The two crickets belong to a genus known for diversity in adult male head shape, which males use in fights with rivals. L. doenitzi, a related species included in the study, displays a clearly different head morphology and is roughly 2 millimeters larger in body length and body part sizes. L. campestris and L. sylvestris, however, show no clear quantitative or qualitative criteria for distinguishing adults. Subtle differences exist, such as a slightly darker body color and a slightly redder abdomen in L. sylvestris, but these are not reliable indicators for species identification. To test whether the two taxa are measurably different, the researchers measured body length, forewing length, femur length, and head width in 80 male and 178 female L. campestris, 16 male and 32 female L. sylvestris, and 20 male and 64 female L. doenitzi individuals. Sexual dimorphism appeared in femur length and head width in all species, but between L. campestris and L. sylvestris, no clear morphological differences were observed.

Genetic analysis told a different story. The team assembled complete circular mitochondrial genomes de novo for all 20 sequenced individuals, annotating all 13 protein-coding genes, two rRNA genes, and 22 tRNA genes at homologous positions, with no premature stop codons or frameshift mutations detected. Between L. campestris and L. sylvestris, the mitochondrial genomes carry 112 fixed differences. Of these, 8 lie in intergenic regions and 39 at the third position of codons, while 53 fixed mutations at codon positions 1 and 2, together with a single fixed insertion/deletion at tRNA-Arg, provide genetic markers that clearly distinguish the two species. Phylogenetic trees inferred from the concatenated amino acid sequences of all 13 mitochondrial protein-coding genes showed the two species as reciprocally monophyletic clades separated by short branch lengths, with the genus Loxoblemmus diverging immediately after the emergence of the subfamily Gryllinae during the Gryllidea radiation.

A crucial question was whether the two crickets are simply geographically isolated populations of one species. The sampling spanned a wide area of the Japanese archipelago without bias, allowing a direct test of isolation by distance. When L. campestris and L. sylvestris were treated as a single population, the correlation between genetic distance and linear geographic distance was essentially zero for mitochondrial genes, with a correlation coefficient of r = −0.006, R² = 0.000042, a Mantel test p-value of 0.437, and a Pearson’s p-value of 0.937. The pairwise genetic distance between species was clearly higher than the distance between strains within either species. The same test on nuclear genes produced a similarly absent correlation, with r = −0.049, R² = 0.00238, Mantel p = 0.678, and Pearson’s p = 0.549. Genetic differentiation between the two lineages is therefore not explained by geographic separation.

The nuclear analysis relied on careful gene filtering because no reference genome exists for the genus Loxoblemmus. The researchers mapped paired-end reads from all 20 individuals to annotated exon sequences of the two-spotted cricket Gryllus bimaculatus, called variants with hard filtering criteria, and selected 1,376 genes that were single-copy in at least 12 of 15 Polyneoptera species from the RefSeq database. After applying depth criteria requiring each gene’s sequence depth to fall within ±5 of both the mean and the median mapping depth, and the difference between mean and median to be within 5, the final set contained 1,021 genes with a mean mapping depth of 17.6 and a median of 15. Population genomic analyses on this set, including principal component analysis and ADMIXTURE analysis, indicated clear genetic differentiation between the species with no obvious evidence of recent admixture.

Genome-wide genetic differentiation between the two species was moderate, with a multilocus weighted FST of 0.196. Genetic diversity estimates were nearly identical between the species: L. campestris showed Watterson’s estimator θW = 0.0121 and θπ = 0.0143 with Tajima’s D = 0.750, while L. sylvestris showed θW = 0.0121, θπ = 0.0141, and Tajima’s D = 0.768. At the gene level, 540 genes showed significant differentiation based on the Benjamini–Hochberg false discovery rate procedure and 161 genes based on the more stringent Bonferroni correction; 29 percent and 53 percent of these genes, respectively, also exhibited low intraspecific genetic diversity with θπ below 0.01. The McDonald–Kreitman test indicated that signals of positive selection across the entire genome were limited, with only a single gene, ATP synthase subunit alpha, showing a significant deviation from neutrality at p < 0.001.

