Tóm tắt Luận án Selection for improving performance of duroc, landrace and yorkshire purebreds raised at dabaco company

Nội dung text: Tóm tắt Luận án Selection for improving performance of duroc, landrace and yorkshire purebreds raised at dabaco company

1. MINISTRY OF EDUCATION MINISTRY OF AGRICULTURE AND TRAINING AND RURAL DEVELOPMENT VIỆN CINSTITUTE OFANIMAL SCIENCES N NUÔI LUU VAN TRANG SELECTION FOR IMPROVING PERFORMANCE OF DUROC, LANDRACE AND YORKSHIRE PUREBREDS RAISED AT DABACO COMPANY ABSTRACT OF PHD THESIS PROMOTER PROF. Dr. DANG VU BINH ASOC. PROF. Dr. LE THI THANH HUYEN HANOI - 2021
2. The thesis is fulfilat: Institute of Animal Sciences PROMOTER Prof. Dr. PhD. DANG VU BINH Asscoc Prof. PhD. LE THI THANH HUYEN Reviewer 1: Assoc Prof. Dr. TRAN HUE VIEN Reviewer 2: Assoc Prof. Dr. PHAM KIM DANG Reviewer 3: Dr. DOAN VAN SOAN The thesis will be defended at a meeting of the Institute-level Thesis Judging Council meeting at: Institute of Animal Science – Thuy Phuong – North Tu Liem – Hanoi on . o’clock;Date Month . Year of Please look for the thesis at: - National Library - Library of Institute of Animal Science
3. LIST OF PUBLISHED WORKS 1. Luu Van Trang, Tran Xuan Manh, Pham Van Hoc, Luu Quang Du, Nguyen Van Khoa and Dang Vu Binh (2021). SELECTION FOR IMPROVING GROWTH PERFORMANCE OF DUROC, LANDRACE AND YORKSHIRE PURBREDS AT DABACO NUCLEAR PIG BREEDING COMPANY. Journal of Animal Science and Technology – Institute of Animal Sciences, No. 123, May 2021, pages 41-52. 2. Luu Van Trang, Tran Xuan Manh, Pham Van Hoc, Luu Quang Du, Nguyen Van Khoa and Dang Vu Binh (2021). SELECTION TO IMPROVE NUMBER OF PIGLET BORN ALIVE OF PURE LANDRACE AND YORKSHIRE SOWS AT DABACO NUCLEAR PIG BREEDING COMPANY. Journal of Animal Science and Technology – Institute of Animal Sciences, No. 123, , May 2021, pages 53-64. 3. Luu Van Trang, Tran Xuan Manh, Pham Van Hoc, Luu Quang Du, Nguyen Van Khoa and Dang Vu Binh (2019). PRODUCTION AND REPRODUCTION PERFORMANCES AND SOME GENETIC PARAMETERS OF MAIN TRAITS OF DUROC, LANDRACE AND YORKSHIRE BREEDS RAISED AT DABACO NUCLEUS BREEDING PIG COMPANY Journal of Animal Science and Technology – Institute of Animal Sciences, No. 100, June 2019, pages 30-43 2
4. INTRODUCTION 1.1. URGENCY OF THESIS SUBJECT Duroc (D), Landrace (L) and Yorkshire (Y) are 3 main pig breeds used in intensive farming systems in most countries around the world as well as in Vietnam. The three pig breeds also participate in all most of the crossbred formula in intensive, semi-intensive systems with different scales in our country. Y and L pigs not only create crossbred sows of these systems, but they are terminal boars to creat crossbreds in livestock smallholder farmers. Meanwhile, D and Piétrain pigs are used to create commercial hybrids of 3 or 4 different breeds. Improving the growth rate of pure Dpigs, as well as reproductive performance of pure L and Y pigs are an important task in the pig production for exotic breeds in our country. Over the years, some breeding centers or breeding companies in our country have built a 3-level breeding pig system, in which the GGP herds of the three pure breeds imported from abroad. Every year, most of these centers or companies have to import a certain number of pure boar and sow pigs from different countries in order to supply genetic resources and improve the quality of the nuclear herds. In recent years, a number of studies on pure exotic pig herds raised in our country have been conducted in order to assess the status of growth and reproductive performance, estimate some important genetic parameters, as well as carry out some selective methods for these breeding pigs (Nguyen Huu Tinh et al., 2013; Nguyen Van Duc, 2015; Trinh Hong Son, 2015; Le Van Sang et al., 2018; Tran Thi Minh Hoang). and cs., 2019a, 2019b, 