Water Research 240 (2023) 120106

Contents lists available at ScienceDirect

Water Research
journal homepage: www.elsevier.com/locate/watres

Microplastics occurrence and fate in full-scale treatment wetlands

A R T I C L E I N F O A B S T R A C T

Keywords: Treatment wetlands (TWs) are an efficient technology for removing microplastics (MPs) from wastewater, ac­
Constructed wetland cording to previous studies. This study investigates the dynamics and fate of MPs in two wastewater treatment
Roots plants (WWTPs) using TWs, one with horizontal subsurface flow (HF) and another with a floating plant system
Vegetation
(FS). Special attention is paid to the retention produced in the sludge and the role of macrophyte roots. The
Nature-based solutions
abundance of MPs in the influent to the WWTPs was on average 20.3 ± 0.85 MP/L and 8.4 ± 1.13 MP/L in HF
Sludge
Polymers and FS respectively, while the effluent had 0.58 ± 0.07 MP/L and 0.17 ± 0.06 MP/L, thus giving overall effi­
ciencies of 97.42% and 98.13%, respectively. In the HF wetland, sludge samples near the inlet and the outlet
were taken, distinguishing between sludge adhered to gravel and sludge attached to roots. In the floating
macrophytes, sludge samples from secondary and tertiary treatments were taken. The results indicate that roots
play a significant role in MPs retention. In the HF wetland, the complex formed by roots and gravel attached
more MPs than gravel alone in the final zone of the wetland. In the FS, roots retained a significant quantity of
MPs, both in the secondary and tertiary treatments, thus giving rise to a sludge less concentrated in MPs. This
study aims to improve the knowledge of MPs behavior and fate in full-scale TWs, providing valuable information
to enhance retention efficiency.

1. Introduction than 150 different plant or animal species (Jabeen et al., 2017), infil­
trating the food chain. Additionally, some studies have shown that the
Plastics have been recognized as a worldwide environmental prob­ presence of MPs in organisms can cause reduction of growth and
lem for decades (Carpenter and Smith, 1972). However, plastic mate­ photosynthesis, hinder reproductive ability, damage intestines, and
rials are very advantageous, being more malleable, durable and cost even cause death (Lu et al., 2016; Wang et al., 2019; Yang et al., 2020).
effective (Andrady and Neal 2009). This is why millions of tons of plastic Nowadays, researchers have found that one of the main sources of
are produced annually and have been increasing over recent years MPs to the marine habitats are wastewaters (Conley et al., 2019). MPs
(Plastic Europe, 2021). As a consequence, a wide range of plastic debris coming from cosmetics or toothpaste that contains microbeads, fibers
are accumulating in the biosphere (terrestrial and marine), which can be from synthetic textiles that are released in washing machines (Boucher
fragmented into smaller pieces due to the influence of different agents and Friot, 2017), and other articles like glitter, tires and cleaners or
(sunlight, wind, mechanical, biological, or chemical forces) (Silva et al., painting products (Sheavly and Register, 2007), reach the wastewater
2018). Plastic particles smaller than 5 mm are considered microplastics treatment plants (WWTP). Once there, most are removed from waste­
(MPs) and are classified into primary and secondary MPs according to water, given that the efficiency varies between 80 and 99% (Prata,
their origin (Thompson et al., 2004). Once in the environment, MPs can 2018). Despite this, even if the MP output concentrations are low, the
act as a means of transportation for other pollutants, such as heavy high discharge flow rates of WWTPs result in a significant number of
metals (Sarkar et al., 2021) or persistent organic pollutants (POPs), MPs being released into the natural environment, causing considerable
among others. Furthermore, due to its small size, MPs can be consumed pollution throughout the day. Moreover, if the WWTPs removal of MPs
by a variety of aquatic fauna. Indeed, MPs have been detected in more are effective, it means that MPs are transferred into sewage sludge (van

* Corresponding author.
E-mail address: [email protected] (C. Hernández-Crespo).

https://doi.org/10.1016/j.watres.2023.120106
Received 3 March 2023; Received in revised form 19 May 2023; Accepted 20 May 2023
Available online 21 May 2023
0043-1354/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-
nc/4.0/).
D.C. Cabrera et al. Water Research 240 (2023) 120106

