1. Selection Lines for Residual Feed Intake (RFI) in Yorkshire Swine Weiguo Cai 1, David Casey 2, Jack Dekkers 1 1 Department of Animal Science, Iowa State University, Ames, IA 50011 and 2 Pig Improvement Company, Franklin, KY 42134

2. Measures of Feed Efficiency FCR = Feed/gain FE = Gain/Feed FCR = Feed/gain FE = Gain/Feed Residual Feed Intake = (Observed FI) – (Expected FI given ADG, BF, BW) (Koch et al., 1963) RFI Factors contributing to RFI Physical activity Physical activity Body temp. regulation Body temp. regulation Basal metabolic rate Basal metabolic rate Digestibility Digestibility Energetic efficiency Energetic efficiency Immune response Immune response Food wastage Food wastage Measurement error Measurement error = FI – b 1 ADG – b 2 BF – b 3 BW.75

3. Objectives Develop lines that differ in FI but have similar ADG, BFDevelop lines that differ in FI but have similar ADG, BF Estimate genetic parametersEstimate genetic parameters Evaluate direct and correlated responses to selectionEvaluate direct and correlated responses to selection Develop a resource population toDevelop a resource population to study the biological / physiological basis of feed intake / efficiencystudy the biological / physiological basis of feed intake / efficiency develop selection tools for feed intake / efficiencydevelop selection tools for feed intake / efficiency Yorkshire RFI Selection Experiment

4. Outline  Feed intake recording in pigs  Design of selection experiment  Results: Direct and correlated responses Estimates of genetic parameters Estimates of genetic parameters  Ongoing and future work

5. Recording individual feed intake in pigs Electronic Feeder Electronic Feeder FIRE © Osborne Inc. FIRE © Osborne Inc. Commercial Feeder Commercial Feeder

6. Feed intake data

7. Extensive edit procedures for feed intake data unexplained transponder zero tag consumption outliers from graphs etc. Developed by David Casey Casey, Stern, Dekkers. 2005. J. Anim. Sci. 83:969–982. Identify errors in visits (16 criteria) Create graphs (each pig) Remove unwanted information Calculate average daily FI Fill-in missing daily FI

8. Design of RFI selection experiment Generation 0 Low RFI line Control line select for RFIGeneration 1 Random selection selection reduce FIGeneration 2 keep adg, bf ~ constant Generation 3 Generation 4

9. Design of RFI selection experiment Generation 2 parity 1 90 boars on FIRE © 90 gilts on FIRE © Random selection Control Line Select Line RFI Generation 1 parity 1 selection Generation 2 parity 2 Select ~12 boars ~70 gilts ~70 gilts Sibs of Selected boars RFI EBV

10. Traits measured − from ~40 to ~115 kg  Electronically measured feed intake (FIRE ©, Osborne Inc.)  1 st parity select boars  2 nd parity select gilts  Ultrasound scan – on-test – mid-test – off-test  Body weight – weekly

11. Average daily gain (ADG)  Weigh pigs weekly  Linear regression of weight on day:  ADG (kg/day)

12. Ultrasound Scan  Backfat (BF) -10th rib  Loin muscle area (LMA)  Intramuscular Fat (IMF)

13. Summary of data across 4 gener. in select line TraitsNMeanPhen.Std RFI (kg/d) 83600.11 ADFI (kg/d) 8361.990.16 ADG (kg/d) 8360.770.08 BF (mm) 8367.952.90 LMA (cm 2 ) 83626.284.04 IMF (%) 5651.740.39 Note: Phen.Std excludes all fixed effects and pen(group) effect Var(RFI) Var(FI) =0.46

14. RFI EBV model (ASREML) FI = group + sex+ pen(group) + b 1 * (onwt * gen) + b 2 * (onage * gen) + b 3 * (offwt * gen) + b 1 * (onwt * gen) + b 2 * (onage * gen) + b 3 * (offwt * gen) + b 4 * (ADG * gen) + b 5 * (BF * gen) + BV RFI + e b 4 * (ADG * gen) + b 5 * (BF * gen) + BV RFI + e ADG and BF pre-adjusted for onwt, onage, offwt Fixed: group and sex Random: pen (group) Note: Metabolic weight not included

