Microplastics affect sedimentary microbial communities and nitrogen cycling

Abstract

Microplastics are ubiquitous in estuarine, coastal, and deep sea sediments. The impacts of microplastics on sedimentary microbial ecosystems and biogeochemical carbon and nitrogen cycles, however, have not been well reported. To evaluate if microplastics influence the composition and function of sedimentary microbial communities, we conducted a microcosm experiment using salt marsh sediment amended with polyethylene (PE), polyvinyl chloride (PVC), polyurethane foam (PUF) or polylactic acid (PLA) microplastics. We report that the presence of microplastics alters sediment microbial community composition and nitrogen cycling processes. Compared to control sediments without microplastic, PUF- and PLA-amended sediments promote nitrification and denitrification, while PVC amendment inhibits both processes. These results indicate that nitrogen cycling processes in sediments can be significantly affected by different microplastics, which may serve as organic carbon substrates for microbial communities. Considering this evidence and increasing microplastic pollution, the impact of plastics on global ecosystems and biogeochemical cycling merits critical investigation.

Fig. 1. Microcosm experimental design.

All five treatments were repeated in triplicate and each microcosm was individually aerated to establish an oxygen gradient in the sediment.

Fig. 2. Principal coordinate analysis of the sediment communities.

Beta diversity was calculated using the Bray–Curtis dissimilarity index, and is plotted for all sample dates: initial (asterisk), 7 days (triangle), and 16 days (circle); and treatments: control (red; CON), PE (green), PVC (purple), PUF (orange), and PLA (blue). Significant effects of the plastic treatment (p = 0.001), day (p = 0.001) and interaction (p = 0.023) were tested by PERMANOVA.

Fig. 3. Bacterial community composition and treatment effects.

Comparison of taxonomic differences (family level) in bacterial communities with different microplastic treatments. Stacked bar plot of the relative abundance of families (>1% abundance) for each plastic treatment (averaged for the three replicates, n = 3 per treatment) for each sediment collection date (0, 7, and 16 days), where CON is the control treatment, a. Families that are significantly different in relative abundance between treatments and controls (averaged across collection dates), determined using DeSeq (α = 0.01), is illustrated in b, showing if a family is significantly greater in one of the plastic treatments (blue) or the control (red).

Fig. 4. Dissolved inorganic nitrogen concentrations in water.

Concentrations (µM) of NO₃⁻ a, NO₂⁻ b, and NH₄⁺ c are shown for each microplastic treatment and control microcosms after 7 and 16 days of incubation (n = 3 per treatment). Error bars are standard error and CON is the control treatment. Initial community (n = 1) concentrations are 0.072, 0.527, and 3.44 µM for NO₃⁻, NO₂⁻, and NH₄⁺, respectively. Statistical analyses can be found in Supplementary Tables 2–4.

Fig. 5. Nitrification and denitrification gene abundances.

The genes encoding ammonia monooxygenase (amoA, a) and nitrite reductase (nirS, b, and nirK,c) were quantified and normalized to 16S rRNA genes. Error bars are standard error (n = 3 per treatment) and CON is the control treatment. Initial community ratios are 1.85e−5, 3.03e−2 and 3.25e−4 for amoA, nirS, and nirK. Statistical analyses can be found in Supplementary Tables 6–8.

Fig. 6. Comparison of potential denitrification rates.

Potential denitrification rate for each treatment in nmol g⁻¹ hr⁻¹, calculated after the end of the experiment (day 17). Error bars are the standard error (n = 6 per treatment) and asterix represent significant difference from the control (p < 0.05; Supplementary Table 9).