The most intriguing result emerged from a comparative genomic framework that evaluates amino acid substitutions in the context of cross-species sequence conservation. Using evolutionary probabilities, which quantify how rarely a specific amino acid is observed at a given site under an inferred neutral model, the team identified 8 sites on 8 genes in L. campestris and 13 sites on 11 genes in L. sylvestris with EP values below 0.05. Enrichment analysis based on KEGG pathway annotations highlighted the MAPK signaling pathway. The only genes containing multiple such evolutionarily rare substitution sites were Raf, the Raf homolog serine/threonine-protein kinase; Sur-8, a leucine-rich repeat scaffolding protein; and Slit, an upstream inhibitory regulator. Assuming that low EP sites are randomly distributed across variable amino acid sites, the probability of these genes containing multiple sites is p < 0.001 after Bonferroni correction. All three genes encode components of the Ras/MAPK signaling pathway, including elements of the core signaling complex in which RAF interacts with RAS1 and SUR-8 acts as a scaffolding protein.

The functional coherence of these independently identified genes is notable. The Ras/MAPK pathway regulates various downstream genes and is thought to influence many cellular processes, including development, particularly eye development, neural function, and immune responses. In Drosophila, Slit is known to be involved in the development of the nervous system and visual system. The genes with multiple low EP sites also showed significantly elevated FST values, converging with the nucleotide-level signal. The authors caution that this pattern is consistent with, but does not demonstrate, coordinated evolution within a functional module, and that whether these substitutions contributed to reproductive isolation or phenotypic differentiation remains to be determined. Using a published orthopteran mitochondrial substitution rate of 14.988 × 10⁻⁹ substitutions per site per year, the divergence between the two lineages was roughly estimated at about 400 thousand years ago.

The study has important limitations that shape how the results should be read. The sample size per species was modest, which may reduce the precision of population genetic statistics based on within-species variation, and gene-level estimates should be interpreted with caution even though multilocus estimates derived from more than 1,000 genes showed consistent patterns. Mapping to exonic sequences of G. bimaculatus may bias against highly divergent regions, producing conservative estimates of genomic divergence, and the framework did not address duplicated genes, non-coding regions, or the link between genetic changes and phenotypic effects, since no transcriptomic data were available. Alternative quartet supports of approximately 0.5 for both species mean that incomplete lineage sorting or low-frequency interspecific gene flow cannot be ruled out, and whether complete reproductive isolation has evolved between the lineages remains an open question.

Even so, the two crickets emerge as a valuable system for studying the earliest molecular events of speciation. Well-known insect models of divergence with gene flow, such as the apple maggot Rhagoletis pomonella and Timema stick insects, are characterized by pronounced ecological differentiation, making it difficult to isolate the earliest genomic signatures of speciation itself. The Hawaiian cricket Laupala offers a speciation axis based on male song differences and female acoustic preference, but this axis is not universal within crickets. L. campestris and L. sylvestris lack clear ecological and behavioral divergence, and their consistent mitochondrial and nuclear differentiation, despite limited protein-level divergence and few morphological differences, supports their treatment as distinct lineages. Another species in the genus, L. appendicularis, harbors three cryptic species in Taiwan, suggesting that similar genomic analyses of other cryptic Loxoblemmus lineages could reveal whether the Ras/MAPK pattern represents a shared axis of divergence or one of multiple alternative routes to speciation.

Subject of Research: Genomic divergence between two cryptic, geographically overlapping cricket species in Japan

Article Title: Molecular Divergence Between Two Geographically Overlapping Cryptic Crickets: Loxoblemmus campestris and L. sylvestris

Article References: Satomura, K., Saito, M., & Ogura, A. (2026). Molecular Divergence Between Two Geographically Overlapping Cryptic Crickets: Loxoblemmus campestris and L. sylvestris. Ecology and Evolution, 16(10), Article e74457. https://doi.org/10.1002/ece3.74457

Image Credits: AI Generated

DOI: 10.1002/ece3.74457

Keywords: speciation, cryptic species, crickets, Loxoblemmus, mitochondrial genomes, population genomics, Ras/MAPK pathway, gene flow, evolutionary probabilities, FST, incomplete lineage sorting, Japan