**3. Results and discussion**

#### **3.1 Toxicology: method used to determine if a site is safe or not**

According to the main quoted authors in the compared articles, it can be stated that Lithner et al. [8] and Everaert et al. [9] laid the foundations for the current estimation of ERA related to MPs. The first authors [8] established a hazard level for each polymer based on the toxicity of their monomers, while the second ones [9] proposed to compare the predicted no-effect concentration (PNEC) (to which the RQ was added later), calculated with a species sensitivity distribution (SSD) analysis (a statistical model that is used for other chemical compounds and nanomaterials [10]) with the predicted environmental concentration (PEC).

Some improvements have been made to the method proposed by Everaert et al. [9]. For example, Besseling et al. [11] limited the articles on which they based their SSD to those that studied endpoints assumed to affect population size; Burns and Boxall [12] used ecotoxicity data limited to 10–5000 μm particle size exposures; and Jung et al. [13] screened the toxicity data used for the SSD to fragment and fiber MPs in the size range of 20–300 μm while measuring (instead of predicting) the MPs concentration in a marine environment.

On the other hand, there have been no important improvements to the path taken by Lithner et al. [8] probably because it is difficult to integrate toxicity information

for all monomers or additives associated with polymers. However, many articles are based on the principles they implemented. The methods used in the reviewed articles will be explained below.

## *3.1.1 Hazard index (HI) and pollutant load index (PLI)*

The hazard index (HI) (Eq. (1)) was proposed by Xu et al. [14].

$$HI = \sum P\_n \* S\_n \tag{1}$$

Where HI is the total polymer risk index caused by MPs, Pn is the specific MP polymer type percentage at each sampling site, and Sn is the polymer hazard score established by Lithner et al. [8], that is based on the toxicity of its monomers. Eq. (1) does not consider the concentration of the polymer; therefore, it would not be correct to use it without a complementary index. For this reason, Xu et al. [14] proposed to incorporate the pollutant load index (PLI), an equation proposed by Tomlinson et al. [15] to observe the pollution difference between zones on the same waterbody.

$$\text{PLI} = \sqrt{\mathbf{C}\_f^i} \tag{2}$$

$$\mathbf{C}\_{f}^{i} = \mathbf{^{C}} \boldsymbol{\%\_{a}} \tag{3}$$

$$\text{PLI}\_{zone} = \sqrt[n]{\text{PLI}\_1 \text{PLI}\_2 \text{PLI}\_n} \tag{4}$$

Where Cf i is the MP concentration factor, Ci is the MP abundance of the sampling site and Coi is the minimum average concentration that may be on the available literature or the minimum concentration found on the study area [15].

Eqs. (1) and (4) were used to develop an Ecological Risk Assessment by 62.3% of the reviewed articles [14, 16–18], but Picó et al. [16] made a modification to Eq. (2); *PLI* <sup>¼</sup> *<sup>C</sup><sup>i</sup> <sup>f</sup>* . These studies could have been comparable easily if all the authors would have used the same Coi (in particles per liter p/L) value or the same equation. This makes the use of data difficult or impossible to compare [14], but some modifications will be made to their calculations, so a comparison is made in the present review.

#### *3.1.2 Potential ecological risk index (RI)*

This factor was introduced to the microplastic risk analysis by Pan et al. [17].

$$E\_r^i = T\_r^i \* C\_f^i \tag{5}$$

$$T\_r^i = {}^{P\_i}\!\!/ {}\_{\mathbb{C}\_i} \* \mathbb{S}\_n \tag{6}$$

$$RI = \sum\_{i=1}^{n} E\_r^i \tag{7}$$

Where Ei <sup>r</sup> is the potential ecological risk index that represents the potential ecological hazard from single MPs polymer, T<sup>i</sup> <sup>r</sup> is the ecotoxicity response factor, Pi is the proportion of each individual MPs polymer type, Sn is the hazard score of MPs polymer type (i) given by Lithner et al. [8], n is the number of types of MPs polymers contained in the sample, and RI is the potential ecological risk from combined MPs

polymers [17, 18]. It is important to highlight that the calculation of RI allows combining the toxic effects of the polymers and the MPs concentration of the study area. Eqs. (5)–(7) were used by 25% of the reviewed articles [17, 18], and provide additional information for future studies to complement HI and PLI (1)–(4).

Eqs. (1)–(7) are based only on the toxicological information of the polymers [8], without considering the additive, size, form, and other aspects that could generate toxicity in an organism. Therefore, these articles will only have a score of 1 in the toxicology column displayed in the next section (**Table 2**).

The interpretation of the described indexes can be observed in **Table 3**.

The hazard level for the HI calculation observed in **Table 3** is based on Lithner et al. [8], which classifies these levels with the toxicity data of the monomers (Sn). It is important to note that a polymer is not easily depolymerized to the monomeric state during use and even less so when discarded in nature [23]. Additionally, this is an issue that the plastic industry has been improving for many years along with increasingly stringent regulations especially for polymers used in the food industry (packaging and others) which are among the most discarded and found in the environment [24]. Still, the authors [8] argue that polymerization reactions during production are rarely complete, therefore, unreacted residual monomers can be found in the polymeric material.

The PLI hazard level was originally developed by Tomlinson et al. [15] for estuarine environments specifically for bioaccumulation of heavy metals on intertidal algae or fauna. Later, Angulo [25] assessed the articles that calculated the PLI for different organisms and assigned them a risk category which is now used for MP concentrations in the water column. This value must be used carefully as it was originally developed for other purposes but using it in a standardized manner (establishing an international Coi or giving it specific thresholds according to the pollution level of the studied waterbody) could yield comparable studies around the world.

The potential ecological risk index and risk (Ei and RI) are a combination of the chemical toxicity coefficient for each polymer type introduced by Lithner et al. [8] and the Cf i (MP concentration factor) introduced by Tomlinson et al. [15] and can be taken as the most complete approach until now. As it is a new perspective only estimated by Refs. [17, 18] the risk categories must be calibrated with studies made around the world in more and less polluted areas.