2019c). Dabaco Nuclear Pig Breeding Company (Dabaco Company) was established in 2010 with a scale of 3,400 sows. The D, L and Y pigs were bred in GGP, GP and PS herds. Researched by Doan Phuong Thuy cs. (2015 and 2016) have determined some genetic parameters of some growth and reproductive traits for 3 purebred breeds raised at Dabaco Company and built a selection orientation for these pure breeds. On the basis of some data from Dabaco Company and other breeding farms, Tran Thi Minh Hoang (2020) also estimated the breeding values of some basic reproductive traits of L and Y sows. However, these studies did not set a specific goal in improving the performance of pure D, L and Y herds raised at Dabaco Company. As a nuclear pig breeder, over the years, Dabaco Company has built a data set with a complete pedigree and individual testing as well as the reproductive 3
5. performance of pureD, L and Y breeds. This study was carried out on the basis of the above data set to estimate some genetic parameters and based on EBV select to improve the performances of 3 pure breeds raised at Dabaco Company. 1.2. STUDY OBJECTIVES General objective To estimate genetic parameters of some growth and reproductive traits, use BLUP method to predicte estimated breeding value (EBV) and selection by EBV to improve the growth performance of pure D, L and Y pigs, improve the reproductive performance of pure L and Y pigs, contribute to meet the requirements of pig production in our country. Specila objectives - Evaluate the results of individual testing, reproductive performance of sows, estimate genetic parameters for some growth and reproductive traits of the GGP herds of D, L and Y breeds raised at Dabaco Company. -Based on EBVselect to improve the growth performance of pure D, L and Y boars; - Based on EBV selectto improve the reproductive perfomance of pure L and Y sows. 1.3. FINDINGS OF THE SUBJECT - Using a large data set collected over a long period of time, estimating heritability, genetic and phenotypic correlation coefficients of some growth and reproductive traits of pureD, L and D pigs; - Using the BLUP method and based on EBV selecting to improve the ADG of pure D, L and Y pigs, and improve the number of piglets born alive per litter (NBA) of pure L and Y sows raised at Dabaco Company. 1.4. SCIENTIFIC AND PRACTICAL VALUES OF THESIS 1.4.1. Scientific value - Estimate some genetic parameters for growth and reproductive traits of pure D, L and Y pig breeds raised in our country; - Use the BLUP method to select and improve the growth performance of D, L and Yboars, improve the reproductive performance of pure L and Y sows raised in our country. 1.4.2. Pratical value - Contribute to improving the performances of pure D, L and Y pig breeds raised at Dabaco Company; 4
6. - Add materials for research and training on exotic pig breeds raised in our country. The thesis included 99 pages, exception of the references and appendices, 34 tables, 8 figures, 130 references including 51 Vietnamese documents and 79 foreign language documents. 2. LITERATURE REVIEW 2.1. SCIENTIFIC BASIS OF THESIS The research issues of the thesis were scientifically based on the characteristics of quantitative traits, genetic parameters, EBV by BLUP, growth and reproductive traits of pigs and other influence factor. 2.2. RESEARCH IN VIETNAM AND OTHER COUNTRIES The thesis has evaluated domestic and foreign research on genetic parameters of growth and reproductive traits of 3 pure D, L and Y breedpigs raised in abroad and in our country. The domestic research works were mainly refer to the reproduction and growth rate of hybrids, there were still few studies on evaluation of pure breeds, genetic characteristics and selection orientations. Therefore, evaluation of growth performance through ADG trait as well as reproductive performance of sows through NBA trait per litter, selection to improve the performances of 3 D, L and Y pig breeds in GGPherds raised at Dabaco Company in this study is very necessary that contribute to improving the genetic potential of exotic breed pigs, meeting the requirements of the pig production in our country. 