National climate change impact assessments underestimate the potential of autonomous adaptation

1 Length of the manuscript: 7296 words + 1500 words (Figures) + 300 words (Table) 1 2 National climate change impact assessments underestimate the 3 potential of autonomous adaptation 4 5 Juliana Arbelaez-Gaviria1,2,3 6 [email protected] 7 Department of Agrosystems and Bioclimatology 8 Zemědělská 1, 61300 Brno - budova A 9 +420 774 510 248 10 11 Esther Boere4,3 12 [email protected] 13 Miroslav Trnka1,2 14 [email protected] 15 Petr Havlík3 16 [email protected] 17 Ian P. Holman1,5 18 [email protected] 19 Paula A. Harrison6 20 [email protected] 21 22 1Global Change Research Institute of the Czech Academy of Sciences, Brno, Czech 23 Republic 24 2Mendel University in Brno, Institute of Agrosystems and Bioclimatology, Brno, Czech 25 Republic 26 3International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria 27 4Institute for Environmental Studies (IVM), VU University Amsterdam, Amsterdam, The 28 Netherlands 29 5Centre for Water, Environment and Development, Cranfield University, Cranfield, UK 30 6Environmental Change Institute, University of Oxford, Oxford, UK 31 32 2 Abstract 33 34 Central Europe is projected to lose up to 25% of its crop productivity by 2050 due to 35 climate change, posing significant challenges to agricultural systems and food security. 36 Effective adaptation strategies must consider not only domestic impacts but also global 37 climate effects, including international trade dynamics. We performed a multilevel analysis 38 of climate change impacts on agriculture, using the Czech Republic, a landlocked, crop 39 production-based economy with an open market, as a case study. We integrated the 40 global biosphere management model (GLOBIOM) with the gridded global crop model 41 EPIC-IIASA. Climate impacts were projected with five global circulation models under 42 three climate scenarios, with and without CO₂ fertilization, and applied in national, EU43 regional, and global productivity change scenarios. Results show that national-only 44 assessments underestimate both risks and opportunities: production is projected to 45 decline by up to 9% when global interactions are excluded, but to increase by up to 8% 46 when trade and market effects are included. Autonomous adaptation mechanisms, such 47 as cropland reallocation, shifts in management intensity, and trade adjustments, buffer 48 biophysical yield losses and improve economic outcomes. Neglecting global interactions 49 in national climate change assessments increases the risk of maladaptation and policy 50 inefficiencies. Incorporating international market linkages enhances the ability to design 51 robust adaptation strategies, enabling countries such as the Czech Republic to maximize 52 resilience while minimizing environmental and socioeconomic trade-offs. 53 54 Keywords 55 56 Climate change, Adaptation, Agricultural trade, Czech Republic, Integrated assessment, 57 Food security 58 59 Introduction 60 61 Climate change is expected to pose substantial agricultural challenges in Central Europe, 62 with potential crop productivity declines of up to 25% by mid-century. (Pörtner et al. 2022). 63 Maize yields can decline by as much as 25%, whereas wheat losses may reach 15%, 64 depending on the extent of CO₂ fertilization between 2040 and 2069 (Eitzinger et al. 2013; 65 Webber et al. 2018). Robust adaptation strategies must be developed to address these 66 anticipated agricultural losses. A key question is whether focusing solely on the direct, 67 national impacts of climate change provides a sufficient foundation for planning and 68 decision-making (Ercin et al. 2021). While national climate change assessments are 69 crucial for designing adaptation strategies, the global nature and interconnectedness of 70 climate change effects and agricultural markets may significantly influence national 71 resilience and adaptation efforts (Ercin et al. 2019). Ignoring the effects of climate change 72 on global agricultural production and international trade when developing national 73 adaptation strategies increases the likelihood of maladaptation; as such, assessments 74 risk underestimating or overestimating the impacts of national climate change (Pörtner et 75 al. 2022). 76 77 Key players in the global agricultural market have relied on national and global agricultural 78 models to assess impacts and develop adaptation plans for instance the United States 79 3 (e.g., Baker et al., 2018), Brazil (e.g., Zilli et al., 2020), and the European Union (EU) (e.g., 80 Blanco et al., 2017). Moreover, 18 countries have started incorporating global frameworks 81 to better understand and address challenges in the agricultural sector and associated 82 linkages to climate mitigation and adaptation, for example, via the FABLE 83 consortium (FABLE 2019). However, in the case of the Czech Republic, climate change 84 impact assessments are based predominantly on countryor region-scale modeling 85 approaches. Although the Czech Republic does not play a dominant role in the global 86 agricultural market, its agricultural sector remains an essential component of the national 87 economy and rural livelihoods (Prochazka et al. 2023). Furthermore, changes in crop 88 suitability and agricultural area expansion may position it as a more significant regional 89 player within the EU at constant food demand (Papadimitriou et al. 2019). 90 91 Czech agriculture is a cereal-based sector, where wheat production represented the 62% 92 of the cereal production in 2024, following by barley with 22% and maize 9% (CZSO 93 2025). The Czech Republic’s agricultural trade is strongly oriented toward the European 94 Union, with roughly four-fifths of all exports directed to EU Member States and only a 95 minor share reaching markets outside the EU (Zábojníková and Kamenický 2024) 96 Germany remains its key trading partner. Despite cereals accounting for more than half 97 of domestic agricultural output, the Czech Republic overall is a net importer of agricultural 98 products (Zábojníková and Kamenický 2024). The impacts of climate change on Czech 99 agriculture have been extensively studied via biophysical models focused on single 100 commodities such as maize (Pavlik et al. 2019), barley, and wheat (Trnka et al. 2004a; 101 Thaler et al. 2012; Eitzinger et al. 2013), as well as livestock (Potopová et al. 2023) and 102 provisioning ecosystem services (Lorencová et al. 2013). Some studies have also 103 modeled multiple key crops (Hlavinka et al. 2015; Pohanková et al. 2022; Pohanková et 104 al. 2024) or analyzed agroclimatic indicators (Eitzinger et al. 2013) at specific sites. 105 Papadimitriou et al. (2019) incorporated transnational market interactions into 106 assessments of the Czech Republic, and a European-scale model was used to simulate 107 variations in imports and exports on the basis of shared socioeconomic pathways (O’Neill 108 et al. 2014). Potopová et al. 2023 used climate projections to determine the future water 109 consumption of livestock in the country and more recently, (Poláková et al. 2025) 110 integrated the feedback loop from local to global by integrating national yield response 111 into a general equilibrium model. Despite their contributions, these studies share common 112 limitations. First, they fail to capture climate change impacts outside their spatial domains, 113 restricting the ability to assess the Czech Republic’s relative competitiveness within the 114 EU. Second, they omit or aggregate agricultural market dynamics beyond Europe, such 115 as international trade, leading to a biased understanding of the country’s autonomous 116 adaptation potential. 