Back

New rice varieties with improved phosphorus-efficiency for low-input smallholder rice production in Africa

Wissuwa, M.; Ranaivo, N. H.; Rakotondramanana, M. F.; Rafaliarivony, S.; Kondo, K.; Ueda, Y.; Dinh, L. T.; Connor, M.; Chin, J. H.; Pariasca Tanaka, J.

2025-12-09 plant biology
10.64898/2025.12.06.692696 bioRxiv
Show abstract

Smallholder farmers in Sub-Saharan Africa frequently produce rice in marginal environments where low soil fertility and other biotic and abiotic stresses limit productivity. Rice varieties developed by centralized breeding under favorable conditions on research stations have often not been adopted by farmers in such marginal environments. Our objective was to develop modern rice varieties adapted to such low-input conditions through combining pre-breeding research with subsequent selection and variety testing directly in the target environment: smallholder farmers fields in Madagascar with phosphorus (P) fixing soils. Two breeding populations were developed for this purpose, one based on marker-assisted introgression of the Pup1 locus into IR64, the second using a donor (DJ123) for internal P utilization efficiency and external P acquisition efficiency. Selection within these populations was conducted in fields managed according to local farmers practice without mineral fertilizer addition. Selected breeding lines underwent government-supervised variety release testing including farmer participatory evaluations in four Malagasy regions between 18-1350 masl altitude, and two lines were released as varieties FyVary32 and VyVary85. Both had between 0.47-0.8 t ha-1 higher grain yield than parent IR64 and local check X265. Yield advantages were stable across a range from 2.1-5.5 t ha-1 (national average: 2.8 t ha-1). Higher yields were accompanied by superior root development and P uptake and by more efficient internal P utilization in FyVary85. Results show that decentralized breeding in marginal environments can produce varieties not only superior in lowest-yielding environments but across a broader range, clearly surpassing national average yields.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.