C4 rice / better photosynthesis
Most crops use an older, wasteful photosynthesis (C3); rebuilding rice or wheat to work like maize (C4) could raise yields by up to half.
Open in the interactive tree →The C4 Rice Project, funded by the Gates Foundation since 2008 and also by the UK government from 2012, aims to give rice the C4 pathway; its researchers suggest it could yield up to 50% more grain with less water and nutrients. A different route, shortening the energy-wasting photorespiration step (RIPE project), raised biomass of field-grown tobacco by up to about 40% in 2019 trials.
As of October 2026
The Oxford-led C4 Rice project (more than 20 research groups in 8 countries) is in its fourth phase, and its stated goal was experimental field plots in Taiwan by 2024; no working C4 rice has been announced as a crop. A simpler intermediate C2 route is under discussion because it needs fewer genetic changes.
What is missing
- The two-cell-type leaf structure (Kranz anatomy) has not been engineered into rice
- Dozens of genes and transporters must work together in the right cells
- Gains seen in greenhouses must survive heat, drought and real field conditions
- Multi-year field trials and regulatory approval for a staple crop
- Long-term funding
Becomes possible once solved
- Higher grain yield per hectare without more land
- Better water and nitrogen use efficiency
- More heat-tolerant staple crops
Open steps
- Kranz leaf anatomy in rice Medium AI leverageFind the regulators that build the two-cell-layer leaf structure of C4 plants and switch them on in rice without harming leaf development.
- Minimal C4 gene and transporter set Medium AI leverageWork out which enzymes and transporters a rice cell needs for a working C4 cycle, and how strongly each must be expressed.
- Faster, more specific Rubisco High AI leverageFind or design Rubisco versions that fix CO2 faster and confuse oxygen less, and that fold and assemble correctly in crop chloroplasts.
- Gains that survive field conditions Low AI leveragePredict whether greenhouse gains of a C4-type rice survive heat, drought and real fields, and which lines deserve multi-year trials first.
Where AI could help
Medium AI leverage. AI can search Rubisco and C4 gene designs, but rice must still be rebuilt, grown and field-tested over many seasons.
- Predict Rubisco kinetics from sequence to pick or design faster variants for crop testing
- Mine gene-expression data for the regulators that build the two-cell Kranz leaf anatomy
- Model C4 or C2 pathway flux to rank which gene combinations are worth building
- Predict plant traits under heat and drought to prioritise lines before field trials
Shown so far
- In November 2025, a PNAS study measured 250+ bacterial and archaeal Rubiscos and used a machine-learning model to predict carboxylation rates for all ~68,000 known sequences. source
- In September 2022, Gaussian-process models predicted Rubisco kinetics from sequence for over 9,000 plant RbcL sequences, matching the known faster rates of C4 and CAM species. source
Prerequisites
- Photosynthesis discovered1779Boosting rice yield means improving how plants turn light, CO2 and water into sugar
- Genetically modified crops1996
- Genome sequencing2003
- CRISPR gene editing2012
Unlocks
- Yield jump from C4 crops2040s?