der Berg et al., 2020). Indeed, the concentration in the sludge is higher 2. Materials and methods
than influent wastewaters (Xu et al., 2020). When the sludge is then
applied in farmlands, it represents a major input to the environment 2.1. Sampling sites
(Hernández-Arenas et al., 2021), for instance MPs may reach water
bodies through runoff from the farmlands. 2.1.1. Carrícola’s WWTP
For a long time, treatment wetlands (TWs) have been investigated as Carrícola’s WWTP (Valencia, Spain) serves a community of 115
a method to remove pollutants in wastewater. TWs are designed systems population equivalents (P.E.). The scheme of this plant is a pretreatment
to treat wastewater and other pollution sources with low maintenance (grids), followed by two Imhoff settlers (18 m3) connected in series with
and operation cost (Vymazal, 2010). They represent a nature-based pipes to the inlet of two horizontal subsurface flow TWs, operating in
solution for wastewater treatment in small villages, for tertiary treat­ parallel, of 200 m2 each and a depth of 0.35 m. They are filled with
ment in bigger plants and for dealing with stormwater (Rousseau et al., coarse gravel (Φ = 40 mm) from the inlet to a distance of one meter and
2008). TWs have demonstrated to be competent in the reduction of MPs then are followed by a mix of coarse and fine gravel (10–25 mm). They
with an efficiency between 88%− 95% when they are used as a tertiary are filled with common reed plants (Phragmites). There is a second TW of
treatment, adding to its already known benefits (Wang et al., 2020; the same configuration (TW2) following TW1 to extend the exposer time
Zhou et al., 2022; Bydalek et al., 2023). Previous researchers have found of the water (Fig. 1).
a good interception of MPs by vegetation in natural wetlands, whether in Previous studies report an average daily inflow of 16 m3/day, a total
leaves, roots or attached in form of biofilm (Yin et al., 2021; Duan et al., suspended solids (TSS) removal efficiency of 98%, with an average value
2021). Macroinvertebrates found in TWs, mainly worms, can also play a of 3.1 ± 3.7 mg/L in the final effluent, and chemical oxygen demand
significant role in the distribution of the MPs (Wang et al., 2020). (COD) removal of 94%, with an average value of 42± 19 mg/L in the
Furthermore, other studies have shown that worms may transport MPs final effluent. The global hydraulic retention time (HRT) is 9 days, on
and break them down even further. Bacteria present in the gut of the average (Hernández-Crespo et al., 2022b).
earthworms digest the MPs into their simpler, more volatile compounds
(Lwanga et al., 2018). 2.1.2. Monasterios WWTP
In addition, there is a need to focus on further improving the legis­ Los Monasterios WWTP (Valencia, Spain) is serving a population of
lative and policy system to eliminate MPs at source (Xu et al., 2022). 1500 PE. The pretreatment consists in a screening and grit chamber. The
Although the literature about microplastics is growing exponentially secondary treatment is an aerated biological reactor with floating hel­
(Zhou et al., 2021), TWs are still a domain that needs to be investigated ophytes (176 m3, 4 m depth). The plants are a mixture of Typha and
to gain a better understanding of the processes and the benefits that they Phragmites situated upon a buoyant structure. This biological reactor
can provide in this field. works in extended aeration mode with sludge recirculation and is fol­
The general objective of this study is to shed more light on the po­ lowed by a sedimentation tank (48 m3). Aeration is intermittent and
tential of TWs to remove MPs from wastewaters and evaluate the role of controlled by dissolved oxygen probes. Finally, it has a tertiary treat­
vegetation roots, as well as the fate of microplastics along the TWs, to ment for disinfection formed by a floating TW (348 m2 of area and a
better understand the complex mechanisms involved. To this end, two volume of 557 m3), composed of vegetation growing on a buoyant
different WWTPs with TWs have been evaluated, including horizontal surface. The buoyant or floating structure is placed into a pond and
subsurface flow wetlands and floating wetlands. To the best of our planted with helophyte (cattails and reed) to create the floating wetland.
knowledge, this is the first study addressing these issues in full-scale The floating structure is the same as in the bioreactor and with the same
TWs. Additionally, this study looks at the degradation of MPs from a plant species. To illustrate the system, two pictures of the structure alone
qualitative point of view. The two main degradation processes for MPs and the structure planted inside the floating TW are included in SI
under research are biodegradation and photocatalysis (Pan et al., 2022). (Fig. S1). Followed by a horizontal subsurface flow TW (68 m2 and 18
It can be hypothesized that MPs in TWs may be degraded by biodegra­ m3 of area and volume respectively) with coarse gravel (2–3 cm) and
dation processes due to the long residence time of solid particles planted with common reed (Fig. 2). It treats a flow of 100–120 m3/day
retained inside TWs and because of the activity of their biological and 100% of the treated water is reused for irrigation of gardens. It has a
community, including bacteria, protozoans and worms, among others. COD and TSS removal efficiencies of 97% and 94%, respectively. The
mean HRT is 1.6 days in the biological reactor, 5.1 days in the floating
TW, and 0.2 days in the horizontal flow TW.

Fig. 1. Scheme of Carrícola’s WWTP, with two horizontal flow TWs in parallel the third in series. With the points to be studied marked with red numbers
(Figure adapted from Hernández-Crespo et al., 2022b).

2
D.C. Cabrera et al. Water Research 240 (2023) 120106

Fig. 2. Scheme of Los Monasterios WWTP, with the points to be studied marked in red numbers.

2.2. Wastewater and sludge sampling 2.3. Microplastics extraction procedure

Water samples were collected from different points in each treatment For the extraction of MPs from the water samples the same procedure
plant. In Carrícola WWTP four sites were sampled: raw wastewater as Wang et al. (2020) was followed. In summary, the beakers were dried
(Point 1), the effluent of the primary treatment (Point 2), the outlet of below 70 ◦ C (Munno et al., 2018), once dried a digestion with H2O2 30%
TW1 (Point 3) and finally the outlet of the treatment plant (Point 4) (see solution was applied and the residue was dried again, and then a
Fig. 1). In Los Monasterios WWTP, three different locations were two-step density separation was performed. First step was done with
sampled: the influent to the treatment plant (Point 1), the outlet of the CaCl2 (1.34 g/cm3) and a second step was done with a higher density
secondary treatment (Point 2), and the outlet of the subsurface TW and solution of KI (1.72 g/cm3) to ensure that all the high density micro­
thereof, the treatment plant (Point 3). plastics were recovered (Zhang et al., 2020). Otherwise, MPs with
Two sampling campaigns were performed in each WWTP. Different higher density than that of the solution could not be counted (Prata
volumes were taken for analysis: 2.5 L for the raw wastewater, 7.5 L for et al., 2019). The vials containing the density solution were cen­
the TW influent, 100 L for the TWs effluent. In Carrícola, just 60 L from trifugated (5 min, 3500 rpm) and the supernatant was filtered through
the effluent of TW2 were taken (Point 4) because of some difficulties in glass-fiber filters (pore size of 1 μm and a diameter of 47 mm). Then, 5
collecting the sample. In the first sampling campaign, all the samples ml of Rose Bengal was applied to stain natural fragments or fibers.
were taken in duplicate with a metal bucket and filtered by a stack of In the case of the sludge samples, the samples were washed over the
sieves (5600 μm, 425 μm, 75 μm, 40 μm) in situ. In the second sampling, sieve of 40 μm to collect the sludge adhered to the roots or the gravel.
one sample was taken at each point because the variability in the du­ After the samples were obtained, the rest of the process is the same as the
plicates in the first sampling was low and only the smallest sieve was water samples, except for the digestion process which required more
used, because these samples were taken for examination with Fourier- volume of H2O2 30% and took more time due to the large amount of
transform infrared spectroscopy (FTIR). The residue remaining on the organic material.
sieves was transferred into glass beakers covered with aluminum foil
and transported to the laboratory for further processing.
2.4. Microplastics counting and characterization
Sludge samples were also collected from different sites. In Carrícola
WWTP one sampling point was at the beginning of the TW1 (2 m from
The filters obtained in the previous step were examined following an
the inlet distribution channel) and another site at the end of the TW1 (3
orderly route: starting at the top from left to right, down and back in a
m before the outlet). In both sites, a depth between 10 and 15 cm was
right to left direction under stereomicroscopes with augmentations be­
excavated and samples were taken in duplicate. Samples with and
tween x20 and x40 (SZ30, Olympus, Japan). Following the guidelines
without roots were taken from the same depth where roots were most
marked by Hidalgo-Ruz et al. (2012), the MPs were identified, also with
developed. At each sampling point, one sludge sample was collected
help of the hot needle test for some particles that were unclear. Upon
from the gravel and a second was collected from sludge attached to the
being touched with the needle the plastic items should curve or melt.
roots excavating in the lower area where a plant was found.
Otherwise, the organic particles usually break under the pressure of the
In Los Monasterios WWTP, two kind of sludge samples were taken
hot needle. During the counting, MPs were classified into fibers, parti­
from the biological reactor (secondary treatment): suspended sludge and
cles, and films. Additionally, they were categorized by size according to
sludge attached to the roots of the floating macrophytes. The sludge
the sieve from which the filter came.
samples were collected during the aeration process to ensure an opti­
As mentioned above, in the second sampling campaign the samples
mally mixed sludge. Roots from the floating TW (tertiary treatment)
were sieved only through the 40 μm sieve, subjected to the same
were also sampled. In all cases, two replicas were extracted.
digestion and density-separation process, and analyzed with micro­
A shovel and a pick were needed for sludge samples. All tools were
–Fourier Transform Infrared Spectroscopy (μFTIR) (Thermo Scientific
made of metal and wood to avoid plastic and furthermore, samples were
Nicolet iN 10). The data base used included the Nicolet polymer and
transported in glass jars covered with aluminum foil to process them in
common materials set in addition to the Hummel polymer library.
the laboratory. Once in the laboratory, both the gravel and roots samples
were placed on a 5 mm sieve and cleaned with distilled water, placing
another sieve of smaller size (40 μm) under the 5 mm sieve to collect the 2.5. Quality assurance and contamination prevention
sludge samples. In the case of the roots, we dug around a plant and
collected all the roots in the area, along with the sludge adhered to them. To avoid contamination during the sampling and experimentation,
quality control measures were adopted. The use of plastic materials was
avoided as much as possible; the beakers were covered by aluminum foil
after sampling and every time that they were stored, another aluminum
foil was placed to protect. During all the steps, including sampling,

