Genetic and genomic selection to reduce boar taint in Danish pigs

Transcription

1 Genetic and genomic selection to reduce boar taint in Danish pigs Anders B. Strathe,, Ingela Velander, Thomas Mark and Haja Kadarmideen Department of Clinical Veterinary and Animal Sciences, University of Copenhagen Danish Agriculture & Food Council, Pig Research Centre

2 Acknowledgements

3 Outline Introduction Performance testing in Denmark Genetic associations to production and reproductive traits in Landrace and Duroc Results from genome wide associations Summary

4 Introduction Stop surgical castration by 8??? animal welfare labour intensive, especially with the use of anesthesia consumer acceptance

5 Deposition in back fat Skatole and indole: Hindgut microbial metabolism Tryptophan is the substrate Fecal like odor Androstenone: Produced in testis Concentrated and secreted in saliva Urine like odor

6 Odours perceived by humans Danish Meat Research Institute (3)

7 So why has nothing happened? Boar taint is heritable Definition of boar taint is inconsistent No common reference method Breeding takes time Unfavourable correlations to other important traits...?

8 Objective To develop a feasible selection strategy against boar taint in Danish pig breeds Performance testing Breeding value estimation Selection Progress!

9 Outline Introduction Performance testing in Denmark Genetic associations to production and reproductive traits in Landrace and Duroc Results from genome wide associations Summary

10 Performance testing Boar taint trait is recorded at teststation All boars, i.e. AI- and slaughter boars Human-nose-score Slaughter boars only trained panel at DC-Ringsted Photometric determination of skatoleeq Boar taint on live AI-boars via biopsi Androstenone, skatole and indole

11 Human nose score Procedure: ml sample bottle 5 g lard (medium size) 75 ml boiling hot water Stand for min Scale: = no boar taint, = week boar taint = strong boar taint Cat. = Threshold

12 Human-nose score Prevalence: Duroc: 5.3% Landrace:.7% Yorkshire: 6.%

13 Heritability of HNS Unitrait model: HNS = Sektion*Panel + Lwgt + Age + Slweek + Animal + e (Fixed effects) (Random effects) Breed Estimate Duroc.8 (.5) Landrace.6 (.7) Yorkshire.9 (.5)

14 Biopsy procedure Biopsy-device from SUISAG Baes et al. Animal. 3. 7:74- ~7 AI-boars without complications Animal care protocol is needed

15 Heritabilities of BT-compounds Multitrait model, i ={Log(Ska), Log(Ind), Log(And)}: Log(BT i ) = Sektion + Lwgt + Age + Animal + e (Fixed effects) (Random effects) Duroc: Log(Ska) Log(Ind) Log(And) Log(Ska). (.7) Log(Ind).78 (.).3 (.7) Log(And).8 (.5). (.4).48 (.9) Yorkshire: Log(Ska) Log(Ind) Log(And) Log(Ska).37 (.4) Log(Ind).85 (.33).9 (.) Log(And).8 (.6).58 (.38).53 (.4)

16 Outline Introduction Performance testing in Denmark Genetic associations to production and reproductive traits in Landrace and Duroc Results from genome wide associations Summary

17 Danish Landrace Boar taint data (SABRE) Skatole(equivalents) 6 boars Androstenone 5 pairs of full sibs 5 Androstenone, ug/g Skatole-equivalent, ug/g

18 Genetic parameters for Landrace Bivariate model: Log(Skatole) = HYS + SLwgt + SLage + Litter + Animal + e Log(Andro) = Herd + SLwgt + SLage + Litter + Animal + e Trait h r g r p Log(skatole).33 (.5) Log(Androstenone).59 (.4).37 (.).6 (.3) Strathe et al. J. Anim. Sci. 3. 9:

19 Production traits: Landrace Multi-trait model for performance traits and boar taint compounds Trait x r g(x, Log(skatole)) r g(x, Log(androstenone)) ADG -.4 (.8). (.) Feed conversion ratio.8 (.9) -.4 (.6) Meat percentage -. (.7) -.8 (.) Low genetic correlations and largely favorable! Strathe et al. J. Anim. Sci. 3. 9:

20 Production traits: Duroc Results from fitting bivariate models Trait x r g(x, ADG) r g(x, Meat percentage) Log(skatole) -.8 (.) -. (.8) Log(Indole) -.4 (.) -.9 (.8) Log(Androstenone) -.7 (.) -.5 (.) HNS.8 (.33) -.33 (.8) In both breeds low genetic correlations Consistent with Dutch results Windig et al. (J. Anim. Sci..9: 9)

