Processing Steps |
- Parameter or Variable: microplastic concentration (measured); Units: counts/individual; Observation Category: in situ; Sampling Instrument: longline with hooks; Sampling and Analyzing Method: Tiger shark stomachs and one spiral valve were opportunistically collected from eight individuals captured in the western North Atlantic Ocean. Seven stomachs (five from Alabama’s Gulf coast, one from near Port Royal, South Carolina, and one from Long Island, New York) and one spiral valve (from Hilton Head, South Carolina) were collected. Stomachs and the spiral valve were dissected whole after securing zip ties on both the cranial and caudal end of the stomach or spiral valve. Specimens were then stored intact in large plastic bags and frozen prior to transportation to the University of Toronto, Canada. Stretch total, fork length, and sex were recorded at the time of sampling. At the University of Toronto, stomachs and the spiral valve were weighed and externally rinsed thoroughly with RO water to remove any potential contamination that may have come from the sample bag. They were then placed in 19 L polypropylene plastic bins for digestion immediately following rinsing. The wet weight (ww) of the tissue was measured using a precision balance (Sartorius Entris® II – Essential Line, Model 3202i-1x). If the mass exceeded the maximum capacity of the balance (3200 g) and the mass could not be measured, the tissue ww was recorded as >3200 g. Specimens and laboratory blanks were fully submerged in 1 μm filtered 20% potassium hydroxide (KOH), with the specimens and KOH solution combined ranging in volume from approximately 1.3 L–12 L. Specimens and blanks were incubated in solution at 45–55 ◦C for 5–7 days, or until completely digested. Digested samples were rinsed through a 25 μm mesh stainless steel sieve and soaked in Contrad® liquid detergent for up to 24 h. Samples were sieved again and placed in a 1.4 g/mL CaCl2 density separation. Floating particles were sieved into three size fractions (>355 μm, 125–355 μm, 45–125 μm). The sieved floating particles for the two larger size fractions (>355 μm, 125–355 μm) were rinsed into glass jars, and the smallest size fraction (45–125 μm) was vacuum filtered onto 20 μm polycarbonate filters. While processing specimens, some suspected anthropogenic particles (i.e., not yet confirmed by chemical identification) were visually observed in the lower portion of the density separation. To better quantify all anthropogenic particles, particles remaining in the lower (dense) portion of the density separation were also sieved and stored in glass jars for analysis with the two larger size fractions (>355 μm, 125–355 μm). However, the lower portion of the density separation contained a large amount of sediment and so it was not feasible to examine on filters for the smallest size fraction (45–125 μm). Therefore, with the exclusion of the lower portion from quantification, the abundance of anthropogenic particles that were quantified in this study likely underestimates total anthropogenic particle contamination. Four laboratory blanks were run in parallel with specimens to estimate procedural contamination and cross-contamination. Under a dissecting microscope (Olympus SZ61 stereo microscope, 6.7X – 45X magnification), suspected anthropogenic particles from all size fractions were quantified visually and described by color and morphology according to a visual identification key. Possible morphologies include fiber, fiber bundle, fragment, film, foam, sphere, pellet, and rubber. A subset of suspected anthropogenic particles (ten particles of each color-morphology combination within each size fraction) were taken from each specimen and mounted on double-sided tape on transparent film for chemical identification (see below). For each specimen, a minimum of ten suspected non-anthropogenic particles were also extracted (across all size fractions) to estimate rates of false negative visual identification. All extracted particles were photographed and measured. Measurements were taken in two dimensions, with the longest dimension representing the length, and the widest dimension perpendicular to the length. The subset of suspected anthropogenic particles was chemically identified to measure the accuracy of our visual identification using attenuated total reflectance - Fourier Transform infrared (ATR-FTIR) and Raman spectroscopy. We selected this method to be representative of broad material type categories (e.g., anthropogenic particles versus non-anthropogenic particles). At minimum, five particles were randomly selected from each color-morphology combination (e.g., black fiber, red fragment, etc.) for chemical identification