3.1. Effect of degradable films on soil properties and yield
UE, CAT, AKP, SC, SWC, ST, pH, and yield were significantly influenced by degradable film mulch (P < 0.01), whereas TN and organic matter content showed no significant differences among the different treatments (Table 1). The sunflower yield under film mulch was higher than that under CK treatment, and the yield magnitude was DF > CF > CK. Specifically, DF increased the yield by 4.92% and 33.49%, respectively, compared with those of CF and CK, and the difference was significant compared with ordinary mulching (P < 0.05). Compared with CK, film mulching significantly increased soil temperature and water content, resulting in heat preservation and moisture retention. Meanwhile, film mulching increased UE, CAT, AKP, SC, SWC, ST, TN, and SOM and decreased soil pH. Degradable film mulching regulated the underground soil microenvironment, provided a sufficient carbon source for microorganisms, and enhanced enzyme activity in the soil, promoting the increase in sunflower yield.
Table 1
Effect of degradable film treatments on soil properties and yield
Soil property and yield | DF | CF | CK | F-value |
UE (ug·(d− 1·g− 1)) | 524.58 ± 9.53 a | 488.47 ± 6.55 b | 376.49 ± 16.22 c | 180.24*** |
CAT(µmoL·(h− 1·g− 1)) | 494.11 ± 13.88 a | 468.91 ± 8.41 b | 341.17 ± 6.94 c | 58.842*** |
AKP (mg·(d− 1·g− 1)) | 1183.85 ± 11.39 a | 909.02 ± 22.59 b | 714.79 ± 14.99 c | 770.62*** |
SC (mg·(d− 1·g− 1)) | 33.28 ± 0.57 a | 32.93 ± 0.28 a | 26.14 ± 0.76 b | 194.64*** |
SWC(%) | 0.20 ± 0 .01 a | 0.19 ± 0.01 a | 0.18 ± 0.02 a | 3.42* |
pH | 8.37 ± 0.07 b | 8.21 ± 0.09 b | 8.63 ± 0.07 a | 27.22*** |
ST(℃) | 30.87 ± 0.25 a | 30.25 ± 0.64 a | 28.87 ± 1.03 b | 8.15** |
SOM(g·kg− 1) | 14.32 ± 0.63 a | 14.14 ± 0.59 a | 13.91 ± 0.38 a | 0.56 |
TN(g·kg− 1) | 0.7 ± 0.020 a | 0.67 ± 0.020 a | 0.63 ± 0.020 a | 7.58* |
Yield(kg·ha− 1) | 4452.13 ± 125.78 a | 3969.68 ± 183.18 b | 3335.17 ± 140.00c | 54.57*** |
Note: DF: degradable film; CF: commonly film; CK: no plastic film; UE: Urease; CAT: catalase; AKP: Alkaline phosphatase; SC: sucrase; SWC: Soil water content; ST: Soil temperature. Different lowercase letters represent significant difference (P < 0.05) |
3.2. Microbial diversity responses to degraded film mulch at different root ecological niches
Analysis of α-diversity of bacterial (Figure S1) and fungal communities (Figure S2) in bulk soil and rhizosphere-related ecological niche (e.g., rhizosphere and endosphere) showed that the Shannon and Chao indices of bacterial and fungal communities in bulk and rhizosphere soil were significantly higher than those in the endosphere (Figures S1 and S2). The degradation of film coverage significantly enhanced the diversity of bacterial microorganisms compared to CK (P < 0.05). In contrast, common film mulching reduced bacterial microbial diversity in both bulk soil and rhizosphere soil, while increasing it specifically in the rhizosphere. Plastic film mulching decreased the diversity of fungal microorganisms, and there was no significant difference among the different treatments, indicating that the effect of film covering on bacteria was higher than that of fungi. Linear regression also found that bacterial and fungal diversities were significantly correlated with the degradation of film mulching treatment in rhizosphere and endosphere, whereas the correlation of fungi was higher than that of bacteria (Fig. 2).