3. CONTENTS AND METHODS 3.1. CONTENTS The thesis was conducted with 3 research topics: - Evaluate productivity and estimate some genetic parameters of main traits of pure D, L and Y pigs raised at Dabaco Company; - Select to improve ADG trait of pure D, L and Y boars raised at Dabaco Company; - Select to improve NBA per litter of pure L and Y sows raised at Dabaco Company. 3.2. MATERIALS AND METHODS 3.2.1. Evaluateproductivity and estimate some genetic parameters of main traits of pure D, L and Y pigs raised at Dabaco Company 3.2.1.1. Materials Research materials were gilt and male pigs, sows in the GGP herd raised at Dabaco Company from 2011 to 2017. 5
7. 3.2.1.2. Methods Male and female in the period ofindividual testing The males and females at the begining of the period of individual testinghad body weight of 25 - 40 kg, corresponding to 70 - 75 days of age, and the end of testing had body weight of 90 - 100 kg, corresponding to 130 - 150 days of age. Pigs were raised in separate groups in closed housing with ventilation and cooling fans in the summer. The individual density was 12 – 15 pigs/barn; 1.5 - 1.8 m2/pig. The pigs were feeding ad libitum and used automatic nipples. Body weigh was weighed on thestartand end date oftesting, lean meat percentage (LM) was determined on the enddate oftesting by indirect method through back fat, loin thickness measured by EXAGO equipment at position P2. Reproductive sows The sows were selected according to the regulations of Dabaco Company, artificially inseminated by double mating method. The diets were according to the technical process of Dabaco Company. Reproductive performance of each sow was considered by the following criteria: date of birth, date of farrowing, mating boar, total number of piglets born per litter (TNB), number of piglets born alive (NBA), body weight of NBA, number of piglets raised per litter, weaning date, number of piglets weaned per litter (NW) and litter weight at birth (LWB) and at weaning (LWW). Calculating methods The monitoring data were imported into Excel 2010 software, excluding values outside the range of Mean ± 3 SD. Statistical model to evaluate the influence of breeds on idividual testing traits: Yijklm = µ + Gi + Sj + YSk + Wl + eijklm in which, Yijklm: phenotypic value of the trait; µ: population mean; Gi: influence of breed; Sj: influence of sex; YSkinfluence of year – season; Wl: influence of body weight atthe test starting; eijklm: random error. Statistical model to evaluate the influence of breeds on reproductive performance traits: Yikmno = µ + Gi + YSk + Bm + Ln + eikmno in which, Yikmno: phenotypic value of the trait; µ: population mean; Gi: influence of breed; YSk: influence of year - season; Bm: influence of the breed of mating boars; Ln: influence of litter; eikmno: random error. Statistical model was used to evaluate the influencing factors for 2 traits of ADG 6