from each specimen. If fewer than five particles in each color-morphology combination were present, all particles were chemically identified. If greater than 50 particles in each color-morphology were present, 10% of the particles were chemically identified up to a maximum of 20 particles per color-morphology. The subsampled particles were proportionally representative of each size fraction, with at least one particle from each size fraction analyzed per specimen whenever possible. For laboratory blanks, all particles were chemically identified. All suspected non-anthropogenic particles removed from specimens and blanks were also chemically identified. For particles larger than ~300 μm, ATR-FTIR with a diamond internal reference (Bruker Ltd., Milton, ON, CA) operating with OPUS – TOUCH software (Bruker Ltd., version 7.8.44) was used. Infrared spectra were collected at a resolution of 4 cm-1 between 4000 and 400 cm-1. Spectra were averaged over 24 scans, and background scans were recorded (24:1 scans). Particle identification was based on existing spectral library databases (Bruker Ltd.; Primpke et al., 2018). If spectra could not be obtained from the examined particle, or if no spectral reference existed, the particles were re-examined using Raman spectroscopy. For several black rubbery fragments, μ-FTIR was used. Spectra were collected with a Nicolet iN10 infrared microscope (Thermo Fisher Scientific – ATR mode; 15X objective, 0.7 numerical aperture), using a germanium ATR crystal and a cooled mercury cadmium telluride single point detector. For each particle, 32 co-added scans and one background spectrum of the crystal were recorded. Spectral resolution was 4 cm-1, and the spectral range used was 4000–675 cm-1. Resulting spectra were matched to reference materials using the FLoPP and FloPP-E libraries and commercial libraries using the OMNIC Picta software (version 9.11.706 – Thermo Fisher Scientific). A combination of Hit Quality Index (HQI) and visual confirmation were used to assign material type matches. Most particles were small (<300 μm) and had to be analyzed using Raman spectroscopy (Horiba Raman XploRA PLUS confocal microscope, Piscataway, NJ, USA) operating with LabSpec6 software (version 6.5.1.24) and equipped with a charge coupled device detector (60 ◦C, 1024 ×256 pixels). Raman spectra were obtained using a 100X LWD objective (NA = 0.8) resulting in laser powers of 11.2 mW and 20.2 mW at 100% filter for the 532 nm and 785 nm lasers, respectively. Spectral resolution ranged from 1.3 cm-1 (785 nm excitation laser, 600 grooves/mm) to 3.3 cm-1 (532 nm excitation laser, 1200 grooves/mm). Spectra acquired via Raman spectroscopy were assigned database matches using the Wiley KnowItAll and ID Expert spectral matching software (KnowItAll Informatics System, 2022; Analytical Edition) from the KnowItAll Raman Spectral Library as well as the Spectral Library of Plastic Particles (SLoPP and SLoPP-E). Microparticles were sorted into material types based on polymer identity and groupings. Manual and automated software baseline correction was applied (e.g., baseline, vertical clipping, intensity distortion, horizontal offset, vertical offset, Raman intensity distortion). Visual confirmation of similar peak alignment and intensity and HQI score were used to assign spectral matches. Anthropogenic particles were assigned polymer identities based on spectroscopy database matches.; Data Quality Method: Throughout collection of specimens and analyses, measures were taken to reduce the potential for procedural contamination. During specimen collection, samples were exposed to air for as little time as possible and kept in sealed containers following collection. In the laboratory, materials were thoroughly rinsed with reverse osmosis water (RO) a minimum of three times prior to use. All materials were covered with lids and/or aluminum foil when not in use to reduce potential airborne contamination. A white cotton lab coat was worn throughout sample processing and quantification. The laboratory is equipped with a HEPA filter to reduce airborne contamination. One laboratory blank was acquired for every two specimens processed, and laboratory blanks were carried through the entire process from sample extraction to quantification and chemical identification to estimate the procedural contamination. Laboratory blanks were processed in the same manner as the shark specimens, using the same materials and containers and the same approximate volumes of added solutions. A representative subsample of all suspected anthropogenic particles was chemically identified using spectroscopy. Blank contamination is reported, but final data is not blank- or spectroscopy-corrected..
|