3.1 Effects of ILs on the growth of soil culturable microorganisms
There was no significant change in the fungi population when the concentration was 1.0 mg/kg and 5.0 mg/kg, while at high concentrations (10.0 mg/kg and 20.0 mg/kg), the fungi population decreased significantly (Fig. 1a-c). The exposure time had no significant effect on the fungi population. The bacteria population decreased as the concentration of ILs increased (Fig. 1d-f), indicating a significant effect of ILs on the growth of bacteria and that this effect was exacerbated at high concentrations and prolonged exposure. Actinomyces in the soil were very sensitive to ILs, and the actinomycete population in the treated group decreased significantly compared to the control group (Fig. 1g-i). There was less variation in the actinomycete population between the high concentration groups treated with ILs (5.0 mg/kg, 10.0 mg/kg and 20.0 mg/kg), while the exposure time had no significant effect on actinomycete population.
Three ILs inhibited the growth of fungi, bacteria and actinomyces in the soil and the results showed that ILs had toxic effects on culturable microorganisms in soil. Different concentrations of ILs showed different degrees of inhibition on the microbes (Zhang et al., 2018), which was consistent with our results. Many ILs are similar in structure to cationic surfactants. They can interact with proteins on the cell membrane of microbial, and then enter the cells, causing microbial respiration to weaken, and cell membrane metabolism to decrease. This may be the reason why the three ILs in this experiment all had inhibitory effects on the growth of soil culturable microorganisms(Cheng et al., 2019).
3.2 Effects of ILs on the activity of soil enzyme
Extracellular enzyme activities (β-glucosidase, urease and phosphatase) in the biogeochemical cycle were used as biological indicators of soil quality(Gougoulias et al., 2014). Dehydrogenase activity was a direct indicator of soil microbial activity because it reflects the biomass of soil microorganisms(Jouni et al., 2023).
There was a significant decrease in the activity of urease in the [C8MIM]Cl and [C8MIM]Br groups at the highest concentration (20.0 mg/kg) compared to the control group from the 20th day of exposure (Fig. S2a-c). No significant effect of [C8DMIM]Br on the activity of urease was observed. Dehydrogenase activity showed a trend of increasing and then decreasing with increasing concentration of ILs on the 10th and 20th day, with only the 20 mg/kg groups showing a significant decrease on the 30th and 40th day (Fig. S2d-f). The dehydrogenase activity of the [C8DMIM]Br group showed a significant decrease on the 40th day in the 20 mg/kg group. Only the higher concentration (10 and 20 mg/kg) on the 20th day showed a significant decrease in acid phosphatase activity under the three ILs groups (Fig. S2g-i). From day 20 of exposure, β-glucosidase activity increased significantly in each concentration group of ILs compared to the control group, with small differences between the different concentrations (Fig. S2j-l).
Exposure of soils to high concentrations of ILs can cause a decrease in the nitrogen content of the soil(Zhang et al., 2014). The ILs stimulated increased dehydrogenase activity at low concentrations and were more toxic at higher concentrations, inhibiting dehydrogenase activity(Campos et al., 2019). The same results were found when studying the effects of chemicals such as herbicides on dehydrogenase activity in the soil(Kaczynski et al., 2016). Some ILs groups affected acid phosphatase activity and thus the nitrogen and phosphorus cycle of the soil(Adetunji et al., 2021). β-glucosidase is a cellulase enzyme, suggesting that ILs may promote the degradation of soil cellulose(Buivydaitė et al., 2023).
3.3 16S sequencing data of soil microorganisms
To investigate the effects of ILs on the microbial community in soil, this study used 16S sequencing technology to analyze the control group, [C8DMIM]Br treatment group, [C8MIM]Br treatment group and [C8MIM]Cl treatment group. Statistics were shown in Table S1. A total of 624947 clean reads were obtained from the 12 samples, where the average length of clean reads was 418 bps, with an average of 52079 clean reads per sample (ranging from 40320 to 58496). To minimize bias due to sequencing depth, samples were drawn flat and a total of 3443 OTUs were defined based on a rule of more than 97% similarity. The rarefaction curves indicated that the sequences obtained reflected the bacterial diversity of each sample (Fig. S3). To describe the increase in total and core species in the samples, a pan/core analysis was carried out and details were shown in Fig. S4 and S5.
