As a globally applied agricultural practice, plastic mulching has produced huge productive and economic advantages including increased yield and improved crop quality since 1960s (Lamont, 2005). It is estimated that the global agricultural films market used for greenhouses, mulching and silage will grow by 59% from 2018 to 2026, especially, plastic mulching account for over 40% (Sintim and Flury, 2017; Sintim et al., 2020). Unfortunately, their improper treatment and environmental accumulation have threatened the development of green and sustainable agriculture in various ways, such as biological accumulation in the food chain, metabolism of soil carbon and nitrogen, and becoming a hotspot of antibiotic resistance genes (ARGs) and pathogens (Ju et al., 2021; Zhu et al., 2021). Biodegradable plastic mulches (BDMs) were introduced in 1990s and have been developed as important substitution for classic plastic mulches in alleviating environmental pollution and reducing labor (Kasirajan and Ngouajio, 2012). Nowadays, global bioplastics production capacity is growing at a considerable rate and is expected to increase from around 2.11 million tons in 2018 to 2.62 million tons in 2023 (Chinthapalli et al., 2019). They usually include two types according to raw materials, namely, bio-based polymers represented by polylactic acid (PLA) and fossil-sourced polymers represented by polybutylene adipate terephthalate (PBAT; Kasirajan and Ngouajio, 2012). As a major advantage, BDMs are usually tilled into soil after their usage and are expected to be gradually biodegraded in situ over a relatively short period. Generally speaking, microorganisms acting on different types of BDMs secrete extracellular hydrolases, which break down the complex polymers into simpler oligomeric and monomeric units. Then, the microorganisms uptake and utilize them for biomass formation and energy production (Sander, 2019; Bandopadhyay et al., 2020). However, the true degradation rate and state of different types of BDMs debris in soil after harvest and their potential ecological risk in comparison with classic plastic mulches have always been controversial (Sintim and Flury, 2017; Bandopadhyay et al., 2018; Sridharan et al., 2021). Meanwhile, it has always been a key issue of concern that microorganisms preferentially adhered to the debris surface may affect the assembly processes of biofilms and lead to a series of ecological heterogeneity (Hou et al., 2021; Sun et al., 2021).

As a new anthropogenic substrate, hydrophobic plastic surfaces can readily be colonized by various microorganisms, constituting a diverse and unique ecological habitat called the “plastisphere” (Zettler et al., 2013; Amaral-Zettler et al., 2020). This region is usually considered as a hot spot for nutrient accumulation, active microbial metabolism, and biodegradation of complex compounds. Moreover, the plastisphere often contain specific functional taxa, particularly the consortia capable of degrading plastics and invasive or pathogenic species. It not only hints at the potential to isolate degradable functional taxa from them, but also raises concerns of ecological risk with respect to aquaculture and human health (Wu et al., 2019; Hou et al., 2021). Therefore, it is of great importance to study the temporal dynamics of plastisphere (Jacquin et al., 2019; Amaral-Zettler et al., 2020; Khan et al., 2020). Through whole-genome sequencing, Bhagwat et al. (2021) investigated the composition and functions of microbes attached to four kinds of plastics. The results suggested that both non-degradable plastics and a kind of BDMs exhibited unique microbial profiles and niche partitioning. All substrates were significantly enriched several microbial genera with functional genes involving in carbon cycling, xenobiotic compound degradation, and plant–pathogen interaction. Recently, a few studies carried out in soils showed similar conclusions as those from aquatic environments. The surface of PE could also act as a distinct microbial habitat in soil and select unique microbiota for enhancing pathways of amino acid metabolism and xenobiotics biodegradation (Huang et al., 2019; Ruthi et al., 2020). Zhou et al. (2021) investigated the microplastisphere of a biodegradable polymer, and the results indicated that the microbial community could utilize the polymer as a carbon source, which had a more active microbial biomass and increased specific microbial growth rate compared with that of the bulk soil. Additionally, considering that various compounds are adsorbed and released during the process of biodegradation, it was expected that a variety of biomarker taxa could be identified to distinguish microbiota between different plastic components (Zettler et al., 2013; Ruthi et al., 2020). For example, PLA/PBAT is degraded into intermediates with different chain lengths or small molecules, such as lactic acid and butanediol, which could be regarded as a nutrient source and may recruit specific microbial taxa (Zumstein et al., 2018). However, it is noteworthy that although the above studies tended to indicate that microbes selectively colonize the plastisphere and form specific biofilms, the dominance of stochastic processes in the community construction on plastics was observed in both aquatic and soil environments (Li et al., 2021a; Zhang et al., 2021; Zhu et al., 2021). Polymer types and sizes, sampling site, or exposure time may all be important factors causing higher stochasticity (Zhang et al., 2021). For example, Zhu et al. (2021) found that specific metabolic pathways selectively drove diverse ecological processes, such as the enrichment of potential pathogens and ARGs in soil, while neutral processes dominated plastisphere community assembly. Overall, current studies have focused on either traditional plastic in aquatic environments or their effects on soil microorganisms and enzyme activities (Sander, 2019; Shruti and Kutralam-Muniasamy, 2019). Little is known about the succession and assembly process of the microbial community during the degradation processes of different plastics, especially BDMs, in soils.

