Genetic approaches to improving lamb survival SBRT, Nottingham - 18-nov-2017 Mark Young CIEL United Kingdom Forbes Brien University of Adelaide Australia
UK Agri-Tech Centres Motivated by vision to drive forward a more efficient, profitable and resilient agri-food supply chain Aim to collaborate & reduce fragmentation in sector
Some of CIEL s Business Members
CIEL s Academic Partners Largest investment in livestock research in a generation Government funding match funded by Partners CIEL manages spend through robust processes with IUK & Universities
CIEL Industry led Focusing on practical innovations Building industry partnerships Supporting innovation projects
CIEL Sheep Facility Investments Total investment c. 1.8 million
CIEL mobile CT scanner (SRUC)
CIEL mobile CT scanner (SRUC)
CIEL mobile CT scanner (SRUC)
CIEL mobile CT scanner (SRUC)
CIEL feed intake facility (AFBI) AFBI picture(s)
CIEL feed intake facility (AFBI) Bon (petit) appétit
CIEL mobile feed intake facility (SRUC)
CIEL mobile feed intake facility (SRUC) Coming soon to a field near you.
Farm Grazing Platform Rothamsted Research, North Wyke
Farm Grazing Platform Rothamsted Research, North Wyke
Farm Grazing Platform Rothamsted Research, North Wyke
Farm Grazing Platform Rothamsted Research, North Wyke
Genetic approaches to improving lamb survival SBRT, Nottingham - 18-nov-2017 Mark Young CIEL United Kingdom Forbes Brien University of Adelaide Australia
What can we learn from other countries?
Outline Importance of lamb survival heritability or heritability times variation? Genetic evaluations around the world The SIL System in New Zealand Non-genetic variation Genetic variation Improving rate of genetic gain NZ Australia Conclusions
Importance of lamb survival Major component of low reproductive efficiency, globally! Cost of lamb losses is high Aus - c.$540m AUD/ yr NZ - c.$300m NZD/ yr UK average is 85% & figures show no detectable improvement in last 40 years (Dwyer et al. 2015)
Extensive conditions Variable, sometimes challenging, environments Typical for NZ, Aus & South Africa, parts of UK Average 75-81% survival based on pregnancy scanning Lower survival for multiples - 1 > 2 > 3 Improved ewe nutrition beneficial Pregnancy scanning key management development Increasing shelter not simple solution
Sheep Ewes lamb & graze on hills, all year round
Non-genetic variation Weather Topography & shelter Nutrition Feeding Litter size Pregnancy scanning Ewe behaviour, maternal-offspring bonding Maternal behaviour score Natural lambing behaviour Management (shepherding, stocking rate) Close supervision Easy-care DNA parentage systems End up breeding from animals with problems?
Less lamb from lowland
more from hills
Ewe BCS at lambing & lamb survival Ewe nutrition during pregnancy Condition Score at Lambing Parity Lamb Survival (%) Low 2.3 Single bearing 85 High 3.2 Single bearing 91 Low 2.2 Twin bearing 57 High 3.2 Twin bearing 71 +6 +14 Data from 13 farms (Behrendt et al. 2011)
Low heritability is it worth bothering? Lamb survival has a low heritability but this is not best way to characterise potential for genetic gain Lamb survival shows a lot of variation Potential for genetic gain is function of heritability AND amount of variation Greater lamb survival has high economic value Incremental gains can deliver significant value, slowly
Genetic evaluations around the world Country Lamb Survival Evaluated When started NZ Yes 2001 Ireland Yes 2008 GE - Base trait(s) Reference Lamb trait direct & maternal Lamb trait direct only Young & McIntyre (2006) McHugh et al. (2014) UK Yes 2016???? Lamb trait direct Ewe trait rearing ability Conington (pers. comm.) Australia Yes Related trait 2018? 2006 Ewe trait rearing ability No. lambs born & weaned Bunter et al. 2017
SIL GE System in NZ Simple definition of lamb survival Did lamb a ewe produced survive to weaning? Birth weight not in evaluation so seldom recorded Extreme data excluded from evaluation Weight of lamb weaned A function of 3 traits with 5 BVs produced Number of lambs born (NLB) Lamb Survival (SUR & SURM) Weaning LW of lamb (WWT & WWTM) Analysis separates ewe & lamb genetic effects Lamb thrift (vigour) Lamb growth from mothering ability from milking ability Evaluation improvements tried, didn t add value Assessing why or when lamb died Fitting different statistical models
Farm size Average of 4,000 stock units (SU) per farm 1 ewe = 1.15 SU, 1 beef cow = 4.5 SU Some farms very large Large Station Discussion Group 16 Farms 560,000 SU
Farm size Average of 4,000 stock units (SU) per farm 1 ewe = 1.15 SU, 1 beef cow = 4.5 SU Some farms very large Paparata Farms Owner - Trevor Johnson 5,860 hectares 63,000 SU 30,400 sheep 6,200 beef cattle Managed as 3 separate farms 2000 ewe stud flock Large Station Discussion Group 16 Farms Paparata Romney 560,000 StudSU 2,000 ewes DNA parentage, 2,700 lambs/year Genetic improvement through selection to improve Number of lambs per ewe Lamb Survival Lamb growth versus adult size Wool production Facial Eczema (disease) Tolerance
Realised genetic gains - NZ Litter size (NLB) increase of 13% in 15 years Survival increase of c.1% in 15 years Survival gains greater in maternals, despite more traits selected for Likely reasons Larger datasets give more accuracy in maternal flocks Strong focus on litter size (NLB) so lamb numbers more accurately recorded 0.14 0.12 0.10 0.08 0.06 0.04 0.02 Maternal flocks - NLB ebv 0.014 0.012 0.010 0.008 0.006 0.004 0.002 Maternal & Terminal flocks - SUR ebv 0.00 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017-0.02 0.000 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017-0.002 Aug-2017 data
Genetic variation Low heritability indicates low additive genetic variation (predictable inheritance) Other genetic effects (unpredictable inheritance) occur to some degree; Dominance - effect of gene depends on what it is paired with Epistasis - interaction between different genes Epigenetic environment affects degree of gene expression Genetic by environment interaction e.g. survival differs for singles versus multiples Can genotype data shed light on non-additive genetic variation?