3
D.C. Cabrera et al. Water Research 240 (2023) 120106

laboratory cotton coat and latex gloves were used. All the working space between 17.2 and 23.4 MPs/L in Carrícola WWTP and between 4.0 and
and equipment were cleaned thoroughly before and after use. 12.8 MPs/L in Los Monasterios WWTP (see Fig. 3). That variability was
At each sampling point, an in situ blank was prepared, and two highly attenuated in the effluent, highlighting the buffering capacity of
laboratory blanks were conducted during the observation of filters to wetlands, a property noted in numerous studies (Hernández-Crespo
quantify MPs pollution in the lab environment. The blanks went through et al., 2022a, b). The effluent concentrations varied between 0.18 and
the same steps as the other samples and were also counted under a 0.98 MPs/L in Carrícola and from 0.04 to 0.29 MPs/L in Los Mon­
stereomicroscope. In the end, nine particles were found in the Carrícola asterios, being significantly lower than the influent concentrations
field blank, and fifteen were found in Los Monasterios in a 10 L distilled (p<0.05). Moreover, it should be noted that the variability of the sam­
water sample. These MP quantities were not subtracted from the sample ples during the same sampling campaign was significantly lower than
results due to uncertainty regarding whether all samples were equally the variability between samplings, as reflected by the error bars of the
contaminated, as most of the contamination in the blanks comes from points in which a duplicate was taken. In connection with the charac­
airborne sources. Instead, the blank concentrations were used as a teristics of both influents, it should be noted that the results of the sec­
detection limit (Smyth et al., 2021), as similar results were found in ond sampling could have been influenced by rainfall events during
previous studies (Conley et al., 2019; Wang et al., 2020). previous days, which could have provoked a flush of the sewage network
In addition, a recovery test was performed. For this, MPs were arti­ (Zhou et al., 2022), causing to a reduction in MPs arrivals on subsequent
ficially prepared from plastic utensils of a different color (clothes, bags days.
and bottle lid) by scraping with a knife or scissors. The test was per­ The MP removal efficiency for Imhoff tanks resulted, on average, in
formed four times with the following quantity of MPs: ten fibers, ten 17.4% for Carrícola. Furthermore, the removal efficiency of TW1 was on
particles and five films with a medium size of 350 μm. The whole pro­ average, 98.6%. It can be seen that in Carrícola the concentration in­
cedure was applied, including sieving, extraction, density separation creases slightly after Point 3, in TW2, which may be caused by a sludge
and visual examination. The following average recovery percentages drain operation of the pipe system, that is carried out monthly, and the
were obtained: 97.5% for fibers, 87.5% for particles/fragments and 75% effluent of this operation is taken to the final part of TW2, which can
for films. As blanks concentrations, this process is carried out to quantify cause a transport of MPs previously sedimented inside the pipes. On the
the imperfections of the method, but not to alter the results since it other hand, the average MP removal efficiency after the secondary
cannot be ensured that it affects all samples equally. treatment in Monasterios was 53.3%. The removal efficiency of the TW
system was on average, 96.2% in Los Monasterios. In terms of global
removal efficiency, Carrícola presented an average of 97.4% and Los
2.6. Statistical analysis Monasterios 98.1%. These global efficiencies are higher than those ob­
tained in small standard (pre-treatment, primary, secondary) WWTPs,
An abundance of MPs were measured as MPs/L in water samples and where efficiencies lower than 90% are typical (Sun et al., 2019).
MPs/g dry weight of sludge in the sludge samples. Statistical analysis Furthermore, in comparison with other studies where TW based tech­
was performed using Statgraphics software Version 18.0 (Statgraphics nology is also used, especially those operated as a tertiary treatment, the
Technologies, Inc., The Plains, Virginia). The concentration, size and removal efficiencies reached in this study are similar and sometimes
shape distributions were compared using parametric tests (ANOVA) if higher, as can be noted in Table 1. Moreover, it has been demonstrated
normality was satisfied and nonparametric otherwise (Kruskal Wallis). that full-scale TW, both working as secondary and tertiary treatment,
In all tests, the results were only considered statistically significant when can efficiently reduce the number of MPs in wastewater.
p value was below 0.05. At first, it was hypothesized that, in a modern and more urban
population such as Los Monasterios, the concentration of MPs in the
3. Results and discussion influent to the WWTP would be higher. However, the influent concen­
trations to Carrícola WWTP were substantially higher. This may be due,
3.1. Wastewater microplastics concentrations and removal efficiencies among other causes, to the fact that it is common to find higher con­
centrations of pollutants in the wastewater of small villages, as they
MP concentrations presented a high degree of variability at the inlet,

Fig. 3. Results of MPs concentration measured in the different points of both WWTPs. For Carrícola: Point 1 influent, Point 2 outlet of the sedimentation tanks, Point
3 outlet of the parallel TWs and Point 4 final effluent. For Monasterios: Point 1 influent, Point 2 outlet of the secondary treatment and Point 3 final effluent.

