The beach sediments used in the comparison of protocols were collected from Guiguiwanen Beach beside the Bolinao Marine Laboratory of The Marine Science Institute, University of the Philippines in Bolinao, Pangasinan ( Figure 1 ) using a small metal shovel and a metal container. The top 5 cm were collected, dried in an oven at 80°C, sieved through a 5.6-mm mesh metal sieve, and labeled as ‘sieved beach sediments’.

Sediments free from plastic contamination, hereinafter referred to as ‘clean’ beach sediments, are essential components of experimental controls for microplastics quantification (i.e., blanks and spiked samples). To determine the most efficient way to prepare ‘clean’ beach sediments, three methodologies were compared. One hundred grams (100 g) of sieved beach sediments were transferred to beakers and aluminum pans, which were then spiked by adding a known number of different common plastic types reported in Philippine-based surveys (PlastiCount Pilipinas, 2022). These include polystyrene (PS) foam, polypropylene (PP) fragment, low-density polyethylene (LDPE) film, high-density polyethylene (HDPE) fragment, polyamide (PA/nylon) fiber, and polyethylene terephthalate (PET) fragment. The microplastics used for spiking were prepared from commonly available consumer plastic items. These were cut and shredded to varying sizes ranging from 1-5 mm and their polymer type verified through Attenuated Total Reflectance - Fourier Transform Infrared (ATR-FTIR) spectroscopy using the Bruker ATR-IR Polymers, Plastics and Additives Library, Kunststoff-Institut Lüdenscheid database and ATR-FTIR Polymer Library (Bruker Optics).

Ten microplastics from each of the six plastic categories were added to each sediment sample (Han et al., 2019). Here, three methodologies for preparing ‘clean’ beach sediments were tested and compared. Method 1 involved the addition of saturated NaCl (ρ = 1.2 g/mL) followed by subsequent vigorous stirring for 2 minutes (Bridson et al., 2020). The mixture was allowed to settle for at least 5 hours before decanting the supernatant to remove floating plastics. The addition of extracting solution, allotment of time for settling, and decantation of the supernatant were repeated thrice. Saturated NaCl solution was used as the density separation solution as it is a more economical and practical approach; higher density solutions are usually expensive and are not cost-efficient for the sole purpose of preparing ‘clean’ beach sediments. On the other hand, Method 2 involved subjecting the beach sediments in aluminum pans to a furnace at 550°C for 4 h (Schütze et al., 2022). Lastly, Method 3 employed a density separation method using saturated NaCl (single extraction) followed by subjecting the beach sediments to a furnace at 550°C for 4 hours (Shim et al., 2016), as shown in Figure 2 . Each procedure was performed in seven replicates.

The methodologies by Thompson et al. (2004), Coppock et al. (2017), and Masura et al. (2015) are commonly used approaches for extracting microplastics from beach sediments but with noted variabilities. To optimize and validate the efficiency of the methods when applied to tropical beach sediments, the three methodologies were compared as shown in Figure 3 .

The most efficient protocol for the preparation of ‘clean’ beach sediments as determined in the previous section was employed. For this part, 100 g of ‘clean’ beach sediments were transferred to a beaker. Ten microplastics from each of the seven plastic categories, namely: polystyrene (PS) foam, polypropylene (PP) fragment, low-density polyethylene (LDPE) film, high-density polyethylene (HDPE) fragment, polyamide (PA/nylon) fiber, polyethylene terephthalate (PET) fragment, and polyvinyl chloride (PVC) shavings were added to each sediment sample ( Figure 4 ). A small amount (~0.02 g) of dried seaweed (Sargassum spp.) and driftwood were also added to the mixture to simulate the presence of some naturally occurring organic material in environmental samples. Each methodology employed triplicate samples.

Thompson et al. (2004) employed density separation using saturated NaCl (ρ = 1.2 g/mL). For this, saturated NaCl was added to the spiked sediment samples and the mixture was stirred for 30 seconds. After allowing the sediments to settle for 2 minutes, the supernatant was filtered, and the residue was collected and counted.