21 Litter size data: Landrace Litter size traits: TBN and LP5 Full and half sib females to the BT boars. parity sows and pure bred litters Trait No Mean SD Min Max TBN LP

22 Model Litter size and BT Let y, y, y 3 and y 4 denote vectors of TNB, LP5, Log(skatole) and Log(androstenone) records with assumptions e e e e c c Z Z a a Z Z d d Z Z s s Z Z p p Z Z b b b b X X X X y y y y s s c c a a d d s s ps ps s R I e G A a d s C I c S I p, ~,, ~,, ~,, ~ N N N N Genetic (co)variances between servicesire, sow and BT compounds Non-existing residual covariances Herd-year-quarter effects and regression on age at first mating Strathe et al. J. Anim. Sci. 3.9:

23 Litter size and BT: Landrace Heritability on the diag. with genetic correlations on the off diag. Trait TNB LP5 Log(skatole) Log(androstenone) Sire Dam Sire Dam TNB LP5 Sire. (.) Dam.36 (.4).9 (.) Sire.7 (.).7 (.). (.) Dam.43 (.5).58 (.5).38 (.3).6 (.) Log(skatole).5 (.).6 (.) -. (.8).3 (.3).33 (.4) Log(androstenone) -. (.7) -.4 (.5) -.4 (.) -. (.7).4 (.4).59 (.3) In general, weak genetic correlations between BT compounds and service-sire fertility Strathe et al. J. Anim. Sci. 3.9:

24 Littersize and BT: Duroc Bivariate models for TNB and boar taint TNB(dam) TNB(S-sire) Log(Andro) Log(Ska) TNB(dam). (.) TNB(S-sire).3 (.).4 (.) Log(Andro) -.8 (.6) -.3 (.9).47 (.8) Second model Log(Ska).9 (.7) -.6 (.8) NA. (.8) Strathe et al. J. Anim. Sci. 3.9:

25 Combine Rep-model () with models (, 3) for boar taint compounds Standard assumptions Model for semen and BT e e e a a a Z Z Z c c Z Z p Z b b b X X X y y y pe a a a c c R I e G A a C I c K I p p, ~ ;, ~ ;, ~ ;, ~ N N N N Genetic (co)variances between semen and BT compounds Non-existing residual covariances Strathe et al. J. Anim. Sci. 3.9:

26 Semen and BT Key genetic correlations Trait x r g(x, Log(skatole)) r g(x, Log(androstenone)) Volume. (.3). (.8) Concentration -. (.3) -.4 (.6) Total sperm -.7 (.3) -.3 (.8) Functional sperm -.6 (.7) -. (.8) High standard errors on genetic correlations Strathe et al. J. Anim. Sci. 3.9:

27 Semen quality and BT Multi-trait liability model for binary semen quality traits Trait x h r g(x, Log(skatole)) r g(x, Log(androstenone)) Normal/abnormal.8 (.).9 (.) -.39 (.5) Motility low/high. (.3) -.8 (.5) -.38 (.9) Low heritabilities for semen quality traits Weak genetic correlations Strathe et al. J. Anim. Sci. 3.9:

28 Semen and BT: Duroc Bivariate repeatability models Trait x h r g(x, Log(Ska)) r g(x, Log(Ind)) r g(x, Log(And)) Motility.4 (.) -.4 (.) -.5 (.) -.6 (.3) Volume.9 (.).5 (.). (.9).5 (.) Again, weak genetic correlations

29 Testosterone and semen Low Test. Line (LTL): 8 ng/ml testosterone High Test. Line (HTL): 44 ng/ml testosterone Walker et al. (J. Anim. Sci. 4. 8:59 63) r p (testosterone; TTS) =. Ren et al. (Reprod. Dom. Anim :93 99)

30 Outline Introduction Performance testing and preliminary results Genetic associations to production and reproductive traits Results from genome wide associations Concluding remarks

31 GWAS for skatole in Landrace Region - Chr. 4 Top-SNP: SIRI94 CYPE Rowe et al. BMC Genomics 4, 5:44

32 GWAS for skatole in Duroc Top-SNP: SIRI94

33 GWAS for androstenone in Duroc Region: Mb on Chr. 6 Candidate genes: SULTA HSD7B Duijvesteijn et al. ()

34 Single-Step GWAS for androsteneone u qdz And ZDZ a Wang et al. Genet. Res.. 94:73 83

35 Summary Selection against boar taint will have Minimal impact on production traits Minimal impact on litter size traits Minimal impact on semen traits GWAS results points to previous mapped regions Breeding is an option Effective online sorting and grading of carcasses Determine the economic weight of the trait Phenotyping costs must be dramatically reduced Breeding will never guarantee % of carcasses, being perceived as free of boar taint