8. during the testing period and back fat at the end of the testing period: Yjklm = µ + Sj + YSk + Wl + ejklm in which, Yjklm: phenotypic value of the trait; µ: population mean; Sj: influence of sex; YSk: influence of year - season; Wl: influence of body weight at the test starting; ejklm: random error. Statistical model was usd to evaluate the influencing factors for 3 traits of litter size of reproductive sows for each breed: Ykmno = µ + YSk + Bm + Ln + ekmno in which, Ykmno: phenotypic value of the trait; µ: population mean; YSk influence of year - season; Bm: influence of the breed of mating boar; Ln: influence of litter; ekmno: random error. For the above statistical models, GLM procedure of SAS 9.1.3 using to evaluate the influence of factors, calculate LSM, SE parameters and compare by Tukey. The pedigree and data files were used PEST to encode data and VCE6 to estimate genetic parameters. 3.2.2. Select to improve ADG of pure D, L and Y breeding boars raised at Dabaco Company 3.2.2.1. Materials The research materials were males and females of 3 pure D, L and Y breds raised for the individual testing at Dabaco Company from 2015 to 2021. 3.2.2.2. Methods - Male and female in the period ofindividual testing The individual testing was in the content of the first study. - The males were selected through 3 stages The selected stages presented in figure 1. In each stage, the best males after individual testing were kept with the percentage of 5% for the next stage. After semen quality checking and mating training, the trained boars were used for insemination in the next stage. - Selection methods Data sets of individual testing were entered into Excel 2010 software, excluding excluding values outside the range of Mean ± 3 SD. The pedigree and data files were used PEST to encode data and VCE6 to estimate genetic parameters. The EBV of ADG was predicted using PEST, then the males which had highest EBV were selected with the percentage of 5%. 7
9. Stage 1 Durocpig Landracepig Yorkshirepig 6/2015 -5/2017 559 males, 395 females 413 males, 1437 females 548 males, 2214 females Selected 5% boars after testing Selected 5% boars after testing Selected 5% boars after testing Stage 2 Durocpig Landrace pig Yorkshire pig 6/2017-12/2018 1118 males, 503 females 527 males, 1323females 935 males, 3118 females Selected 5% boars after testing Selected 5% boars after testing Selected 5% boars after testing Stage 3 Duroc pig Landracepig Yorkshire pig 1/2019 -2/2021 835 males, 653 females 339 males, 689 females 719 males, 2024 females Figure 1. Diagram of selection stages for males - Evaluating selective results Selection results were evaluated through phenotypic value, EBV, accuracy of EBV and genetic tendency of the ADG trait over the selection periods. 3.2.3. Select to improve the NBA of pure L and Y sows raised at Dabaco Company 3.2.3.1. Materials The research materials were L and Y sows raised at Dabaco Company from 2015 to 2021. The sows were evaluated and selected through 2 stages with the number of sows and the number of litters according to the diagram in figure 2. In each phase, 40% of the sows with the highest breeding value on NBA were selected for the next stage. Stage 1 Landrace pig Yorkshire pig 6/2015-5/2017 222 sows, 667 litters 342 sows, 1055 litters Selected 40% of sows based on EBV Selected 40% of sows based on EBV Stage 2 Landrace pig Yorkshire pigs 6/2017-2/2021 134 sows, 361 litters 244 sows, 766 litters Figure 2. Diagram of selection stages for sows 3.2.3.2. Methods - Sow selection method The data on litter size traits of each sow were entered into Excel 2010 software, excluding values outside the range of Mean ± 3 SD. The pedigree and data files were used PEST to encode data and VCE6 to estimate genetic parameters. The EBV of NBA was predicted using PEST, then the 8