Coppock et al. (2017) made use of a Sediment-Microplastic Isolation (SMI) unit, shown in Supplementary Figure 1 , to facilitate the density separation using a solution of ZnCl₂ (ρ = 1.5 g/mL). The SMI unit was initially primed through repeated washing using ZnCl₂ solution. Spiked sediment samples were then added to the SMI unit followed by the addition of ZnCl₂ solution. A magnetic stir bar was also added to agitate the mixture for 5 minutes followed by the settling process for another 5 minutes. To remove possible trapped air bubbles in the sediments, three short pulses were introduced through the magnetic stir bar. The mixture was then left to stand. When the supernatant was clear with sediments, the valve was closed, and the mixture in the upper half of the SMI unit containing the floating particles was transferred to a beaker and subsequently filtered. The residue was collected and counted.

The procedure by Masura et al. (2015) with slight modifications as sugested by Besley et al. (2017) was employed for the third set of spiked samples. The general workflow involved an initial two-round density separation step through the addition of ZnCl₂ solution (ρ = 1.5 g/mL), followed by wet peroxide oxidation to remove organic matter, then another two-round density separation using ZnCl₂ to facilitate the isolation of microplastics from the remaining sediments. The detailed flowchart is shown in Supplementary Figure 2 .

This study employed protocols to minimize and account for the possibility of contamination. As much as possible, the use of plastic materials was avoided during sampling and laboratory processing. Gloves were worn throughout the experiment. Laboratory work was performed in the Microplastics Quantification, Identification, and Biodegradation (MicroQuIB) facility in the Bolinao Marine Laboratory (BML) of the UP MSI, a dedicated clean facility for plastics research equipped with an air purifier to minimize the possibility of air contamination. Blanks for air contamination were also set up. Prior to the conduct of the experiment, the working table was wiped with a paper towel and 70% ethanol; lint remover was also used on the clothes and lab gowns of the researchers before entering the laboratory. Samples being processed were covered with aluminum foil or glass petri plates whenever possible. Lastly, reagents used in the experiment were pre-filtered through GF/C filters (1.2-µm pore size) prior to usage.

Efficiencies of methodologies were expressed through percent removal (for preparation of controls) and percent extraction efficiencies. To test for significant differences between treatments, one-way analysis of variance (ANOVA) was employed at α = 0.05 and implemented in PAST version 4.13 (Hammer et al., 2001). Tukey’s pairwise comparison was used as the post-hoc test.

Experimental controls, both positive (spiked samples) and negative (blanks), are essential QA/QC measures in any type of experiment. In microplastics research, spiked samples are used to assess how well the methodology was performed (Miller et al., 2017). Spiking involves the addition of a known amount of microplastics to ‘clean’ beach sediments. Spiked samples then undergo the usual extraction protocol and the amount of microplastics recovered by the end is noted; a high recovery rate serves as an assurance of the quality of the conduct of the experiment (Primpke et al., 2022). Additionally, spike-recovery tests may also be used, as in this experiment, to validate the effectiveness of methods in extracting microplastics from environmental matrices (Way et al., 2022). On the other hand, blanks are employed to account for extraneous contamination from the environment, reagents, and equipment used in processing samples (Dawson et al., 2023). Sample blanks essentially contain everything in the sample matrix except the analyte, in this case, microplastics (Cantwell, 2019). This allows for a realistic estimation of contamination as the sediment samples are being processed. However, obtaining a reference material in the form of ‘clean’ beach sediments is very difficult and not commercially plausible (Silva et al., 2018; Cadiou et al., 2020). While some studies have employed their own protocol for ‘cleaning’ beach sediments (Shim et al., 2016; Bridson et al., 2020; Schütze et al., 2022), no controlled comparisons have yet been made that can inform future harmonization efforts. Thus, there is a need to optimize towards an efficient method for removing plastic contamination in beach sediments.