Improving genetic evaluations for lamb survival Better data Improve prediction models Use other predictors
Improving genetic evaluations for lamb survival Better data Record all lamb deaths Use weaning information to update (SIL System Strategy) Use reproduction data to fill in gaps At pregnancy scanning, discriminate litter size among multiple carrying ewes and record foetal age Consider recording reasons for lamb deaths Have complete pedigrees (value of DNA parentage!) Progeny test young rams? Lifts accuracy of BVs Mandatory sire paternity testing?
Improving genetic evaluations for lamb survival Better data Improve prediction models Direct and maternal effects Permanent environmental effects Account for randomness of effects Separate on basis of litter size Higher heritability in twins than singles Other smart statistical approaches Be aware variability in data structure impacts on analyses Must test theory to assess impact before making changes to evaluation How do we cope with this? Can we combine merit into one BV or are they different traits? Indicates a GxE effect What does on farm observation tell us?
Improving genetic evaluations for lamb survival Better data Improve prediction models Use other predictors Lambing ease/ difficulty, birth assistance Lamb vigour score Maternal behaviour or agitation scores Time lamb behaviours Lamb rectal temperature Birth weight?
Potential additional predictors Time point Pre-mating AI Mating Mid-pregnancy New trait Ewe LW, ewe BCS Ewe LW, ewe BCS Foetal scan number, ewe LW, ewe BCS Lambing Birth weight Lambing ease Birth cost score Lamb vigour Maternal behaviour score Rectal temperature Skeletal dimensions Death Autopsies
Potential predictors of lamb survival Brien et al. (unpublished) Trait Visual Scores of: Abbreviation Heritability (%) Genetic correlation (%) Lambing ease LE 3-37 Maternal behaviour MBS 25-23 Overall lamb vigour OBV 11-35 Objective measures of: Rectal temperature RT 5 +74 Time taken to bleat BLT 4-43 Time taken to follow FOLL 7-52
Indicator traits for lamb survival Brien et al. 2010; Brien et al. (unpublished) 0.3 0.25 Heritability 0.2 0.1 0.03 0.04 0.05 0.07 0.11 0 Lambing Ease Bleat Rectal Temp Follow Birth VigourMat. Score
Indicator traits for lamb survival Brien et al. 2010, Brien et al. (unpublished) Genetic Correlation 1 0.8 0.6 0.4 0.2 0-0.2-0.4-0.6-0.23-0.35-0.37-0.43-0.52 0.74
Potential gains in lamb survival DP+ index Brien et al. (unpublished) Base + one extra predictor Genetic gain over 10 years in LSW (lambs weaned/100 born) 2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 1.13 +68% +61% +40% +37% +30% +10% 0.00 Base B+RT B+FOLL B+OBV B+MBS B+BLT B+LE Selection Scenario
Potential gains in lamb survival DP+ index Brien et al. (unpublished) Genetic gain over 10 years in LSW (lambs weaned/100 born) 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 1.13 +154% Base + multiple predictors +173% +172% +174% +199%
Potential gains in lamb survival DP+ index Brien et al. (unpublished) Base + most practical predictors Genetic gain over 10 years in LSW (lambs weaned/100 born) 2.50 2.45 2.40 2.35 2.30 2.25 2.20 2.15 2.10 Compared to base gain of 1.13 +130% +113% +98% RT,OBV RT,OBV,MBS RT, OBV, MBS, LE Selection Scenario
Improving rate of genetic gain for lamb survival Slow genetic gain due to low heritability but high variability partly offsets this Can select in both sexes, which raises selection intensity need complete data on lamb losses All animals up for selection are alive! Family information on lost lambs informs BV estimation so pedigree critically important Intervention at lambing inhibits genetic gain (& may make things worse?) Must challenge animals in flocks focused on genetic improvement interventions act against this
How a genetic approach to improve lamb survival can suit extensively-run flocks Focus on preparation before lambing Optimise ewe nutrition, health (vaccination & anthelmintic) Provide adequate pasture & shelter for lambing Manage ewes in litter size classes? Optimise flock size & density Management during lambing Commercial flocks - little intervention, less mis-mothering, low labour costs In studs need pedigree - lambing rounds, DNA parentage, Pedigree Matchmaker Optional recording of indicator traits near birth to boost accuracy Need easy/quick to measure traits
Conclusions Existing tools can deliver genetic gain in lamb survival Good data critically important to maximise gains All deaths recorded, accurate pedigree Some lambing management systems may act against selection to improve lamb survival More sophisticated models may increase accuracy more development work needed! Genomics may offer most value by targeting non-additive genetic variation