The assessment of microplastic and microfibres in freshwater systems through different sampling methods reveals causes of incomparability.

Around twenty years of studies on microplastic pollution have revealed a major environmental concern. However, far from understanding the presence of microplastics in environmental matrices, abundances among studies differ highly. This is not only caused by the inherent variability of this pollution in aquatic ecosystems, but also because the use of different methodologies adds large uncertainties. This study assesses microplastics data and examines the differences induced by the methods used. A literature mining was performed in Web of Science to find relevant studies on microplastics in freshwater aquatic ecosystems worldwide. Out of 501 relevant (peer-reviewed) articles found in freshwater systems, 200 articles were selected for analysis, i.e., those offering data results per sample rather than summarizing per areas or studies. Such selection comprised 4297 samples from freshwater systems in the five continents. A wide range of concentrations of microplastics was detected worldwide (spanning 8 orders of magnitude). Grouping microplastic concentrations by sampling methods (nets, pumps, and bulk sampling) narrowed the variability distributions, particularly for nets. To elucidate the driving variables behind these changes, factors associated to each method were examined, showing that the main differences in the methods and concentrations obtained were related to the amount of water volume sampled, the mesh size (or minimum size reported), and whether microfibres were considered in the studies. Concentrations were highly and negatively correlated with the volume sampled (cor = -0.82; p < 0.001). This pattern was maintained within each sampling method. Differences of several orders of magnitude were found in the abundances obtained depending on the volume sampled, irrespective of the sampling instrument used. While the typical particle size distribution indicates that the smaller the particles, the larger the number, this was not the case when lower sampling volumes (< 0.1 m³) were grouped by minimum size reported. Furthermore, analysis by particle type (microplastics particles versus microfibres) showed a predominance of microplastics particles in the higher volume samples, while this was not observed in the lower volume samples. Depending on the method used, when microfibres are reported, the variability in abundances may not reflect environmental distributions, adding large variability and differences in particle size distributions and type of microplastics. Results obtained from lower volume sampling may be biased, e.g., influenced by cross-contamination of microfibres, because small variations in particle counts could magnify errors when extrapolated to larger volumes. This study shows that concentrations of microplastics can be comparable, regardless of sampling approach used, if the limitations of the methodology are known in relation to the volume sampled, the size spectrum reported and whether microfibres are counted.