In order to ensure the repeatability and comparability, a pot experiment within a greenhouse was carried out in our study. PE mulch and four types of BDMs were buried into ceramic pots with 12 biological duplications and extracted for five consecutive months. We chose not only pure PLA and PBAT, but also two mixtures of these compounds by considering real agricultural application. Base on the experimental design, we proposed to address the following four scientific questions. (i) Regarding differences in composition of the microbial communities, does time or composition of the plastic have a greater impact? (ii) What kind of dynamic changes occur in the composition and functions of the attached microbiota during the degradation period, and what are their potential ecological risks? (iii) What are the biomarker taxa in the plastisphere to distinguish BDMs with PE mulch? and (iv) What are the assembly processes of microbial communities colonized on the plastic films, and how do they change over time?

In order to determine the fate of both PE and BDMs in soils, it is necessary not only to characterize the species composition and assembly mechanism within the plastisphere, but also to reveal the succession of functional microbial taxa which drove the degradation. To date, studies on the plastisphere of PE and BDMs in soils have been relatively scarce (Ruthi et al., 2020; Zhou et al., 2021). In this study, we successionally surveyed bacteria in the plastisphere of PE, PLA, PBAT, and two mixed component mulches in the potted soils over a 5 months period. Exploring the temporal differences and similarities among conventional PE and four types of BDMs deepened our understanding of the functional potential and ecological impact of the plastisphere in soils.

Previous studies have shown that different plastic substrates could recruit various microbial communities to adhere and form biofilms. Meanwhile, the biofilms showed dynamic changes with the residence time of plastics in aquatic and soil environments (Jacquin et al., 2019; Wu et al., 2019; Ruthi et al., 2020). However, these studies usually demonstrated only the temporal changes of microbiota on single non-degradable plastics or BDMs individually, without determining whether the differences were caused by the residence time in soil or the plastic composition (Bandopadhyay et al., 2020; McGivney et al., 2020). Here, our study showed that composition of the plastic components and length of incubation were both important driving forces of microbiota dissimilarity and that both had significant effects (Table 1). When PE samples were considered, the difference in the composition of the plastics was greater than that of sampling month, while the difference in microbial community caused by incubation time of the four BDMs in soil was greater than that caused by different components (Table 1). Since these five types of plastics used in this study are widely applied in agricultural mulching, whether these results are applicable to other non-degradable plastics and BDMs remain to be investigated.

