Phenomics

Scaling methane phenotyping in beef cattle: integrating high-throughput methane sensors into feed intake nodes for simultaneous measurement of methane and feed intake

---

Abstract

Understanding methane emissions from cattle is relevant in the context of understanding contributions to agricultural greenhouse gas emissions and the energetic loss to the animal. Traditional methane measurements, such as respiration chambers and tracer techniques, have low throughput or are labour-intensive, and spot-sampling systems depend on voluntary animal visits, resulting in inconsistent data capture. As a result, phenotypic datasets of methane currently available remain small, constraining genetic evaluations and breeding programs aimed at impacting methane emissions. Additionally, concurrent measurement of feed intake is of particular importance in understanding factors associated with differences in enteric methane emissions, thus increasing technical complexity and hardware requirements. A solution is to integrate methane sensors into feed intake systems, thus enabling automated, high-throughput measurements during routine feeding. Vytelle, in collaboration with Integrity Communication Solutions (CO, USA), developed a methane sensor designed for co-deployment with Vytelle SENSE feed intake nodes. The sensor leverages nondispersive infra-red (NDIR) detection systems to measure methane concentrations in air samples drawn from the feed through every 10-seconds during feeding events, defined as periods when the animal is actively consuming feed. First-generation sensors were deployed at a commercial bull testing facility in Alberta, Canada, and later an updated version was installed at West Virginia University's (WVU) central testing station (WV, USA). Thirty-four animals were tested at WVU for 50 days, and 44 animals for 37 days at the commercial site. After quality control, which excluded feeding events shorter than 60 seconds and with fewer than four methane measurements, the final data set had 15,026 feeding events at WVU and 10,257 events at the commercial facility. Methane phenotypes were calculated as the area under the curve (AUC) per second for each qualifying feeding event. An animal's trial phenotype was estimated using a mixed model, with AUC per second as the response variable and node, day, and hour of day as fixed effects, and animal as a random effect. Repeatability estimates indicated that 29 days of measurement were needed to determine the methane phenotype with an accuracy of 0.8. Validation of this phenotype was assessed by correlating with measured average daily feed intake (ADFI). The methane phenotype showed a moderate correlation with ADFI at WVU (r=0.47, 95% CI:0.16-0.70). The commercial site using an early first-generation sensor version showed similar trends (r=0.26), but with greater variability. Although absolute methane levels differed between locations because of independent sensor calibrations, both sites showed similar biological patterns and repeatability. The results show that integrating methane sensors with feed intake nodes enables high-throughput phenotyping suitable for genetic evaluations across various production environments.

Keywords: 2026

How to Cite:

Thekkoot, D., Gillespie, J., Powell, T. & Osterstock, J., (2026) “Scaling methane phenotyping in beef cattle: integrating high-throughput methane sensors into feed intake nodes for simultaneous measurement of methane and feed intake”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2283304. doi: https://doi.org/10.31274/wcgalp.23496

Rights: 1