117 118 Building on the approaches of Baker et al. (2018) and Papadimitriou et al. (2019), a 119 comprehensive, multilevel framework for assessing the Czech agricultural sector’s 120 autonomous adaptation response to climate change is proposed in this study. We 121 hypothesize that global climate change impacts and international market dynamics play 122 critical roles in shaping the effectiveness of national adaptation strategies. Specifically, 123 the Czech agricultural sector’s adaptation potential is evaluated by integrating national, 124 regional, and global impacts through a combination of globally consistent models: a partial 125 equilibrium model of agriculture and forestry and a gridded global crop model. By 126 4 employing this approach, we aim to provide context-specific insights into the interactions 127 between national and global factors, laying the foundation for more robust adaptation 128 strategies and policies. While the primary focus is on climate change impacts on Czech 129 agricultural indicators—such as production, consumption, and prices—scenarios 130 encompassing direct climate change impacts on Czech agriculture are compared with 131 scenarios that incorporate indirect effects through regional and global productivity 132 changes. Unlike previous studies, this research explicitly captures interactions between 133 the Czech Republic and the rest of the world, offering a comprehensive understanding of 134 systemic vulnerabilities and adaptation opportunities. 135 136 Methods 137 138 We apply the global biosphere management model (GLOBIOM) (Havlík et al. 2014), a 139 partial equilibrium model that represents the global agriculture, forestry, and bioenergy 140 sectors. The model was enhanced for the European Union by incorporating updated land 141 cover and land use information from the best available European datasets (Frank et al. 142 2015). Commodity markets and international trade are represented for 57 economic 143 regions, one for each EU member state and the UK, and 29 additional regions outside the 144 EU. Within each region, a representative consumer optimizes consumption on the basis 145 of preferences and commodity prices, while producers maximize margins, and GLOBIOM 146 is used to solve for the market equilibrium scheme that achieves overall welfare 147 maximization. 148 149 The supply side of the model follows a bottom-up approach using detailed spatial data for 150 land cover, land use, management systems, and biophysical and technical costs. 151 Environmental impacts such as greenhouse gas and nutrient emissions are also 152 integrated into the model. The EU28 is represented at the NUTS2 level, ensuring fine153 scale detail. Crop, livestock, and forest production activities are considered via biophysical 154 modeling frameworks. Primary forest productivity and harvesting costs are estimated via 155 the Global Forest Model (G4M) (Kindermann et al. 2008). EPIC is used to compute crop 156 productivity, fertilizer requirements, and irrigation management practices. The European 157 crop sector is modeled via crop rotations for 18 key crops, derived from EUROSTAT 158 statistics at the NUTS2 level, with the CropRota model (Schönhart et al. 2011). The 159 livestock sector and its production system parameters are modeled with the RUMINANT 160 model (Herrero et al. 2013). Six dynamically modeled land-use types (cropland, 161 grassland, short-rotation tree plantations, managed forests, natural forests, and other 162 natural land) can be converted on the basis of the demand and profitability of land-based 163 activities. Within Europe, no deforestation for agricultural expansion is assumed due to 164 restrictive land-use legislation (Bauer et al. 2004). Additional information about the global 165 and European versions of GLOBIOM were presented by Havlík et al. (2014) and Frank et 166 al. (2015), respectively. 167 168 GLOBIOM has been widely applied to assess climate change impacts and mitigation 169 pathways at the global (Nelson et al. 2014; Hasegawa et al. 2018; Fujimori et al. 2019) 170 and EU level, including the recent impact assessment of the European Commission’s Fit171 for-55 package (EC 2021). Unlike models that aggregate countries into broader regional 172 5 blocks, GLOBIOM explicitly represents market relationships among EU Member States, 173 making it particularly suitable for national-scale analyses (Frank et al. 2015) . It is also 174 included among the IPCC’s Integrated Assessment Models (IAMs), where it complements 175 the MESSAGE model by representing the land-based mitigation sector (Krey et al. 2020) 176 and climate impacts (Awais et al. 2024) . Beyond food production, GLOBIOM incorporates 177 land competition with forestry as well as demand for feed and bioenergy (Havlík et al. 178 2011; Havlík et al. 2014), enabling analysis of cross-sectoral trade-offs and co-benefits. 179 Its detailed representation of agricultural commodities, including wheat, barley, and maize, 180 provides a robust basis for evaluating cereal-based agricultural systems such as those in 181 the Czech Republic. 182 183 Autonomous adaptation 184 185 Climate change adaptation refers to “the process of adjustment to actual or expected 186 climate and its effects” (Pörtner et al. 2022). The adjustment can be explicitly planned or 187 occur spontaneously, triggered by farmers or market changes as a response to climate 188 change—referred to as autonomous adaptation (Pörtner et al. 2022; Maskell et al. 2025). 189 In GLOBIOM, autonomous adaptation to climate-induced changes in crop yields can be 190 explored through adjustments in production, consumption, and trade patterns. Supply191 side adaptation occurs through land reallocation by expanding cropland into other land 192 cover types, altering crop shares at the national level, or shifting between low-input and 193 high-input management systems (Leclère et al. 2014). Consumers adapt by modifying 194 both the quantity and structure of food consumption on the basis of price signals (Mosnier 195 et al. 2014). International trade serves as another crucial adaptation mechanism. Climate196 induced changes in productivity may shift comparative advantages across regions, 197 enabling trade to redistribute surplus production from favorable regions to deficit regions 198 (Janssens et al. 2020). In GLOBIOM, economic regions adjust trade quantities and trading 199 partnerships to buffer productivity shocks and maintain market balance. 200 201 Scenario design 202 203 The global gridded crop model EPIC-IIASA (Balkovič et al. 2013) was run in conjunction 204 with five distinct global circulation models (GCMs) from the Coupled Model 205 Intercomparison Project Phase 6 (CMIP6) (O’Neill et al. 2016; Jägermeyr et al. 2021). 