Large-scale Genomic Analysis Settles Debate on Insect Flight Origins and Palaeoptera Monophyly
Background
Acquiring flight capability was the key driver behind the overwhelming success of insects in Earth's ecosystem. Winged insects, the Pterygota, first appeared during the Carboniferous period of the Paleozoic Era and underwent explosive diversification by dominating terrestrial and aerial environments. However, the debate surrounding when and through which evolutionary pathways they completed their flight structures has remained unresolved for the past century.
At the center of this debate are the primitive flying insects, Ephemeroptera (mayflies) and Odonata (dragonflies). Unlike Neoptera, which comprise the majority of modern insects and can fold their wings backward over their bodies, these two groups share the unique anatomical trait of being unable to fold their wings. Taxonomists have grouped them into the Palaeoptera, but three conflicting hypotheses exist regarding whether these two orders form a monophyletic group derived from a single common ancestor.
Previous studies relied on morphological traits or the analysis of a limited number of genetic markers. Morphological studies either supported the hypothesis of grouping dragonflies closer to Neoptera due to differences in flight muscle attachment, or placed Ephemeroptera as the most basal group based on differences in aquatic respiratory organs during the larval stage. Early molecular phylogenetic analyses also added confusion, producing conflicting cladograms depending on the samples and analysis models due to base composition bias and Long-Branch Attraction (LBA). To establish the root of the insect evolutionary tree, it was essential to build an advanced genomic analysis framework capable of removing evolutionary artifacts using taxonomic sampling on an unprecedented scale.
Key Findings
An international collaborative research team conducted a phylogenomic analysis by integrating large-scale genomic and transcriptomic data from 48 species of Ephemeroptera and 52 species of Odonata across 34 families worldwide, along with major taxa within the Neoptera. To prevent phylogenetic distortion, a set of approximately 2,800 single-copy ortholog gene sets was selected to reconstruct the amino acid-based phylogenetic tree.
The analysis results strongly supported the Palaeoptera hypothesis, with Ephemeroptera and Odonata forming a single clade with 100% statistical confidence (bootstrap value 100). To mitigate phylogenetic distortion frequently encountered during phylogenetic analysis, the team applied a combination of the CAT-GTR model, which accounts for site heterogeneity, and non-homogeneous amino acid substitution models. Based on comparative verification models, it was clearly revealed that the previous grouping of dragonflies with Neoptera was an artificial connection caused by rapid nucleotide substitution rates.
Distinct evolutionary signatures were also identified in the evolutionary patterns of flight-related gene clusters. When the team traced the lineages of Sallimus, a flight muscle elastic protein, and proteins related to wing cuticle formation, they observed that specific gene duplication and domain rearrangement events were concentrated during the ancestral stage of Odonatoptera. This strongly supports the idea that the primitive flight system emerged through a single successful innovation between the Late Devonian and Early Carboniferous periods, approximately 390 million years ago.
Significance and Outlook
This study holds great value in fundamental biology, as it concludes the century-long debate over the monophyly of Palaeoptera in insect phylogenetics using advanced genomic statistical models. By clearly establishing that the origin of insect flight was established in a single common ancestor, this study provides a definitive reference point for tracing how the arthropod exoskeleton underwent genetic reorganization to transform into wings for flight.
However, the temporal discrepancy between the fossil record and molecular clock estimates remains a challenge to be solved. A time gap of approximately 20 million years exists between the dates of primitive insect wing fossils confirmed in the Paleozoic and the divergence points indicated by genomic data. A subsequent challenge is to conduct integrative phylogenetic research that combines fossil-derived morphological data and high-resolution micro-CT data of soft-tissue-preserved fossils with this genomic matrix.
Proceedings of the National Academy of Sciences, Volume 123, Issue 42, October 2026. SignificanceThe question of whether Palaeoptera (Ephemeroptera + Odonata) forms a monophyletic group has drawn considerable attention in insect phylogenetic studies. Leveraging the most comprehensive taxonomic sampling to date alongside refined ...
Elucidating the genomic architecture of primitive flying insects provides a useful blueprint for the fields of biomimetic robotics and advanced materials development. Dragonflies and mayflies boast simultaneous capabilities for ultra-low-power precision flight, high-speed rapid turns, and gliding, which are difficult to implement in modern mechanical systems. The structure of flight muscle proteins specific to Palaeoptera and the gene networks for wing chitin complex synthesis identified by the research team provide the core molecular design for producing high-elasticity, lightweight biomaterials. This dielectric data can be directly applied to the production of flexible wing hinges for Micro Air Vehicles (MAV) or the synthesis of artificial joint materials capable of withstanding extreme aerodynamic fatigue.
Immediate applications are also possible in agricultural ecosystem monitoring and the development of aquatic biomarkers. Transcriptome maps of mayfly and dragonfly larvae, which are sensitive to changes in water quality, serve as a reference library for identifying genes that act as indicators of climate change stress. If the detoxification enzyme and stress protein expression pathways of model organisms that respond to specific chemical pollutants are constructed into biosensor chips, it will lead to the development of high-sensitivity environmental diagnostic kits for real-time detection of river ecosystem disturbances.