Results summarized in Figure 5 showed that two of the three protocols tested exhibited high efficiencies in removing plastic contamination in sediments, allowing for its use in experimental controls. Specifically, only Method 1 exhibited relatively low removal efficiencies, which also varied depending on polymer type. Method 1 operated on the principle of physical separation through density differences among the sediment matrix, extracting solution, and microplastics. The low average removal efficiency of nylon fibers (76.2%) was observed, which may be related to its morphology; the small diameter of fibers relative to its length makes it very difficult to extract, as seen in many spike-recovery tests (Crichton et al., 2017; Cashman et al., 2020). The density of nylon (1.13 – 1.15 g/mL) is also near the maximum density of saturated NaCl solution (1.2 g/mL) ( Table 1 ), and the presence of additives could have increased the fibers’ density towards or beyond 1.2 g/mL, lowering its removal efficiency (Nuelle et al., 2014). Lastly, the fibers’ susceptibility to the effects of static electricity facilitates its adherence to the sides of the glass container, which could have resulted in a lower removal rate (Catarino et al., 2017). Meanwhile, PET fragments also recorded a very low average removal efficiency (18.1%), owing to its higher density (1.37 – 1.39 g/mL) than the extracting solution. It is for this reason that other extracting solutions with a higher density such as zinc chloride (ZnCl₂), sodium iodide (NaI), and zinc bromide (ZnBr₂) are generally considered for microplastics extraction, albeit deemed impractical for the sole purpose of preparing ‘clean’ beach sediments for controls due to their higher costs (Nuelle et al., 2014; Prata et al., 2019a). Removal efficiencies beyond 100% may be attributed to errors in counting during spike addition, since the thinness of LDPE films makes it easy to mistake two films as one. Also, some of the less rigid microplastics (i.e., LDPE films and PS foams) are prone to fragmentation during processing, resulting in higher counts after the extraction (Cashman et al., 2020).

Methods 2 and 3, both of which relied on high temperatures in a furnace to facilitate the removal of microplastics from beach sediments, showed 100% removal efficiency for all polymer types ( Figure 5 ). Thermal decomposition of the polymer types used in this experiment occurs at temperatures between 350°C and 500°C (Beyler and Hirschler, 2002; Dümichen et al., 2017; Johnson et al., 2020; Ni et al., 2020). However, to account for the possible increase in thermal decomposition temperature due to the presence of additives in plastics, this study increased the employed temperature up to 550°C. Since Method 2 showcased a 100% removal efficiency with just a single step compared to Method 3, it was considered as the most efficient way of preparing ‘clean’ beach sediments for experimental controls in microplastics extraction.

Once ‘clean’ beach sediments are available for experimental controls, extraction of microplastics from environmental samples, along with spike-recovery tests, can then be conducted. However, due to the high degree of variability among properties and composition of microplastics, it is also reasonable to expect certain differences in recoveries depending on the methodology employed. For instance, Cashman et al. (2020) in a similar study noted that microplastics larger in size are generally well-extracted, which may point towards observer bias as an important factor affecting counts among studies. Additionally, matrix characteristics such as organic matter content may also influence the quality of extraction (Cashman et al., 2020).

Here, the modified protocol of Masura et al. (2015) exhibited an extraction efficiency of 100% for all plastic types except PA/Nylon (93.33%), as shown in Figure 6 . This observation seems to be likely associated with the use of a heavier extraction solution (ZnCl₂, ρ = 1.5 g/mL) compared to Thompson et al. (2004), and the implementation of wet peroxide oxidation to digest organic matter. This is in congruence with the findings of other studies, suggesting that the removal of organic matter contributes to higher percent recovery during extraction, as organic matter can interact and aggregate with microplastics (Cashman et al., 2020; Constant et al., 2021). The main advantage of this methodology is that it significantly reduces the amount of contaminating organic matter that may complicate further processing and analyses of microplastics, without damaging the integrity of the polymers (Prata et al., 2019b). This is especially important if other methods aside from the traditional, microscopy-based counting (i.e., Nile Red staining and spectroscopic methods) will be employed, as the presence of organic matter, particularly biofilms, can interfere with the signals from the microplastics. As shown in Figure 7 , notable amounts of organic matter (i.e., Sargassum spp. and driftwood) remained in samples that did not undergo a digestion step. Meanwhile, the sample that underwent wet peroxide oxidation showed less organic matter, mostly in the form of undigested driftwood. Hydrogen peroxide (H₂O₂) excels in digestion of plant organic matter, although studies have noted that animal tissues and cellulosic materials such as driftwood are not removed as efficiently (Dyachenko et al., 2017; Herrera et al., 2018; Prata et al., 2019b). For environmental samples containing substantial amounts of animal tissues, a digestion protocol with 10% KOH has been suggested (Prata et al., 2019b).