206 The climate scenarios considered in this study were obtained from the Inter-Sectoral 207 Impact Model Intercomparison Project (ISI-MIP) and its latest protocol, ISIMIP3b—both 208 subsets of CMIP6 (Eyring et al. 2016). ISI-MIP provides consistent projections for 209 evaluating climate impacts on agriculture. For the biophysical climate change impacts on 210 productivity, we used three climate scenarios from the most recent CMIP6 ensemble, 211 combining Shared Socioeconomic Pathways (SSPs) and Representative Concentration 212 Pathways (RCPs): SSP1–2.6, SSP3–7.0, and SSP5–8.5 (Gidden et al., 2019), simulated 213 with five general circulation models (GCMs): GFDL-ESM4, IPSL-CM6A-LR, MPI-ESM1214 2-HR, MRI-ESM2-0, and UKESM1-0-LL. Supplementary Table 1 provides further details 215 about each model. The impacts of climate on crop productivity were estimated via EPIC216 IIASA for four key crops (maize, rice, soy, and wheat). Productivity for 17 additional crops 217 (e.g., barley, silage maize, cotton, and sugar beet) was computed on the basis of their 218 C3/C4 photosynthesis pathways, following the approach of Janssens et al. (2020) 219 6 (Supplementary Table 2). The EPIC-IIASA projections were available at a 0.5 x 0.5 220 degrees resolution and upscaled to 2 × 2 degrees cells, matching the resolution of 221 GLOBIOM’s land units, using a weighted average based on the respective crop areas in 222 the year 2000. As GLOBIOM explicitly accounts for societal changes in land-based 223 sectors, we used SSP2 as the non-climate-change baseline, enabling us to isolate the 224 effects of climate from those of socio-economic change and to examine the autonomous 225 adaptation response in the absence of mitigation policies (O’Neill et al. 2016). Livestock 226 impacts were modeled indirectly through changes in feed production rather than explicit 227 productivity impacts. In the baseline scenario without climate change, exogenous yield 228 improvements originated solely from long-term technological developments. 229 230 To simulate differences in national, regional, and global climate impacts, each RCP-GCM231 CO₂ scenario was first analyzed considering national impacts, followed by regional and 232 global impacts. In the national scenarios, productivity changes were applied only to Czech 233 production systems, keeping yields elsewhere consistent with socioeconomic 234 assumptions. In the regional scenarios, changes extended to the EU27, including the UK. 235 Both scenarios accounted for endogenous changes in global productivity and market 236 interactions. In the global scenario, productivity impacts were applied across all the 237 regions modeled (Table 1.). This scenario design isolates the effects of countryand 238 region-scale assessments from global-scale impacts, enabling comparisons of climate239 induced productivity changes. All the scenarios included the same autonomous 240 adaptation options, although economically optimal adaptations differed on the basis of 241 whether national, regional, or global effects were modeled. A comparison of the results 242 across these scenarios revealed key differences in national crop production patterns and 243 autonomous adaptation responses across market indicators. 244 245 Climate Impact Scenario Regional Extent Rationale Climate Scenarios GCMs National Czech Republic Climate impacts are applied to crop productivity in the Czech Republic. The rest of the world retains SSP2 productivity levels. SSP1–2.6 w/o CO₂ SSP1–2.6 w/ CO₂ SSP3–7.0 w/o CO₂ SSP3–7.0 w/ CO₂ SSP5–8.5 w/o CO₂ SSP5–8.5 w/ CO₂ GFDL-ESM4 IPSL-CM6ALR MPI-ESM1-2HR MRI-ESM2-0 UKESM1-0LL Regional EU27 + UK Climate impacts are applied to crop productivity in the EU27 and the UK. The rest of the world retains SSP2 productivity levels. Global World Climate impacts are applied to global crop productivity. 7 Table 1. Climate impact scenarios assessed in this study, showing the regional extent, rationale, climate scenarios, 246 and CMIP6 general circulation models (GCMs) used to evaluate biophysical climate change impacts on crop 247 productivity. 248 249 250 Results 251 252 Our results focus on projected relative changes to the no-climate-change scenario for 253 different agricultural indicators in the Czech Republic, the EU28, and globally by 2050. 254 The biophysical effects of climate change on yields (see Figure 1a) vary from -27% to 6% 255 across crops, scenarios, and climate models. Compared with the maize yield, the wheat 256 yield declines less severely in these scenarios, ranging from -5% (RCP 8.5) to 6% (RCP 257 7.0), whereas the maize yield declines by -22% (RCP 8.5) to 5% (RCP 2.6). Overall, the 258 effects of climate change on crop yields in the Czech Republic follow a pattern similar to 259 that observed for wheat (−5% to +10%), reflecting the dominance of C3 crops in Czech 260 agricultural production. Wheat (C3) and maize (C4) both show the largest yield declines 261 under the high-emission scenario RCP8.5 without the CO₂ fertilization effect. The wheat 262 and maize yields decrease from -12% to 2% and from -27% to 3%, respectively, when 263 CO2 fertilization effects are neglected. UKESM1-0-LL consistently projects the most 264 negative impacts, while GFDL-ESM4 shows the most positive impacts across crops and 265 climate scenarios. The large variation among GCMs can be attributed to differences in 266 their CO₂ concentration pathways in the CMIP6 experiment and their climate sensitivity. 267 By the end of the century, UKESM1-0-LL registers the highest temperature increase, 268 whereas GFDL-ESM4 shows the lowest across all climate scenarios (Jägermeyr et al. 269 2021). 270 271 272 8 273 Fig. 1. Impacts of climate change on crop yields and agricultural indicators in 2050 in the Czech Republic under 274 alternative climate and impact scenarios. (a) Biophysical yield changes relative to a no-climate-change baseline (%) 275 simulated by EPIC-IIASA for wheat, maize and aggregated crops under SSP1-2.6, SSP3-7.0 and SSP5-8.5, each with 276 and without CO ₂ fertilization. Bars show multi-model means (CMIP6 ensemble average) and symbols denote individual 277 general circulation models (GFDL-ESM4, IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, UKESM1-0-LL). (b) 278 Changes (%) in sectoral indicators, yield, harvested area, production, consumption, prices and net trade, under national, 279 regional (EU) and global climate-impact contexts. Boxplots indicate interquartile ranges, whiskers show ranges 280 excluding outliers, black lines denote medians and filled dots mark means. Results aggregate 30 climate scenarios (n 281 = 30) 282 9 283 Economic response to the effects of climate change in the Czech Republic 284 285 Figure 1b shows an overview of how the biophysical effects of climate change on yields 286 propagate across agricultural indicators under different climate impact scenarios. 287 GLOBIOM transfers the initial effect on yields to the response of agricultural indicators via 288 supply and demand adjustments to agricultural production in the country. Markets, 289 production, and consumption patterns adjust to the assumed yield and trade conditions, 290 with the goal of maximizing total economic surplus by 2050 globally, including in the Czech 291 Republic. Producers respond to climate change primarily through agricultural area 292 expansion, which averages 3%, rather than intensified management practices (-0.08%) 293 (see Figure 1b, global climate impact scenario and the olive-green point). The area 294 changes remain within ±5% across all scenario dimensions, reaching up to 14% under 295 the most extreme scenario (RCP 8.5 without CO2 fertilization effects). In contrast, yield 296 changes are small, are centered mostly at approximately 0% and show limited variability 297 across scenarios. The combined area increase and stable yield result in a mean 298 production increase of 3%, with changes within ±10%. Consumption in the Czech 299 Republic is stable across most scenarios, and only a small change (3%) is expected under 300 RCP 8.5 without CO2 fertilization effects. Net trade (calculated as exports minus imports) 301 displays the largest variation of all the indicators, with changes ranging from -8% to 27%, 302 depending on the scenario. On average, net trade increases 10%, indicating a net export 303 surplus in response to climate change effects. 