Genomic evidence reveals taxonomic oversplitting in a wing dimorphic grasshopper complex Tetrix japonica

Accurate species delimitation is fundamental to biodiversity research but remains challenging in taxonomically controversial groups exhibiting morphological plasticity. The Tetrix japonica complex (Orthoptera: Tetrigidae) includes widespread populations and several nominal species defined primarily by variable traits such as wing dimorphism. The validity of these taxa and the genera based upon them (Alulatettix and Formosatettix) is highly debated. To resolve these boundaries, we performed whole-genome resequencing anchored to a chromosome-level reference genome.

We analyzed 1,095,722 high-quality SNPs from 50 specimens collected across East Asia. Population genomic analyses, including Principal Component Analysis (PCA) and ADMIXTURE (optimal K = 1), revealed a unified gene pool with no significant genetic substructure corresponding to current taxonomic classifications. Genetic differentiation was low (global F_ST near 0.03) and significantly correlated with geographic distance (Isolation by Distance), indicating that spatial separation rather than reproductive isolation drives genetic variation. A Random Forest classifier achieved only 16.7% accuracy in distinguishing nominal species, confirming that genomic data cannot support the current morphological taxonomy. Phylogenomic reconstruction (IQ-TREE) further demonstrated that all these included specimens are tightly related. Key taxa such as A. wudangshanensis and F. gonggashanensis were recovered as nested lineages deep within the T. japonica complex, clustering with geographically proximate populations rather than forming distinct generic clades.

Our findings demonstrate that T. japonica represents a single, widespread metapopulation with high morphological plasticity. Diagnostic characters previously used to erect separate species and genera, particularly wing morphology and flightlessness, are reinterpreted here as intraspecific polymorphisms. Our genomic data provide molecular support for the previous hypothesis that A. wudangshanensis, F. shennongjiaensis, and F. gonggashanensis belong to the T. japonica complex. This study highlights the necessity of integrating phylogenomics with machine learning to correct historical oversplitting in Orthoptera taxonomy.