Meanwhile, the relatively high extraction efficiency for Thompson et al. (2004) despite only using saturated NaCl solution may be a result of the reduced heteroaggregation between the sediment matrix and the microplastics, as induced by ‘cleaning’ the beach sediment samples through a furnace prior to mixing with the microplastics. Generally, suspended sediments can form heteroaggregates with microplastics and facilitate their sinking, resulting in a lower extraction efficiency (Li et al., 2019). However, studies have shown that at temperatures beyond 500°C, the physical properties of the soil are altered and that its aggregative properties are disrupted, allowing for a more efficient extraction through decantation (Zihms et al., 2013). This was also visually noted during the performance of the experiment. Overall, these observations may point towards the potential influence of matrix properties (e.g., sediment grain size, organic matter content) on the efficiency of extraction, which was also validated by similar studies (Cashman et al., 2020; Constant et al., 2021).

Interestingly, the overall mean extraction efficiency for the protocol of Coppock et al. (2017) was lower (90.48 ± 2.97%) compared to Thompson et al. (2004) (95.71%) despite using a heavier extracting solution, albeit statistically insignificant (p = 0.09) ( Supplementary Figure 3 ). This also manifested individually across polymer types such as PP (96.67% vs. 100%), LDPE (93.33% vs. 100%), PA/Nylon (70% vs. 90%), and PVC (80% vs. 93.33%). One possible explanation for the lower recovery of Coppock et al. (2017) may be related to the equipment and the interactions it facilitates through its use. For instance, it was visually noted that the unrecovered microplastics were either adhered to the inside walls of the SMI unit or trapped within the ball valve. This was the case despite priming the SMI unit with ZnCl₂ solution prior to extraction, as suggested by Coppock et al. (2017), and despite repeated washing after extraction. The same observation was reported by Nel et al. (2019) wherein they suggested not utilizing the ball valve at all and detaching the upper chamber for a more thorough rinsing than when it is attached. The adherence of microplastics to the walls of the SMI unit may be a function of hydrophobic and/or electrostatic interactions and the morphology of the microplastics (Lechthaler et al., 2020; Al Harraq and Bharti, 2022). Among the microplastics used in recovery tests, LDPE films, PVC shavings, and nylon fibers had the highest surface area-to-volume ratio, making them more susceptible to the effects of hydrophobic and/or electrostatic forces (Enders et al., 2020). Thus, it is certainly possible that morphology plays an important role in facilitating interactions between the surface of the plastic and other components during processing. In some studies, modifications were introduced to the SMI unit mostly by replacing the original plastic PVC design into other materials (e.g., glass, aluminum) to avoid plastic contamination and hydrophobic interactions; however, aluminum-based SMIs are known to be susceptible to corrosion when ZnCl₂ is used as the extracting solution (Nakajima et al., 2019; Lechthaler et al., 2020). On another note, a higher extraction efficiency for PET (96.67%) was observed for Coppock et al. (2017), which was expected given that PET is less dense than the ZnCl₂ extracting solution. Comparing the overall mean extraction efficiencies among the three protocols tested ( Supplementary Figure 3 ), the modified methodology of Masura et al. (2015) was considered the most effective protocol for microplastics extraction from beach sediments.

