Drivers of variability in greenhouse gas footprints of crop production

Abstract

Identifying drivers of farm-level greenhouse gas (GHG) footprints of crop production can reveal opportunities to improve farming practices and enable more targeted GHG mitigation strategies. Although many studies evaluated the GHG footprints of crop production, differences between and within crops have not been systematically evaluated for a large number of farms so far. Here, we evaluated possible sources of variability in GHG footprints (in terms of kg CO2-eq/kg crop produced) of 26 crops, grown in compliance with Unilever's Sustainable Agriculture Code, using data from 4565 farms in 36 countries from 2013 through 2016. We quantified crop-farm-specific GHG footprints based on four components: (i) emissions from electricity use, (ii) emissions from fossil fuel (petrol and diesel) use, (iii) emissions from crop and pruning residue application, and (iv) emissions from fertilizer use. On average, fertilizer use contributed most to the GHG footprint for 23 out of the 26 crops in our dataset. We further found that variability in GHG footprints was smaller between crops (45%) than within crops (55%). Regression modelling revealed that on average 44% of the GHG footprint variability within crops could be attributed to (a selection of) three explanatory variables, i.e., yield, area of production, and year of production. Of these, yield was the most important explanatory variable. Lower GHG footprints were associated with higher yields for 24 out of the 26 crops. Relationships with area and year of production were less clear, and directions of the relationships were more variable between crops. Strategies to improve fertilizer use efficiencies while maintaining or increasing yields are preferable in a GHG reduction programme.