Microplastics in soil affect plant-soil systems depending on their shape and polymer type. However, previous research has not yet considered the effects of degraded plastics, which are the plastic materials actually present in the environment. We selected 8 microplastics representing different shapes (fibers, films and foams) and polymer types, and exposed them to UV-C degradation. Each microplastic was mixed with soil at a concentration of 0.4% (w/w). The phytometer Daucus carota grew in each pot. At harvest, soil properties and plant biomass were measured.

Photodegradation altered microplastics physical and chemical properties, impacting plant-soil systems. Microplastics degradation effects on plant and soil were observed with fibers and foams, but there were negligible effects with films. The latter could be explained by the polymer structure of films and manufacturer’s additives, potentially delaying their degradation.

Degraded fibers increased soil respiration more than their non-degraded counterparts, as photodegradation increased the positive effects of fibers on soil water retention. The emergence of oxygenated groups during degradation may have increased the hydrophilicity of fibers, enhancing their ability to retain water. Degraded foams increased soil respiration, which could be related to the possible leaching of organic substances with lower partition coefficients, which may promote soil microbial activity.

By contrast, degraded foams decreased soil aggregation, likely as degradation produced larger holes increasing their permeability. Also, the increase of hydrophilic molecules could have decreased soil particle cohesiveness. Degraded fibers and foams increased shoot and root mass as a result of microplastic effects on soil properties. Photodegraded microplastics affected root traits, which could be linked to microplastic effects on soil water status and plant coping strategies.

Plant species and microplastic selection

We selected Daucus carota as a phytometer, which is a herbaceous plant typical of grassland ecosystems (Federal Agency for Nature Conservation, 2019) that shows clear responses to microplastics in soil (Lozano et al., 2021b, 2022). Also, we selected 8 secondary microplastics widely used in daily life, which represent three microplastic shapes: fibers, films and foams and 6 polymer types: polyamide (PA), polyester fibers made to at least 85% of polyethylene terephthalate (PET; Council Directive, 2011), polypropylene (PP), low density polyethylene (PE), polyethylene terephthalate (PET), polystyrene (PS) and polyurethane (PU) (Figure 1).

Microplastic degradation

Plastics were exposed to UV-C degradation (254 nm irradiation) by using a photodegradation chamber with three 36W UV-C lamps (Figure S1). The average incident energy in the chamber was of 20.98 Wm^-2 (photometer; item number HD 2302.0, DeltaOHM), which simulated UV-C wavelengths. UV-A (present in solar light) and UV-C (used for shorter photodegradation experiments due to its higher energy) are ranges of the ultraviolet wavelength spectra that produce similar final outputs (e.g., breaking of polymer chains, volatile organic compounds production, reactive oxygen species or surface cracking) (Wu et al., 2023; de Freitas et al., 2022), and therefore could be compared. Each plastic type was photodegraded during two weeks; except PET film, which was photodegraded during four weeks as these films showed an extremely slow rate of degradation after FTIR analyses. Microplastics were randomly distributed in the chamber and their position was shifted twice during the photodegradation time. Then, plastic was manually cut with scissors and an upper size of 5.0 mm length for fibers and 5.0 mm² area for films and foams was established. Foams were cut by using a Philips HR3655/00 Standmixer (1400 W, ProBlend 6 3D Technologie, Netherlands) and sieved through a 4 mm mesh.

Soil preparation

We collected dry sandy loam soil (Albic Luvisol; 0.07% N, 0.77% C, pH 6.66) from a dry grassland plant community located in Dedelow, Brandenburg, Germany (53 37’ N, 13 77’ W). The soil was sieved (4 mm mesh size), homogenized, and mixed with each microplastic type at a concentration of 0.4% (w/w). Thus, 0.76 g of each microplastic type were mixed into 190 g of soil for each pot (4 cm diameter, 21 cm height, 200 ml). Microplastics were manually separated and mixed with the soil during 1 min in a large container, before placing it into each pot, to help provide an equal distribution of microplastics throughout the soil. The same handling was applied to the control soils (without microplastics) to provide the same disturbance level.

Experimental design

In September 2020, soil mixed with microplastics was pre-incubated without plants for 42 days in a glasshouse with a daylight period set at 12 h, 50 klx, and a temperature regime at 22/18 ॰C day/night with a relative humidity of 40%. This incubation time allowed for the interaction between soil microbial communities and microplastic particles, and the potential leaching of microplastic chemical substances into the soil. Pots were watered with 50 ml of tap water, and covered to avoid evaporation but allowing aeration. Seeds of Daucus carota were surface-sterilized with 4% sodium hypochlorite for 5 min and 75% ethanol for 2 min and then thoroughly rinsed with sterile water. Seedlings of similar size were transplanted into pots 3 days after germination, with one single seedling per pot. Then, plants grew for 43 days and were watered every third day with 40 ml during the first 3 weeks, and then every second day with 30 ml of tap water, to maintain water holding capacity at ~70%. The experimental design consisted of 8 microplastic types (3 fibers, 3 films, 2 foams) x 2 degradation levels (non-degraded, degraded) x 7 replicates = 112 pots. Twelve additional pots were established as control without microplastics. All pots were randomly distributed in the glasshouse chamber, and their position shifted twice during the experiment to homogenize environmental conditions. All plants survived until the end of the experiment. At harvest, plants were separated into above and belowground parts; soil was air-dried and stored at 25 ॰C for soil aggregation analyses, while fresh soil samples were used to measure soil respiration.

