Research report: A community-level view of Wolbachia DNA positivity across mosquito assemblages in eastern China

A field survey of 2,434 mosquitoes from three localities in eastern China found Wolbachia DNA in 419 specimens, an overall detection frequency of 17.2% (95% confidence interval, 15.8–18.8%). The pooled frequency concealed substantial variation among mosquito assemblages: positivity was 11.3% in Huai’an, Jiangsu, 60.8% in Lu’an, Anhui, and 90.0% in Xuzhou, Jiangsu. The study used taxon-resolved screening and molecular profiling to examine whether these differences could be explained by the species composition of the local mosquitoes.

Compositional calculations indicated that host composition accounted for much of the contrast, although a locality-associated component remained for some hosts. The collections were opportunistic, unbalanced among sites and conducted in different months, so the results characterize these sampled assemblages rather than establishing independent geographic effects. The collection comprised nine mosquito taxa and was strongly dominated by five species: Culex pipiens, Anopheles sinensis, Culex tritaeniorhynchus, Armigeres subalbatus and Aedes albopictus together represented 98.6% of all specimens.

Huai’an supplied 2,196 mosquitoes, or 90.2% of the total, and contained seven taxa. Its assemblage was dominated by Cx. pipiens, with 1,314 specimens, followed by An. sinensis with 440 and Cx. tritaeniorhynchus with 362. Lu’an also contained seven taxa, but Ae. albopictus, Ar. subalbatus and Cx. tritaeniorhynchus were its most abundant members.

Xuzhou produced only 90 mosquitoes from three taxa, including 86 Ar. subalbatus. This marked variation in community structure means that a site-level percentage combines the detection frequencies of individual hosts with the abundance of those hosts in the local collection. Among taxa represented by more than 100 mosquitoes, Ar. subalbatus had the highest detection frequency, with 110 of 134 specimens positive, or 82.1% (95% CI, 74.7–87.7%).

Aedes albopictus followed with 60 of 114 positive, or 52.6% (43.5–61.6%), while Cx. pipiens had 224 of 1,316, or 17.0% (15.1–19.1%). Wolbachia DNA was rarely detected in An. sinensis, with five of 442 mosquitoes positive, and Cx. tritaeniorhynchus, with one of 394. Culex pipiens, Ar. subalbatus and Ae. albopictus together accounted for 394 of all 419 positive specimens.

Coquillettidia crassipes, Cx. sitiens and Mansonia uniformis also had high apparent detection frequencies, but each was represented by fewer than 10 mosquitoes, so the investigators reported positive and tested counts rather than treating those values as reliable prevalence estimates. The researchers calculated the positivity expected from each locality’s species mix by weighting global taxon-specific detection frequencies by local composition. Expected values were 12.4% for Huai’an, 50.9% for Lu’an and 78.9% for Xuzhou, compared with observed frequencies of 11.3%, 60.8% and 90.0%.

The modeled composition closely tracked the observed Huai’an value and accounted for most of the elevation at the other two sites, leaving residuals of approximately 10 percentage points in Lu’an and Xuzhou. Host-resolved tests confirmed locality-associated heterogeneity for Ae. albopictus and Ar. subalbatus but not Cx. tritaeniorhynchus. For Ae. albopictus, positivity was 14.3% in Huai’an and 81.5% in Lu’an; for Ar. subalbatus, it was 28.6%, 68.3% and 93.0% across the three sites.

The apparent Cx. pipiens contrast depended on two positive specimens from Lu’an and did not remain significant after correction across all taxa with multi-site data. Beyond detection screening, the investigators sequenced the Wolbachia surface-protein gene, wsp, and five standard multilocus sequence typing loci: gatB, coxA, hcpA, ftsZ and fbpA. A mosquito counted as positive when wsp or at least one MLST locus amplified and yielded an unambiguous sequence.

Of the 419 positive specimens, 331 produced wsp sequences and 106 had usable sequence data at all five MLST loci. The wsp analysis resolved two major lineages corresponding to Wolbachia supergroups A and B. Supergroup A formed a compact group associated with Ar. subalbatus from Lu’an and Xuzhou and shared a single pattern across the four analyzed hypervariable regions of wsp.