10. sows which had highest EBVs were selected with the percentage of 40%. - Evaluated slection results The selection results were evaluated through phenotypic value, EBV, accuracy of EBV and genetic tendency of the NBA trait over the selection periods. 4. RESULTS AND DISCUSSION 4.1. GROWTH, REPRODUCTIVE PERFORMANCES AND SOME GENETIC PARAMETERS OF MAJOR CHARACTERISTICS OF PURE D, L, AND Y PIGS RAISED AT DABACO COMPANY 4.1.1. Growth performance and influent factors Table 1. Results of individual testing for D, L and Y D L Y n LSM SE n LSM SE n LSM SE Age at starting to 2799 79.11a 0.12 3586 76.77c 0.11 5766 77.51b 0.09 test (day) Body weight at 2799 32.11c 0.08 3586 33.88a 0.07 5766 33.28b 0.06 starting to test (kg) Number of 2799 76.97a 0.16 3586 72.37b 0.14 5766 72.49b 0.12 days to test (day) Body weight at 2799 94.35a 0.17 3586 93.89ab 0.15 5766 93.53b 0.12 the end to test (kg) ADG 2799 812.83b 1.92 3586 832.95b 1.68 5766 834.36a 1.43 (g/day) Back fat (mm) 1071 11.42b 0.44 1551 12.18a 0.44 2584 12.21a 0.44 Loin muscle 1071 58.01a 0.86 1551 55.43b 0.86 2584 55.27b 0.85 thickness (mm) LM (%) 1071 60.32a 0.41 1551 59.12b 0.41 2584 59.06b 0.41 Note: LSM values in the same row with different a, b superscripts were statistically different (P<0.05) The individual testing data (Table 1) showed that: The ADG was highest in Y pig, the difference was statistically significant (P<0.05) in comparision with L and D pigs. The BF of Dpigs was the lowest, their loin thickness and lean percentage were also the highest, the differences were statistically significant (P<0.05). Both of these indicators were at a much better level than the economic-technical norm for exotic boars checked by the MARD, as well as higher than the recent publications on the growth performances of exotic pigs raised in our country. However, the ADGs of all three pig breeds raised in Dabaco were still lower 9
11. than many published documents in developed countries. The data in table 2 showed that: The heritabilities for ADG and LM were 0.35 – 0.43 and 0.48 – 0.52, respectively. The heritabilityerrors for these two traits was relatively low. The heritabilites of these 2 traits were relatively high compared to some research by the domestic and foreign authors. According to Ha Xuan Bo et al. (2014), heritabilities for ADG and LMof stress resistant Piétrain pigs were 0.31 and 0.19, respectively. Ngo Thi Kim Cuc et al. (2015) showed that the heritabilites of D, Piétrain and L breeds on ADG were 0.30; 0.29 and 0.32, respectively. Table 2. Genetic parameters for ADG and LM Breed Traits n ADG LM ADG 2799 0.43 ± 0.06 0.18 ± 0.09 D LM 1071 -0.08 0.50 ± 0.08 ADG 3586 0.41 ± 0.05 0.08 ± 0.10 L LM 1551 0.09 0.52 ± 0.06 ADG 5766 0.35 ± 0.04 0.31 ± 0.09 Y LM 2584 0.09 0.48 ± 0.05 Note: The diagonal elements were the heritability (h2 ± SE), the elements above the diagonal were the genetic correlation coefficient (rA ± SE), the elements below the diagonal were the phenotypic correlation coefficient (rP) Tomka et al. (2010) suggested that the heritability for ADG was ranged from 0.13 to 0.23. Radović et al. (2013) confirmed that the heritability for ADG of L pigs raised in Serbia was low (0.11) and lean percentage was high (0.63). However, there have also been some publications showing that the heritability for these two traits was fluctuated in a wide range and in many cases at a high level. Cluster (2010) collected 19 published papers on the heritability for ADG in pigs with ad libitum and semi-restricted feeding regimes were 0.03 and 0.49, respectively and the average was 0.29, and the heritabilities of8 documents of restricted diets were 0.14 - 0.76; the average was 0.30. Szyndler-Nędza et al. (2010) reported that heritability for ADG of D, L and Large White boars raised in the Netherlands were 0.472; 0.421 and 0.345, respectively. Radović cs. (2013) reported that the heritability for LMof L pigs raised in Serbia was 0.63. According to the National Swine Improvement Federation of America (2019), the heritability for ADG was 0.30. Dong et al. (2019) reported that Large White pigs at the age of 100 kg reached the heritability of 0.22. The reason for the large difference between the research results was that the 10
