Solène Fresco's thesis defense

The defense will take place in French at the INRAE center in Jouy-en-Josas, Bâtiment 440 - Amphithéâtre Jacques Poly, as well as online, on Wednesday March 26 at 9 a.m.

The thesis is entitled: “Towards genetic selection to reduce enteric methane emissions from French dairy cattle.”

It was written under the supervision of :
- Didier BOICHARD, Research Director, UMR GABI, INRAE Jouy-en-Josas
- Pauline MARTIN, Research Associate, UMR GABI, INRAE Jouy-en-Josas

She was supervised by :
- Sébastien FRITZ, Genetics Team Manager, Eliance

She will be evaluated by : 
- René BAUMONT, Research Director, INRAE Université Clermont-Auvergne, Rapporteur
- Sandrine GRASTEAU, Research Director, INRAE Université de Tours, rapporteur
- Adrien BUTTY, Scientific Executive, Qualitas AG (Switzerland), examiner
- Xavier ROGNON, Professor, INRAE Université Paris-Saclay, examiner
- Flavie TORTEREAU, Research Engineer, INRAE Université de Toulouse, examiner

Summary

Agricultural methane emissions, particularly those linked to ruminant enteric fermentation, represent a major challenge in the fight against climate change. In this context, it has become crucial to understand, measure and reduce these emissions in order to limit the environmental impact of dairy cattle farming. One solution is to select individuals that are genetically less methane-emitting.
The first step in this project was to deepen our understanding of methane emissions during lactation, using direct measurements on experimental farms with GreenFeed devices. These analyses revealed that methane emissions vary throughout the day according to the rhythm of feed distribution. Three methane traits expressed in different units were analyzed: in g/d, in g/kg of milk corrected for protein and butyrate (PLC), and in g/kg of dry matter ingested. Each shows different correlations with traits of interest such as milk production, dry matter intake, body weight or body condition score. 

However, direct measurements of methane emissions are not suitable for the large-scale phenotyping required for genetic selection. This is why we have developed equations for predicting methane emissions from the mid-infrared (MIR) spectra of milk.We developed three equations, one for each methane trait, based on reference data.A fourth prediction was obtained by multiplying the prediction in g/kg PLC by the PLC corresponding to the spectrum used for the prediction. In parallel, two equations predicting methane in g/d and developed in other projects were made available to us, bringing to six the number of predicted methane characters studied in this thesis. These equations were applied to MIR spectra routinely collected on commercial farms belonging to the milk recording scheme and supplied by livestock advisory companies.
We obtained 13,721,855 methane predictions for Holsteins, 12,183,188 for Montbeliarde and 929,638 for Normande.However, a complementary study showed that predictions were less reliable at the beginning and end of lactation, and only predictions corresponding to MIR spectra collected between 70 and 200 days of lactation were used for genetic analysis.
These predictions were used as the phenotype for estimating genetic parameters of methane emissions.The heritability of the traits was moderate, demonstrating the feasibility of their selection. However, genetic correlations between methane traits are low to moderate, which raises the question of which criterion to select, and implies that the consequences of selection will vary according to the methane trait(s) included in the selection objectives.The genetic determinism of predicted methane emissions is polygenic, and we have been able to identify promising candidate genes.Genetic correlations between methane traits and traits of interest are mostly unfavorable, but low to moderate. 

They therefore make it possible to integrate one or more methane traits into selection objectives.
The pilot genomic evaluations have shown satisfactory results, highlighting the variability of indexes for the various methane traits, as well as the existence of young bulls that improve them.Moreover, the accuracy of the indexes is sufficient to deploy the evaluation on a routine basis. 
All the results obtained in this thesis confirm the technical feasibility of genetic selection to reduce methane emissions from dairy cattle, which the industry is expected to implement in 2025.