| 隨機小語 |
灰心是動搖的開端，動搖是失敗的近鄰。
[易普生] |
|
第二十九卷(2000)
台灣藍瑞斯、約克夏與杜洛克登錄豬活仔數之育種價預測
李世昌(1) 張秀鑾(1) 吳明哲(1) 黃鈺嘉(1) 賴永裕(1) 王旭昌(2) 宋永義(3)
(1)行政院農業委員會畜產試驗所 (2)財團法人中央畜產會 (3)國立台灣大學畜產系
應用種豬登錄資料庫之種母豬分娩資料(自1989年4月至2000年1月間分娩者)，計包括33,931頭登錄母豬之112,000胎分娩記錄；其中藍瑞斯、約克夏與杜洛克頭數分別為17,127、6,287與10,517頭，相對應之分娩胎數為56,687、20,947與34,366胎。繁殖性狀之遺傳評估模式係針對登錄豬之活仔數產仔能力為觀測值，而以品種、出生年別與季節、分娩該胎之年別與季節、以及分娩該胎之產次為固定效應，母豬分娩該胎之日齡與日齡平方為共變數，母豬之累加性遺傳效應與殘差等為逢機效應；其中季節分為熱(每年5至10月)與涼(每年11至隔年4月)季，產次則分為初產與經產。應用前述動物混合模式所得之累加性遺傳效應與殘差之限制極大似然估值(REML)換算之遺傳變異率估值為0.237，隨後應用於前述遺傳評估模式進行固定效應之最佳線性無偏估計(BLUE)檢定與登錄種豬產活仔數之遺傳值最佳線性無偏預測(BLUP)。分娩時，經產母豬顯著地較初產者多0.79±0.02頭活仔數(P<0.001)；出生於涼季之母豬於分娩時顯著地較熱季出生者多產0.05±0.02頭活仔豬(P<0.01)；分娩季節效應亦有相同的趨勢(0.10±0.01頭，P<0.001)。此外，登錄種豬產活仔數之遺傳值最佳線性無偏預測(BLUP)與五十大排行榜，以及前述三品種之二十大排行榜，均已於網際網路查詢系統(http://www.angrin.tlri.gov.tw/)中，供業者選種用。
關鍵語：育種價、最佳線性無偏預測、登錄豬。
PREDICTION OF BREEDING VALUES FOR REGISTERED STOCKS IN
TAIWAN LANDRACE, YORKSHIRE AND DUROC PIGS
S. C. Lee(1), H. L. Chang(1), M. C. Wu(1), Y. C. Huang(1), Y. Y. Lai(1), H. C. Wang(2) and Y. Y. Sung(3)
(1)Taiwan Livestock Research Institute, Council of Agriculture
(2)National Animal Industry Foundation (3)National Taiwan University
There were 112,000 litters produced during April 1989 and January 2000 from 33,931 registered sows in Taiwan Swine Herdbook used in the study, including 56,687, 20,947 and 34,366 litters from 17,127 Landrace, 6,287 Yorkshire and 10,517 Duroc sows, respectively. Litter size born alive was the reproductive trait considered. Genetic evaluation model included breed, season and year of sow at birth, season and year as well as parity of sow at farrowing as fixed effects, age in days at farrowing and its quadratic as covariates, additive genetic effects and error as random effects. There were two classes in sows' birth and farrowing seasons, hot and cool seasons. The former was defined as from April to October and the latter was for the rest months of the year. Parity of sows at farrowing was also classified as two levels, first parity and parous sows. Estimates of heritability from additive genetic variances via REML for the trait evaluated was 0.237, which was then implemented in the animal mixed linear model to estimate and test the corresponding fixed effects (BLUE) and to predict the random additive genetic values of animals (BLUP). Parous sows produced 0.79±0.02 piglets more than those of first parity sows (P<0.001). Also, sows born and littering at cool season had 0.05±0.02 (P<0.01) and 0.10±0.01 (P<0.001) piglets more than those of born and littering at hot season ones, respectively. Furthermore, BLUP of additive genetic values and top 50 across breeds ranking for registered stocks as well as top 20 within breed ranking for the trait evaluated were listed in the search system with URL (http://www.angrin.tlri.gov.tw/) available for selection purpose.
Key Words: Breeding value, BLUP, Registered pigs.
  
|
|
|
|
|