4
D.C. Cabrera et al. Water Research 240 (2023) 120106

Table 1 machines and using natural fiber clothing, among other measures
MP removal efficiency (%) in other WWTPs with TWs. SFTW for surface flow (Prata, 2018). Likewise, simple actions such as removing fibers from our
TW; HSFTW for horizontal subsurface Flow TW; VFTW for vertical subsurface feet before entering the shower can help to reduce the fiber content in
flow TW. wastewater. In this sense, treating greywater separately could help to
Study type Design type Removal efficiency (%) Reference reduce the MP load into wastewater, as it is considered a significant
Laboratory SFTW 81.61 Chen et al. (2021) source of microfiber pollution (Sotiropoulou et al., 2023).
Laboratory HSFTW 100 Chen et al. (2021) In addition, it is necessary to consider the important effect that the
Field VFTW 83–89 Zhou et al. (2022) biofilm plays in the retention of MPs in TWs, since retention and
Field SFTW 27.3–60 Zhou et al. (2022) adsorption are important mechanisms for pollutant removal in wetlands
Laboratory VFTW 100 Wang et al. (2021)
Field HSFTW 88 Wang et al. (2020)
(Hernández-Crespo et al., 2017). The colonization of MPs by biofilm is a
Field HSFTW 45–100 Wei et al. (2020) facilitator mechanism for the retention of MPs in wetlands, increasing
Field HSFTW 72–68 Long et al. (2022) their surface area and their density. Moreover, the biofilm adhered to
Field SFTW 95 Bydalek et al. (2023) the vegetation and the roots of the wetlands, also improves retention.
This effect is of great importance in the retention of fibers. Due to their
high surface area to volume ratio, the fibers are more likely to be
usually use a lower quantity of water and there are less uses (commer­
colonized by microorganisms than other forms of MPs (Chen et al.,
cial, industry…) that can dilute these pollutants (Huertas et al., 2013).
2021). This colonization also becomes very important when dealing
This can be corroborated with the COD and TSS concentrations of the
with low density polymers (less than water) that might escape from
influents of each WWTP: while in Los Monasterios we found average
settlers by flotation. During this study, fibers colonized by biofilm could
values of 357 and 82 mg/L respectively, in Carrícola values of 794 and
be observed (Fig. S2), which could indicate a first symptom of potential
266 mg/L were registered. Similarly, this was confirmed in other studies
biodegradation, a mechanism described as feasible by previous studies
such as Wei et al. (2020), who analyzed different treatment plants
(Du et al., 2021). Signs of degradation in the MPs, such as bites, flakes,
located in different areas: near cities, in rural areas and in mountainous
etc., were also observed during this study after passing through the
areas far from large urban centers. Their results indicated that, waste­
wetlands (Fig. S3). In addition, the fibers are susceptible to entangle­
water concentration was higher in less urbanized areas. However, it
ment, which is positive because it helps to improve their retention in the
should be noted that other studies in WWTPs in the city show very wide
wetland gravels and their sedimentation, since their density and surface
ranges (Qiu et al., 2020). Therefore, MPs concentration does not depend
area increase (Fig. S4).
on the served population (Mahon et al., 2017), greater attention should
The results show a clear trend regarding the categorization by MPs
be paid to the lifestyle of each community.
size. This was one of the initial hypotheses: as MPs pass through the
The shape and size distribution of MPs in wastewater are shown in
treatment system, especially in the TW, their size distribution shifts to­
Figs. 4 and 5. Regarding the shape distribution, the plastics found were
wards smaller dimensions. In both cases, as shown in Fig. 5, MPs
divided into groups: fibers, particles, and films. As in other contempo­
collected in the larger sieve were reduced in proportion as they pro­
rary studies (Long et al., 2022; Wang et al., 2020), the predominant form
gressed through the system and in turn, the smaller MPs increase in
was the fiber, in both influent and effluent. The removal efficiency was
percentage after passing through the TWs. This may be explained by a
97.5% for fibers in both WWTPs. Films were the second most typical
number of mechanisms: physical filtration, size reduction by earthworm
form in both WWTPs, which is not in accordance with other studies, that
digestion and size reduction due to the mechanical wear suffered by the
show it is usually the smallest proportion (Bydaleck et al., 2023; Wang
MPs as they pass through the gravel. Therefore, a design recommenda­
et al., 2020; Zhou et al., 2022). Regarding particles, their removal is
tion that can be drawn from the results is the desirability of reducing the
complete, which may be because particles are more likely to be retained
size of the filter material in the end zone of the wetlands, in order to
by the filtering effect of TW. Given that fibers are the most abundant
retain these small size MPs, or design a multistage system varying the
fraction, it is advisable to carry out campaigns to raise awareness of this
filter porous material. This will ensure that MPs are removed at different
problem. Previous studies have proposed installing filters in washing
stages of the process, thereby increasing the overall efficiency of the CW.

Fig. 4. Distribution by form (% in bars) and concentration (MPs/L) of each one (labels on bars) of the MPs in water sampled from different points in both WWTPs.

5
D.C. Cabrera et al. Water Research 240 (2023) 120106

Fig. 5. Percentage of MPs classified by size in each point of both WWTPs.