3.4 Analysis of Alpha diversity of soil microbial communities
Alpha diversity analysis was used to explore the abundance and diversity of soil microbial communities after exposure to different ILs(Yan et al., 2021), and the results were shown in Fig. S6 and Table S2. The two treatment groups, [C8DMIM]Br and [C8MIM]Cl, showed an increasing trend in community diversity relative to the control group (CK), while the [C8MIM]Br group showed a decreasing trend compared to CK, according to the Sobs, Ace, and Chao indices. There was a decreasing trend in Simpson index in all ILs treatment groups except CK, indicating an increase in community diversity in all ILs treatment groups(Qiu et al., 2022). There was no significant change between the CK and experimental groups in the Shannon and Coverage indices.
The higher species abundance of the microbial community in the [C8DMIM]Br group compared to the [C8MIM]Br group indicated that the ILs with double methyl substituents are less toxic than the ILs with single methyl substituent at the same concentration(Lian et al., 2022). The increased species abundance in the [C8MIM]Cl group compared to the [C8MIM]Br group may be due to the self-protective mechanism to maintain community homeostasis by increasing the abundance of community species to enhance certain functions when toxicity increases to a certain value (Liu et al., 2018)。
3.5 Analysis of the species composition of soil microbial communities
We detected 33 species of bacteria at the phylum level by analyzing the species community composition. Of these, Actinobacteria, Proteobacteria, Acidobacteriota, Firmicutes, and Chloroflexi were the dominant species in the soil microbial community, with the remaining bacteria of less than 1% being placed within the “other”, as shown in Fig. 2a. Meanwhile we carried out the analysis of the species community composition at the family level, the results of which are shown in Fig. 2b.
The percentage of Actinobacteria and Proteobacteria in the microbial community decreased significantly after exposure to ILs, especially in the [C8MIM]Cl group. The percentage of Actinobacteria in the control group was 35.51%, whereas it decreased to 23.74% after [C8MIM]Cl exposure. Proteobacteria accounted for 30.64% in the control group and 24.21%, 24.98% and 25.70% in the [C8DMIM]Br, [C8MIM]Br and [C8MIM]Cl groups, respectively. Acidobacteriota became significantly higher in the microbial community with 4.38% in the control, 14.72% in [C8DMIM]Br, 17.42% in [C8MIM]Br and 19.28% in [C8MIM]Cl. The remaining major dominant species, Firmicutes and Chloroflexi, decreased and increased their proportion in the microbial community following soil exposure to ILs, respectively, with both the [C8MIM]Br and [C8MIM]Cl groups showing a greater degree of change than the [C8DMIM]Br group.
We presented the species shared by the control and experimental groups as well as the endemic species at the family level in a Venn diagram (Fig. 2c). A total of 291 families of core microbiota were detected, of which 184 families were shared and 18 species of bacteria were unique to the control group at the family level, with the top three being Selenomonadaceae, Enterococcaceae and Veillonellaceae. The [C8DMIM]Br, [C8MIM]Br and [C8MIM]Cl groups had 11, 13 and 16 endemic bacteria, respectively.
The results were clustered by similarity in species abundance between samples, presenting information on the species composition and species abundance of the microbial community, as shown in Fig. 2d. The control and experimental groups clustered clearly apart and differed significantly. Several of the bacteria with the greatest difference in species abundance were Micrococcaceae, Pyrinomonadaceae, Streptomycetaceae, Blastocatellaceae, Vicinamibacteraceae, KD4-96, and MB-A2-108.