With the wide usage of BDMs in agriculture, the potential ecological risks of BDMs in soil have been of general interest (Bandopadhyay et al., 2020; Qin et al., 2021). Our results of species composition suggested that the microbiota of the five plastics were similar at the phylum level, but zooming in the order level, microbes of BDMs and PE showed significantly different patterns of composition and changes (Figure 2). Similar to our results, several studies have reported that the predominant bacterial orders of BDMs, Burkholderiales and Pseudonocardiales, were closely related to plastic degradation (Pérez-Pantoja et al., 2012; Ruthi et al., 2020). They have been observed to play a major role in the degradation of PLA. These suggested that BDMs were capable of enriching for specific plastic degrading bacteria, which was consistent with our results of the functional predictions (Figure 3). Some their genera have also been cultured to further explore potential metabolic pathways (Jeszeová et al., 2018). The protein RPA1511 isolated from Rhodopseudomonas palustris has the catalytic triad residues, which are important for the hydrolysis of both monoester and polyester substrates. It was considered to be critical for the hydrolysis of PLA and was active in catalyzing solid PLA with the production of lactic acid monomers, dimers, and larger oligomers (Hajighasemi et al., 2016; Gricajeva et al., 2021). Moreover, some functions related to the carbon, nitrogen, and sulfur metabolic pathways were revealed by FAPROTAX, which indicated that BDMs may affect the biogeochemical cycling of elements in agroecosystems. Considering the residual BDMs in the farmland as a carbon input, their ecological effects need to be assessed separately in terms of their degradation in the relatively short period and their continued accumulation in the long-term mulch. In particular, we were interested in the enriched functional taxa related to degradation and pathogenicity, as both of these groups are of wide concern and may have a direct impact on the ecological risk of agroecosystems. For example, Xanthomonadales, as an important and widely studied plant pathogen in agricultural production (Bayer-Santos et al., 2019), had been found on all plastic mulches. The above findings are supported by recent studies based on metagenomic and metatranscriptomic technologies in aquatic environments (Seeley et al., 2020; Bhagwat et al., 2021; Wright et al., 2021).

However, knowledge about the microbial taxa for several plastics remains limited, especially lack of qualitative comparison between non-degradable PE and BDMs (Bhagwat et al., 2021). Here, we particularly emphasized the comparative analysis using samples of PE as the control group, which provided us with an accurate assessment of whether the use of BDMs in agricultural production is effective in addressing the soil environmental contamination and ecological risks caused by PE. The response ratio results further suggested that distinct functional groups of degradation can be found on PE and BDMs though some taxa were shared, which could due to different degradation products and mechanisms (Figure 3). As found in previous studies, some cultured taxa were observed to have broad degradation capabilities of plastics, and some specific degrading genera may represent exclusive degradation mechanisms and metabolic pathways for one of plastic components (Pérez-Pantoja et al., 2012; Jeszeová et al., 2018; Butbunchu and Pathom-Aree, 2019). For example, various strains of Pseudomonas can secrete exopolysaccharides attachment to plastic mulch and have a broad range of enzymes, which are promising candidates for biodegradation of both PE and BDMs (Wilkes and Aristilde, 2017; Bandopadhyay et al., 2020), which can also be observed in the species composition of our study. In addition, it was surprising that BDMs seem to have an overall significantly higher (>95%) pathogenic functions for plant, animal, and human-associated pathogens compared to PE (Figure 3). This was consistent with the result of species composition at the order level (Figure 2B). Considering the degradation cycle of BDMs in soils, it is reasonable to assume that the residence of BDMs debris may cause more negative agricultural production impacts and serious ecological risks in the relatively short term (Moreno et al., 2017; Qin et al., 2021).

To further distinguish the representative taxa on BDMs in soils, a high-resolution RF model was carried out to identify individual bacterial species that could be sought to discriminate different plastic components. The results (Figure 4) indicated that most of the identified biomarkers were highly congruent for PLA/PE and PBAT/PE at the species (ASV) level. A possible explanation is that BDMs or their metabolites provide suitable ecological niches for the bacterial members that readily colonize the plastisphere. In addition, it was worth noting that all ASVs that specifically belonged to PLA and PBAT could be classified to the dominant order Burkholderiales. It has been widely reported that Burkholderiales had a specialized ability for decomposing recalcitrant polyaromatic hydrocarbons or act as a plant or animal pathogen (Pérez-Pantoja et al., 2012; Sasaki et al., 2021). This was consistent with the results of species classification and functional prediction (Figures 1, 2). For the temporal heatmap, it was obvious that the biomarkers of PE and BDMs varied greatly with sampling month. This may be due to the fact that the effects brought by the plastic components were reflected not only in the differentiation of the entire bacterial community (Figure 1), but also in the impact on these specific biomarker species.