304 305 To investigate the drivers of heterogeneity in the selected indicators induced by climate 306 change, an in-depth analysis with GLOBIOM is needed. Figure 2 shows the univariate 307 regression lines of the selected indicators plotted against the biophysical effect on yields. 308 The slope coefficient reflects the local response, and it can be understood as the ability of 309 a variable to change, interpreted as adaptive capacity. A value of 1 can be interpreted as 310 a percentage change in the impact of climate change on yields in response to an 311 equivalent percentage change in a given indicator. The intercept coefficient can be 312 interpreted as a local change driven by indirect climate change effects and price effects 313 transmitted by international markets. An intercept value other than 0 suggests that local 314 changes arise from effects in other regions transmitted via price effects through 315 international trade (Nelson et al. 2014). The slope and intercept coefficients for each 316 climate impact scenario are also reported in Supplementary Table 3. 317 318 Yield in the Czech Republic appears unresponsive in terms of productivity management. 319 With a slope close to 1 and an intercept close to 0, there is no additional compensation 320 through management for climate change impacts on yield (see Figure 2c). The yield 321 shows a slight local effect on crop reallocation between C3 and C4 crops due to differential 322 climate change impacts (Supplementary Table 5). The area shows a negative relationship 323 between biophysical productivity and area, indicating a strong response in the Czech 324 Republic. The area of specific crops is expected to decrease as the productivity of the 325 crops increases. Climate change has led to a decrease in the cultivated areas of most 326 impacted crops, and losses in production have been offset by imports from more favorable 327 areas in the EU28. The same inverse relationship and market reallocation trend are shown 328 16 516 Fig. 5 Projected bilateral trade flows of crops aggregated in 2050 under the RCP8.5 scenario (in million tons) 517 across four climate impact scenarios. (a) No climate change, (b) National impact, (c) Regional impact, and (d) Global 518 impact. The colors represent regions, with the Czech Republic (green) as the focal area. The thickness of the connecting 519 lines indicates the trade volume. 520 Czech Republic remains among the top-performing countries in terms of this crop. Austria, 521 Germany, and Italy are identified as the largest importers of Czech crop commodities. 522 In contrast, Czech imports from neighboring countries in east-central Europe range 523 between 1.8 and 1.9 million tons in the no climate change scenario and global climate 524 17 impact scenarios, respectively. The RCA values of potatoes and maize are low, with the 525 value for potatoes below 1 for the Czech Republic and that for maize lower than those for 526 leading key players in the EU28, such as Germany, Slovakia, Hungary, Slovenia, and 527 Romania. The Czech Republic lags behind both the EU28 and global RCA levels, 528 highlighting its limited competitiveness in terms of maize production, with a value of 529 approximately 1. A comparative disadvantage is projected for potatoes in the Czech 530 Republic, where countries such as Poland and Belgium exhibit strong comparative 531 advantages both in the EU28 and globally. Slovakia and the Netherlands are the leading 532 exporters to the Czech Republic, where maize and potatoes represent 75% of the total 533 imports. When climate change impacts are isolated to the Czech Republic, the country is 534 projected to decrease the total export of crops by 6% and increase the total import of crops 535 by 5% compared with that in the no climate change scenario. When climate change 536 impacts are applied globally, the total export of crops increases by 6%, with total imports 537 projected to decrease by 5% by mid-century. Supplementary Tables 5 and 6 show the 538 values for each commodity. 539 540 Comparison with global and European results 541 Compared with the European Union, the Czech Republic faces more considerable 542 projected biophysical yield reductions compared to the global average. The EU28 region 543 is projected to experience comparatively smaller reductions, mostly between –5% and – 544 15%, while the global impacts are expected to be even less severe, typically between – 545 5% and –10% (Supplementary Figs. 2–4, 6). Projected biophysical yields for wheat are 546 more resilient than those for maize, ranging from –15% to 1% in the EU28 and from –10% 547 to 3% globally, although variability increases under high-emission scenarios (RCP8.5) 548 without CO₂ fertilization, for which the Czech Republic decline is up to –12%. The average 549 biophysical yield changes for the EU28 mask the heterogeneous impacts of climate 550 change among countries. As the level of warming increases from SSP1-2.6 to SSP5-8.5, 551 yield impacts become more extreme, particularly for some southern and eastern 552 European countries. Compared with other EU28 countries, the Czech Republic 553 experiences relatively modest yield changes, similar to those in other Central European 554 countries, such as Poland (Supplementary Fig. 6). The negative extreme effects are 555 dominated by C4 crops such as maize, with negative effects also expected in most other 556 EU28 countries, with the most severely affected countries being Italy, France, Croatia, 557 and Slovenia. In contrast, wheat shows a mixed pattern of impacts, with positive effects 558 in some western and east-central European countries and negative effects in some 559 northern and eastern European countries (Supplementary Figs. 3 and 6). 560 The larger reductions in the EU28 compared with the global average can be explained by 561 both climatic and structural factors. Climatic conditions in southern and eastern Europe 562 amplify negative impacts, while more temperate regions in central and western Europe 563 sometimes benefit. At the global level, however, trade reallocation across continents helps 564 buffer production losses, dampening the overall average. Crop type sensitivity further 565 explains the differences: C4 crops such as maize respond more negatively to high 566 temperatures, whereas C3 crops such as wheat display a broader range of outcomes. 567 568 18 Supplementary Figs. 8–10 show the economic responses to the effects of climate change 569 globally and in the EU28 under regional and global scenarios. Like the Czech Republic, 570 the EU28 is projected to experience more variability and severe impacts than those 571 observed globally. In contrast, price fluctuations stand out, with changes ranging from – 572 7.5% to +4.5% regionally and –5.0% to +3.2% globally. The consumer response becomes 573 more relevant at the EU28 level than at the Czech Republic level, with changes ranging 574 from –4% to 2% under regional impact scenarios and from –2.8% to +1.3% in the global 575 impact scenario. Yield and production are positively correlated with biophysical yields, 576 which are driven by climate impacts in the EU28 and market interactions. An inverse 577 relationship is observed for the EU28, as for the Czech Republic. Both climate and market 578 effects are greater in the global impact scenario than in the regional impact scenario. 