚忠),LI JY(李俊雅),et al.Genetic and statistical analysis between production traits and secondary traits in Chinese simmental(中國西門塔爾牛次級性狀與生產(chǎn)性狀的遺傳統(tǒng)計分析)[J].Acta Veterinaria et Zootechnica Sinica (畜牧獸醫(yī)學報),2008,39(8):1025-1032.

[19]CAO HZ (曹洪戰(zhàn)),WU CX (吳常信),SHI SK (師守堃).Estimation of the heritability of back fat alive on doublemuscle Large Yorkshire swine by MTDFREML (利 用MTDFREML 估 計大約克雙肌臀種豬活體背膘厚遺傳力)[J].Animal Husbandry & Veterinary Medicine(畜牧與獸醫(yī)),2003,35(1):6-7.