The research team, led by Zhenhua Lu and Zheng Li of Yunnan University together with collaborators including Laura Ellen Rose of Heinrich-Heine-Universität Düsseldorf, took advantage of a natural experiment that evolution has run repeatedly across the rice genus Oryza. Within this genus, several closely related species pairs exist in which one member is perennial and the other is annual. The researchers compared three such pairs: Oryza rufipogon, a perennial wild relative of cultivated rice, against its annual counterpart Oryza nivara; Oryza glumaepatula, a perennial South American wild rice, against the African cultivated annual Oryza glaberrima; and Oryza meyeriana against Oryza brachyantha. By lining up the genomes of species that have independently converged on annuality, the team could search for genetic changes that recur across these transitions, a far stronger signal than any single species comparison could provide.

The central logic of the study is elegantly simple. If a gene family is retained in every perennial species examined but is absent from every annual species, then its loss is unlikely to be random noise. Across the three pairs, the researchers identified exactly 91 gene families that fit this pattern: present in all three perennials, missing from all three annuals. These 91 families represent the strongest candidates for genetic changes that accompanied or enabled the perennial-to-annual transition, and they now form a curated shortlist for anyone seeking the molecular basis of plant life history.

What do these lost genes actually do? Functional annotation and prediction of the 91 families suggested that they are not a random assortment of housekeeping genes. Instead, they map onto a battery of physiological systems that one would expect to matter for a plant that must survive year after year. The categories include phenology, the timing of developmental events such as flowering and dormancy; energy allocation, the decisions about where captured carbon and nutrients are invested; and source–sink dynamics, the coordinated movement of photosynthate from producing tissues such as leaves to consuming tissues such as roots, seeds and storage organs. A perennial plant must budget its resources across multiple reproductive cycles, and the study suggests that annual species shed genes that help manage that long-horizon accounting.

To probe function more deeply, the team went beyond sequence annotation and performed comparative co-expression analysis. Genes rarely act alone; they operate in regulatory networks whose activity patterns reveal their roles. By comparing co-expression relationships between perennial and annual Oryza species, the researchers found that many of the 91 candidate gene families likely participate in orchestrating processes that differ systematically between the two life histories. Three themes stood out: reproduction, root development, and responses to environmental stimuli. Each makes intuitive sense for the perennial-to-annual shift. Annual plants pour resources into a single, terminal flowering event, while perennials must regulate reproduction so that it can recur; perennial roots, which anchor the plant through multiple seasons and store reserves, require developmental programs that annuals can afford to simplify; and a plant that lives for years must continuously adjust to a changing environment rather than timing its entire life cycle to a single season.

The study also examined the regulatory architecture surrounding these candidate genes, analyzing promoter cis-elements and protein motifs, and documented pseudogenes in Oryza nivara, hints of genes decaying into nonfunctional remnants in the annual lineage. While the paper presents these as candidate-level evidence rather than proven causation, the convergence of several independent lines of evidence, presence-absence patterns, functional annotation, enrichment analyses, and co-expression comparisons, strengthens the case that gene loss played a real role in the emergence of annuality in Oryza.

Perhaps the most important conceptual takeaway is that the transition to annuality appears to be evolutionarily complex rather than the product of one or a few master mutations. The 91 gene families touch on a wide range of biological processes, and the co-expression results imply that annual species have rewired regulatory networks across reproduction, root biology and environmental response. This complexity carries a direct message for breeders: improving perennial rice will likely require attending to many aspects of perennial physiology in combination, rather than fixing a single perenniality gene and expecting the whole life-history package to follow. The authors argue that a combinatorial view of perenniality is needed to improve the overall agronomic performance of perennial rice varieties.

The practical stakes are considerable. Perennial rice has already demonstrated potential to alleviate the multidimensional sustainability problems of current rice agriculture, reducing the need for yearly tillage, planting and the associated labor and inputs. But breeding better perennial varieties has been hampered precisely by the limited knowledge of perenniality mechanisms that this study begins to address. By supplying a defined set of candidate gene families, the work gives geneticists concrete targets for functional validation, and gives breeders molecular markers that could be tracked as they combine perennial traits from wild relatives such as Oryza rufipogon with the yield and grain quality of cultivated rice.

Methodologically, the study showcases the power of treating evolution’s replicates seriously. Any single perennial-annual pair might differ for reasons unrelated to life history, such as local adaptation or domestication bottlenecks. But requiring a gene family to be absent from all three annual lineages while present in all three perennials filters out much of that noise. The team supplemented genome comparisons with RNA-sequence data from public archives, measured expression in counts per million, and built co-expression networks in the perennial O. rufipogon that they then connected to modules in cultivated annual rice, allowing them to ask whether the regulatory neighborhoods of candidate genes had been reorganized during the transition. Enrichment analyses against Gene Ontology, Trait Ontology, KEGG pathways, Pfam protein domains and EggNOG orthology groups rounded out the annotation pipeline.

The work, which received support from China’s National Key Research and Development Program, the National Natural Science Foundation of China, Yunnan provincial research programs, the New Cornerstone Science Foundation and Germany’s Excellence Strategy, opens a clear path forward. The next step is experimental: knocking out or transferring individual candidate genes in controlled backgrounds to test whether they genuinely influence perennial traits such as regrowth after harvest, ratooning ability or resource storage. If even a fraction of the 91 families proves functional, the study will have converted a long-standing evolutionary puzzle into an actionable breeding toolkit, bringing the vision of rice fields that never need replanting closer to routine agricultural reality.

Subject of Research: Comparative genomics of perennial-to-annual evolutionary transitions in Oryza rice species

Article Title: Mining genetic changes underlying perennial-to-annual transitions in Oryza with comparative genomics

Article References: Lu, Z., Zhou, Y., Zheng, H., Yang, G., Rose, L. E., Hu, F., & Li, Z. (2026). Mining genetic changes underlying perennial-to-annual transitions in Oryza with comparative genomics. BMC Plant Biology. https://doi.org/10.1186/s12870-026-10041-7

Image Credits: AI Generated

DOI: 10.1186/s12870-026-10041-7

Keywords: Oryza, perennial rice, comparative genomics, gene families, life history evolution, annual plants, co-expression analysis, crop breeding, genome evolution, plant genetics, sustainable agriculture, source-sink dynamics