How much greenhouse gases are emitted by agriculture ?

In a rapid­ly chan­ging world, the prio­ri­ty for the agri­cul­tu­ral sec­tor is to feed more people. At the same time, the sec­tor is trans­for­ming itself by adap­ting to cli­mate change, or even miti­ga­ting it, through a varie­ty of dif­ferent mea­sures : the reduc­tion of green­house gas (GHG) emis­sions, the sto­rage of car­bon in soil and ener­gy pro­duc­tion. In France, the imple­men­ta­tion of all these mea­sures – pro­mo­ted by the Natio­nal Low Car­bon Stra­te­gy – would lead to a 46% reduc­tion in green­house gas emis­sions lin­ked to agri­cul­ture by 2050¹.

On a glo­bal level the “agri­cul­ture, fores­try and other land use” sec­tor is res­pon­sible for 23% of anthro­po­ge­nic GHG emis­sions, i.e. 12 GtCO₂ equivalent/year². Most of these emis­sions are due to either agri­cul­tu­ral emis­sions of methane (CH₄) (4 GtCO₂ equivalent/year) and nitrous oxide (N₂O) from nitro­gen fer­ti­li­sa­tion (2.2 GtCO₂ equivalent/year), or to land use changes and defo­res­ta­tion, which release 5.2 Gt of car­bon dioxide (CO₂) per year.

Accor­ding to a Euro­pean Com­mis­sion report³, pre­ci­sion far­ming could reduce GHG emis­sions from Euro­pean agri­cul­ture by 1.5 to 2%. This is main­ly based on variable rate appli­ca­tion sys­tems, which deli­ver a dose of fer­ti­li­ser adap­ted to the needs of the plants, thus redu­cing the asso­cia­ted N₂O emis­sions. Other tools that can reduce GHG emis­sions are self-gui­dance devices for agri­cul­tu­ral machi­ne­ry, through impro­ved dri­ving that reduces fuel consumption.

Pre­ci­sion far­ming allows for indi­vi­dua­li­sed inputs to a plant or ani­mal accor­ding to its needs. It is based on an “observation/diagnosis/prediction/action” cycle that relies on infor­ma­tion and com­mu­ni­ca­tion tech­no­lo­gies. Satel­lite data, increa­sin­gly sup­ple­men­ted by on-board sen­sors, are used to mea­sure plant defi­cien­cies, par­ti­cu­lar­ly in field crops. This data is then inte­gra­ted into agro­no­mic models that pro­vide recom­men­da­tions for fer­ti­li­ser appli­ca­tions at variable rates, depen­ding on the posi­tion within the plot. Simi­lar deci­sion sup­port tools are also used in live­stock far­ming to avoid over­fee­ding cat­tle, limi­ting manure and thus CH₄ emis­sions.

Digi­tal tech­no­lo­gy suf­fers from a signi­fi­cant lack of use. In Europe, as lit­tle as 22% of farms use variable rate fer­ti­li­ser appli­ca­tion tools. In France, only 10% of cereal farms have adop­ted them.

Seve­ral fac­tors explain this. First­ly, the return on invest­ment is not always clear­ly eva­lua­ted. These tech­no­lo­gies and ser­vices are cost­ly, and far­mers need to know the bene­fits – whe­ther eco­no­mic, envi­ron­men­tal, or rela­ted to per­cei­ved use­ful­ness. In the Occi­ta­nie region, we have set up the Occi­ta­num Living Lab to test these tools on dif­ferent farms and eva­luate the bene­fits and costs they entail.

The use­ful­ness of this tool depends on how it is inte­gra­ted into the wor­king envi­ron­ment. This is why it is impor­tant to encou­rage co-desi­gn that brings toge­ther manu­fac­tu­rers and far­mers to pro­duce tools that are adap­ted to far­mers’ needs. They can thus be sim­pler to use and adap­ted to the work car­ried out in the field. Howe­ver, there are still a num­ber of obs­tacles : lack of trai­ning in the agri­cul­tu­ral sec­tor in gene­ral, ideo­lo­gi­cal oppo­si­tion, ques­tions about data secu­ri­ty, etc.

No, tech­no­lo­gy is not the solu­tion, but it is a part of it. Rather, it is the changes in agri­cul­tu­ral prac­tices, which tech­no­lo­gy will faci­li­tate, that reduce the impact on the envi­ron­ment. Tech­no­lo­gy accom­pa­nies these changes, to help them deve­lop on a lar­ger scale, for example.

I am tal­king about agroe­co­lo­gy. This approach consists of pro­mo­ting a balance in the sys­tem with the help of eco­lo­gi­cal pro­cesses, without che­mi­cal inputs, unlike conven­tio­nal agri­cul­ture. For example, mono­cul­tures can be repla­ced by a mix­ture of spe­cies, which reduces the need for inputs.

But agroe­co­lo­gy is a more com­plex far­ming sys­tem. On the one hand, it requires close atten­tion to plant and ani­mal health to anti­ci­pate and treat the pro­blem qui­ck­ly. Tech­no­lo­gi­cal tools can help to detect pro­blems ear­ly : opti­cal sen­sors for plant health, connec­ted insect traps to detect pests, or ani­mal move­ment sen­sors to moni­tor their health. On the other hand, mixing plant spe­cies requires pre­ci­sion sowing, even in the middle of a pre­vious crop. Pre­ci­sion see­ders make it easier to do this, while avoi­ding tur­ning over the soil and relea­sing CO₂ into the atmosphere.

This is a ques­tion that the scien­ti­fic com­mu­ni­ty is begin­ning to address, but an assess­ment of their envi­ron­men­tal foot­print through life cycle ana­ly­sis has not yet been made. Even so, the GHG savings from digi­tal tools are expec­ted to be much higher than their actual eco­lo­gi­cal foot­print.  Never­the­less, we must conti­nue to take mea­su­re­ments in order to get an accu­rate pic­ture of the envi­ron­men­tal benefits.

The cen­tral issue is data flow. We are not yet at the stage of big data in agri­cul­ture, but the ques­tion needs to be asked before the data explodes. We will need to consi­der the choice of data to be kept, the form of data sto­rage, deve­lop­ment of fru­gal algorithms…

While until now digi­tal tech­no­lo­gies have main­ly been concer­ned with eco­no­mic gains and com­fort, which are the main concerns of ope­ra­tors, their contri­bu­tion to and impact on cli­mate change are now beco­ming increa­sin­gly important.