579 The overall effect of climate change impacts in the EU28 is a decrease in production 580 despite shifts in management systems and area expansion. Changes in global indicators 581 are relatively minor, remaining mostly within ±1%, except for prices, which display greater 582 variability, reaching up to 3%. Globally, trade adjustments and production compensate for 583 the regional effects of climate change, yet crop prices are expected to surge. As in the 584 EU28 and the Czech Republic, yield and production have a positive relationship with 585 biophysical yields, whereas area has an inverse relationship. Strong reallocation patterns 586 across regions, both in terms of management productivity and less so in terms of area, 587 help buffer losses in production due to climate change impacts globally. 588 Discussion 589 We applied a multilevel framework using two globally consistent models, GLOBIOM and 590 EPIC-IIASA, to evaluate how climate change affects Czech agriculture and how the 591 country responds through autonomous adaptation. Unlike earlier Czech studies such as 592 Pohanková et al. (2022), which focused on biophysical outputs for specific crop rotations, 593 and other field-based projections, our approach links biophysical yield impacts with 594 economic and trade dynamics at the national scale. This allows us to move beyond single595 crop or site-level insights (e.g., Hlavinka et al., 2015) and align our analysis with broader 596 global frameworks. While previous intercomparison studies such as Jägermeyr et al. 597 (2021) established similar patterns globally, our study uniquely traces how yield changes 598 in the Czech Republic are transmitted through international markets to shape production, 599 trade, and competitiveness. Importantly, we incorporate the latest CMIP6 projections (Gier 600 et al., 2024), providing a more realistic representation of carbon–nitrogen interactions and 601 land-use dynamics. Our contribution also complements emerging protocols that explicitly 602 link local processes with global dynamics, such as the framework of Poláková et al. 603 (2025). In this context, our study demonstrates the novelty of situating Czech agriculture 604 within a multilevel framework, revealing adaptation opportunities and risks that remain 605 hidden in national-only assessments. 606 Our projections show that maize is more vulnerable to climate stress than wheat, 607 particularly under high-emission scenarios, consistent with Eitzinger et al. (2013) and 608 Trnka et al. (2018). This aligns with Pohanková et al. (2022) and Muench et al. (2024), 609 who emphasized that potential yield gains depend on management practices and farmer 610 adoption of adaptation. We also find that national-scale assessments may overestimate 611 19 local yield shocks while underestimating the buffering role of trade. Similar outcomes have 612 been observed in Brazil (Zilli et al., 2020), Gambia (Carr et al., 2024), the UK (Challinor et 613 al. 2016), and Ireland (Adenaeuer et al., 2023). Importantly, our findings highlight 614 transnational climate risks: yield shocks abroad propagate through trade and prices to 615 influence Czech production and competitiveness. This echoes studies on Europe’s cross616 border vulnerabilities showing that droughts or losses outside the EU can significantly 617 affect its food security and economy (Ercin et al., 2019; Ercin et al., 2021). 618 Land expansion emerged as the dominant autonomous adaptation strategy, especially 619 under global scenarios where price signals are transmitted via trade. This reflects Czech 620 Republic’s relatively favorable land base, which is less affected by drought than in 621 neighboring countries (Eitzinger et al., 2013). Yet the scope for expansion is limited: under 622 the Common Agricultural Policy (CAP), conversion of permanent grassland is prohibited 623 in protected areas and heavily restricted elsewhere (Ministry of Agriculture of the Czech 624 Republic 2022), and expansion would carry environmental costs including biodiversity 625 loss, greenhouse gas emissions, and reduced ecosystem services (Lorencová et al., 626 2013; Papadimitriou et al., 2019). Thus, while land expansion provides an immediate 627 buffer against yield shocks, it is unlikely to be sustainable. In practice, adaptation will rely 628 more on reallocating within existing arable land, maintaining ecological areas, adopting 629 soil-conserving practices, and adjusting trade. 630 Reliance on a narrow set of commodities also increases vulnerability. Wheat, barley, and 631 rapeseed dominate Czech exports (e.g., beer, feed), and while they are well captured in 632 GLOBIOM, the model does not differentiate organic versus conventional farming. This is 633 important as organic farming is projected to reach 21% of land by 2028, and policy 634 measures aim to strengthen fruit, vegetable, hops, wine, and apiculture sectors (Ministry 635 of Agriculture of the Czech Republic 2022). Planned adaptation is therefore being 636 reoriented toward soil, water, and biodiversity outcomes while sustaining competitiveness. 637 Crop diversification, combined with sustainable intensification, should complement land 638 expansion to enhance resilience and long-term competitiveness. 639 The implications of our results extend beyond production to the policy and institutional 640 dimensions of adaptation. CAP regulations protect grasslands, wetlands, and ecological 641 features, making large-scale expansion legally and economically difficult (Ministry of 642 Agriculture of the Czech Republic 2022). Farmers also face financial and practical 643 barriers, such as high upfront investments, uneven advisory support, and uncertainty over 644 climate and markets. Consequently, realistic adaptation pathways in the Czech Republic 645 will depend on CAP-compatible strategies such as reallocating existing arable land, 646 adopting soil-conserving practices, and diversifying into resilient or higher-value crops. At 647 the EU scale, production reallocation among member states helps buffer localized shocks 648 but generates distributional consequences across regions. Globally, trade integration 649 stabilizes supply and prices but exposes small open economies like the Czech Republic 650 to risks from regulatory mismatches, or sudden disruptions. National-only assessments 651 that ignore these dynamics risk maladaptation by overstating self-sufficiency and 652 underestimating the benefits and trade-offs of global integration. By embedding Czech 653 agriculture in a multilevel framework, our study shows that effective adaptation requires 654 20 attention to both domestic and transnational dimensions, providing a stronger foundation 655 for policies that enhance resilience while minimizing unintended trade-offs. 656 Conclusion 657 658 Our study contributes to the growing body of research that moves beyond isolated yield 659 projections toward systemic, multiscale assessments of agricultural resilience. By 660 situating Czech agriculture within a trade-mediated global context and complementing 661 recent advances in local-to-global modeling, we provide a novel perspective that better 662 captures both the opportunities and risks of autonomous adaptation. The results highlight 663 the importance of integrating global agricultural impacts and trade dynamics into national 664 climate change assessments. Accounting for international market interactions reveals a 665 greater adaptive capacity for the Czech Republic than suggested by national-only 666 analyses, particularly through trade-driven responses and land-use reallocation. However, 667 heavy reliance on land expansion raises sustainability concerns, underscoring the need 668 for policies that balance adaptation with mitigation and environmental protection. These 669 findings reinforce the value of multiscale approaches for informing robust adaptation 670 planning. Policymakers should prioritize strategies that leverage trade and market 671 responses while advancing sustainable intensification and resource-efficient practices. 