What a Pest Moth’s Gut Bacteria Reveal About the Power of Plant Diets

Inside the midgut of a caterpillar, a hidden ecosystem rises and falls with every bite of food. A new study of the meadow moth, Loxostege sticticalis, one of the most destructive migratory pests of grasslands and crops in Asia, shows that simply changing what the larvae eat can dramatically reshape the bacterial communities living in their guts. The research, published in Frontiers in Insect Science, compared larvae raised on three botanically distinct host plants and an artificial diet, and found that each diet left a recognizable fingerprint on the insects’ internal microbial world. The findings add an important piece to one of the most intriguing puzzles in modern entomology: whether the gut microbes of herbivorous insects are passive passengers picked up from food, or active partners that help their hosts exploit an extraordinary range of plants.

Loxostege sticticalis, commonly known as the meadow moth or beet webworm, is a polyphagous caterpillar, meaning it feeds on plants spanning multiple botanical families. It migrates across vast distances between China, Russia, and Mongolia, and its outbreaks can devastate alfalfa, sugar beet, and other crops. That dietary flexibility makes it an ideal subject for studying how diet shapes gut microbiota. Insects’ gut bacteria have been implicated in digestion, nutrient acquisition, immune regulation, pathogen resistance, and the detoxification of plant defensive chemicals. Yet unlike termites, whose microbial partners are indispensable for breaking down wood, many caterpillars appear to harbor sparse, transient bacterial communities acquired from their environment rather than stable, inherited symbioses. Where the meadow moth falls on this spectrum has been unclear.

A team at Beijing Forestry University’s School of Grassland Science, led by Dong Li, Jianyu Wang, Dandan Zhang, Weichan Cui, and Haijun Xiao, set out to fill that gap. They established a laboratory colony of the moth from adults collected in Hohhot, in Inner Mongolia, and reared it under tightly controlled conditions of temperature, humidity, and light. Newly hatched larvae from the same parental cohort were randomly assigned to one of four diets: leaves of Chenopodium album, a common weed; Medicago sativa, alfalfa; Beta vulgaris, sugar beet; or a standard artificial diet. Alfalfa and sugar beet seedlings were grown hydroponically to standardize their conditions, and only young, undamaged, insect-free leaves were offered. Each diet was treated as a composite exposure, encompassing both the plant’s own traits and any microorganisms riding along on or inside the leaves.

When the larvae reached their fifth and final instar, the researchers selected externally healthy individuals, sterilized their cuticles with ethanol, and dissected out the midgut, the central digestive compartment where most nutrient absorption occurs. For each diet, they pooled midguts from twenty larvae into a single biological replicate, producing three pooled replicates per treatment. DNA was extracted and the V3–V4 hypervariable region of the bacterial 16S rRNA gene was amplified and sequenced on an Illumina platform. The team processed more than 1.2 million raw reads, retaining nearly a million high-quality sequences that resolved into 1,738 unique amplicon sequence variants across the dataset. Extraction blanks and no-template PCR controls produced no amplification, reducing concerns about contamination.

The starkest result emerged from the diversity analysis. Larvae fed alfalfa carried significantly richer and more diverse midgut bacterial communities than those on any other diet, as measured by the Chao1 richness estimator and the Shannon diversity index. In contrast, larvae fed Chenopodium album showed a comparatively simplified community, with fewer dominant taxa. The artificial diet and sugar beet groups fell in between. At the phylum level, two bacterial groups dominated across all treatments: Bacillota and Pseudomonadota, a pattern consistent with previous surveys of lepidopteran guts. But the proportions shifted with diet. Chenopodium-fed larvae were strongly dominated by Bacillota, while alfalfa-fed larvae displayed a broader taxonomic profile with a notable increase in Actinomycetota, a group known for versatile metabolism, including the degradation of complex compounds and the production of secondary metabolites.