Among other general recommendations to ensure the long-term effec­ goodness of fit (see R2 in Fig. 6). Even those data for which the rela­
tiveness of TWs as optimize the hydraulic load, monitor and maintain tionship was weak, when included in this graph, are within the point
the system. If smaller MPs are kept within the TWs, it may promote their cloud formed by the set of values.
biodegradation potential. Specifically, the MPs can be consumed by the A very interesting parameter to quantify is the number of MPs that
macroinvertebrates inhabiting the wetlands, such as earthworms, as each treatment plant is releasing into the natural environment per
observed by Wang et al. (2021). The microorganisms present in both the inhabitant, which is known as emission of MPs per capita. In this study
earthworms and the wetlands could then facilitate the biodegradation only for Carrícola this parameter could be calculated, in the case of Los
process, as reported by Lwanga et al. (2018) and Yuan et al. (2020). Monasterios, there was no precise information available to calculate this
Furthermore, previous studies paid attention to finding a correlation parameter. With a flow rate of 14,000 L/d and a population of 115 PE.,
between TSS and MPs, because if a strong correlation is found, the the emission rate would be 71 MPs/capita/day. The data obtained are
concentrations of TSS could be used as an indicator of MPs content. lower than those showed in other studies, such as that of Wang et al.
Currently, the TSS variable has not yet been used as an indicator, (2020), where emissions of MPs per capita were also calculated,
because there are a variety of opinions on the matter, and the re­ resulting in 428 MPs/capita/day, for a population of 494 PE. These
lationships obtained are different in each study, hence there is still no variations may be due to the different forms of consumption in the
agreed empirical equation. There are studies that found a strong corre­ different societies studied.
lation (Wang et al., 2020; Zhou et al., 2022), while others did not find a
strong enough correlation to use as an indicator (Bydalek et al., 2023). In
this study, the correlation is weak (Fig. 6) although statistically signif­ 3.2. Sludge microplastics retention
icant (p<0.05), which gives an indication of the interest in pursuing such
relationships. A graph including the data from this study and previous Different sludge samples were taken to check the role of the roots to
ones has been made in order to further analyze this correlation. As can capture MPs, both in TW and by introducing floating macrophytes in the
be seen in Fig. 6, the results of several studies fit this linear relationship bioreactor.
in a statistically significant way (p<0.05) and with a considerably high Regarding the fate of MPs in Carricola’s TW, Fig. 7 indicates a higher
accumulation of MPs in the inlet zone. This is an accordance with the

Fig. 6. Relationship between TSS and MPs from different studies. Data from Carrícola and Los Monasterios WWTPs obtained in this study. Data from other studies:
Zhou et al., 2022; Wang et al., 2020; Long et al., 2019; Bydalek et al., 2023.

6
D.C. Cabrera et al. Water Research 240 (2023) 120106

Fig. 7. Concentrations measured in Carrícola and in Los Monasterios sludge, for the different sampling points, measured in MPs per dewatered gram of sludge.
Gravel inlet, for the average of the samples taken from gravels in the TW1 inlet. Gravel outlet, for the average of the samples taken from gravels in the TW1 outlet.
Roots inlet, for the average of the samples taken from roots in the inlet. Roots outlet, for the average of the samples taken from roots in the outlet of the TW. Sludge
Reactor, for the suspended sludge in the bioreactor; Root Reactor, for the sludge taken from the roots inside the bioreactor; Root 1 TW, for the sludge taken from the
roots in one corner of the floating TW; Roots 2 TW, for the sludge taken from the roots in the other corner of the floating TW.

higher accumulation of TSS in the inlet zone of TWs (García et al., 2003, dry sludge respectively in one WWTP in Italy and another one in China.
2005). Similar values are reported by Li et al. (2018), where 79 sludge samples
As it can be seen in Fig. 7 for Carrícola, the initial hypothesis cannot were collected from 28 different WWTPs in China, obtaining results
be accepted at the inlet of the TW. Comparing both inlets, gravel sludge ranging from 1.6 to 56.4 MPs/g dry sludge.
had a higher concentration of MPs per dry gram of sludge. This may be With regards to levels in roots, other authors (Yin et al., 2021; Duan
due to the ease of flow through the lower part of the wetland, below the et al., 2021), reported amounts of 4 MPs/g and 5.11±2.95 MPs/g of dry
roots, which is why more MPs are dragged to that area. Moreover, in this sludge respectively, although those studies were carried out in a natural
zone a high quantity of solids and biofilm are present, so the gravel can environment, which is likely to be less polluted. However, the concen­
be as efficient as roots. Conversely, in the outlet, the concentration of trations at the outlet do resemble those found in the natural environ­
MPs of the sludge attached to the roots was higher than in gravel, so here ment in these studies and are even lower.
the hypothesis is fulfilled. It must be highlighted that the difference
between roots and gravel concentration is, in relative terms, greater at
3.3. Microplastic polymer types in treatment wetlands systems
the outlet than in the inlet. This shows the relevant role of roots in
enhancing the retention mechanism. (Gersberg et al., 1986).
Seven different types of polymers were the most commonly found in
In the case of Los Monasterios, Fig. 7 shows the high difference, more
both water and sludge samples, including polyethylene terephthalate
than double, that exists between the concentrations measured in the root
(PET), polyethylene (PE), polypropylene (PP), Polyvinyl chloride (PVC),
sludge and the sludge suspended in the bioreactor. This highlights the
polyurethane (PUR), Polytetrafluoroethylene (PTFE), Polytrimethylene
positive role of vegetation, since the concentration of MPs in the sludge
terephthalate (PTT), other types of polymers were also found to a lesser
is being reduced thanks to the capture by macrophytes, thus helping to
extent (OTHER) (e.g. polystyrene, polyester, polyamide) (see Fig. 8).
reduce the load of MPs introduced in the environment when sewage
The most dominant component in water samples were PET, which is in
sludge is used in agriculture, apart from other additional benefits pro­
third place overall. PP and PVC were the most predominant polymers in
vided by the macrophytes in the reactor (Gersberg et al., 1986).
the sludge, being significant in both WWTPs, as in other studies (Lu
Regarding the root samples from the floating TW, the results indicate
et al., 2022; Huang et al., 2020).
that roots continue playing a crucial role in the tertiary treatment,
The higher proportion of PP and PVC in the sludge is in line with the
trapping a noticeable quantity of MPs. In the floating TW, intercepted
production data of plastics, since the most produced plastics are, in this
MPs could remain attached to the roots for a long time, since roots are
order, PP, PE, and PVC, mainly for their uses as packaging and in the
not removed when the macrophytes are harvested, this being a positive
construction and automotive industry (Plastic Europe, 2021), followed
factor to facilitate their potential degradation.
by PET. So, it is not surprising that they are the most common, as in
A recently published study that characterized the sediment of five
other sediment studies where PP prevails (Yuan et al., 2020; Lu et al.,
TWs in Australia (Lu et al., 2022), reported values of MPs in the sedi­
2022; Olesen et al., 2019).
ments varying between 736 ± 335 and 3480 ± 4330 MP/kg dry sedi­
In the water samples (Fig. 8), a large amount of PET is found, espe­
ment. It is clear that sediment and sludge are not directly comparable
cially in Carrícola where it reaches almost 60% of the total polymers
but serve as a reference. In this case study, higher values of MPs in the
found in the influent. PET is a material widely used in textiles and
sludge were observed, but also the input loads to the TWs were signif­
packaging, so it is very common to find, in addition to PE, which is also
icantly higher.
very important in the Monasterios influent, which may be contributing a
Within the research conducted, there is a wide range of results, also
large amount of the fibers found in the water samples. It should be noted
depending on many factors, such as the type of TW operation, popula­
that both materials are the most predominant after passing through the
tion served, size of filter used, as mentioned by Hidalgo-Ruz et al.
primary treatments, as can be seen in points 2. A high percentage of PVC
(2012), among others. Higher values than those obtained in this study
is also found in both treatment plants at the inlet, although this type of
can be found in Magni et al. (2019) and Liu et al. (2019). Both authors,
polymers is satisfactorily removed along the WWTP. It should be noted
reported values of 113±57 MPs/g dry sludge and 240.3 ± 31.4 MPs/g
that point 3 is not very relevant due to the small quantities of MPs found