It is important to acknowledge that the microplastics used in these experiments were neither weathered nor biofouled, which could possibly be more representative of microplastics in the natural environment. However, a sufficient representation was achieved by sourcing microplastics from commonly available consumer plastic items, as listed among the most common plastic types documented in baselining surveys in the Philippines (PlastiCount Pilipinas, 2022). Also, the beach sediment type of the sampling site may not capture the diversity of beach sediment types in the tropics. Lastly, the effect of the furnace on the integrity of the sediment was beyond the scope of this study.

Global awareness on microplastics pollution only started in the 1970s, when researchers were supposedly sampling for Sargassum spp. yet came across a lot of small plastic pellets in their neuston nets (Carpenter and Smith, 1972). Since then, more and more studies have documented the presence of microplastics in different environments around the world, a testament to its ubiquity (Rochman, 2018). As a response, global policy-making bodies have listed marine plastic pollution among their priority concerns (Bank et al., 2021). The United Nations Sustainable Development Goal 14: Life Below Water (UN SDG 14) identified pollution by marine plastics under Target 14.1.1b, albeit only to a limited degree for microplastics (Walker, 2021). The United Nations Environment Assembly (UNEA) in March 2022 adopted a resolution to work towards an “international legally binding instrument” to combat plastic pollution, recognizing microplastics as part of the equation (United Nations Environment Assembly resolution 5/14, 2022). Meanwhile, the detailed G20 Report on Actions Against Marine Plastic Litter acknowledges the need to establish harmonized monitoring efforts to better assess the efficiency of countermeasures being employed (MoEFCC, 2023). While these are laudable achievements in terms of global policies, the translation of these frameworks into actions on the regional and local settings remains suboptimal, as microplastics research still suffer drawbacks due to the lack of harmonization in sampling, processing, and characterization (Hartmann et al., 2019).

For Southeast Asia and the Philippines, the field of plastics research is still considered in its infancy. A deeper look into the status of microplastics research in the Philippines reveals the considerably varied approaches among baselining efforts (Galarpe et al., 2021). Table 2 demonstrates how protocols for processing and characterizing microplastics from beach sediments differ among local researchers, which is also prevalent in other countries in the region (Omeyer et al., 2022).

There are several challenges to the pursuit of harmonization in the Philippines, the most obvious one being the valid technological constraints (Galarpe et al., 2021). The cost of reagents and equipment for microplastics quantification and characterization is an important consideration. For instance, NaCl is a much cheaper option as the salt for density separation than ZnCl₂ or NaI; however, using NaCl runs the risk of underestimating the heavier plastic types given its lower density at saturation (ρ = 1.2 g/mL). Other extracting reagents such as oil are also being explored as a cheaper and greener alternative to dense salt solutions, although extraction efficiencies remain variable across studies; not to mention its potential interference to subsequent spectroscopic analyses (Bellasi et al., 2021; Constant et al., 2021). Despite the array of options available for extraction solutions, it can be argued that harmonization in this aspect is still possible even in consideration of the cost. The data from the PlastiCount Pilipinas portal (https://plasticount.ph), which is the most comprehensive database of all plastics research in the Philippines to date (Alindayu et al., 2023), suggest a significantly higher abundance of lighter macroplastics such as PE and PP compared to the heavier plastics such as PET and PVC. Assuming that most of the microplastics in Philippine beaches originated from the fragmentation and degradation of these macroplastics (Van Emmerik and Schwarz, 2019), this data could possibly imply that using saturated NaCl as a density separation solution may provide a sufficient estimate of microplastics pollution, at least for the lighter plastic types. Therefore, the comparison between studies employing different extracting solutions (saturated NaCl vs. ZnCl₂) can then still be valid even in just a few plastic types (i.e., plastics with ρ ≤ 1.2 g/mL), as long as the other steps in the protocol were done similarly and that limitations have been acknowledged and considered.

Another technological constraint is the limited access to well-designed and equipped laboratories for microplastics research. This is important, as air contamination of microplastics is a well-established phenomenon that is known to affect the quality of the results (Prata et al., 2021). Access to equipment for spectroscopic methods such as Fourier Transform Infrared (FTIR) spectroscopy and Raman spectroscopy, commonly employed to determine polymer composition, is also limited to very few laboratories or facilities. However, with the proper implementation of experimental controls, background contamination and procedural errors can be adequately accounted for. It is important to note that the harmonization being put forward is geared towards adopting a possible array of options, albeit limited, instead of a one size fits all-approach, in consideration of the diversity of local environmental contexts in the Philippines as well as the differences in resource availability and accessibility.

