3.1 Effect of GR on soil enzyme activities
The soil enzyme activities are considered as excellent indicators of changes of soil microbial activity under GR stress. Fig. 1 summarize the enzyme activity of soil exposure to GR through different adding methods. A decrease in the enzyme activities of urease and FDA esterase for single exposure after 4 days of incubation following the last application of GR to soil, which is consistent with the studies of some studies (Ren et al. 2015). Due to the high surface-to-volume and low homeostatic capacity of microorganisms, their activity decreased a little under high concentration of GR exposure for a short time(Shrestha et al. 2013). However, up to 21days, the enzyme activities of these two enzymes for repeated exposures were increased by approximately 20%, and the enzyme activities under the two exposures increased more than under three exposures. This is probably because that the acute and high concentration of GR exposure repressed the growth of some soil microorganisms which produced these enzymes, and the chronic and low concentration of GR proliferated to a much greater extent than the bacteria that were inhibited. As the increase of exposure time, the gradual growth of the relevant microorganisms narrowed the gap between the single and repeated exposures after exposure for a longer time. A possible explanation is that the soil microbes start to tolerate and adapt the inhibitor after some time which is consistent with the environ-mental change by transformation of metabolism (Simonin et al. 2015). In the course of the whole experiment, the change of urease activities under GR stress for different exposure time were greater than that of the FDAE activities. Urease is a representative extracellular enzyme and the key source of bacterial nitrogen, appears to be more sensitive to pollution than FDAE activities which represent overall microbiological activity of soil. And with the extending of incubation time, GR will likely increase soil nitrogen cycling, and urease activities significantly increased in the soil community.
3.2 Effect of GR on soil bacterial community alpha diversity
Based on high-throughput sequencing, the alpha diversity (Shannon index) of different treated soil were displayed in Fig. 2. The alpha diversity at the beginning of exposure to GR through different methods was almost identical because the bacterial community was from the same soil sample. Although the final concentration of GR was the same, alpha diversity followed an order of GR31>GR21>GR11 after 4 days incubation. This result demonstrated that high concentration of GR exposure for the short term decreased the diversity within a specific bacterial community, and in the low concentration range, soil microbial diversity increased, so the diversity under repeated exposures increased more than under single exposure after 4 days incubation. When the incubation time reached 21 days, the alpha diversity of repeated exposures soils reached a maximum, which indicated that the repeated exposures exhibited pronounced increased after a relatively longer time. When the incubation time was prolonged to 60 days, the alpha diversity decreased. This is due to that the relatively tolerant and resistant species became more abundant and occupied the niche of the sensitive species that were eliminated, leading to a lesser diverse (Song et al. 2018). These results are consistently with previous work that the soil bacterial alpha diversity under exposure to GR gradual decrease when the exposure time was longer than 60 days, regardless of the GR concentration (Wu et al. 2020). Overall, the results suggested that with the extension of exposure time the repeated exposures have more positive effect on the soil alpha diversity.
3.5 Effect of GR on soil bacterial community Beta diversity
In order to compare similarities of soil bacterial community compositions under different exposure method, the Bray-Curtis distance of soil samples were compared relative to control. From Fig. 3, the repeated exposures of GR played a stronger effect on soil bacterial communities as compared to the single exposures after 4 days incubation, while this effect became weak gradually with the extension of incubation time. The principal co-ordinates analysis (PCoA) based on Unweighted-UniFrac distance which incorporates phylogenetic distance into relative abundance measurement was also conducted to study the difference of soil bacterial community. The PCoA results showed that, the soil bacterial community after exposure to GR in a short-term time were totally separated from the control, especially for the soil under single exposure. And the soil bacterial communities changed constantly with the growth of culture time, and the distinction between the single exposure and repeated exposures got smaller. In addition, after a longer period, the sample from the control group, the repeated GR exposure were clustered with single GR exposure, suggesting similar microbial community structures. Furthermore, after 60 days incubation, the soil bacterial communities of all the soil samples were totally different form that of the initial samples. And adonis analysis on the bacterial community further confirmed that the difference in soil bacterial communities under different incubation time to GR was significant (R2=0.573, p=0.001).
Different exposure model induced certain shifts in the soil bacterial community composition, some core taxa still remained under all treatments. Via high-throughput sequencing, the Venn diagrams showed that a total of 2334 OTUs, 2965OTUs and 2910 OTUs were detected under the different exposure time. The 10.92% 6.35%,11.7% and 12.68% of all OTU were specific to control, 1 exposure, 2 exposures and 3 three exposures after 4 days incubation. However, as the incubation time lengthened to 60 days, these data were 9.68% 8.05%, 9.43% and 9.87%, respectively. These results indicated that the in the short term, adding a certain amount of graphene to a soil at a time may inhibit the growth of some unique species. Together, both beta diversity and core taxa analysis suggest that, yet in the long run no matter what exposure scenario is, and GR could promote soil bacteria to a certain extent.
3.5 Effect of GR on soil bacterial populations at phylum level
To further identify the changes in the bacterial community of soil exposure to GR through different adding method, the sample level clustering tree and the relative abundance(RA)of phylotypes was summarized at the phylum level based on the pyrosequencing data. According to the Figure 4A, the soil bacterial communities of single exposure were obviously different from that of repeated exposures when the incubation time was lower than 21 days, and the difference grow small with the extending of incubation time. And the soil bacterial community of all the soil including control changed when the incubation time reached 60 days according to the sample level clustering tree. Figure 4B revealed that there was high bacterial diversity in all treated soil and control samples, and 85% of the total bacterial counts could be represented by eight predominant bacterial phyla among which Proteobacteria, Acidobacteria, Planctomycetes, Actinobacteria, and Bacteroidetes are the dominant phyla in agricultural soils(Buckley and Schmidt 2003). The Proteobacteria is the most predominant phylum in the soil bacterial community (33.62%-52.01%), which consists of a majority of Gram-negative bacteria(Mai-Prochnow et al. 2016). And the RA of Proteobacteria in soil under repeated exposure increased more than that under single exposure after 4 days incubation, the gap had been narrowed with the extension of time. Normally, the Gram-negative bacteria have a higher tolerance to external stress than Gram-positive bacteria because of their outer membrane and cell wall(Premanathan et al. 2011). And the phylum Proteobacteria could participate in degradation of carbonaceous compounds(Cebron et al. 2008; Spain et al. 2009). So, an increase in RA of Proteobacteria under multiple exposure to GR reflected its high resistance and high carbon mineralization rate in soil. It is obvious that Proteobacteria and Acidobacteria were the primary phylum under a short period (4 and 21 days). The acidobacteria and actinobacteria were slightly reduced under GR exposure compared to control, and the rest of phylum did not change significantly under any of the GR exposure way, except for these three phyla. Intriguingly, with the extension of time, up to 60 days, for all the soil sample, not only the Proteobacteria increased, but also the actinobacteria increased for all the soil samples. The phylum Actinobacteria has the potential to degrade lignocellulose; some tentative cellulose degrading enzymes were found in Actinobacteria genomes( Tetrovsky et al. 2014). So, the lignocellulose degradation rate for all soil increased because of the potential to degrade lignocellulose of phylum Actinobacteria. Based on sample level clustering tree and the relative abundance༈RA༉of phylotypes, it could be inferred that graphene changed the soil bacterial community to a certain extent and the changes became weaker or disappeared with extending of exposure time no matter what exposure scenarios were.