Light Rewrites the Microbial Communities That Colonize Buried Plastic

Twelve months in glass jars have revealed something quietly dramatic about the microbes that gather around buried plastic. Researchers at Wesleyan University filled 36 modified Winogradsky columns with soil from a decommissioned landfill in Portland, Connecticut, seeded half of them with sterilized polyethylene strips, and let half of each group sit in ambient laboratory light while the rest were wrapped in aluminum foil. When the team sequenced the communities that had formed, light and dark columns had traveled down distinctly different ecological paths, and the light-exposed communities carried a markedly larger inventory of genes associated with plastic-degrading enzymes.

The experiment, published in Frontiers in Environmental Science, exploited a classical microbiology tool with a modern twist. Winogradsky columns recreate the vertical gradients of oxygen, nutrients and light found in natural soils within a transparent vessel. The team layered landfill soil in 500-milliliter jars at the 100, 200 and 300 milliliter marks, topped the columns with Bushnell Haas broth, an inorganic medium with no carbon source, and slid polyethylene strips horizontally at each layer interface. Crucially, the strips were positioned within the soil rather than against the illuminated glass wall, so the plastic itself received no direct light. Any effects of illumination on the strip communities were therefore indirect, mediated by the microbes living near the jar walls.

The soil itself told a story about plastic persistence. The researchers collected their inoculum in fall 2022 from a legacy landfill used only between 1962 and 1988, digging to roughly 1.5 meters, where they encountered intact and partially degraded plastic sheets still in large pieces after four decades of burial. Working alongside local authorities, they extracted soil from around these materials, anticipating that resident microbes would have adapted to life alongside synthetic polymers. Global plastic production has surpassed 8,300 million metric tons, with up to 60 percent accumulating in landfills or the environment, where it fragments into microplastics and nanoplastics that threaten soils, groundwater and ecosystems.

After a year of incubation, the taxonomic differences between light and dark columns were unmistakable. In the 16S rRNA profiles, dark columns were dominated by Pseudomonadota, Bacteroidota, Acidobacteriota and Patescibacteria, while light exposure promoted photoautotrophic Cyanobacteriota and Chloroflexota. Cyanobacteriota appeared only under light conditions, and were slightly more abundant on plastic strips than in surrounding soil. Chloroflexota showed a striking substrate dependence: in soil without plastic they were more abundant in the dark, but on plastic strips they increased under light. Dark conditions instead favored anaerobic Bacillota and Bacteroidota. Pseudomonadota reached their highest abundance on plastic strips in dark conditions, at just over 50 percent of the community.

The metagenomic data, gathered from the plastic treatments only, echoed these patterns. Pseudomonadota remained the most abundant phylum, and Actinomycetota, Gemmatimonadota and Cyanobacteriota all showed positive log2 fold changes under light, while Bacteroidota and Pseudomonadota were relatively more abundant in the dark. The authors note that the two sequencing approaches diverged for a few taxa, consistent with known methodological differences such as primer bias, 16S copy-number variation and the greater sensitivity of shotgun sequencing to archaea and rare lineages. They treat the amplicon data, with 110 samples after quality filtering, as the replicated backbone of the study, and the metagenomes as a complementary window onto functional genes.

Diversity results came with an important caveat about statistical power. In the fully replicated 16S dataset, neither species richness nor Shannon diversity differed significantly between light and dark conditions or between plastic and soil substrates, with p-values ranging from 0.381 to 0.902 and Cohen’s d effect sizes near zero. Higher diversity under light appeared only in the smaller, exploratory metagenomic dataset of three samples per condition, where effect sizes were large but replication minimal. The team is careful not to overstate this exploratory signal, and the most defensible conclusion from the diversity analysis is that light reshaped community composition rather than uniformly raising or lowering its diversity.

The functional findings are where the study becomes most provocative. Matching assembled gene names against curated databases, the researchers identified 353 putative plastizymes, plastic-degrading enzyme genes, in their metagenomes. Light-exposed plastic treatments carried an average of 228 such genes compared with 125 under dark conditions. Nearly every plastizyme category was more frequent in the light, and oxidoreductases showed the strongest response, with roughly 16,500 genes under light versus about 14,800 in the dark, an increase of about 11 percent that was statistically significant. This matters because oxidoreductases are implicated in the oxidative transformation of polyethylene, and oxidative steps are thought to govern the pace of subsequent depolymerization by hydrolases and other enzymes, which also increased modestly under light.

Light also acted as a regulatory signal in the metagenomes. Circadian rhythm pathways showed the largest shift of any category, with an average log2 fold change of about 3.0 under light, followed by flavone and flavonol biosynthesis at roughly 2.5 and phototransduction pathways at about 2.0. Light-sensing pathways rose by about 1.2 and pigment biosynthesis by about 0.4, while oxidative stress response, energy metabolism and light-harvesting pathways barely moved. The authors favor a cautious interpretation of the circadian signal: because metagenomics measures gene abundance rather than transcription, the pattern most likely reflects the higher relative abundance of phototrophic taxa such as Cyanobacteriota that carry these genes, rather than activation of a regulatory program within existing residents.

The limitations are clearly acknowledged. The study measured genetic potential, not enzyme expression or actual polymer breakdown; no mass loss or molecular weight changes were recorded, and confirming active degradation would require metatranscriptomics, proteomics, enzyme assays and polymer characterization. The metagenomic analysis covered only six samples, all from plastic treatments, and the plastizyme assignments relied on name matching rather than sequence-level validation, so some matched genes may be versatile homologs with no activity on polyethylene. The columns, moreover, cannot fully reproduce the heterogeneity of an active landfill, and the communities represent an experimentally assembled plastisphere seeded by legacy soil rather than the original buried-sheet community. Even so, the work suggests that even limited illumination can enrich communities with enhanced potential for oxidative polyethylene transformation, an insight that could inform light-informed bioremediation strategies, particularly for plastics near landfill surfaces, while deep, permanently dark strata remain beyond the reach of these light-driven effects.

Subject of Research: Light exposure reshapes microbial community assembly on plastic in Winogradsky columns

Article Title: Light exposure reshapes microbial community assembly on plastic in Winogradsky columns

Article References: Olabemiwo, F. A., Huang, Y., Thompson, M., Ryan, S., Omar, H., Bashaw, T., Kraut, K., Arevalo, P., Nazzaro, V., & Cohan, F. M. (2026). Light exposure reshapes microbial community assembly on plastic in Winogradsky columns. Frontiers in Environmental Science, 14, Article 1824803. https://doi.org/10.3389/fenvs.2026.1824803

Image Credits: AI Generated

DOI: 10.3389/fenvs.2026.1824803

Keywords: Light, exposure, reshapes, microbial, community, assembly, plastic, Winogradsky, columns, scientific research