Towards quality-assured measurements of microplastics in soil using fluorescence microscopy.

Fluorescence microscopy is increasingly seen as a fast, user-friendly, and high-throughput method for detecting microplastics (MPs) in soil; however, its effectiveness across diverse MP types and soil properties remains underexplored. This study tested a fluorescence microscopy-Nile red (NR) staining approach on eight MP types, covering both biodegradable and non-biodegradable plastics, in three size ranges (≤ 150 µm, 100-250 µm, 500-1000 µm) across loamy, clayey, and sandy soils. Each sample, processed in triplicate, underwent a relatively quick and straightforward extraction procedure involving density separation, organic digestion, and NR staining, followed by fluorescence and bright-field microscopy. A new digital image analysis pipeline using Image J was developed to expedite and (semi)automate MP quantification. Recoveries ranged from 80% to 90% for MPs with a Feret diameter of 500-1000 µm, regardless of soil type. In contrast, the recovery of smaller MPs (Feret dia. ≤ 250 µm) varied depending on the soils and plastic types: recoveries for low-density polyethylene (LDPE) reached 85% in sandy soil and 90% in loamy soil, whereas those for biodegradable polybutylene adipate terephthalate/polylactic acid (PBAT/PLA) were only 60% and 10%, respectively. The lowest recovery rate was observed in clayey soil and for biodegradable plastics. The method was tested on non-agricultural soil samples, yielding a MP mean number concentration of 20.7 ± 9.0 MPs/g for MPs sized from dia. ≥ 25 µm, comparable to Fourier transform infrared (FPA-µ-FTIR) results of 13.1 ± 7.3 MPs/g (p > 0.05). We conclude that fluorescence microscopy with NR staining and automated particle quantification offers a time-efficient, reproducible, and accurate method for MP detection in light-textured soils, whereas limitations remain for reliable MP analysis in clay-dominated soils.