The rest is methane. This can be released when manure is stored in effluent ponds. But in New Zealand, where most cows and sheep leave their droppings where they stand, most methane emissions come from the animals' mouths as part of normal digestion.
The simple solution to reduce greenhouse gas emissions from agriculture is to reduce agriculture, either by reducing intensity or reducing overall output.
In New Zealand, where agricultural exports made up half of all commodity exports in 2013, it's also a deeply unpopular option and, given the increasing global population, it may also be impractical to reduce food supply.
The New Zealand Agricultural Greenhouse Gas Research Centre (NZAGRC), backed by $48.5m in government funding over ten years, is instead tasked with finding direct ways to reduce emissions without reducing agricultural output.
NZAGRC Director Harry Clark says the centre is currently negotiating its next round of research contracts. The last round included work to develop both a vaccine and compounds that could be added to feed to reduce methane emissions.
"We've shown they work in the lab," Clark says. "We now need to demonstrate they work in an animal."
Other projects include breeding sheep and cattle that produce lower emissions. The centre is also screening plants for their ability to reduce nitrous oxide emissions in a bid to find a natural alternative to dicyandiamide (DCD), and is looking at how plants with different attributes, such as deep roots, can increase the amount of carbon stored in soil.
Lincoln University Professor of Biogeochemistry Leo Condron has been studying soil chemistry for three decades. He believes there's "incredible potential" to use land management to alter CO2 and nitrous oxide emissions in New Zealand.
Some farm management techniques are now well-proven. Only use as much fertiliser as you need and target it well. Don't send animals onto wet paddocks.
Others seem to have potential. One of Condron's studies involved applying biochar - charcoal made from waste biomass - to the kind of soil that's found under dairy land. The result? A "massive" reduction, both in the lab and in the field, in nitrous oxide loss.
Biochar doubles as a waste management and carbon sequestration tool as it stabilises and stores the carbon from the materials such as sawdust used to make it. The catch, Condron says, is that the best results in the test came from applying the equivalent of 30 tonnes of biochar per hectare, which isn't currently practical.
"At the end of the day, it has to be economic," he says. "For individual farmers, it's never going to be economic for a single farmer to make biochar. How do they make it?"
The same issue comes up on a larger scale: "If someone is buying a hectare of land for $40,000, they need to get a reasonable return on that money. Because the value of land is so high in dairying, you have to drive the production to try to get the returns."
Again, one way to reduce nutrient loss or emissions from agriculture is simply to produce less. "That's one of the arguments for organic production," he says.
Clark agrees that it's a difficult problem. "Agriculture is a major driver of economic growth and with the demand of meat and milk products, there are large opportunities for expansion of output," he says.
"This poses challenges for greenhouse gas emissions, reducing emissions against a backdrop of economic output. It's a very difficult challenge."
As the latest IPCC report makes clear, though, finding a speedy solution is crucial.
The Climate Change Solutions series is a joint project between Element and Lincoln University, which is intended to illuminate a pathway for a sustainable future.
