S. Vanderick1,2, F.G. Colinet1, A. Mineur1, R.R Mota1, Genetic parameters of novel mid-infrared predicted milk traits in three dual-purpose cattle breeds S. Vanderick1,2, F.G. Colinet1, A. Mineur1, R.R Mota1, N. Gengler1, H. Hammami1 1 Gembloux Agro-Bio Tech, ULiège, B-5030 Gembloux, Belgium 2 sylvie.vanderick@uliege.be EAAP 2018 Meeting Dubrovnik

EAAP 2018 Meeting Dubrovnik Dual-purpose cattle Dual-purpose cattle in Europe play a very important role because one individual cow combines the two essential characteristics: production of milk and production of meat. The main dual-purpose cattle breed in Walloon Region of Belgium is Dual-Purpose Belgian Blue, which is a local breed, but there are also MON and NOR animals. Dual-purpose Belgian Blue (dpBB) Montbéliarde (MON) Normande (NOR) EAAP 2018 Meeting Dubrovnik

EAAP 2018 Meeting Dubrovnik Dual-purpose cattle For breeding or management purposes, we need to collect relevant phenotypes in routine on a large-scale … but this can be a challenge especially for these small sized cattle populations, even more in organic farming which are pasture-based systems! Costly Not repeatedly collected even on-farm level Difficulty in gathering relevant large-scale data in routine Small sized cattle populations (Organic) pasture based production systems EAAP 2018 Meeting Dubrovnik

Usefulness of mid-infrared spectra (MIR) Milk samples (milk payment, milk recording) Raw data = MIR spectra Calibration equations Quantification wide range of novel phenotypes ! MIR analysis FOSS Since recently, mid-infrared spectrometry has been recognized as a rapid and cost-effective tool to collect routinely a wide range of accurate of milk-based biomarkers. Therefore, MIR can be also useful for dual-purpose cattle… in order to get accurate and relevant data which could be helpful for breeding and management purposes… EAAP 2018 Meeting Dubrovnik

EAAP 2018 Meeting Dubrovnik Objectives Genetic parameters of 39 novel MIR predicted milk traits How MIR traits can predict longevity at early stages? Any help from genomics? Our objective was three-fold, First, genetic parameters of 39 novel MIR traits were estimated for each of the 3 dual-purpose cattle breeds. Then, in the context of development of breeding programs for small sized populations, it was important to see how these novel MIR traits can be useful to assess longevity at early stages. And finally, also in line with the topic of this session, we investigated how genomics can be helpful for the predictions of longevity with those novel MIR traits. EAAP 2018 Meeting Dubrovnik

EAAP 2018 Meeting Dubrovnik Data dpBB MON NOR Cows in production 2,988 1,330 621 1st lactation test-day 21,287 10,062 4,637 2nd lactation test-day 11,771 6,716 2,532 3rd lactation test-day 6,246 4,217 1,370 Pedigree 7,744 3,807 2,304 For the genetic parameters estimation, test-day records from the first-three lactations for the 3 breeds were used. As you can see, these data are limited with almost 3000 cows in production for the original national breed, about 1300 cows in production for MON and only 600 cows for NOR EAAP 2018 Meeting Dubrovnik

MIR predicted milk traits Cheese properties: 8 Groups of fatty acids (FA): 14 Minerals: 5 Protein components: 8 Ketone bodies: 3 Methane (CH4) single-trait multi-lactation random regression test-day model Genetic parameters of the following MIR traits were computed within breed with a single-trait multi-lactation random regression test-day model. Some of those traits are quite familiar – such as milk fatty acids, cheese technological properties or Methane – and others are less - minerals, ketone bodies… Genetic parameters were also computed for conventional production traits (milk, fat, protein contents and % + SCS) EAAP 2018 Meeting Dubrovnik

