Introduction

Decomposition of agricultural non-CO2 mitigation

We now want to focus on the 2050 time horizon and elaborate in detail how and at what costs agriculture could contribute to climate stabilization. At the 100 $/tCO2eq carbon price considered by IPCC (2014), we find agricultural non-CO2 emissions could be reduced by up to 2.6 GtCO2eq in 2050 considering both technical and structural supply side mitigation options as well as consumers’ response to price changes. This corresponds to non-CO2 emission reductions of almost 40% relative to the baseline. Around 70% of the potential mitigation originates from a reduction of CH4 emissions and 30% from reductions in N2O. Figure 2 presents the agricultural mitigation potential in 2050 by region (2a) and mitigation mechanism (2b) as a function of carbon price. Asia and Latin America offer particularly significant potential for emission reductions, while developed regions like North America, Europe and Oceania can contribute to a much lesser extent due to their lower baseline GHG emission intensity and limited share of global non-CO2 emissions by 2050. Technical options can provide direct emission reductions of around 0.85 GtCO2eq/year at a carbon price of 100$/tCO2eq, accounting for around 33% of the total agricultural mitigation globally. However, including indirect effects of these options through related productivity increases adds 0.15 GtCO2eq/year in mitigation and increases the share to 38% of total global mitigation. For example, propionate precursors or anti-methanogen vaccination do not only reduce CH4 emissions from enteric fermentation through improved digestibility but also enhance animal productivity. Hence, more livestock products may be produced with less emissions as the emission intensity (emissions per unit of output produced) decreases. Adoption of these technical options would require investment and operation costs of around 13 bn $/year globally by 2050 (12 bn$ in 2030), the majority arising in emerging and developing regions like Asia (almost half) or Sub-Saharan Africa (10%) while only one quarter of the costs occur in developed countries (Europe, Oceania, and North America). Structural adjustments, i.e., shift from rather GHG inefficient extensive grazing systems toward mixed grass–cereal feeding systems, and could contribute 1.0 GtCO2eq/year and consumer response to prices will add another 0.6 GtCO2eq/year at 100 $/tCO2eq in 2050. The cost-efficient contribution of the different mitigation options to the total potential varies across regions and at different carbon prices, hence “one size fits all” policies will not enable to realize mitigation potentials cost-efficiently across regions. Figure 3 provides the disaggregated regional mitigation potentials for carbon prices of 40 and 100$/tCO2eq. Technical mitigation options contribute most significantly (in relative shares) in the developed regions of Europe (80% of total potential, 100 $/tCO2eq) and North America (50% of total potential, 100$/tCO2eq), mainly through the adoption of highly (cost-) efficient (large-scale) manure management and nitrogen fertilization technologies. In the remaining regions, technical options account for around one quarter of the total mitigation potential. While at lower carbon prices mainly improved rice management options and anaerobic (large-scale) digesters are being adopted, improved fertilization management becomes profitable with rising carbon prices. In Asia, improved rice management such as switching to dryland rice (with residue incorporation) and reduced nitrogen fertilization, offers opportunities to significantly reduce CH4 emissions of up to 0.3 GtCO2eq/year at 100 $/tCO2eq (more than 50% emissions reduction) from flooded rice paddies. Similarly, Hussain et al.22 highlighted the significant potential for GHG reduction, comparing to traditional rice production systems through improved tillage, irrigation, and fertilization practices of up to 67%. Structural adjustments such as shifts in production systems or relocation through international trade account for around 39% (1.0 GtCO2eq/year) of the total mitigation potential at 100$/tCO2eq. Especially in Latin America and East Asia, transition of livestock production systems may significantly reduce non-CO2 emissions. We find mitigation of up to 0.5 GtCO2eq/year in 2050 at 100 $/tCO2eq through the decrease of ruminant production in tropical areas and shift to mixed-cereal feeding systems in temperate areas with higher productivities. In Latin America, this development coincides with a decrease in extensive grassland-based systems. The importance of this transition of ruminant livestock production systems for climate change mitigation is also acknowledged by other studies3,9,24. Many of these structural changes are highly cost-efficient. Thus, the mitigation from structural adjustments tends to account for a larger share of total agricultural mitigation potential at lower carbon prices, i.e., structural adjustments provide about half of the total mitigation at 40$/tCO2eq.

Reduction in consumption levels due to price increases accounts for around 24% (0.6 GtCO2eq/year) of the mitigation potential at 100 $/tCO2eq. Carbon price induced commodity price increases drive consumers to reduce their consumption levels of GHG-intensive products in our modeling framework, mainly in Latin America and South Asia. Global average calorie intake decreases from around 3300 kcal/capita/day to around 3200 kcal/capita/day (−3%) in 2050 while average agricultural commodity prices increase by 18%. However, calculated mitigation potentials take into account a food security constraint, which limits the number of people undernourished in each region. In developed regions such as Europe and North America, consumers are less impacted due to more efficient production systems and therefore less significant price increases, as well as lower demand elasticities related to higher income levels. As the mitigation potential from technical and structural options becomes exhausted with increasing carbon price, additional mitigation is mainly achieved through demand side adjustments as consumption decreases with increasing prices. Across the three mitigation wedges, the livestock sector accounts for around 70% (at 100$/tCO2eq) of the total potential, mostly coming from structural adjustments and reduction in consumption levels. This highlights the importance of livestock for climate change mitigation and its role as a land use driver and source of non-CO2 emissions.

Potential synergies for CO2 mitigation

We find that a mitigation policy targeting only non-CO2 emissions from agriculture yields synergies with CO2 mitigation through avoided land use change. Results show additional reduction of 0.7 GtCO2eq/year (100 $/tCO2eq) of CO2 in 2050 due to land sparing through productivity increases, and reduces consumption and production levels. This is in line with other studies who find co-benefits of non-CO2 mitigation through intensification for CO2 emissions from land use change12,25. Synergies with CO2 mitigation come on top of the 2.6 GtCO2eq/year reduction in CH4 and N2O and play an important role especially in regions with high land use change emissions such as in Sub-Saharan Africa, Latin America, and Southeast Asia. As GHG-intensive products such as ruminant meat become relatively more expansive with increasing carbon prices especially in regions with low productivity, expansion of extensive pastures decreases significantly in developing regions (Fig. 4). Globally, around 35 million hectare (Mha) of cropland and 225 Mha of pastures are freed up from agricultural use by 2050 at a carbon price of 100$/tCO2eq on non-CO2 emissions only which could even provide further co-benefits for carbon sequestration in soil and biomass through revegetation and afforestation12.

Discussion

We identified the mix of mitigation strategies that are cost-effective at a given carbon price across regions with developed regions predominantly employing technical options while structural changes through transition toward more intensive but GHG efficient agricultural production systems are projected to be the main source of emission reductions in developing regions. To realize the mitigation potentials presented in this study, several adoption barriers such as lack of education and infrastructure, poor access to markets or land tenure insecurity2 will still have to overcome, which will require immediate attention by policy makers. Educating farmers about the positive impacts of, e.g., increasing nitrogen use efficiency or conservation tillage, on individual farmers’ welfare, could speed up the adoption of these mitigation practices26,27. Even though results imply positive synergies of non-CO2 mitigation for CO2 emissions from land use change, further research is needed to consider explicitly the impact on soil carbon given its importance for the carbon cycle15. For example, Sanderman et al.28 found a strong correlation between soil carbon loss and agricultural land degradation and restoration may enable co-benefits for CO2 sequestration and non-CO2 mitigation in certain areas15,29.

Data availability

The authors declare that the main data supporting the findings of this study are available within the article and the supplementary information. Additional data are available upon request from the authors.