12. populations had different gene frequencies, data sources and different calculation methods. The genetic correlation coefficient between ADG and LM was low, ranged from 0.08 to 0.31 and error was larger than the ones of the heritability. There was not almost phenotypic correlation between the two traits (the phenotypic correlation coefficient only ranged from -0.08 to 0.09). The reason may be due to the LM was estimated by the formula from the data of back fat and loin thickness using Exago ultra-sound equipment. The data in Table 3 show that: In general, most of the traits in a litter of D sows were lower than that of L and Yorkhire sows, especially number of newborn piglets, NBA, NW (P<0.05). the Y sows were superior to L sows in terms of NBA, but the difference in the NW was not statistically significant. The number of weaned piglets/sow/year of these 3 sows were 21.96; 24.34 and 24.66, respectively. Table 3. Reproductive performance of D, L and Ysows D L Y n LSM ± SE n LSM ± SE n LSM ± SE Age at first 802 377.33a ± 0.86 1097 365.81b ± 0.76 1405 365.19b ± 0.66 farrowing (days) Parity interval 1908 149.59b ± 0.94 4256 151.45a ± 0.89 4554 150.06b ± 0.89 (days) TNB 2779 9.85c ± 0.06 5847 11.58b ± 0.04 6252 12.02a ± 0.04 NBA 2779 9.02c ± 0.06 5847 10.41b ± 0.04 6252 10.70a ± 0.04 Number of weaned 1482 22.72b ± 0.06 3673 22.99a ± 0.04 4227 23.06a ± 0.03 days (days) NW 1482 9.00b ± 0.04 3673 10.10a ± 0.02 4227 10.14a ± 0.02 Body weight 2757 1.51a ± 0.003 5791 1.49b ± 0.002 6205 1.40c ± 0.002 of piglet at birth (kg)Body weight of piglet at 1482 5.77c ± 0.03 3673 6.21a ± 0.02 4227 5.93b ± 0.02 weaning (kg) Note:Lowest mean square values (LSM) values in the same row with different a, b superscripts were statistically different (P<0.05). Exception of NBA for D sows, in general, the reproductive performance of these sows exceed the requirements of economic and technical norms for original breed pigs, as well as higher than reproductive performance of exotic sows raised in our 11
13. country which has been recently published by studies. Paura cs. (2014) reported that the first farrowing ages of Latvian L and Ypigs were 359.0 and 375.9 days, respectively; NBA in litter 1 was 9.3 and 10.1, respectively; in litter 2, 10.4 and 10.2, respectively. Meanwhile, Ye et al. (2018) evaluated 14097 Y sows with 40262 litters in Guangdong province, China, and showed that number of newborn piglets and NBA were 13.84 and 12.22 pigs, respectively. Estimation of the heritabilities and repeatabilities for the number of newborn piglets, NBA and number of weaned piglets per litter of the sows of all three breeds (Table 4) had low values and to the extent that virtually all literature has confirmed. The results of heritability estimation of L and Y pigs for NBA were: 0.12 and 0.14 respectively (Nguyen Huu Tinh and Nguyen Thi Vien, 2011). Trinh Hong Son et al. (2014) reported that the heritabilities for NBA and weaned piglet of VCN03 pigs were 0.19 and 0.11, respectively. Le Van Sang et al. (2018) reported that the heritabilities of VCN03 for number of newborn piglets and NBA were 0.26 and 0.13, respectively. Table 4. Genetic parameters for liter size traits Number of Number of weaned Breed Traits NBA newborn piglets piglets TNB (n = 2779) 0.14 0.92±0.02 0.88±0.05 0.11±0.02 0.13 D NBA (n = 2779) 0.83 0.91 ± 0.03 0.09±0.02 NW (n = 1482) 0.69 0.85 0.12 0.10 ± 0.02 0.20 TNB (n = 5847) 0.94±0.03 0.76±0.17 0.09±0.03 L NBA (n = 5847) 0.88 0.16 0.80±0.17 0.06±0.02 NW (n= 3673) 0.54 0.66 0.05 0.03±0.01 0.24 TNB (n= 6252) 0.89±0.04 0.60±0.12 0.17±0.10 Y NBA (n = 6252) 0.87 0.20 0.81 ± 0.19 0.12±0.14 NW (n = 4227) 0.54 0.66 0.10 0.06±0.09 Note: The diagonal elements were the repeatability (in italic) and the heritability (h2 ± SE), the elements above the diagonal were the genetic correlation coefficient (rA ± SE), the elements below the diagonal were the phenotypic correlation coefficient (rP) 12