7
D.C. Cabrera et al. Water Research 240 (2023) 120106

Fig. 8. Components of the MPs found, in percentage, in water samples at both monitoring locations.

there. highest presence is PVC, one of those with the highest density, in both
The polymer most adhered to roots, both in Carrícola’s inlet roots WWTPs. Other studies reported different results, where the PE and PP
and in the Monasterios TW roots (Fig. 9) is PP, the lowest density fragments were predominantly recorded in the sediment, hence the
polymer registered. This highlights the important role of the roots in the density cannot be considered as the only factor in the distribution of MPs
retention of lower density MPs, which could otherwise be escaping from in the medium (Lu et al., 2022), although in this case it does seem to be
the WWTPs by flotation. In addition, the roots seem to play a key role in relevant. Moreover, it should be noted that not all studies use the same
intercepting PET, which is usually referred to as polyester in textile extraction procedure, and this could lead to different results.
products. This coincides with the higher proportion of fibers in this
sludge samples (Fig. S5) as well as in the effluent samples (Fig. 4). On the
other hand, in the suspended sludge the polymer compound with the

Fig. 9. Components of the MPs found, in percentage, in sludge samples at both monitoring locations.

8
D.C. Cabrera et al. Water Research 240 (2023) 120106

Conclusions Bydalek, F., Ifayemi, D., Reynolds, L., Barden, R., Kasprzyk-Hordern, B., Wenk, J., 2023.
Microplastic dynamics in a free water surface constructed wetland. Sci. Total
Environ. 858, 160113.
The following conclusions can be deduced from the study carried Carpenter, E.J., Smith, K.L., 1972. Plastics on the Sargasson sea surface. Science 175
out: (4027), 1240–1241.
Chen, Y., Li, T., Hu, H., Ao, H., Xiong, X., Shi, H., Wu, C., 2021. Transport and fate of
microplastics in constructed wetlands: a microcosm study. J. Hazard. Mater. 415,
- Treatment Wetlands (TWs) have significant potential for reducing 125615.
the discharge of MPs into the natural environment, functioning as Conley, Kenda, Clum, Allan, Deepe, Jestine, Lane, Haven, Beckingham, Barbara, 2019.
both secondary and tertiary treatment. Wastewater treatment plants as a source of microplastics to an urban estuary:
removal efficiencies and loading per capita over one year. Water Res. X 3, 100030.
- Fibers are the most common form of influent and effluent of both Du, H., Xie, Y., Wang, J., 2021. Microplastic degradation methods and corresponding
WWTPs, presenting the greatest challenge for MP pollution control. degradation mechanism: research status and future perspectives. J. Hazard. Mater.
Actions aimed at reducing their arrival at WWTPs, such as washing 418, 126377.
Duan, J., Han, J., Cheung, S.G., Chong, R.K.Y., Lo, C.M., Lee, F.W.F., Xu, S.J., Yang, Y.,
machine filters and promoting the use of natural textiles, are highly Tam, N.F., Zhou, H.C., 2021. How mangrove plants affect microplastic distribution
recommended. in sediments of coastal wetlands: case study in Shenzhen Bay, South China. Sci. Total
- Smaller particles predominate at the final stage, and it is recom­ Environ. 767, 144695.
Garcı́a, J., Ojeda, E., Sales, E., Chico, F., Pı́riz, T., Aguirre, P., Mujeriego, R., 2003. Spatial
mended to have a final zone or wetland cell with a finer filtering variations of temperature, redox potential, and contaminants in horizontal flow reed
material to further reduce MP emissions. Once these smaller MPs are beds. Ecol. Eng. 21 (2–3), 129–142.
retained inside the wetland, their biodegradation becomes more García, J., Aguirre, P., Barragán, J., Mujeriego, R., Matamoros, V., Bayona, J.M., 2005.
Effect of key design parameters on the efficiency of horizontal subsurface flow
feasible.
constructed wetlands. Ecol. Eng. 25 (4), 405–418.
- MPs are largely concentrated in the sludge from sewage treatment Gersberg, R.M., Elkins, B.V., Lyon, S.R., Goldman, C.R., 1986. Role of aquatic plants in
plants. Roots play a positive role in capturing more MPs, both in TWs wastewater treatment by artificial wetlands. Water Res. 20 (3), 363–368.
and in floating macrophytes in the bioreactor. Therefore, it is Hernández-Arenas, R., Beltrán-Sanahuja, A., Navarro-Quirant, P., Sanz-Lazaro, C., 2021.
The effect of sewage sludge containing microplastics on growth and fruit
important to maintain an adequate vegetation density in TWs and to development of tomato plants. Environ. Pollut. 268, 115779.
install floating macrophytes in the biological reactors of WWTPs. Hernández-Crespo, C., Gargallo, S., Benedito-Durá, V., Nácher-Rodríguez, B., Rodrigo-
- MPs and TSS exhibit similar behavior in the TWs, being mostly Alacreu, M.A., Martín, M., 2017. Performance of surface and subsurface flow
constructed wetlands treating eutrophic waters. Sci. Total Environ. 595, 584–593.
retained in the initial zone, and are significantly correlated. Further Hernández-Crespo, C., Oliver Rajadel, N., Peña, M., Añó-Soto, M., Martín Monerris, M.,