Given the understandable difficulties in harmonizing methodologies due to a lot of things that need to be considered, a more readily applicable aspect of harmonization can be implemented in terms of the categories used for data reporting. Microplastics extracted from environmental samples are usually described in terms of color, morphology/shape, size, and polymer type (Rocha-Santos and Duarte, 2015). In the Philippines, the most common categories employed in beach sediment studies are summarized in Table 2 . Harmonization in the categories used in data reporting will allow for a more effective comparison among studies and integration of results; in fact, this is already recognized by other regions globally, such as the European Union, whose guidelines are currently being updated to incorporate harmonized categories for data reporting (Bäuerlein et al., 2023).

There is also the aspect of harmonization in expertise and skill. Even if methodologies and data reporting protocols are harmonized, a high degree of variability between the researchers’ skills due to differences in the degree of training, experience, and expertise in executing protocols could lead to difficulties in comparability of results (Cadiou et al., 2020; Kotar et al., 2022; Piccardo et al., 2022). This highlights the need to conduct interlaboratory calibration – a harmonization exercise commonly employed as part of QA/QC measures to evaluate the reliability of results across participating laboratories not just within the Philippines but also at the regional level (Tsangaris et al., 2021). To do this, an uncontaminated sediment sample and a sediment sample spiked with a known amount of microplastics, are sent to participating laboratories for processing and subsequent quantification. The results returned by the laboratories are then collated to assess reproducibility of methodologies and comparability of results. Studies that have conducted interlaboratory calibration noted that the operators’ experience seems to influence the quality of extraction, particularly in smaller size classes (Cadiou et al., 2020; Piccardo et al., 2022). An interlaboratory study by Cadiou et al. (2020) revealed that underestimation due to losses of microplastics during handling and processing may be a much greater source of error than overestimation due to extraneous contamination among the participating laboratories. These kinds of insights emphasize the importance of interlaboratory comparisons and consequently, the necessity of an efficient methodology to prepare ‘clean’ beach sediments, which this study was able to achieve.

The challenges faced by microplastics research in the Philippines is a problem shared with other developing countries not just in Southeast Asia but also worldwide (Hartmann et al., 2019; Curren et al., 2021; Omeyer et al., 2022). The paucity of baselines brought about by unharmonized protocols impede our progress to a better understanding of the problem (Omeyer et al., 2022). As a consequence, efforts to address plastic pollution at the policy level suffer from the lack of a realistic direction. In the Philippines, for example, the overarching goal of the National Plan of Action for the Prevention, Reduction, and Management of Marine Litter (NPOA-ML) was set to “Zero waste to Philippine waters by 2040”, yet there is a clear recognition of the lack of comprehensive baselines with which the evaluation of progress is dependent on (Department of Environment and Natural Resources – Environmental Management Bureau, 2021). The problem of the lack of harmonization is likewise recognized on the regional level through Component II of the ASEAN Regional Action Plan for Combating Marine Debris in the ASEAN Member States (2021-2025) under “Research, Innovation, and Capacity Building” (ASEAN Secretariat, 2021). For this, the academe and research institutions will play huge roles in providing guidance and support in the form of science-based recommendations. In the Philippines, a network of plastics researchers in the Philippines was just recently launched as the Plastics Research Network (PlaReNet) Philippines (PlastiCount Pilipinas, 2022), with the objective of assisting in the harmonization of plastics research efforts in the country and aiding the implementation of the NPOA-ML. On the regional level, an international network of experts on marine plastic pollution published a comprehensive set of research agenda and recommendations on tackling marine plastics in the region (Omeyer et al., 2022). This study hopes to contribute to these undertakings by providing information on the potential methodologies that may be considered for harmonized protocols as evaluated on tropical beach sediment samples.