15. Selection response for RFI and FI 1.1  p

16. Correlated responses for ADG and BF 0.3  p

17. Correlated responses for LMA and IMF 0.1  p

18. Heritabilities (on diagonal) genetic (below diagonal) phenotypic (above diagonal) Heritabilities and correlations Heritabilities and correlations Based on ASREML 2-trait analysis 0.53 IMF 0.62 0.62 LMA 0.61 0.61 BF 0.33 0.33 ADG 0.40 0.40 FI 0.33 0.33RFIIMFLMABFADGFIRFI

19. Heritabilities (on diagonal) genetic (below diagonal) phenotypic (above diagonal) Heritabilities and correlations Heritabilities and correlations Based on ASREML 2-trait analysis 0.53 IMF 0.07 0.62 0.62 LMA 0.04-0.05 0.61 0.61 BF 0.18 0.11 0.11 0.29 0.29 0.33 0.33 ADG 0.10-0.03 0.44 0.44 0.65 0.65 0.40 0.40 FI 0.04-0.10-0.01-0.01 0.51 0.51 0.33 0.33RFIIMFLMABFADGFIRFI

20. 0.53 0.26 0.26-0.06 0.27 0.27 0.23 0.23 0.15 0.15IMF 0.07 0.62 0.62-0.05 0.19 0.19-0.10-0.26LMA 0.04-0.05 0.61 0.61 0.40 0.40 0.55 0.55-0.01BF 0.18 0.11 0.11 0.29 0.29 0.33 0.33 0.74 0.74 0.15 0.15ADG 0.10-0.03 0.44 0.44 0.65 0.65 0.40 0.40 0.71 0.71FI 0.04-0.10-0.01-0.01 0.51 0.51 0.33 0.33RFIIMFLMABFADGFIRFI Heritabilities (on diagonal) genetic (below diagonal) phenotypic (above diagonal) Heritabilities and correlations Heritabilities and correlations Based on ASREML 2-trait analysis

21. Design of direct line comparison Generation 3 Generation 3 Parity 2 Parity 2 Selection line n=50 Control line n=38 88 gilts on FIRE identify extremes at 70 kg (gene expression)

22. Trait Line difference (select-control) Line difference (select-control) estimated based on EBV in selection line Directcomparison RFI, g/d -124-93*** FI, g/d -175-123*** ADG, g/d -24 -22 NS BF, mm -0.77 0.063 NS *** significant at P<.01, NS Not significant at P<.10 Estimates of line differences differences

23. Summary u 46% of variation in FI is not related to ADG and BF = RFI u RFI has substantial heritability (0.33) u 4 generations of selection for RFI has resulted in significant change in RFI (93-124 g/d) u RFI selection has slightly reduced ADG and BF

24. Yorkshire selection & control lines for Residual Feed Intake Integration of Functional Genomics and Quantitative Genetics to Improve Feed Efficiency in Pigs Dekkers, Tuggle, Anderson, Nettleton, Honavar Iowa State University Rekaya, Barb, Hausman University of Georgia and USDA-ARS-RARC NRI Grant no. 2005-35604-15618

25. Gene expression studies Low RFI gilts Select line Restr. feeding SR Ad Libitum SA Restr. feeding CR Ad Libitum CA High RFI gilts Control line Liver Fat Gene Expression Microarray Measure the level of expression of >20,000 probes for genes across genome Source: http://www.affymetrix.com Image from Hybridized GeneChip RNA 4444 ~70 kg Limitednumbers

26. Biological basis of differences in RFI Residual Feed Intake = (Observed FI) – (Expected FI given ADG, BF, BW) (Koch et al., 1963) RFI New grant funded by National Pork Board and Iowa Pork Producers Association and Iowa Pork Producers Association Factors contributing to RFI Physical activity Physical activity Body temp. regulation Body temp. regulation Basal metabolic rate Basal metabolic rate Digestibility Digestibility Energetic efficiency Energetic efficiency Immune response Immune response Food wastage Food wastage Measurement error Measurement error Richardson and Herd 2004

27. Ann CrockJennifer KerkmanMatt Wolfe Jay Lampe Bryce MartinBenny Mote Napapan PiyasatianRyan VoylesDoug Newcom Jeremy BurkettNick BerryTom Baas Sender LkhagvadorjLong Qu Chris Tuggle Yangfang WangDan NettletonLloyd Anderson PIC/Genus for providing FIRE © feeders ISU Center for Integrated Animal Genomics Monsanto Choice Genetics USDA-CSREES NRI Grant no. 2005-35604-15618 Acknowledgements Weiguo Cai, David Casey John Newton and personnel at ISU Bilsland farm