study of their relationship can provide an estimation tool for MPs 2022a. Valorisation of drinking water treatment sludge as substrate in subsurface
content. Pooling data from different studies can provide a more solid flow constructed wetlands for upgrading treated wastewater. Process Saf. Environ.
Protection (158), 486–494.
relationship. Hernández-Crespo, C., Fernández-Gonzalvo, M.I., Miglio, R.M., Martín, M., 2022b.
- The importance of root retention in capturing lower-density MPs that Escherichia coli removal in a treatment wetland-pond system: a mathematical
could escape from WWTPs is once again demonstrated in polymer modelling experience. Sci. Total Environ., 156237
Hidalgo-Ruz, V., Gutow, L., Thompson, R.C., Thiel, M., 2012. Microplastics in the marine
retention. Additionally, the correlation between the different mate­ environment: a review of the methods used for identification and quantification.
rials observed (PET and PE) in the water samples and the main forms Environ. Sci. Technol. 46 (6), 3060–3075.
(i.e., fibers) can be appreciated. Huang, Y.L., Tian, M., Jin, F., Chen, M.Y., Liu, Z.G., He, S.Q., Li, F.X., Yang, L.Y.,
Fang, C., Mu, J.L., 2020. Coupled effects of urbanization level and dam on
microplastics in surface waters in a coastal watershed of Southeast China. Mar.
Declaration of Competing Interest Pollut. Bull. 154, 111089.
Huertas, R., Marcos, C., Ibarguren, N., Ordás, S., 2013. Guía Práctica Para la Depuración
The authors declare that they have no known competing financial de Aguas Residuales en Pequeñas Poblaciones. NIPO, DL: VA 3-2013, 283-12-005-9.
Jabeen, K., Su, L., Li, J., Yang, D., Tong, C., Mu, J., Shi, H., 2017. Microplastics and
interests or personal relationships that could have appeared to influence mesoplastics in fish from coastal and fresh waters of China. Environ. Pollut. 221,
the work reported in this paper. 141–149.
Li, X., Chen, L., Mei, Q., Dong, B., Dai, X., Ding, G., Zeng, E.Y., 2018. Microplastics in
sewage sludge from the wastewater treatment plants in China. Water Res. 142,
Data availability 75–85.
Liu, X., Yuan, W., Di, M., Li, Z., Wang, J., 2019. Transfer and fate of microplastics during
Data will be made available on request. the conventional activated sludge process in one wastewater treatment plant of
China. Chem. Eng. J. 362, 176–182.
Long, Z., Pan, Z., Wang, W., Ren, J., Yu, X., Lin, L., Lin, H., Chen, H., Jin, X., 2019.
Microplastic abundance, characteristics, and removal in wastewater treatment plants
Acknowledgements in a coastal city of China. Water Research 155, 255–265. https://doi.org/10.1016/j.
watres.2019.02.028.
Long, Y., Zhou, Z., Yin, L., Wen, X., Xiao, R., Du, L., Zhu, L., Liu, R., Xu, Q., Li, H.,
Darío Calzadilla had a scholarship funded by Cathedra Aguas de Nan, R., Yan, S., 2022. Microplastics removal and characteristics of constructed
Valencia. Qintong Wang is financially supported by the China Scholar­ wetlands WWTPs in rural area of Changsha, China: a different situation from urban
WWTPs. Sci. Total Environ. 811, 152352.
ship Council (CSC) by a CSC PhD grant (Ref. 201906690045). The au­
Lu, H.C., Ziajahromi, S., Locke, A., Neale, P.A., Leusch, F.D., 2022. Microplastics profile
thors would like to express their gratitude to Carricola’s municipality in constructed wetlands: distribution, retention and implications. Environ. Pollut.
and Urbanization Los Monasterios (Puçol) for their cooperation in the 313, 120079.
study. Lu, Y., Zhang, Y., Deng, Y., Jiang, W., Zhao, Y., Geng, J., Ding, L., Ren, H., 2016. Uptake
and accumulation of polystyrene microplastics in zebrafish (Danio rerio) and toxic
effects in liver. Environ. Sci. Technol. 50, 4054.
Supplementary materials Lwanga, E.H., Thapa, B., Yang, X., Gertsen, H., Salánki, T., Geissen, V., Garbeva, P.,
2018. Decay of low-density polyethylene by bacteria extracted from earthworm’s
guts: a potential for soil restoration. Sci. Total Environ. 624, 753–757.
Supplementary material associated with this article can be found, in Magni, S., Binelli, A., Pittura, L., Avio, C.G., Della Torre, C., Parenti, C.C., Gorbi, S.,
the online version, at doi:10.1016/j.watres.2023.120106. Regoli, F., 2019. The fate of microplastics in an Italian wastewater treatment plant.
Sci. Total Environ. 652, 602–610.
Mahon, A.M., O’Connell, B., Healy, M.G., O’Connor, I., Officer, R., Nash, R., Morrison, L.,
References 2017. Microplastics in sewage sludge: effects of treatment. Environ. Sci. Technol. 51
(2), 810–818.
Andrady, A.L., Neal, M.A., 2009. Applications and societal benefits of plastics. Phil. Munno, K., Helm, P.A., Jackson, D.A., Rochman, C., Sims, A., 2018. Impacts of
Trans. R. Soc. B 364, 1977–1984. temperature and selected chemical digestion methods on microplastic particles.
Boucher, J., Friot, D., 2017. Primary Microplastics in the Oceans: A Global Evaluation of Environ. Toxicol. Chem. 37 (1), 91–98.
Sources. IUCN, Gland, Switzerland. Olesen, K.B., Stephansen, D.A., van Alst, N., Vollertsen, J., 2019. Microplastics in a
stormwater pond. Water (Basel) 11 (7), 1466.