At the genus level, the differences became even more distinctive. Alfalfa-fed larvae showed elevated abundances of Curtobacterium, Methylobacterium, Glutamicibacter, Bifidobacterium, Streptococcus, and Corynebacterium. The artificial diet group was characterized by Carnobacterium, Serratia, and Providencia, while sugar beet-fed larvae carried more Klebsiella, Stenotrophomonas, and Enterobacter in some replicates. A machine-learning-based technique called LEfSe, which identifies taxa that distinguish groups by both statistical significance and effect size, pinpointed Actinomycetota-related features, including Corynebacterium sputi, as characteristic of the alfalfa group, and Bacillota-related features as characteristic of the artificial diet group. No taxa met the thresholds in the Chenopodium or sugar beet groups, which the authors caution reflects the criteria applied rather than proof that such taxa are absent.

The statistical picture is nuanced and the authors are careful not to overstate it. A principal coordinate analysis based on Bray–Curtis dissimilarities showed that the first two axes explained 34.6 percent and 25.9 percent of the variation, with samples distributed differently across treatments. A PERMANOVA test on the full dataset found a significant overall association between diet and community composition, explaining roughly 72 percent of the variation. However, exact pairwise permutation tests between any two diets did not reach significance, each yielding a P value of 0.100, a limitation the researchers attribute to having only three pooled replicates per group. In other words, the evidence supports a diet-associated difference in community structure overall, but cannot yet single out which specific pair of diets differs.

Using PICRUSt2, a tool that predicts metabolic potential from taxonomic data, the team also found diet-associated differences in predicted functional profiles. Pathways related to amino acid metabolism and the metabolism of terpenoids and polyketides were predicted to be more abundant in alfalfa-fed larvae, while Chenopodium-fed larvae showed higher predicted abundances of carbohydrate, nucleotide, and glycan metabolism pathways. Differences also appeared in the metabolism of cofactors and vitamins and in xenobiotics biodegradation, the latter being particularly intriguing because xenobiotic metabolism could, in principle, relate to processing plant defensive compounds. Comparable diet-driven shifts in predicted functions have been reported in other polyphagous moths, including the oriental fruit moth Grapholita molesta and the fall armyworm Spodoptera frugiperda, suggesting the pattern is not unique to the meadow moth.

The authors are explicit about the boundaries of interpretation, and this restraint is what makes the study scientifically valuable. Because the plant-associated microbiota on leaf surfaces and inside tissues were never characterized, and because the nutritional and chemical composition of the diets was not quantified, the study cannot determine whether the observed gut differences arose from direct acquisition of food microbes, from selective conditions imposed by plant chemistry, or a combination of both. Previous work suggests these mechanisms are not mutually exclusive: in fall armyworm, even axenic larvae fed gamma-irradiated maize still showed genotype-dependent gut communities, implying that plant traits alone can filter gut bacteria. Moreover, PICRUSt2 predictions are inferences, not measurements of genes, transcripts, or enzyme activities, and benchmarking studies have shown their accuracy varies considerably across sample types.

What the study delivers is a rigorous descriptive baseline: clear evidence that botanically distinct diets are associated with distinct midgut bacterial communities in a major migratory pest, along with a shortlist of candidate taxa and pathways for future functional work. The researchers call for follow-up studies that integrate plant microbiota profiling, plant chemistry, metagenomics, metabolomics, and measurements of host performance to distinguish microbial input from dietary filtering and to test whether any of these bacteria genuinely contribute to host-plant adaptation. For now, the meadow moth’s gut offers a compelling reminder that in the ecology of herbivorous insects, the line between what an insect eats and what lives inside it is drawn in shifting sand, and that understanding pest biology may require reading the microbial script written by every meal.

Subject of Research: Diet-associated variation in the midgut bacterial communities of a polyphagous moth larva

Article Title: Diet-associated variation in the midgut bacterial communities of Loxostege sticticalis larvae fed botanically distinct host plants

Article References: Li, D., Wang, J., Zhang, D., Cui, W., & Xiao, H. (2026). Diet-associated variation in the midgut bacterial communities of Loxostege sticticalis larvae fed botanically distinct host plants. Frontiers in Insect Science, 6, Article 1949678. https://doi.org/10.3389/finsc.2026.1949678

Image Credits: AI Generated

DOI: 10.3389/finsc.2026.1949678

Keywords: Loxostege sticticalis, gut microbiota, 16S rRNA sequencing, host plants, polyphagous pest, midgut bacteria, Bacillota, Actinomycetota, PICRUSt2, alfalfa, Lepidoptera, diet-microbiota interactions