High levels of species diversity and abundance were considered important features of ecosystem integrity, which increased the resilience of soil microbial communities to disturbance(Zheng et al., 2023). ILs altered the species composition of microorganisms in the soil and reduced the proportion of dominant bacterial species, resulting in a lower abundance of bacterial species in the experimental group than in the control group. Similar results were obtained when studying the effects of nanopolystyrene addition on soil microbial community composition, indicating ILs have a similar potential effect on soil microorganisms as nanopolystyrene (Lehmann, 2020; Zou et al., 2022).
3.6 Comparative analysis of soil microbial communities
Similarities in the composition of the soil microbial community can be studied by principal component analysis (PCA) of the control and experimental groups (Fig. 3) (Alsayegh et al., 2021). The control (CK) and [C8DMIM]Br groups (BrS) had the most overlap in the figure, followed by the [C8MIM]Br groups (BrD). The control and [C8MIM]Cl groups (Cl) had no overlap. There was more overlap between the three experimental groups. This meant that CK was the most similar to BrS in terms of community composition and the least similar to Cl, with significant differences between CK and experimental groups. The results of comparative analysis of soil microbial communities showed that the ILs treatment had a significant effect on the soil microbial community composition, and the ranking of the effect was: [C8MIM]Cl༞[C8MIM]Br༞[C8DMIM]Br.
3.7 Differences in the species composition of soil microbial communities
The control and experimental groups were tested for significant differences and the results were shown in Figs. 4 and 5. At the phylum level, 15 phyla were significantly different between the control and experimental groups (p < 0.05), the top eight being Actinobacteriota, Proteobacteria, Acidobacteriota, Firmicutes, Chloroflexi, Myxococcota, Bacteroidota, and Gemmatimonadota, with the top five phyla being the dominant species of bacteria.
A two-by-two analysis of variance within the experimental groups was performed (Fig. 6) and the two groups with the greatest differences in bacteria at the family level were found to be the [C8DMIM]Br and [C8MIM]Cl groups, followed by the [C8DMIM]Br and [C8MIM]Br groups, according to the values of the p-value.
The species abundance of some microbiome had a significant impact on the differential effect, and LEfse analysis allowed the identification of such microbiome (Fig. S7), with a greater number of such microbiome indicating a greater difference. A comparison of the three ILs groups with the control group revealed zero in the [C8DMIM]Br group, two in the [C8MIM]Br group and five in the [C8MIM]Cl group. At the family level, 9 down-regulated biomarkers and 6 up-regulated biomarkers were found in the control group compared to the [C8DMIM]Br group; 10 down-regulated biomarkers and 5 up-regulated biomarkers compared to the [C8MIM]Br group; 7 down-regulated biomarkers and 6 up-regulated biomarkers compared to the [C8MIM]Cl group. ILs with single methyl substituent caused greater differences in the composition of the soil microbial community than those with double methyl substituents(Wang et al., 2020). The three ILs caused significant changes in the species composition of the soil microbial community, and the three ILs caused different levels of variation(Chu et al., 2022a).
3.8 Predictive analysis of the function of soil microbial communities
Bacterial gene function was predicted from 16S rRNA sequencing data by using PICRUSt. The control group showed six different categories of KEGG pathways at the first level compared to the experimental groups, which included ‘Metabolism’, ‘Genetic Information Processing’, ‘Environmental Information Processing’, ‘Cellular Processes’, ‘Human Disease’, and ‘Organic Systems’. Among these, ‘carbohydrate metabolism’ was the most abundant burst, followed by the KEGG pathway associated with ‘amino acid metabolism’ (Table S3). In addition, COG abundance statistics tables and abundance plots of COG categories (Table S4 and Fig. 7) can be seen in the COG functions showing that 23 functions produced differences in the experimental groups compared to the control group. Similar results were seen when soils are contaminated with polycyclic aromatic hydrocarbon(Li et al., 2019), and ILs may inhibit the metabolic activity of microorganisms if they enter the soil, thus affecting soil quality.