Identification of Microplastics
According to Hidalgo-Ruz et al. (2012) visual sorting to separate potential microplastics from other organic or inorganic material in the sample residues is an obligatory step for the identification of microplastics. If large microplastics are the target of a study this can be done by visual inspection (Morét-Ferguson et al. 2010) whereas smaller microplastic particles should generally be sorted out under a dissection microscope (Doyle et al. 2011). Sorting of aqueous samples can be facilitated by the use of sorting chambers (e.g. Bogorov counting chamber). Generally, if no more accurate methods (e.g. FTIR or Raman spectroscopy) are used to verify synthetic polymer origin of potential microplastic particles the
visual identification should not be applied to particles <500 µm as the probability of a misidentification is very high. Hidalgo-Ruz et al. (2012) thus suggest an even higher size limit of 1 mm for visual identification. According to Norén (2007), selection of particles according to standardized criteria in connection with a strict and conservative examination reduces the possibility of misidentification. He suggests the following criteria: (1) no structures of organic origin should be visible in the plastic particle or fibre, (2) fibres should be equally thick and have a threedimensional bending to exclude a biological origin, (3) particles should be clear and homogeneously colored, (4) transparent or whitish particles must be examined under high magnification and with the help of fluorescence microscopy to exclude a biological origin (Norén 2007). General aspects that are used to describe visually sorted microplastics are source, type, shape, degradation stage, and color of the particles (Hidalgo-Ruz et al. 2012).
It is strongly recommended to subsequently analyze sorted particles by techniques that facilitate a proper identification of plastics (Hidalgo-Ruz et al. 2012; Dekiff et al. 2014) because the quality of the data produced by visual sorting depends strongly on (1) the counting person, (2) the quality and magnification of the microscope and (3) the sample matrix (e.g. plankton, sediment, gut content). Another fundamental drawback of visual sorting is the size limitation, i.e. particles below a certain size cannot be discriminated visually from other material or be sorted because they are unmanageable because of their minuteness. Furthermore, visual sorting is extremely time-consuming. In summary, even an experienced person cannot discriminate all potential microplastic particles unambiguously from sand grains, chitin fragments, diatom frustule fragments, etc. Thus the error rate of visual sorting reported in the literature ranges from 20 % (Eriksen et al. 2013a) to 70 % (Hidalgo-Ruz et al. 2012) and increases with decreasing particle size.