14. Pholsing et al. (2009) reported that NBA of Large White pigs raised in Thailand was 0.11. According to Chansomboon et al. (2010) heritability for litter size traits of L pigs raised in Thailand was ranged from 0.05 to 0.06; repeatabilities of these traits was ranged from 0.15 to 0.18. According to Ye et al. (2018), the heritability for TNB and NBA of Y sow was 0.07 and 0.06, respectively; and repeatabilities were 0.17 and 0.14, respectively. Genetic correlation coefficients between litter size traits were positive values and higher than phenotypic correlation coefficients. Thus the influence of raising environment reduced the values of genetic correlation. Some studies in domestic as welll as foreign also obtained similar results. 4.2. SELECTION TO INCREASEBODY WEIGHT OF D, L, AND Y BOARS 4.2.1. Selection to increase ADGof D boars The data in table 5 showed that: At different stages of selection, ADG and LM of D pigs were relatively high level for male and female. In general, the ADG as well as LM in the latter period were higher than that of the previous period. The individual testing data at each stage were used to estimate genetic parameters and EBV for each stage. Table 5. Results of individual testing of Dpigs through selected stages Stage Indexes Paramatters Female Male Total ADG n 395 559 954 (g/pig/day) Mean±SE 806.83±4.86 820.96±4.25 815.11±3.20 1 n 252 361 613 LM (%) Mean±SE 60.33±0.16 60.52±0.12 60.44±0.10 ADG n 503 1118 1621 (g/pig/day) Mean±SE 807.77±2.33 828.20±2.07 822.17±1.57 2 n 455 817 1272 LM (%) Mean±SE 60.45±0.08 60.51±0.06 60.49±0.05 ADG n 653 835 1488 (g/pig/day) Mean±SE 823.47±3.84 838.99±3.74 832.18±2.71 3 n 255 528 783 LM (%) Mean±SE 60.92±0.11 61.04±0.07 61.00±0.06 The results of the selection for ADG over the stages (Table 6) showed that the BLUP selection had given an accuracy ranging from 77 to 86% for D boar. If only based on the phenotypic value for the trait, selection accuracy only achieved from 73 to 75% for D boar. Thus, using BLUP had increased the accuracy of selection by 4 to 11% for D boar. 13
15. Table 6. Results of D male selection through selection stages Selective rate (%) Stage Indexes 5 10 15 100 The phenotypic value of 940.68 903.20 881.70 820.96 ADG (g/pig/day) 1 EBV of ADG 80.70 66.04 56.17 9.81 Accuracyof EBV (%) 83.46 83.32 83.28 83.20 The phenotypic value of 941.52 908.39 888.99 828.20 ADG (g/pig/day) 2 EBV of ADG 89.57 79.33 72.96 15.22 Accuracyof EBV (%) 77.20 77.47 77.24 77.15 The phenotypic value of 1006.03 913.43 892.14 838.99 ADG (g/pig/day) 3 EBV of ADG 98.14 73.09 61.56 16.21 Accuracyof EBV (%) 85.98 85.94 85.88 85.75 The genetic tendency for male with selective rate of 5% as well as for the whole herd was in an increasing trend through the selection stages, that was clearly shown in Figure 3. Annual genetic tendency for ADG in D maleswas 4.71 g/head/day. Figure 3. Genetic tendency ofADG over the selective stages of D males (left: 5% selective rate, right: whole herd of boar) 14
16. Table 7. LSM values of ADG for Dpig over the selection stages Selection stages Statistical parameters Female Male n 395 559 1 LSM±SE (g/day) 807.07b ± 5.77 819.97c ± 4.25 n 503 1118 2 LSM±SE (g/day) 810.08b ± 3.95 830.53b ± 2.07 n 653 835 3 LSM±SE (g/day) 824.71a ± 4.99 836.87a ± 3.74 Note: Least square mean (LSM) in the same column with different a, b superscripts were statistically different (P<0.05) The LSM of ADG for D pig in table 7 provided a more accurate evaluation of selective efficiency over the three stages. Overall, ADG at stage 3 in both male and female was always highest and the difference was statistically significant compared with stages 1 and 2 for the males. Particularly for the females, the difference between stage 2 and stage 1 was not statistically significant, because the selection was done only for the males. 