9
D.C. Cabrera et al. Water Research 240 (2023) 120106

Pan, Y., Gao, S.H., Ge, C., Gao, Q., Huang, S., Kang, Y., Wang, A.J., 2022. Removing Vymazal, J., 2010. Constructed wetlands for wastewater treatment. Water (Basel) 190
microplastics from aquatic environments: a critical review. Environ. Sci. (3), 69–96.
Ecotechnol., 100222 Wang, W., Ge, J., Yu, X., 2019. Bioavailability and toxicity of microplastics to fish
PlasticsEurope, 2021. Plastics - the Facts 2021. Retrieved from. https://plasticseurope. species: a review. Ecotoxicol. Environ. Saf. (November), 109913
org/knowledge-hub/plastics-the-facts-2021/. Wang, Q., Hernández-Crespo, C., Santoni, M., Van Hulle, S., Rousseau, D.P., 2020.
Prata, J.C., 2018. Microplastics in wastewater: state of the knowledge on sources, fate Horizontal subsurface flow constructed wetlands as tertiary treatment: can they be
and solutions. Mar. Pollut. Bull. 129 (1), 262–265. an efficient barrier for microplastics pollution? Sci. Total Environ. 721, 137785.
Prata, J.C., da Costa, J.P., Duarte, A.C., Rocha-Santos, T., 2019. Methods for sampling Wang, Q., Hernández-Crespo, C., Du, B., Van Hulle, S.W., Rousseau, D.P., 2021. Fate and
and detection of microplastics in water and sediment: a critical review. TrAC Trends removal of microplastics in unplanted lab-scale vertical flow constructed wetlands.
Analyt. Chem. 110, 150–159. Sci. Total Environ. 778, 146152.
Qiu, R., Song, Y., Zhang, X., Xie, B., He, D., 2020. Microplastics in urban environments: Wei, S., Luo, H., Zou, J., Chen, J., Pan, X., Rousseau, D.P., Li, J., 2020. Characteristics
sources, pathways, and distribution. Microplast. Terrestr. Environ.: Emerg. Contam. and removal of microplastics in rural domestic wastewater treatment facilities of
Major Challenges 41–61. China. Sci. Total Environ. 739, 139935.
Rousseau, D.P., Lesage, E., Story, A., Vanrolleghem, P.A., De Pauw, N., 2008. Xu, D., Yin, X., Zhou, S., Jiang, Y., Xi, X., Sun, H., Wang, J., 2022. A review on the
Constructed wetlands for water reclamation. Desalination 218 (1–3), 181–189. remediation of microplastics using constructed wetlands: bibliometric, co-
Sarkar, D.J., Sarkar, S.D., Das, B.K., Sahoo, B.K., Das, A., Nag, S.K., Manna, R.K., occurrence, current trends, and future directions. Chemosphere, 134990.
Behera, B.K., Samanta, S., 2021. Occurrence, fate and removal of microplastics as Xu, Q., Gao, Y., Xu, L., Shi, W., Wang, F., LeBlanc, G.A., Cui, S., An, L., Lei, K., 2020.
heavy metal vector in natural wastewater treatment wetland system. Water Res. 192, Investigation of the microplastics profile in sludge from China’s largest Water
116853. reclamation plant using a feasible isolation device. J. Hazard. Mater. 388, 122067.
Sheavly, S.B., Register, K.M., 2007. Marine debris & plastics: environmental concerns, Yang, X., He, Q., Guo, F., Sun, X., Zhang, J., Chen, M., Vymazal, J., Chen, Y., 2020.
sources, impacts and solutions. J. Polym. Environ. 15, 301–305. Nanoplastics disturb nitrogen removal in constructed wetlands: responses of
Silva, A.B., Bastos, A.S., Justino, C.I.L., da Costa, J.P., Duarte, A.C., Rocha-Santos, T.A.P., microbes and macrophytes. Environ. Sci. Technol. 54 (21), 14007–14016.
2018. Microplastics in the environment: challenges in analytical chemistry – A Yin, L., Wen, X., Huang, D., Zeng, G., Deng, R., Liu, R., Zhou, Z., Tao, J., Xiao, R., Pan, H.,
review. Anal. Chim. Acta 1017, 1–19. 2021. Microplastics retention by reeds in freshwater environment. Sci. Total
Smyth, K., Drake, J., Li, Y., Rochman, C., van Seters, T., Passeport, E., 2021. Bioretention Environ., 148200
cells remove microplastics from urban stormwater. Water Res. 191. Yuan, J., Ma, J., Sun, Y., Zhou, T., Zhao, Y., Yu, F., 2020. Microbial degradation and
Sotiropoulou, M., Stefanatou, A., Schiza, S., Petousi, I., Stasinakis, A., Fountoulakis, M., other environmental aspects of microplastics/plastics. Sci. Total Environ. 715,
2023. Removal of microfiber in vertical flow constructed wetlands treating 136968.
greywater. Sci. Total Environ. 858, 159723 https://doi.org/10.1016/j. Zhang, L., Xie, Y., Liu, J., Zhong, S., Qian, Y., Gao, P., 2020. An overlooked entry
scitotenv.2022.159723. pathway of microplastics into agricultural soils from application of sludge-based
Sun, J., Dai, X., Wang, Q., van Loosdrecht, M.C., Ni, B.J., 2019. Microplastics in fertilizers. Environ. Sci. Technol. 54 (7), 4248–4255.
wastewater treatment plants: detection, occurrence and removal. Water Res. 152, Zhou, M., Wang, R., Cheng, S., Xu, Y., Luo, S., Zhang, Y., Kong, L., 2021. Bibliometrics
21–37. and visualization analysis regarding research on the development of microplastics.
Thompson, R.C., Olsen, Y., Mitchell, R.P., Davis, A., Rowland, S.J., John, A.W., Environ. Sci. Pollut. Res. 1–15.
McGonigle, D., Russell, A.E., 2004. Lost at sea: where is all the plastic? Science 304, Zhou, X., Zhao, Y., Pang, G., Jia, X., Song, Y., Guo, A., Wang, A., Zhang, S., Ji, M., 2022.
838–839. Microplastic abundance, characteristics and removal in large-scale multi-stage
van den Berg, P., Huerta-Lwanga, E., Corradini, F., Geissen, V., 2020. Sewage sludge constructed wetlands for effluent polishing in northern China. Chem. Eng. J. 430,
application as a vehicle for microplastics in eastern Spanish agricultural soils. 132752.
Environ. Pollut. 261, 114198.

10