Agriculture, bioenergy, and water implications of constrained cereal trade and climate change impacts

systems and its interaction with the trade constraints, rather than to quantify a comprehen- sive range of potential global climate change impacts. Like the global zero cereal import dependency constraint, this should be considered as a thought experiment for exploring the multi-sector dynamics, although our model can be readily applicable to other scenarios. In total, we explore four model scenarios considering the impact of global zero cereal import dependency and climate change. The four scenarios are: 1. Reference (RF), which is the GCAM baseline model outputs run in 5-year time steps up to 2050; 2. Import Dependency (ID), where zero cereal import dependency constraint is imposed for all GCAM regions globally upon the baseline, starting in 2025; 3. Climate Change (CC), where the climate change impact on crop yield for all GCAM crops across all GCAM regions is imposed upon the baseline beginning in 2020; 4. Import Dependency under Climate Change (IC), which incorporated both ID and CC to study the model responses under the influence of both global zero cereal import depen- dency and climate change. GCAM solves for all periods and all scenarios. That is, every market across different scales (e.g., regional market and global market) in the model is solved so that supply and demand are equal. 3 Results 3.1 Cereal Import Dependency Ratio (CIDR) Fig 1a shows the 2015 (base year) global pattern of CIDR. The CIDR changes between 2015 and 2050 are within ± 14 percentage points for all regions except Brazil and Australia_NZ

(a GCAM aggregate region of Australia and New Zealead; S2 Table in S1 File), neither of which depends on cereal imports and have large decreases in their cereal import depen- dency over the period (S1 Fig in S1 File). In general, regions in South America are projected to have decreasing cereal import dependency over time, while regions in Europe are pro- jected to increase their dependence on imported cereals in 2050, relative to 2015 (S1 Fig in S1 File). In 2015, the largest cereal consumers are China, USA, and India (Fig 1b). They are also the top cereal producers, with China having the highest CIDR of the three at 2%, and USA and India showing negative CIDR. Regions at the higher end of CIDR include Japan and the Mid- dle East (Fig 1c). A table with more information on major cereal importers and exporters is included in the SI (S3 Table in S1 File). 3.2 Global results In ID, every region is constrained to cereal import dependency less than or equal to zero beginning in 2025. As a result, no region’s cereal imports exceed exports, with all regions hav- ing a CIDR of zero in ID and IC. Global trade of cereals decreases under these scenarios (Fig 2a). In 2025, global total import of cereals in ID and IC decreases to nearly half of the trade quantity in RF, with larger decreases in later periods (Fig 2a). Relative to the effects of ID, the change in global total cereal import quantity due to climate change impacts on agricultural yields, both CC and IC, is very small (Fig 2a). Despite large decreases in cereal trade in ID, total cereal production and consumption

decrease only 3.1%-4.1% over the 2025–2050 period, while climate change accelerates the PLOS ONE Agriculture, bioenergy, and water implications of constrained cereal trade and climate change impacts PLOS ONE | https://doi.org/10.1371/journal.pone.0291577 September 15, 2023 5 / 18