3.3. Microbial community responses to degradable film mulch at different root ecological niche
The dominant bacterial phyla in different ecological niches are Proteobacteria, Actinomyces, and Firmicutes (Figure S3a–c), and the dominant fungal phyla are Ascomycota, Olpidiomycota, Basidiomycota, and Mortierellomycota (Figure S3d–f). The main factors affecting the dominant phyla of bacteria and fungi were SWC, SC, UE, AKP, and CAT (Figure S2), indicating that degradation membrane treatment was beneficial for water retention and microbial reproduction. PERMANOVA confirmed that ecological niche variation (bulk soil, rhizosphere, and rhizosphere) was the main factor influencing the diversity of bacterial and fungal communities (interpretive variation > 60%, P < 0.001), and the effect on bacteria was higher than that on fungi, whereas degradation film mulch had no significant effect on bacteria but had a significant effect on fungi. However, PERMANOVA of different ecological niches showed that degradable film mulch significantly affected bacterial and fungal endosphere community diversity (P < 0.001) (Table 2).
PCoA was used to analyze the effects of plastic film mulch on bacterial (Fig. 3a–d) and fungal (Fig. 3e–h) β-diversities in different ecological niches. There were significant differences in β-diversity between bulk soil and rhizosphere communities, whereas there were no significant differences between rhizosphere and endosphere (P < 0.05; Fig. 3). Significant differences in bacterial and fungal communities under different degradation membrane treatments indicated that degradable film mulch would significantly affect the changes in soil microbial community structure.
The random forest was used to identify bacterial and fungal biomarkers and environmental drivers covered by degradable film mulch (Fig. 4). The top 20 ASVs of bacteria (Fig. 4a) and fungi (Fig. 4d) were analyzed. ASV2121 (Amycolatopsis) and ASV2102 (unclassified_Bacteria) had significant differences in bacterial communities and were significantly enriched under degradable film mulch. ASV2121 (Amycolatopsis) was positively correlated with soil characteristics. In particular, there was a significant positive correlation with UE, CAT, and AKP (Fig. 4c). ASV6 (Schizothecium) exhibited significant differences in fungal communities and is significantly concentrated under CF. Through random forest analysis, the contribution of the top 20 ASVs of bacteria was > 50% (Fig. 4b) and that of the top 20 ASVs of fungi was > 60% (Fig. 4e). ASV556 (Fusarium) showed a significant negative correlation with UE, CAT, AKP, SC, SWC, ST, SOM, and TN (Fig. 4f), and pH was the key driving factor affecting the abundance of the top 20 ASVs in bacteria and fungi.
Table 2
PERMANOVA of the effects of degradable film mulch on fungal communities in bulk soil, rhizosphere, and endosphere.
Factors | Bacterial community | Fungal Community |
R2 | p-value | R2 | p-value |
Bulk soil | 0.674 | 0.001 | 0.702 | 0.001 |
Rhizosphere | 0.923 | 0.001 | 0.868 | 0.001 |
Endosphere | 0.265 | 0.133 | 0.957 | 0.001 |
Degradable mulching film | 0.025 | 0.664 | 0.214 | 0.013 |
Ecological niche | 0.914 | 0.001 | 0.608 | 0.001 |
3.4. The assembly process of root microbial communities with different ecological niches
The whole community zero model analysis showed that the ecological assembly of bacterial and fungal communities under plastic film mulch was composed of selection and neutral processes, whereas neutral processes such as diffusion restriction and drift dominated bacterial and fungal community structures under each treatment (Fig. 5a and d). Compared to CK, the assembly process of bacterial communities in the laminating treatment increased the heterogeneous selection process and reduced the homogeneous diffusion process, whereas CK significantly increased the diffusion restriction process of fungi (Fig. 5b). By calculating the βNTI value based on the ASV abundance matrix to evaluate the ecological assembly process of the microbial community, the proportion of |βNTI| > 2 of the bacterial community under DF was also higher than that under the no film mulch, indicating that the composite film treatment would promote the transformation of the microbial community assembly process to a deterministic process (Fig. 5b and e). For different ecological niches, the proportion of rhizosphere |βNTI| > 2 was higher than that of bulk soil and endosphere (Fig. 5c and f), indicating that rhizosphere ecological assembly gradually changed from a random process to a deterministic process.