3.2.2. Selection to increase ADGfor L pigs Table 8 showed the different growth performance of L pig through selection stages 1,2 and 3. Similar to the results of individual testing in D pig, both ADG and LM of L pig were gradually increased through the selection stages. Table 8. Results of performance testing of L pigs through selected stages Stage Indexes Paramatters Female Male Total ADG n 1437 413 1850 (g/pig/day) Mean±SE 833.29±2.83 842.57±3.25 835.36±2.49 1 n 646 227 873 LM (%) Mean±SE 59.14±0.10 59.17±0.10 59.15±0.09 ADG n 1323 527 1850 (g/pig/day) Mean±SE 846.85±2.38 862.71±2.51 851.37±3.23 2 n 561 362 923 LM (%) Mean±SE 60.13±0.10 60.52±0.09 60.28±0.12 ADG n 689 339 1028 (g/pig/day) Mean±SE 860.98±3.91 876.27±3.91 866.02±5.82 3 n 344 196 540 LM (%) Mean±SE 60.50±0.10 60.81±0.12 60.61±0.14 15
17. Table 9. Results of L male selection through selection stages Selective rate (%) Stage Indexes 5 10 15 100 The phenotypic value of 923.25 918.05 907.24 842.57 ADG (g/pig/day) 1 EBV of ADG 71.73 61.46 54.59 0.17 Accuracyof EBV for ADG (%) 69.13 68.66 68.62 68.18 The phenotypic value of 927.82 922.50 910.38 862.71 ADG (g/pig/day) 2 EBV of ADG 76.64 51.91 14.36 7.25 Accuracyof EBV for ADG (%) 71.75 71.67 71.79 70.85 The phenotypic value of 932.26 925.91 918.39 876.27 ADG (g/pig/day) 3 EBV of ADG 120.99 131.55 125.20 10.42 Accuracyof EBV for ADG (%) 73.61 73.27 73.20 71.96 The data in table 9 showed thatthe phenotypic values and EBVs of ADG for L males was keptfor breeding increased through each selection stage. The accuracy of EBV wasranged from 68 to 74%. If only based on phenotypic value to selection, the accuracy of EBV achieved only from 57 to 62%. Thus, using BLUP had increased the accuracy of EBV by 11 - 12% for L males. The EBV of ADG in stage 3 compared with stage 1 given an annual genetic tendency for this trait was 13.31 g/pig/day in L pigs. Figure 4. Genetic tendency of ADGthrough selection of L males (left: 5% selection rate, right: whole herd) 16
18. The genetic tendency for selected males to be kept for breeding at the rate of 5%, as well as for the whole individual testing herd for productivity was on an increasing trend through the selection stages, that clearly showed in figure 4. Table 10. LSM ofADG for L pigs over the selection stages Selection stages Statistical parameters Female Male n 1437 413 1 LSM±SE (g/day) 835.24c ± 3.34 844.07c ± 6.39 n 1323 527 2 LSM±SE (g/day) 849.90b ± 3.02 860.31b ± 4.56 n 689 339 3 LSM±SE (g/day) 865.79a ± 4.28 878.30a ± 6.60 Note: Least square mean (LSM) values in the same column with different a, b superscripts were statistically different (P<0.05). The data in table 10 provided a more accurate evaluation of selection efficiency through the selection stages. The LSM values on body weight gain of males as well as females were always highest at stage 3, the difference was statistically significant compared with stage 1 and stage 2 (P<0.05). The LSM of ADG for L males and females were increased gradually from stage 1 to stage 3, the difference was statistically significant (P<0.05). 4.2.3. Selection to increase ADG of Ypigs Results of individual testing for Ypigs through selection stages were presented in table 11. Table 11. Results of performance testing of Y pig through selection satges Stage Indexes Paramatters Female Male Total ADG n 2214 548 2762 (g/pig/day) Mean±SE 849.06±2.22 856.38±4.80 850.51±2.03 1 n 1229 313 1542 LM (%) Mean±SE 58.82±0.08 59.50±0.13 58.96±0.07 ADG n 3118 935 4053 (g/pig/day) Mean±SE 864.36±1.75 873.17±2.95 866.39±1.53 2 n 879 451 1330 LM (%) Mean±SE 60.16±0.07 60.27±0.10 60.20±0.06 ADG n 2024 719 2743 (g/pig/day) Mean±SE 876.50±2.27 884.64±4.21 878.63±2.06 3 n 945 447 1392 LM (%) Mean±SE 60.30±0.06 60.45±0.11 60.35±0.05 17