Disentangling the assembly mechanisms of biofilm microbial communities is a longstanding issue of ecologists. To date, studies have explored the dynamics of microbial communities inherited from conventional non-degradable plastics in aqueous environments by means of null modeling-based or neutral theory-based approaches. Most studies demonstrated the dominant role of stochastic processes in plastisphere microbiota (Sun et al., 2021; Zhang et al., 2021; Zhu et al., 2021), and a few suggested a higher influence of deterministic ecological processes (Hou et al., 2021; Li et al., 2021a). None of previous studies presented the microbial assembly of BDMs in soils. Although our results on species composition and function tended to suggest strong selection as the driver of community construction, the iCAMP results demonstrated a higher proportion of stochastic processes (dispersal plus drift) than deterministic processes (selection) for both BDMs and PE (Figure 5; Supplementary Figure S4). A smaller selective role may imply that the plastisphere microbial community was more resistant to changes in plastic components and environmental factors. Many previous studies have found that properties of plastics and environmental factors largely affected the composition and assembly mechanism of microbial communities (Li et al., 2021a; Sun et al., 2021). One of the possible explanations for the lower selection observed in our study is that the samples were only collected in our potted soils. The scale of sampling may influence the importance of stochastic and deterministic processes in shaping microbial communities (Chase, 2014; Zhou and Ning, 2017). It was reported that narrow scales tended to maintain a lower level of environmental heterogeneity, which may reduce the impact of selection and increase the role of stochastic processes (Garzon-Lopez et al., 2014; Wang et al., 2020b; Sun et al., 2021). These studies corroborated with our findings on BDMs. Therefore, sampling scale effects need to be considered when interpreting the assembly patterns of microbial communities on plastic in future studies.

Another interesting result was the opposite trends of bacteria on BDMs and PE mulches in the selection and drift processes over time (Figure 5; Supplementary Figure S4). The potential reason of their opposite trends may be the temporal scale effects of low migration and preferential colonized selection effects within plastisphere communities in soil. The proportion of selection on BDMs and PE was similar in the initial month, which may be due to the fact that BDMs had not yet started to degrade and both of them were present as common adherents. Thereafter, the selection stress on PE mulch consistently increased over time (Figure 5). This was possibly attributed to the microbial preferential colonization and selection effects that emerged in PE as a non-degradable stable plastic, resulting in enhanced selection (Wang et al., 2020a; Li et al., 2021a; Sun et al., 2021). In contrast, with the continuous degradation and fragmentation of BDMs over time, the trend of the microbial survival space (colonized area provided by the substrate) and proportion of selection process were consistently decreasing (Sander, 2019; Sintim et al., 2020). It is probably due to the spatial separation of the plastisphere microecology and niche caused by degradation, causing an increase in the stochastic process (Zhou and Ning, 2017; Sun et al., 2021). This may further elucidate the deep mechanism by which non-degradable PE and BDMs have different patterns of community construction. Moreover, our findings also indicated that understanding the pattern of temporal turnover was of great significance in revealing the community assembly of plastisphere.

In this work, we focused on five plastic mulches (especially BDMs) in soil and performed a time-series pot soil trial to track the changes in species composition and function of plastic-associated microbiota during degradation. Samples with similar plastic composition showed more similar changes at the phylum and order levels, while BDMs and PE were capable of enriching for specific plastic degrading bacteria which may imply that they had different degradation mechanisms. For future studies, combining multi-omics approaches, such as macroproteomics, metaproteomics, and culturomics, together may be a trend to provide comprehensive information on the biodegradation mechanisms and metabolic pathways. Enriched degradation-related and potential pathogen-related functional groups were discussed in this report, which provided a reference for assessing the potential ecological risk of BDMs debris in agricultural fields. BDMs and PEs showed a predominance of stochastic processes and displayed opposite patterns of community assembly over time. Sampling and temporal scale effects demonstrated the possible underlying mechanisms of the phenomenon. Therefore, future studies need to take these factors into account when focusing on the succession and assembly mechanisms in the plastisphere of BDMs. More soil types and locations also need further investigation in the experimental design to exclude the influence of discrepancies in physicochemical properties on the results.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

ZJ, DJ, and YD conceived and designed the experiments. ZJ performed the experiments. ZJ, XD, KF, SL, and SG analyzed the data. ZJ and YD wrote the paper. All authors read and approved the final manuscript.

This study was supported by the National Natural Science Foundation of China (nos. 31861133002 and 41977122).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.