The NCM was used to predict the relationship between occurrence frequency and relative abundance of ASV in subcommunities of three different ecological niches in bulk soil, rhizosphere, and endosphere and in all root datasets (Fig. 5g–n). Results showed that the NCM successfully estimated most relationships between the occurrence frequency of ASV and its relative abundance change, with high interpretation rates (R2) in bulk soil, rhizosphere, endosphere, and sum, indicating that random processes are important for the formation of microbial community assembly in different ecological niches. Moreover, the Nm values of bacteria and fungi in bulk soil (Nm = 1074.81 and 371.14) were higher than those in the rhizosphere (Nm = 1022.74 and 491.49) than in endosphere (Nm = 678.13 and 224.26), indicating that the species diffusion of microorganisms in bulk soil was significantly higher than that in rhizosphere and endosphere.
3.5. Microbial co-occurrence network changes at different root spatial locations
Symbiotic networks of bacterial and fungal communities in bulk soil, rhizosphere, and endosphere were constructed under the degradable film mulch (Fig. 6). There were more nodes and edges of the bacterial community network, rhizosphere (nodes 197 and edges 4157) and endosphere (nodes 200 and edges 4181) than bulk soil (nodes 200 and edges 2908), and more nodes and edges of the fungal community network, rhizosphere (nodes 199 and edges 2984) than bulk soil (nodes 200 and edges 1727) and endosphere (nodes 118 and edges 788). It was shown that the complexity of the rhizosphere soil bacterial and fungal network was higher than that of bulk soil and endosphere (Table S3). The modular degree of bacterial and fungal community networks was the largest in the endosphere, and the positive correlation edges of fungi were significantly more than those of bacteria, indicating that fungi had a more obvious positive influence on the microbial network. Bacillus, Anaeromyxobacter, and Sphingomonas belong to the dominant genus of bulk soil bacteria. Pseudoxanthomonas, Rhodonellum, and Amycolatopsis belong to the rhizosphere dominant bacteria, and Aquiflexum, Aminobacter, and Aromatoleum belong to the endosphere core bacteria. Olpidium, Spizellomyces, Lecythophora, and Dichotomopilus belong to the dominant genus of bulk soil fungi. Wardomyces, Olpidium, and Schizothecium belong to the dominant rhizosphere fungi, whereas Sarocladium, Madurella, and Striaticonidium belong to the core endosphere fungi (Table S4).
The subordinate level analysis reveals the changes in the bacterial network graph (Fig. 4) after the degradable film cover was applied. The degradable film mulching significantly increased the number of edges in the bacterial network, decreased the modularity index, increased the average degree, and increased the network connectivity. The changing trend of a fungal network diagram differed from that of bacteria. Degradation film treatment can reduce the number of fungal network diagram edges, increase the proportion of degradable film mulching positive correlation edges, and positively regulate the change of microbial network. Moreover, degradation film treatment can significantly increase the average network diameter and reduce the modularity index and average degree of the network. It was suggested that degradation membrane covering reduced the network complexity of fungi (Figure S4; Table S5).
According to ZiPi most ASVs in these co-occurring networks are classified as peripheral nodes. Module hubs, connectors, and networks were defined. The total number of key bacterial dominant species in hubs was 273 (e.g., Helianthus, Marmoricola, Chryseolinea, Massilia, Rhizobiaceae, Microbulbifer, Rhizobiaceae, Amycolatopsis, Sphingomonas, and Bacillus) (Table S6), The total number of fungal dominant species was 141 (e.g., Coprinopsis, Gibberella, Pseudeurotium, Golovinomyces, Kernia, Septoglomus, Olpidium, and Pyxidiophora) (Table S6).
3.6. Effects of different ecological niches in roots on increasing yield driven by microorganisms
Structural equation and random forest synthesis (Fig. 7) were used to analyze the effects of degradable film mulching on soil physicochemical properties, enzyme activity, and microbial diversity of different ecological niches of bacteria and fungi on yield. Results showed that soil physical and chemical characteristics significantly affected soil enzyme activity, and soil enzyme activity significantly affected microbial diversity, particularly fungal diversity, and further significantly affected yield. The random forest further verified that the main microbial driving factors affecting yield were bacterial bulk soil diversity and fungal endosphere diversity, and the main soil driving factors were UE, AKP, CAT, SC, pH, and ST. Analysis of the direct, indirect, and total effects of yield contribution also verified that the main influencing factors of yield were soil physical and chemical properties and soil enzyme activity.