Abstract :
[en] Livestock production constitutes a critical component of global food security, yet it faces persistent challenges of production efficiency, environmental sustainability, and animal health. In high-yielding dairy cows, early lactation is marked by negative energy balance (NEB), when the energy demands of milk synthesis exceed energy intake from feed. This metabolic state induces excessive mobilization of body reserves, disruption of metabolic homeostasis, impaired fertility, and heightened disease susceptibility. These consequences not only impose substantial economic costs but also raise animal welfare concerns, thereby undermining the sustainability and social acceptability of modern dairy systems. Improving energy balance is therefore essential for production efficiency, animal well-being, and long-term industry viability. However, genetic selection against NEB has been constrained by the difficulty of large-scale phenotyping. This thesis aimed to develop and evaluate genetic strategies for characterizing and improving energy balance in dairy cows by integrating large-scale genomic data with innovative proxy phenotypes.
First, measured reference NEB (RNEB) was compared with predicted NEB (PNEB), a novel energy deficiency score (EDS), 15 biomarkers, and 3 production traits, showing moderate to high phenotypic correlations. Genetic parameters of 20 traits were estimated using a 20-traits repeatability model, with heritabilities ranging from 0.16 to 0.38. Genetic correlations between logit-transformed PNEB (LPNEB) and the studied traits varied from –0.60 to 0.87, with the strongest associations for logit-transformed EDS (LEDS) (0.85) and non-esterified fatty acids (NEFA) (0.87). Collectively, the 19 traits explained 89% of the genetic variance of LPNEB, with biomarkers alone accounting for 82%, LEDS alone 65%, and NEFA alone 62%. Recursive modeling further identified eight traits, including NEFA and LEDS, as highly dependent on LPNEB, underscoring their potential as robust biomarkers.
Second, SNP-based genomic correlation analyses, extended to the chromosomal level, revealed strong associations between LPNEB, LEDS, and metabolic biomarkers, particularly NEFA. Independent contribution analyses confirmed NEFA as the primary driver of both LPNEB and LEDS, while chromosomal scans identified BTA19 and BTA25 as candidate regions underlying NEB regulation.
Third, single-step genome-wide association analyses with a 50-SNP sliding window, using 30,634 records from 25,287 first-parity cows and 566,170 SNPs from 3,757 genotyped animals, identified the top genomic regions for LPNEB and LEDS. Three regions (BTA1, BTA5, and BTA16) were shared between the two traits, alongside unique loci supporting distinct genetic architectures. Candidate gene analyses revealed 17 genes for LPNEB and 10 for LEDS, with 6 shared. Functional enrichment highlighted roles in energy metabolism for LPNEB, and additional neuronal signaling pathways for LEDS. QTL enrichment indicated associations with fertility and somatic cell score.
In conclusion, this thesis explored the genetic correlations between LPNEB, LEDS and multiple proxy traits, metabolic biomarkers, and production traits. The results revealed strong associations with key indicators such as NEFA and identified genomic regions and candidate genes underlying NEB regulation. These findings enhance understanding of the genetic mechanisms of LPNEB and LEDS. Moreover, the results provide evidence supporting the implementation of the MIR-based indicator LEDS as a reliable and scalable tool for improving resilience and energy efficiency in dairy cows through genomic selection and metabolic monitoring.