672 Overemphasis on self-sufficiency risks underestimating adaptation potential and 673 increasing vulnerability, whereas trade-based strategies can buffer national shocks, 674 enhance resilience, and optimize resource use. For small, open economies such as the 675 Czech Republic, recognizing the interplay between domestic responses and transnational 676 climate risks will be critical for achieving sustainable and effective adaptation. 677 678 Acknowledgments 679 680 We extend our gratitude to the Integrated Biosphere Futures (IBF) group at the 681 International Institute for Applied Systems Analysis (IIASA) for their invaluable support 682 and expertise, which significantly contributed to the development of this study. We also 683 wish to honor the memory of Hind Rajab and Professor Refaat Alareer. Their lives and 684 work continue to inspire us to pursue research that advances understanding, equity, and 685 resilience in the face of global challenges. 686 687 Funding 688 689 This study was financially supported by Zhodnocení vlivu klimatických změn na 690 zemědělství ve Střední a Východní Evropě v kontextu globálních podmínek (SP2210041 691 - AF-IGA2021-IP015) and by the Ministry of Education, Youth and Sports of the Czech 692 Republic (grant AdAgriF - Advanced methods of greenhouse gases emission reduction 693 and sequestration in agriculture and forest landscape for climate change mitigation 694 (CZ.02.01.01/00/22_008/0004635). 695 696 Author contributions 697 698 21 JAG, PH and MT developed the conceptual framework; JAG and EB developed the 699 scenarios, methodological framework and improved model (GLOBIOM). JAG wrote the 700 initial manuscript and performed the data analysis. MT, PH, EB, IPH and PAH edited 701 and commented on the manuscript. IPH and PAH contributed to the discussion with 702 JAG. MT supervised the project. 703 704 Conflict of interest 705 The authors declare that they have no conflicts of interest. 706 707 Data availability 708 The code and data used in the scenario analyses are available from the corresponding 709 author upon request. 710 711 22 References 712 Awais M, Vinca A, Byers E, Frank S, Fricko O, Boere E, Burek P, Poblete Cazenave M, 713 Kishimoto PN, Mastrucci A, Satoh Y, Palazzo A, McPherson M, Riahi K, Krey V 714 (2024) MESSAGEix-GLOBIOM nexus module: integrating water sector and 715 climate impacts. Geosci Model Dev 17(6):2447–2469. 716 https://doi.org/10.5194/gmd-17-2447-2024 717 Baker JS, Havlík P, Beach R, Leclère D, Schmid E, Valin H, Cole J, Creason J, Ohrel S, 718 McFarland J (2018) Evaluating the effects of climate change on US agricultural 719 systems: sensitivity to regional impact and trade expansion scenarios. Environ 720 Res Lett 13(6):064019. https://doi.org/10.1088/1748-9326/aac1c2 721 Balassa B (1965) Trade Liberalisation and “Revealed” Comparative Advantage. Manch 722 Sch 33(2):99–123. https://doi.org/10.1111/j.1467-9957.1965.tb00050.x 723 Balkovič J, van der Velde M, Schmid E, Skalský R, Khabarov N, Obersteiner M, Stürmer 724 B, Xiong W (2013) Pan-European crop modelling with EPIC: Implementation, up725 scaling and regional crop yield validation. Agric Syst 120:61–75. 726 https://doi.org/10.1016/j.agsy.2013.05.008 727 Blanco M, Ramos F, Van Doorslaer B, Martínez P, Fumagalli D, Ceglar A, Fernández FJ 728 (2017) Climate change impacts on EU agriculture: A regionalized perspective 729 taking into account market-driven adjustments. Agric Syst 156:52–66. 730 https://doi.org/10.1016/j.agsy.2017.05.013 731 Challinor A, Adger WN, Di Mauro, M (2016) International dimensions. In: UK Climate 732 Change Risk Assessment Evidence Report. London 733 CZSO (2025) Czech Statistical Office. In: Czech Stat. Off. 734 https://vdb.czso.cz/vdbvo2/faces/en/index.jsf?page=statistiky&katalog=30840. 735 Accessed 2 Sept 2025 736 Eitzinger J, Trnka M, Semerádová D, Thaler S, Svobodová E, Hlavinka P, Šiška B, 737 Takáč J, Malatinská L, Nováková M, Dubrovský M, Žalud Z (2013) Regional 738 climate change impacts on agricultural crop production in Central and Eastern 739 Europe – hotspots, regional differences and common trends. J Agric Sci 740 151(6):787–812. https://doi.org/10.1017/S0021859612000767 741 Ercin E, Chico D, Chapagain AK (2019) Vulnerabilities of the European Union’s 742 Economy to Hydrological Extremes Outside its Borders. Atmosphere 10(10):593. 743 https://doi.org/10.3390/atmos10100593 744 Ercin E, Veldkamp TIE, Hunink J (2021) Cross-border climate vulnerabilities of the 745 European Union to drought. Nat Commun 12(1):3322. 746 https://doi.org/10.1038/s41467-021-23584-0 747 Eyring V, Bony S, Meehl GA, Senior CA, Stevens B, Stouffer RJ, Taylor KE (2016) 748 Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) 749 23 experimental design and organization. Geosci Model Dev 9(5):1937–1958. 750 https://doi.org/10.5194/gmd-9-1937-2016 751 FABLE (2019) Pathways to Sustainable Land-Use and Food Systems. International 752 Institute for Applied Systems Analysis (IIASA) and Sustainable Development 753 Solutions Network (SDSN), Laxenburg and Paris 754 Frank S, Schmid E, Havlík P, Schneider UA, Böttcher H, Balkovič J, Obersteiner M 755 (2015) The dynamic soil organic carbon mitigation potential of European 756 cropland. Glob Environ Change 35:269–278. 757 https://doi.org/10.1016/j.gloenvcha.2015.08.004 758 Fujimori S, Hasegawa T, Krey V, Riahi K, Bertram C, Bodirsky BL, Bosetti V, Callen J, 759 Després J, Doelman J, Drouet L, Emmerling J, Frank S, Fricko O, Havlik P, 760 Humpenöder F, Koopman JFL, van Meijl H, Ochi Y, Popp A, Schmitz A, 761 Takahashi K, van Vuuren D (2019) A multi-model assessment of food security 762 implications of climate change mitigation. Nat Sustain 2(5):386–396. 763 https://doi.org/10.1038/s41893-019-0286-2 764 Hasegawa T, Fujimori S, Havlík P, Valin H, Bodirsky BL, Doelman JC, Fellmann T, Kyle 765 P, Koopman JFL, Lotze-Campen H, Mason-D’Croz D, Ochi Y, Pérez Domínguez 766 I, Stehfest E, Sulser TB, Tabeau A, Takahashi K, Takakura J, van Meijl H, van 767 Zeist W-J, Wiebe K, Witzke P (2018) Risk of increased food insecurity under 768 stringent global climate change mitigation policy. Nat Clim Change 8(8):699–703. 769 https://doi.org/10.1038/s41558-018-0230-x 770 Havlík P, Schneider UA, Schmid E, Böttcher H, Fritz S, Skalský R, Aoki K, Cara SD, 771 Kindermann G, Kraxner F, Leduc S, McCallum I, Mosnier A, Sauer T, Obersteiner 772 M (2011) Global land-use implications of first and second generation biofuel 773 targets. Energy Policy 39(10):5690–5702. 