These efforts were predicted to have a greater impact than simply encouraging people to eat less meat, according to the new study, published today in AGU Advances.
The paper's results could help inform future climate policy, and the methods developed in the study would allow countries to make up-to-date estimations of their methane emissions from livestock.
The authors emphasised that these improvements should not come at the expense of the environment, which can occur in factory farming.
"We do not endorse the industrial livestock system for methane mitigation, because it causes many other environmental problems like pollution, failed manure management and land-use changes for grain and high-quality fodder," Jinfeng Chang, an environmental scientist at Zhejiang University and first author of the new study, said.
"There are many other more sustainable ways to improve efficiency."
Globally, raising animals for milk, meat and eggs accounted for one-third of human methane emissions.
Dairy and beef cows were the top contributors, due in part to their large numbers.
Also, cows were ruminant animals, like buffalo, sheep and goats, and the microbes in their guts produced methane as a by-product of breaking down their food.
According to the Food and Agriculture Organization of the United Nations (FAO), methane emissions from livestock rose more than 50 per cent between 1961 and 2018, and were expected to continue to rise as demand for animal products increases, especially in countries with growing populations and income.
In the study, the authors created new estimates of global methane production from livestock using the most recent methods proposed by the Intergovernmental Panel on Climate Change (IPCC).
Next, they calculated the production efficiency for each country, using livestock production data from the FAO.
The analysis looked at all types of animal products, including milk and meat from cattle, buffaloes, goats and sheep, pork products, poultry and eggs.
From 2000-2018, even as total emissions rose, the intensity of emissions from most types of livestock fell globally as production became more efficient.
This drop resulted from advanced breeding practices and improvements in nutrition, which created animals that yielded more milk and meat.
When the authors looked at methane emissions under future scenarios, they saw that, while eating less meat will help, continuing these gains in production efficiency has even more potential for cutting methane, especially in countries with low efficiency and high future production.
They calculated that under a "Business-As-Usual" socioeconomic scenario, agricultural improvements in the top 10 countries with the greatest potential to reduce methane could account for 60-65 per cent of the decrease in global methane emissions by 2050 from increasing efficiency.
"They made it absolutely clear that improving production efficiency has much greater mitigating effect than demand-side efforts, particularly in low-income countries," Ermias Kebreab, (an animal scientist at the University of California, Davis, who was not involved in the research) said.
Kebreab agreed that increasing efficiency without causing environmental damage was possible and that some regions could improve their efficiency by up to 20 per cent.
With cattle, for example, cows adapted to the local environment could be bred with high-yield breeds to increase meat and milk production.
Software in the local language could then formulate a balanced diet to support the crossbred cattle using feed available in the region.
"It's just it's a matter of resource allocation."
Organisation/s: Zhejiang University (China), International Institute for Applied Systems Analysis (Austria), Peking University (China), California Institute of Technology (US), Université Paris-Saclay (France), Commonwealth Scientific and Industrial Research Organization (Australia)
Funder: Jinfeng Chang is supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (Grant No. XDA26010303). Philippe Ciais acknowledges support from the CLAND Convergence Institute of the French National Research Agency (ANR). Mario Herrero acknowledges funding from the Bill and Melinda Gates Foundation through the MERLIN project (INV-023682). The authors declare no conflict of interest.