A mechanistic framework to quantify sampling uncertainty of soil microplastics: integrating spatial heterogeneity and...

Area of Science:

• Environmental Science
• Soil Science
• Analytical Chemistry

Background:

• Soil microplastic (MP) contamination is widespread, but data on its statistical representativeness is lacking.
• This knowledge gap hinders accurate environmental surveys and risk assessments of soil pollution.
• Current understanding of sampling biases in soil MP analysis is limited.

Purpose of the Study:

• To develop a mechanistic framework for quantifying soil microplastic sampling uncertainty.
• To assess the impact of particle size, contamination levels, and sampling strategies on data representativeness.
• To provide guidance for improving the accuracy of soil MP measurements.

Main Methods:

• Integration of a multinomial patch model for spatial heterogeneity in field sampling.
• Application of a Poisson model to account for laboratory subsampling uncertainty.
• Simulation using realistic ranges of MP contamination, particle size, patch volume, and sampled mass.

Main Results:

• Fine microplastics (≤100 μm) require at least 100g of soil for <20% uncertainty.
• Coarse microplastics (≥500 μm) exhibit high variability (>100%) and significant false-negative probabilities (up to 80%) under low abundance and limited sample volumes.
• Laboratory subsampling with small aliquots inadequately reduces uncertainty for larger particles at low contamination levels.

Conclusions:

• Sampling strategies significantly influence the statistical representativeness of soil microplastic data.
• Particle size and sampled mass are critical factors determining data accuracy and reliability.
• The developed framework offers practical guidance for designing more robust soil microplastic sampling protocols.