Insights from a Long-Term Outdoor Mesocosm Study: eDNA Metabarcoding Reveals Exacerbated but Transient Impacts from a Nanoenabled Pesticide Formulation (Nano-TiO2-Coated Carbendazim) on Freshwater Microbial Communities

Abstract

Fungicides currently encompass the second-most-used class of agricultural pesticides globally. Residues are frequently detected in freshwater, leading to undesired ecological impacts. Nanoenabled pesticide formulations have recently gained prominence in the scientific literature and have been suggested to exhibit favorable properties over conventional pesticide formulations by facilitating reductions in emissions toward nontarget locations. However, data on unintended effects on nontarget aquatic communities are scarce, especially concerning microbial communities. In this study, long-term effects of nano titanium-dioxide- (nTiO₂)-coated carbendazim and its constituents on (pelagic) freshwater microbial communities in simulated agricultural ditches were investigated over a period of 14 weeks using environmental DNA (eDNA) metabarcoding. Impacts on bacterial diversity (α and β) were observed 2 weeks after the treatment application and most pronounced in the nTiO₂-coated carbendazim treatment, followed by its active substance, i.e., noncoated carbendazim. The observed patterns possibly imply that nTiO₂-coated carbendazim imposed more pronounced and potentially delayed or extended effects compared to the noncoated form of carbendazim. Bacterial communities also proved to be resilient under the tested conditions as they returned to the control-state within 5 weeks after the treatment application. Overall, our data suggest that eDNA metabarcoding data on microbial communities can help uncover time-dependent effects of nanoformulated pesticides.

Figure 1

Bacterial community composition data per treatment over the course of the experiment. (A) Relative read abundance (RRA) visualized per phylum—or class within phylum for Proteobacteria—for the eight most abundant groups. Each bar represents a treatment within a time point, showing the average RRA from the five replicates. Treatments are labeled as following: z = control, a = nTiO₂, b = carbendazim, c = nTiO₂ and carbendazim, d = nTiO₂-coated carbendazim; (B) boxplot visualizing family richness per treatment over time. Significant differences between treatments within time point 2 are indicated by letters, with boxes sharing a letter being not significantly different from one another; (C) boxplot visualizing read abundance-based Shannon–Weiner index scores per treatment over time. Values are based on summed reads per family; (D) PCoA plots visualizing Sørensen (dis)similarity. Centroids represent mean PCoA scores per treatment and polygons are drawn around PCoA scores of individual replicates (i.e., ditches). Boxplots and PCoAs can be found enlarged in the Supporting Information.

Figure 2

Phytoplankton community composition data per treatment over the course of the experiment. (A) Relative read abundance (RRA) visualized per phylum—or class within phylum for Ochrophyta—for the eight most abundant groups. Each bar represents a treatment within a time point, showing the average RRA from the five replicates. Treatments are labeled as following: z = control, a = nTiO₂, b = carbendazim, c = nTiO₂ and carbendazim, d = nTiO₂-coated carbendazim; (B) boxplot visualizing order richness per treatment over time; (C) boxplot visualizing read abundance-based Shannon–Weiner index score per treatment over time. Values are based on summed reads per order; (D) PCoA plots visualizing Sørensen similarity. Centroids represent mean PCoA scores per treatment and polygons are drawn around PCoA scores of individual replicates (i.e., ditches). Boxplots and PCoAs can be found enlarged in the Supporting Information.