Microplastics characterization

We measured the water absorption capacity of microplastic pieces of 6 x 6 cm² following the UNE EN ISO 62:2008 standard. Plastic was immersed in distilled water for 24 hours and the changes in weight and thickness were determined. Water absorption was measured on films and foams. Due to the flexibility and brittleness of single fibers after degradation, it was not possible to assess their water absorption capacity. Identity of the plastics (i.e., polymer type) was confirmed by using a FTIR Jasco ATR-FTIR-4100 (Jasco International Co. Ltd., Tokyo, Japan), reflection mode, from 4000 to 600 cm^-1, an average of 32 scans performed with the resolution of 4 cm^-1. For each plastic type, we measured two samples three times.

Soil Measurements

Soil aggregation: We measured water-stable soil aggregates (WSA) following a protocol by Kemper and Rosenau (1986), modified by Lehmann et al (2019). Briefly, we placed 4.0 g of dried soil (<4 mm) on sieves with a mesh size of 250 µm. Soil was rewetted with deionized water by capillarity and inserted into a sieving machine (Agrisearch Equipment, Eijkelkamp, Giesbeek, Netherlands) for 3 min where the agitation and re-wetting caused the treated aggregates to slake. Then, we dried and weighed the water-stable fraction (dry matter) and subsequently, we extracted the coarse matter, which was also dried at 60 °C for 24 h. Soil aggregation was calculated as: WSA (%) = (dry matter - coarse matter)/(4.0 g - coarse matter).

Soil respiration: We measured soil respiration via infrared gas analysis. We placed 25 g of fresh soil in individual 50 ml centrifuge tubes (Sarstedt AG & Co. KG, Nümbrecht Germany, item number 62.548.004) whose lids were modified to control gas exchange via a rubber septum (Supelco, Darmstadt, Germany, item number 27235 U). We measured CO₂ concentration (ppm) at two time points: First, we flushed the tubes with CO₂ free air for five minutes to measure CO₂ concentration at time zero. Then, soil samples were incubated at 20°C for 24 h and we measured CO2 concentration for the second time. At both times, we took a 1-mL air sample and injected it to an infrared gas analyzer (LiCOR- 6400XT photosynthesis system, Li-Cor Biosciences, Lincoln, NE). Measurements were obtained every ~2 secs. The difference between the maximum and the minimum value (peak) was converted to ppm using the calibration equation (ppm= -467 +195.18 peak).

Plant measurements

Plant biomass and root morphological traits

At harvest, roots were carefully removed from the soil and gently washed by hand in order to measure morphological traits of fine roots (i.e., <2 mm in diameter which included mostly first to third order roots). We measured length, surface area, volume and root average diameter on a fresh sample (random portion of the middle part of the root system) using the WinRhizoTM scanner-based system (v.2007; Regent Instruments Inc., Quebec, Canada). We calculated different root morphological traits: specific root surface area (SRSA; cm² mg⁻¹), specific root length (SRL; cm mg⁻¹), root average diameter (RAD; mm) and root tissue density (RTD; root dry weight per volume mg cm⁻³). Shoot and root mass were measured after drying samples at 60 °C for 72 h.

Statistical analyses

We performed principal component analysis (PCA) on our response variables for each microplastic shape either degraded or non-degraded, using the function “prcomp” and “fviz_pca” from the R package “factoextra” (Kassambara and Mundt, 2017). Ellipses in the graph grouped the treatments with a confidence level of 0.95. To validate differences between treatments we used the function “manova” from the R package “MASS” (Venables and Ripley, 2002). We tested the null hypothesis by means of the Pillai trace statistic, which is robust to violation of multivariate normality and homogeneity of the variance-covariance matrix (Quinn and Keough, 2002).

We performed linear models to test the effect of microplastic degradation on plant-soil variables. Microplastic shape, polymer type and degradation were considered as explanatory variables. Polymer was nested within microplastic shape as each shape had different polymer types, and was included as a random factor in the model (Schielzeth and Nakagawa, 2013) as follows: ~ A*B +(1|A/C), with A representing ‘Shape’, B ‘Degradation’, and C ‘Polymer’. Response variables were log-transformed to fulfill linear models assumptions. The function “lmer” from the “lme4” R package was used in the mixed models. Then, we implemented the function “emmeans” from the eponymous R package to define pairwise comparisons. “Tukey” tests were used to compare each degraded microplastic type (shape and polymer) with its non-degraded counterpart, while the “Dunnett” test was used to compare each them with the control. Secondly, polymer and degradation were considered as fixed factors in the linear model. Residuals were checked to validate assumptions of normality and homogeneity and when necessary, we implemented the function “varIdent” from the “nlme” R package (Pinheiro et al., 2021) to account for heterogeneity in the treatment. Then, we implemented the function “glht” and the “Tukey” or “Dunnett” test from the “multcomp” R package to compare among treatments (Bretz et al., 2011). Statistical analyses were done in R 4.2.3 (R Core Team, 2023).