Conventional Production traits Cheese properties Ketone bodies In this slide, results for average daily h² over the lactation of each trait are summarized (means through the first-three lactations of these average daily h² for each trait) . Breeds are differentiated by colours and shapes of symbols. CH4 Conventional Production traits Cheese properties Ketone bodies Groups of FA Minerals Proteins EAAP 2018 Meeting Dubrovnik

Conventional Production traits Cheese properties Ketone bodies These values tended to be higher for dpBB (0.13-0.64) than for MON and NOR (0.03-0.60) breeds. Some large differences between breeds have been observed… especially between dpBB and Normande breed for milk composition mainly. CH4 Conventional Production traits Cheese properties Ketone bodies Groups of FA Minerals Proteins EAAP 2018 Meeting Dubrovnik

Conventional Production traits Cheese properties Ketone bodies The estimated values of h² for methane make sense with a range of 0.30 to 0.45, even if the calibration equation was not validated for these breeds! CH4 Conventional Production traits Cheese properties Ketone bodies Groups of FA Minerals Proteins EAAP 2018 Meeting Dubrovnik

Conventional Production traits Cheese properties Ketone bodies It’s important to note that, given the limited number of data, the results have to be taken carefully … especially those for Normande breed where a high sampling error is expected with only 600 cows in production. CH4 Conventional Production traits Cheese properties Ketone bodies Groups of FA Minerals Proteins EAAP 2018 Meeting Dubrovnik

Conventional Production traits Cheese properties Ketone bodies However, we can say/summarize that most MIR traits are moderately to highly heritable but most of them might not have obvious economic value (in Walloon Region of Belgium) for those dual-purpose breeds in conventional and even more in organic farming systems. Therefore, how can we use them ? Why not to predict lowly heritable traits or traits difficult to measure as for example longevity? CH4 Conventional Production traits Cheese properties Ketone bodies Groups of FA Minerals Proteins EAAP 2018 Meeting Dubrovnik

Prediction of longevity at early stages ? production & novel MIR traits Longevity, is by its association to animal robustness an important part of any breeding objective, also in dual-purpose cattle and even more in organic breeding systems. However, information on longevity records can only be measured after an has done its productive life. Therefore, predictors of longevity traits could be very useful in order to breed for improved longevity at early stages. It’s important to note that longevity is defined here in terms of number of lactations done by an animal! EAAP 2018 Meeting Dubrovnik

Prediction of longevity at early stages ? 21,287 test-day records 2,988 primiparous dpBB cows 7,744 animals in pedigree 314 genotyped animals (108 ♀ and 206 ♂ ) EBV & GEBV for production & novel MIR traits Using TD records from dpBB primaparous cows, EBV for conventional production traits and for some novel MIR traits were computed. Similarly, GEBV were also computed from genomic information of about 300 genotyped animals. EAAP 2018 Meeting Dubrovnik

Prediction of longevity at early stages ? Different scenarios without and with genomics Partial least squares analyses (high correlated traits) 21,287 test-day records 2,988 primiparous dpBB cows 7,744 animals in pedigree 314 genotyped animals (108 ♀ and 206 ♂ ) EBV & GEBV for production & novel MIR traits Different scenarios including or not genomic information were tested to predict longevity. To this end, PLS analyses were performed in order to avoid major multicollinearity issues due to the fact some traits are highly correlated. EAAP 2018 Meeting Dubrovnik

Longevity from MIR traits? Scenarios EBV (r) GEBV (r) S1: Milk, fat & protein yields 0.25 0.19 S2: Acetone, citrates & BHB 0.22 0.37 S3: S1 +S2 0.34 0.41 S4: S1 + UFA + SCFA + MCFA 0.42 0.29 S5: S2 + UFA + SCFA + MCFA S6: S3 + UFA + SCFA + MCFA 0.51 0.47 The different scenarios were evaluated in terms of predictive ability of novel traits towards longevity, The left column showing the results in terms of EBV, the right when using genomically enhanced EBV Based on genotyped dpBB bulls with ≥ 10 daughters (N=160) EAAP 2018 Meeting Dubrovnik