774 https://doi.org/10.1016/j.enpol.2010.03.030 775 Havlík P, Valin H, Herrero M, Obersteiner M, Schmid E, Rufino MC, Mosnier A, Thornton 776 PK, Böttcher H, Conant RT, Frank S, Fritz S, Fuss S, Kraxner F, Notenbaert A 777 (2014) Climate change mitigation through livestock system transitions. Proc Natl 778 Acad Sci 111(10):3709–3714. https://doi.org/10.1073/pnas.1308044111 779 Herrero M, Havlik P, Valin H, Notenbaert A, Rufino MC, Thornton PK, Blummel M, 780 Weiss F, Grace D, Obersteiner M (2013) Biomass use, production, feed 781 efficiencies, and greenhouse gas emissions from global livestock systems. Proc 782 Natl Acad Sci 110(52):20888–20893. https://doi.org/10.1073/pnas.1308149110 783 Hlavinka P, Kersebaum K, Dubrovský M, Fischer M, Pohanková E, Balek J, Žalud Z, 784 Trnka M (2015) Water balance, drought stress and yields for rainfed field crop 785 rotations under present and future conditions in the Czech Republic. Clim Res 786 65:175–192. https://doi.org/10.3354/cr01339 787 24 Jägermeyr J, Müller C, Ruane AC, Elliott J, Balkovic J, Castillo O, Faye B, Foster I, 788 Folberth C, Franke JA, Fuchs K, Guarin JR, Heinke J, Hoogenboom G, Iizumi T, 789 Jain AK, Kelly D, Khabarov N, Lange S, Lin T-S, Liu W, Mialyk O, Minoli S, Moyer 790 EJ, Okada M, Phillips M, Porter C, Rabin SS, Scheer C, Schneider JM, Schyns 791 JF, Skalsky R, Smerald A, Stella T, Stephens H, Webber H, Zabel F, Rosenzweig 792 C (2021) Climate impacts on global agriculture emerge earlier in new generation 793 of climate and crop models. Nat Food. https://doi.org/10.1038/s43016-021-00400794 y 795 Janssens C, Havlík P, Krisztin T, Baker J, Frank S, Hasegawa T, Leclère D, Ohrel S, 796 Ragnauth S, Schmid E, Valin H, Van Lipzig N, Maertens M (2020) Global hunger 797 and climate change adaptation through international trade. Nat Clim Change. 798 https://doi.org/10.1038/s41558-020-0847-4 799 Kindermann G, McCallum I, Fritz S, Obersteiner M (2008) A global forest growing stock, 800 biomass and carbon map based on FAO statistics. Silva Fenn 42(3). 801 https://doi.org/10.14214/sf.244 802 Krey V, Havlik P, Kishimoto PN, Fricko O, Zilliacus J, Gidden M, Strubegger M, 803 Kartasasmita G, Ermolieva T, Forsell N, Gusti M, Johnson N, Kikstra J, 804 Kindermann G, Kolp P, Lovat F, Mc DL, Min J, Pachauri S (2020) MESSAGEix805 GLOBIOM 806 Leclère D, Havlík P, Fuss S, Schmid E, Mosnier A, Walsh B, Valin H, Herrero M, 807 Khabarov N, Obersteiner M (2014) Climate change induced transformations of 808 agricultural systems: insights from a global model. Environ Res Lett 809 9(12):124018. https://doi.org/10.1088/1748-9326/9/12/124018 810 Lorencová E, Frélichová J, Nelson E, Vačkář D (2013) Past and future impacts of land 811 use and climate change on agricultural ecosystem services in the Czech 812 Republic. Land Use Policy 33:183–194. 813 https://doi.org/10.1016/j.landusepol.2012.12.012 814 Maskell G, Shukla R, Jagannathan K, Browne K, Ulibarri N, Campbell D, Franz C, Grady 815 C, Joe ET, Kirchhoff C, Madhavan M, Michaud L, Sharma S, Singh C, Orlove B, 816 Nagle Alverio G, Ajibade I, Bowen K, Chauhan N, Galappaththi E, Hudson AJ, 817 Mach K, Musah-Surugu J, Petzold J, Reckien D, Schauberger B, Segnon A, Van 818 Bavel B, Gornott C (2025) Dichotomy or continuum? A global review of the 819 interaction between autonomous and planned adaptations. Ecol Soc 30(1):art18. 820 https://doi.org/10.5751/ES-15335-300118 821 Ministry of Agriculture of the Czech Republic (2022) Strategický plán společné 822 zemědělské politiky České republiky na období 2023–2027 (verze 6). Prague 823 Mosnier A, Obersteiner M, Havlík P, Schmid E, Khabarov N, Westphal M, Valin H, Frank 824 S, Albrecht F (2014) Global food markets, trade and the cost of climate change 825 adaptation. Food Secur 6(1):29–44. https://doi.org/10.1007/s12571-013-0319-z 826 25 Nelson GC, Valin H, Sands RD, Havlík P, Ahammad H, Deryng D, Elliott J, Fujimori S, 827 Hasegawa T, Heyhoe E, Kyle P, Von Lampe M, Lotze-Campen H, Mason d’Croz 828 D, van Meijl H, van der Mensbrugghe D, Müller C, Popp A, Robertson R, 829 Robinson S, Schmid E, Schmitz C, Tabeau A, Willenbockel D (2014) Climate 830 change effects on agriculture: Economic responses to biophysical shocks. Proc 831 Natl Acad Sci 111(9):3274–3279. https://doi.org/10.1073/pnas.1222465110 832 O’Neill BC, Kriegler E, Riahi K, Ebi KL, Hallegatte S, Carter TR, Mathur R, van Vuuren 833 DP (2014) A new scenario framework for climate change research: the concept of 834 shared socioeconomic pathways. Clim Change 122(3):387–400. 835 https://doi.org/10.1007/s10584-013-0905-2 836 O’Neill BC, Tebaldi C, van Vuuren DP, Eyring V, Friedlingstein P, Hurtt G, Knutti R, 837 Kriegler E, Lamarque J-F, Lowe J, Meehl GA, Moss R, Riahi K, Sanderson BM 838 (2016) The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6. 839 Geosci Model Dev 9(9):3461–3482. https://doi.org/10.5194/gmd-9-3461-2016 840 Papadimitriou L, Trnka M, Harrison P, Holman I (2019) Cross-sectoral and trans-national 841 interactions in national-scale climate change impacts assessment—the case of 842 the Czech Republic. Reg Environ Change 19(8):2453–2464. 843 https://doi.org/10.1007/s10113-019-01558-9 844 Pavlik P, Vlckova V, Machar I (2019) Changes to Land Area Used for Grain Maize 845 Production in Central Europe due to Predicted Climate Change. Int J Agron 846 2019:1–9. https://doi.org/10.1155/2019/9168285 847 Pohanková E, Hlavinka P, Kersebaum KC, Nendel C, Rodríguez A, Balek J, Dubrovský 848 M, Gobin A, Hoogenboom G, Moriondo M, Olesen EJ, Rötter R, Ruiz-Ramos M, 849 Shelia V, Stella T, Hoffmann MP, Takáč J, Eitzinger J, Dibari C, Ferrise R, 850 Bohuslav J, Bláhová M, Trnka M (2024) Expected effects of climate change on 851 the soil organic matter content related to contrasting agricultural management 852 practices based on a crop model ensemble for locations in Czechia. Eur J Agron 853 156:127165. https://doi.org/10.1016/j.eja.2024.127165 854 Pohanková E, Hlavinka P, Kersebaum K-C, Rodríguez A, Jan Balek, Bednařík M, 855 Dubrovský M, Gobin A, Hoogenboom G, Moriondo M, Nendel C, Olesen JE, 856 Rötter RP, Ruiz-Ramos M, Shelia V, Stella T, Hoffmann MP, Takáč J, Eitzinger J, 857 Dibari C, Ferrise R, Bláhová M, Trnka M (2022) Expected effects of climate 858 change on the production and water use of crop rotation management 859 reproduced by crop model ensemble for Czech Republic sites. Eur J Agron 860 134:126446. https://doi.org/10.1016/j.eja.2021.126446 861 Poláková J, Potopova V, Kristkova ZS, Weststrate J, van Zeist W-J, Oosterwijk A, 862 Kolářová M, Viera MD, Zahradníček P, Štěpánek P, Bunnefeld N, Dettenhofer M, 863 Manikas I (2025) From local to global and from global to local: Designing the 864 protocol to model agriculture and climate resilience. Environ Sustain Indic 865 27:100855. https://doi.org/10.1016/j.indic.2025.100855 866 32 yes RCP 8.5 0,00 0,00 2,33 0,18 -0,20 no RCP 8.5 0,00 -1,68 3,21 -2,87 1,43 yes RCP 8.5 GFDL-ESM4 0,00 0,00 1,98 0,15 -0,96 IPSL-CM6A-LR 0,00 0,00 2,33 0,18 -1,04 MPI-ESM1-2-HR 0,00 0,00 1,86 0,14 -0,14 MRI-ESM2-0 0,00 0,00 1,11 0,09 -1,34 UKESM1-0-LL 0,00 -1,68 3,56 -2,84 1,65 yes RCP 2.6 Mean GCM ensemble Poultry meat 0,00 0,00 0,11 0,01 0,19 no RCP 2.6 0,01 0,00 -0,21 -0,04 0,98 yes RCP 7.0 -0,01 0,00 7,31 0,86 -0,11 no RCP 7.0 0,03 0,00 0,46 0,00 2,38 yes RCP 8.5 0,02 0,00 0,80 0,06 0,37 no RCP 8.5 0,00 0,00 -0,20 -2,70 3,03 yes RCP 8.5 GFDL-ESM4 0,07 0,00 2,55 0,17 0,09 IPSL-CM6A-LR 0,04 0,00 0,11 -0,07 0,76 MPI-ESM1-2-HR 0,07 0,00 0,72 -0,04 0,20 MRI-ESM2-0 -0,03 0,00 0,11 0,08 -0,61 UKESM1-0-LL 0,03 -0,98 0,50 -2,66 2,69 973 Supplementary Table 7 | Changes in average cropland, grassland and other natura land areas for the 974 Czech Republic and the EU28 a relative to the no-climate-change baseline across RCP and GCM 975 scenarios Results are presented for scenarios where climate impacts are applied only to the Czech 976 Republic (national), the EU28 (regional) and to the entire world (global). 