Supergroup B was more heterogeneous, with one large cluster detected in all three localities and in Ae. albopictus, Cx. pipiens and Cx. quinquefasciatus, while another cluster was restricted to Ae. albopictus from Lu’an. Two were treated as database-consistent profiles, while 28 could not be matched confidently to existing sequence types. The unmatched profiles were assigned to evidence classes according to whether every allele was confirmed by an exact full-length match, whether a call required partial sequence or represented a divergent allele, or whether a profile rested on a provisional singleton.

Twenty-one of the 28 unmatched profiles occurred only once. Sensitivity analyses showed that incomplete sequencing could account for part of the apparent diversity: 15 profiles remained when all five loci had at least 90% of their canonical sequence length, and 10 remained at a 95% threshold. Frequent profiles were more robust, including one unmatched profile recovered from 35 mosquitoes, with 22 of those samples providing exact full-length matches at all five loci.

Most inferred profiles remained associated with a single mosquito taxon. Cross-host sharing was concentrated among frequent profiles rather than spread according to host relationships. An exploratory reconstruction using 500 stochastic character maps placed Ae. albopictus at the root of the sampled profile tree with a posterior probability of 0.994.

Excluding Ae. albopictus-associated profiles shifted the most probable root state to Ar. subalbatus, with a probability of 0.686. Because the reconstruction was sensitive to uneven taxon sampling and based on partial markers, the investigators described it as a summary of the sampled tree rather than evidence for an ancestral host origin. Network and incompatibility analyses detected some discordance among loci but did not demonstrate genomic recombination.

A NeighborNet analysis represented conflicting relationships among marker combinations as reticulations. An exploratory RDP5 screen identified 12 putative breakpoint or incongruence intervals involving 11 of the 30 profiles, and 11 intervals crossed at least one locus boundary. Since the five MLST loci occupy separate genomic locations and were concatenated only for analysis, such intervals could not be interpreted as continuous genomic recombination tracts.

A four-gamete scan found 32.8% of informative biallelic site pairs incompatible, compared with a 95.8% expectation under free association. Incompatibility was more frequent between loci than within loci, but most of that difference disappeared after excluding fbpA, the marker with the poorest sequence recovery. The observations were compatible with mosaic-like marker structure, but could not distinguish biological recombination, horizontal allele transfer, mixed bacterial templates or sequencing artifacts.

Several limitations define what can be inferred from the three-site field survey. PCR detection demonstrates the presence of Wolbachia DNA, not a stable, vertically transmitted infection; low-titer or transient detections, environmental material and bacterial fragments integrated into mosquito genomes can all produce positive amplification. Many MLST amplicons were incomplete, and direct Sanger sequencing may not resolve mixed Wolbachia templates within individual mosquitoes.

Opportunistic collection, unequal trapping effort, different sampling months and highly unequal host representation limited separation of locality from host effects. Marker profiles also do not establish compatibility phenotypes, transmission efficiency or pathogen-blocking properties. Within these boundaries, the study shows that community-level Wolbachia surveillance requires taxon-resolved reporting rather than reliance on one pooled percentage.

It also documents an incompletely sampled operational allelic reservoir, particularly in Ar. subalbatus and Ae. albopictus, that can guide fuller molecular and functional investigation.

Subject of Research: Community-level Wolbachia DNA detection, host association and marker diversity in eastern Chinese mosquito assemblages

Article Title: A community-level view of Wolbachia DNA positivity across mosquito assemblages in eastern China

Article References: Yang, P., Ni, W., He, F., Yan, Q., Ni, Z., Peng, X., Gao, T., Lv, M., Yang, S., Zhang, R., Liang, L., Hu, W., & Kang, Y.-J. (2026). A community-level view of Wolbachia DNA positivity across mosquito assemblages in eastern China. Frontiers in Microbiology, 17, Article 1906293. https://doi.org/10.3389/fmicb.2026.1906293

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1906293

Keywords: Wolbachia, mosquitoes, eastern China, community composition, DNA detection, host association, multilocus sequence typing, molecular ecology, microbial symbiosis, community-level, view, positivity