**Background:** Psyllids are phloem-feeding insects that can transmit phytoplasmas, intracellular bacteria causing plant diseases such as Pear Decline. The insect microbiome can influence physiology and vector competence. Three sympatric pear psyllid species in Central Europe — Cacopsylla pyri, C. pyricola, and C. pyrisuga — differ in voltinism, overwintering strategy, and vector competence for 'Candidatus Phytoplasma pyri'. C. pyri and C. pyricola are polyvoltine (3–5 generations/year) and overwinter predominantly on pear trees, while C. pyrisuga is univoltine with an obligate overwintering association with conifers. This study aimed to characterize and compare their microbiomes and investigate the impact of host ontogeny, seasonal generations, and geography.
**Methods:** A total of 151 psyllid individuals (C. pyri n=36, C. pyricola n=79, C. pyrisuga n=39) and three egg samples of C. pyrisuga were collected from February 2020 to February 2021, primarily from a pear orchard in Starý Lískovec (Brno, Czech Republic), with additional specimens from other Czech localities, Italy, and France. DNA was extracted and the V3–V4 region of the 16S rRNA gene was amplified using primer pair 341F/805R and sequenced on Illumina MiSeq (2×300 bp). Reads were processed with QIIME2/dada2, taxonomy assigned via Silva v132. Alpha diversity (Chao1, Shannon) was compared using t-tests with Monte Carlo permutations. Beta diversity was assessed via PCoA (Bray-Curtis) and ANOSIM, with finer-scale differences evaluated using NMDS and EnvFit. Indicator species analysis (IndVal) and differential abundance (edgeR) identified taxa driving species differences. A maximum likelihood phylogeny of dominant Enterobacteriaceae ASVs was constructed using IQ-TREE.
**Key Results:** Sequencing yielded 2,352–59,068 high-quality reads per sample (mean 24,080), clustered into 4–72 ASVs per sample (mean 15.88). The three psyllid species harbored significantly different microbiomes (ANOSIM R=0.927, p=0.0001; 40.50% of variation explained). At the co-occurrence locality (CZ2), only 7 of 1,046 ASVs were shared among all three species. Shannon diversity was significantly higher in C. pyri (mean 2.39) compared to C. pyricola (mean 2.02; t-test p=0.006) and C. pyrisuga (mean 2.14; p=0.042), but Chao1 richness did not differ. All species harbored Carsonella ruddii (0%–21.8% of reads due to primer bias) and a dominant Enterobacteriaceae endosymbiont. C. pyri and C. pyricola shared a closely related Enterobacteriaceae taxon (Group1, 58.90% and 89.90% of reads, respectively) that formed a clade with endosymbionts of C. jukyungi (BS=97), C. myrthi, and Anomoneura mori. C. pyrisuga carried a distinct Sodalis-related Enterobacteriaceae (Group2, 87.56% of reads) clustering with endosymbionts of C. burckhardti and Cardiaspina maniformis (BS=96). Indicator species analysis identified 10 taxa: six for C. pyri (Group1 pyri, Wolbachia, Streptococcus, Pediococcus, Microbacterium, Fimbriimonas), two for C. pyricola (Group1 pyricola, Gluconobacter), and two for C. pyrisuga (Group2 pyrisuga, Rickettsia). Five additional taxa were differentially abundant (Bradyrhizobium, Methylobacterium, Sphingomonas, Variovorax, Pajaroellobacter). In C. pyri, microbiome composition varied significantly with the interaction of developmental stage and generation (EnvFit r²=0.243, p=0.004) and with locality (r²=0.330, p=0.001). Wolbachia relative abundance varied by locality (CZ1: 20.19%, CZ2: 65.42%, CZ3: 3.02%, Italy: 0.10%) and by ontogeny at CZ2 (summer adults: 84.43%, overwintering adults: 70.29%, summer immatures: 41.54%). In C. pyricola, developmental stage significantly affected microbiome composition (r²=0.089, p=0.002), and the interaction of stage and generation was also significant (r²=0.164, p=0.005). Immatures had higher richness and diversity than adults (Chao1 p=0.001; Shannon p=0.002). Wolbachia was abundant only in spring-form immatures (83.19% after removing dominant endosymbionts), while autumn-form immatures harbored Gluconobacter (24.01%), Pseudomonas (15.58%), and Serratia (13.46%). Phytoplasma was detected in one C. pyricola female (93% of reads). In C. pyrisuga, no significant effects of developmental stage, seasonality, or geography on microbiome composition were found. Eggs showed distinct clustering by NMDS. Overwintered adults (remigrants) uniquely harbored Arsenophonus (41.74%), and Rickettsia reached 79.35% in immatures from CZ3.
**Clinical Implications:** This study provides foundational knowledge of the microbiomes of three sympatric pear psyllid vectors with different vector competences for 'Ca. Phytoplasma pyri'. The distinct endosymbiont profiles — particularly the shared Enterobacteriaceae in C. pyri and C. pyricola versus the Sodalis-related symbiont in C. pyrisuga — may relate to differences in vector efficiency. The presence of Wolbachia in C. pyri (varying by locality and ontogeny) and Arsenophonus and Rickettsia in C. pyrisuga warrants further investigation into their potential roles in modulating phytoplasma transmission. The stability of the C. pyrisuga microbiome across life stages and localities contrasts with the variability in the polyvoltine species, suggesting that host biology (voltinism, overwintering strategy) strongly shapes microbiome structure. These insights could inform future strategies for managing Pear Decline by targeting symbiont–vector interactions.