977 Czech Republic Cropland area [%] Grassland area [%] Other natural land areas [%] Climate impact scenario National Regional Global National Regional Global National Regional Global Climate Scenario RCP 2.6 2.09 3.01 -0.07 -0.04 0.00 0.00 -12.68 -18.33 0.42 RCP 2.6 wo CO2 0.22 3.51 4.64 0.02 -0.04 -2.15 -1.31 -21.34 -24.40 RCP 7.0 2.67 0.47 -0.57 -0.04 0.01 0.02 -16.23 -2.87 3.41 RCP 7.0 wo CO2 0.04 4.44 7.11 0.00 -2.13 -2.15 -0.26 -23.32 -39.50 RCP 8.5 2.66 3.83 3.38 -1.54 -2.04 -2.04 -13.49 -19.73 -16.99 RCP 8.5 wo CO2 0.25 7.17 11.29 -0.02 -2.15 -2.15 -1.50 -39.83 -64.99 Average 1.32 3.74 4.30 -0.27 -1.06 -1.41 -7.58 -20.90 -23.68 EU28 Cropland area [%] Grassland area [%] Other natural land areas [%] Climate impact scenario National Regional Global National Regional Global National Regional Global Climate Scenario RCP 2.6 0.47 0.71 0.41 -0.08 -0.04 0.02 -4.82 -5.90 -14.37 RCP 2.6 wo CO2 0.37 0.90 1.63 -0.06 -0.11 -0.10 -3.24 -9.97 -19.85 33 RCP 7.0 0.43 0.44 0.01 -0.04 -0.08 -0.03 -3.91 -7.50 -33.04 RCP 7.0 wo CO2 0.26 1.35 1.97 -0.04 -0.13 -0.12 -2.43 -15.96 -4.21 RCP 8.5 0.58 0.88 0.75 -0.10 -0.08 -0.04 -5.92 -11.24 -5.67 RCP 8.5 wo CO2 0.44 1.96 3.36 -0.05 -0.14 -0.19 -4.13 -22.40 -10.28 Average 0.43 1.04 1.35 -0.06 -0.10 -0.08 -4.07 -12.16 -14.57 978 979 Supplementary Table 8 | Average crop production and price impacts for the Czech Republic and the 980 EU28 relative to the no climate change SCENARIO across RCP and GCM scenarios. Changes in 981 production (top) and price (bottom) for major agricultural commodities are presented for scenarios where 982 climate impacts are applied only to the Czech Republic (national), the EU28 (regional) and to the entire 983 world (global). 984 Production Czech Republic EU28 Climate impact scenario National Regional Global National Regional Global Wheat 2.33 3.08 2.82 0.83 0.66 0.28 Maize -0.08 -0.80 -0.68 -0.52 -0.46 -1.41 Barley 5.27 7.52 6.47 0.03 -0.40 -2.62 Rapeseed 3.82 5.50 5.06 1.37 0.90 2.07 Potatoes -0.31 -0.22 0.15 -0.13 -0.41 -0.35 Prices Czech Republic EU28 Climate impact scenario National Regional Global National Regional Global Wheat 1.20 -1.10 -1.79 0.37 0.54 0.39 Maize 1.98 3.05 2.71 9.81 1.45 6.92 Barley 0.32 2.74 1.21 0.61 1.49 -0.55 Rapeseed 0.89 1.49 1.91 0.88 1.61 1.80 Potatoes 0.18 0.92 1.27 0.15 0.29 0.34 985 986 34 987 Supplementary Table 9 | Intercept and slope coefficients for Supplementary Fig. 9 for the European 988 Union + UK. 989 Climate impact scenario National Regional Global Variable Intercept Slope Intercept Slope Intercept Slope Yield -0.14 2.48 -0.07 0.69 -0.06 0.73 Area 0.42 -0.79 0.40 -0.25 0.15 -0.44 Production 0.29 1.71 0.34 0.45 0.11 0.32 Consumption 0.24 0.16 0.18 0.15 0.33 0.20 Exports 0.56 5.24 -1.65 0.53 -2.81 0.33 Imports 0.37 -0.23 -1.91 -0.47 -1.73 -0.07 Prices 0.49 -1.63 0.69 -0.36 0.19 -0.58 990 Supplementary Table 10 | Intercept and slope coefficients globally for Supplementary Fig. 9 991 Climate impact scenario National Regional Global Variable Intercept Slope Intercept Slope Intercept Slope Yield 0.01 -10.24 -0.09 -0.05 -0.32 0.28 Area -0.04 10.08 0.04 0.14 0.43 -0.16 Production -0.03 -0.15 -0.05 0.09 0.10 0.13 Consumption -0.03 -0.15 -0.05 0.09 0.10 0.13 Prices 0.06 9.17 0.16 0.04 -0.56 -0.49 992 993 35 Supplementary Figures 994 995 996 997 998 999 1000 Supplementary Fig. 1| Regional classification of European countries and the rest of the world for the 1001 international trade analysis. The map categorizes the regions into the Czech Republic (green), Central 1002 East (pink), West (blue), North (purple), South (orange), and Rest of the World (yellow). 1003 1004 36 1005 Supplementary Fig. 2 | Evolution of cereal and oilseed trade in the Czech Republic under alternative 1006 climate impact scenarios. Panels show projected imports and exports of (a) wheat, (b) rapeseed, and (c) 1007 barley between 2020 and 2050. Lines indicate mean changes across GCMs, with colors representing 1008 climate scenarios (RCP2.6, RCP7.0, RCP8.5). Shaded ribbons denote the range of uncertainty across 1009 GCMs. Results are shown for three levels of climate impact aggregation: National (CZ), Regional (EU28), 1010 and Global (World). 1011 1012 37 1013 Supplementary Fig. 3 | Percentage change in biophysical crop aggregated yields under three Shared1014 Socioeconomic Pathways (SSP)- Representative Concentration Pathways (RCPs), calculated as 1015 averages of General Circulation Models (GCMs): (a) SPP12.6 (b) SSP3 - 7.0, and (c) SSP5 - 8.5 with 1016 CO2 effect. The x-axis represents individual countries and regions grouped geographically, while the y-axis 1017 shows percentage yield changes relative to a non-climate-change scenario. The colors of the bars 1018 represented EU28 regions used in the analysis of international trade. 1019 38 1020 Supplementary Fig. 4 | Percentage change in biophysical wheat aggregated yields under three Shared1021 Socioeconomic Pathways (SSP)- Representative Concentration Pathways (RCPs), calculated as 1022 averages of General Circulation Models (GCMs): (a) SPP12.6 (b) SSP3 - 7.0, and (c) SSP5 - 8.5 with 1023 CO2 effect. The x-axis represents individual countries and regions grouped geographically, while the y-axis 1024 shows percentage yield changes relative to a non-climate-change scenario. The colors of the bars 1025 represented EU28 regions used in the analysis of international trade. 1026 1027 39 1028 Supplementary Fig. 5 | Percentage change in biophysical maize aggregated yields under three Shared1029 Socioeconomic Pathways (SSP)- Representative Concentration Pathways (RCPs), calculated as 1030 averages of General Circulation Models (GCMs): (a) SPP12.6 (b) SSP3 - 7.0, and (c) SSP5 - 8.5 with 1031 CO2 effect. The x-axis represents individual countries and regions grouped geographically, while the y-axis 1032 shows percentage yield changes relative to a non-climate-change scenario. The colors of the bars 1033 represented EU28 regions used in the analysis of international trade. 1034 1035 1036 40 1037 Supplementary Fig. 6 | Revealed Comparative Advantage (RCA) for five key agricultural commodities (a: 1038 Wheat, b: Maize, c: Barley, d: Rapeseed, e: Potatoes) across the European Union (EU) countries, the 1039 EU28 region, and global averages under various climate impact scenarios. The RCA index reflects the 1040 relative export competitiveness of a country in a specific crop, with values above 1 indicating a 1041 comparative advantage. Different climate impact scenarios are represented by distinct symbols and 1042 different climate change scenarios by distinctive colors 1043 1044 41 1045 Supplementary Fig. 7 | Projected biophysical yield changes (%) with respect to the non climate change 1046 scenario for wheat, maize, and aggregate crops in the European Union COUNTRIES plus United Kindom, 1047 and globally under different climate scenarios and General Circulation Models (GCMs). Bars represent 1048 GCM-averaged values, and individual points correspond to specific GCMs 1049