Crop Yields in a Warming World
The food connection
This is the most direct food connection imaginable: warming attacks the yield of the exact plants that feed the most people. The geography of the damage is deeply unequal. AR6 projects that the heaviest losses fall on the tropics and subtropics — the regions already closest to crop heat limits and least able to afford adaptation — while some high-latitude regions (parts of Canada, Russia, northern Europe) may see neutral or even improved conditions for certain crops. For African staple crops specifically, AR6's synthesis (Figure 9.22) projected median yield losses worsening from roughly −10% at 2°C of warming to around −23% at 4°C even with adaptation measures, and substantially worse without them. The crops that gain are not generally the crops that the hungry depend on.
The human cost
Heat is not a gentle dial on crop growth; past a plant's optimum, yields fall, and they fall faster the hotter it gets. The most-cited synthesis — Zhao and colleagues' 2017 study in the Proceedings of the National Academy of Sciences, which pooled four independent methods (grid models, point models, statistical regression, and field-warming experiments) — found that, absent CO₂ fertilization, adaptation, and breeding, each additional degree Celsius of global mean warming would reduce global yields of wheat by about 6.0%, rice by 3.2%, maize by 7.4%, and soybean by 3.1%. A 2025 threshold analysis in Scientific Reports, drawing on more than 8,700 crop-model simulations from over 200 studies compiled for the IPCC's Fifth and Sixth Assessment Reports, refined this picture: wheat loses about 6.1% per degree below a 2.38°C warming threshold but 8.2% per degree above it; rice loses about 1.1% per degree until a 3.13°C threshold, after which losses jump to 7.1% per degree; maize shows no threshold and loses roughly 4.0% per degree throughout. The lesson of the thresholds is the dangerous part: the damage is not linear. Beyond certain points, the same extra degree does far more harm.
The IPCC's Sixth Assessment Report (Working Group II, 2022) complicates the raw heat numbers with the offsetting effect of carbon-dioxide fertilization, which boosts the growth of "C3" crops like wheat and rice more than "C4" crops like maize. Frances Moore's analysis, cited in AR6, estimated that warming over 1961–2017 cut yields of the three staples by an average of about 5.3% (maize 5.9%, wheat 4.9%, rice 4.2%) before accounting for CO₂; once CO₂ fertilization is included, the net global effect may have been close to zero so far. A separate study by Iizumi and colleagues found CO₂ gains offsetting heat losses for wheat and rice but not for maize and soybean, which showed net declines. This is the honest state of the science: warming is unambiguously harmful to crop physiology, the harm is partly masked today by CO₂ and by farmer adaptation, and the masking is expected to weaken as warming accelerates past crop thresholds.
This entry is about projected, not recorded, cost, and honesty requires saying so. The IPCC and FAO do not publish a single "deaths from yield loss" figure, because famine is mediated by markets, governance, and conflict, not by yield alone. What the evidence supports: drought-related yield losses have already affected roughly three-quarters of the global harvested area in recent decades (UNDP summary of AR6); hundreds of millions of people in South Asia, Sub-Saharan Africa, the Middle East, and the Sahel live in the regions projected to lose the most; and the FAO recorded 733 million people facing hunger in 2023. The mechanism by which warming kills is the mechanism documented throughout this section — not direct starvation in the field, but price spikes, export bans, and the collapse of purchasing power among the poor. The yield numbers are the fuse; the human cost is lit by politics.
Political & economic context
The countries that emitted the carbon are overwhelmingly not the countries that will lose the most yield. This is the central injustice of climate and food: the Sahel, which contributes a negligible share of historical emissions, sits among the most exposed regions on earth, while the high-latitude potential "winners" are largely industrialized emitters. Within agriculture, the beneficiaries of warming-driven volatility are often grain traders and commodity speculators, for whom price swings are opportunity; the losers are smallholders without irrigation, crop insurance, or political voice.
Historical legacy
The genuine scientific debate is not whether warming harms crops — that is settled — but how much the offsetting factors (CO₂ fertilization, adaptation, breeding) will continue to compensate. Optimists point to the historical record, in which global yields kept rising through decades of warming. Pessimists point out that this rise came from technology and inputs that masked an underlying climate drag, that CO₂ fertilization plateaus and degrades nutritional quality (lower protein, zinc, and iron in CO₂-enriched grain), and that adaptation has limits which the thresholds above make explicit. The weight of AR6 evidence favors the pessimists for the second half of the century.
Food culture legacy
Staple shifts reshape cuisines. As traditional growing zones become marginal, cultures face the slow loss of foods central to identity: heat-stressed Basmati and other aromatic rices in South Asia, climate-squeezed coffee and cacao belts, the northward retreat of wine and olive regions. Conversely, drought-tolerant "orphan crops" long dismissed as poverty food — millets, sorghum, fonio, teff — are being revalued precisely because they tolerate the conditions that staples cannot. The UN's designation of 2023 as the International Year of Millets was, in part, a climate-adaptation statement dressed as a culinary celebration.
Reference notes
Cross-link to: The Bengal Famine of 1943, The Irish Potato Famine, The Dust Bowl (all Food, War & Peace). Related entries: Millets, Grains & Seeds (Legumes, Grains & Seeds document); Rice Varieties of the World (heat-vulnerable aromatic varieties). Related cuisines: South Asian, West African, Sahelian. Content advisory: standard section tag; flag projections clearly as projections in any user-facing summary. Suggested cross-link anchor: "orphan crops / climate-resilient staples."
---