Longevity from MIR traits? Scenarios EBV (r) GEBV (r) S1: Milk, fat & protein yields 0.25 0.19 S2: Acetone, citrates & BHB 0.22 0.37 S3: S1 +S2 0.34 0.41 S4: S1 + UFA + SCFA + MCFA 0.42 0.29 S5: S2 + UFA + SCFA + MCFA S6: S3 + UFA + SCFA + MCFA 0.51 0.47 Results were at the first glance surprising as the ability to predict was decreasing for traditional traits under genomics. However different authors found that genomic correlation at the basis of this can also be different from results based on pedigree relationships. Based on genotyped dpBB bulls with ≥ 10 daughters (N=160) EAAP 2018 Meeting Dubrovnik

Longevity from MIR traits? Scenarios EBV (r) GEBV (r) S1: Milk, fat & protein yields 0.25 0.19 S2: Acetone, citrates & BHB 0.22 0.37 S3: S1 +S2 0.34 0.41 S4: S1 + UFA + SCFA + MCFA 0.42 0.29 S5: S2 + UFA + SCFA + MCFA S6: S3 + UFA + SCFA + MCFA 0.51 0.47 Ketosis related traits showed promising results as did some major groups of fatty acids. Based on genotyped dpBB bulls with ≥ 10 daughters (N=160) EAAP 2018 Meeting Dubrovnik

Longevity from MIR traits? Scenarios EBV (r) GEBV (r) S1: Milk, fat & protein yields 0.25 0.19 S2: Acetone, citrates & BHB 0.22 0.37 S3: S1 +S2 0.34 0.41 S4: S1 + UFA + SCFA + MCFA 0.42 0.29 S5: S2 + UFA + SCFA + MCFA S6: S3 + UFA + SCFA + MCFA 0.51 0.47 All scenarios behaved much worse in a GEBV setting when production traits were added. Based on genotyped dpBB bulls with ≥ 10 daughters (N=160) EAAP 2018 Meeting Dubrovnik

MIR traits ~ useful in 3 dual-purpose cattle breeds Take home messages Novel MIR predicted milk traits for dual-purpose cattle: heritable  selection promising early indicators of longevity (for dpBB) Use of genomics: Slight increase of reliabilities (not shown) GEBV for novel MIR traits tend to be better predictors of longevity In conclusion, the take home messages are: MIR traits could be useful for dual purpose cattle breeds… Indeed, MIR traits are heritable and then could be used for breeding and management purposes. They showed also promising results to predict longevity at early stages for dual purpose Belgian Blue breed. Moreover, using genomic information, allow to increase slightly reliabilities and GEBV for MIR traits tend to be better predictors of longevity MIR traits ~ useful in 3 dual-purpose cattle breeds EAAP 2018 Meeting Dubrovnik

EAAP 2018 Meeting Dubrovnik Acknowledgements EAAP 2018 Meeting Dubrovnik

S. Vanderick1,2, F.G. Colinet1, A. Mineur1, R.R Mota1, Genetic parameters of novel mid-infrared predicted milk traits in three dual-purpose cattle breeds S. Vanderick1,2, F.G. Colinet1, A. Mineur1, R.R Mota1, N. Gengler1, H. Hammami1 1 Gembloux Agro-Bio Tech, ULiège, B-5030 Gembloux, Belgium 2 sylvie.vanderick@uliege.be EAAP 2018 Meeting Dubrovnik

Conventional Production traits Cheese properties Ketone bodies In this slide, results for average daily h² over the lactation of each trait are summarized (means through the first-three lactations of these average daily h² for each trait) . Breeds are differentiated by colours and shapes of symbols. Conventional Production traits Cheese properties Ketone bodies Groups of FA Minerals Proteins EAAP 2018 Meeting Dubrovnik

R² of calibration equation min max Cheese properties 0.44 0.78 Groups of FA 0.77 0.99 Minerals 0.82 Proteins 0.69 Ketone bodies 0.62 0.89 CH4 0.70 EAAP 2018 Meeting Dubrovnik