The Sichuan Basin is one of the four main basins in China (28°10′-32°25′N and 101°56′-108°32′E) (Ran and Xinyue, 2018). The Sichuan Basin covers an area of 2.6×10⁵ km², with an average elevation of 400 m, which is home to around 90 million people. Most of the terrain is composed of plains and hills, with an average annual temperature between 16 and 18 degrees Celsius. This region belongs to the subtropical monsoon climate zone. Most crops planted on arable land are rice, rapeseed, maize, and sweet potatoes. There are several types of arable soil, such as dry field, paddy field (http://ir.imde.ac.cn/: records of Sichuan soil varieties) and according to the second national soil census of China, 77.75% of the soil types in the Sichuan basin are mainly paddy soil and purple soil. We collected soil samples in May and August 2021, respectively, all the sample sites were recored by GPS ( Figure 1 ). 54 samples (0-2 cm) were collected in sterile bags and stored in dry ice boxes, transported to laboratory, and stored in a -80°C freezer. Soil samples at 0-20 cm depth were collected for determination of soil physical and chemical characteristics.

pH was determined by the potentiometric method, with specific steps as follows:weigh 10 g air-dried soil sample, add 25 mL of deionized water (ddH₂O), stir for one minute, and measure after 30 minutes using a pH meter. The Kjeldahl method, the sodium hydroxide melting molybdenum antimony colorimetric method, and the external-heat potassium dichromate oxidation-colorimetric method, with specific operations referring to HJ717-2014, GB8937-88, and GB9834-1988, were utilized to determine the total nitrogen (TN), total phosphorus (TP), and organic matter (OM) of soil samples. Soil IBM was determined by using an X-ray diffractometer (XRD). A sufficient amount of a sieved 200-mesh sample was used to form an oriented sheet. Then adjust the Cu target to 2.2 kW, working voltage to 60 kV, current to 50 mA, scanning range to 5°–80°, and scanning speed to 5°/min (Harris and Norman White, 2008). IBM’s logo appears in the program Jade 9.0. The semi-quantitative XRD “adiabatic method” calculation formula for IBM’s mass fraction is as follows: (Chung, 1974):

(W_x: Mineral phase mass fraction, I: Mineral peak intensity, K: Mineral phase RIR value, ∑ i = A N I i K A i : There are N sums of phases I/K).

DNA from soil samples was extracted using a soil DNA kit (BIO101, MP Biomedicals, US). The quantity and molecular size of the extracted DNA were detected by the Nanodrop NC-2000 UV spectrophotometer (Thermo Scientific, USA) and 1.2% agarose gel electrophoresis before further Analysis. The V1-V3 regions of bacterial 16S rRNA gene for PCR amplification. The upper and lower primers are 27F (5^’-AGAGTTTGATCCTGGCTCAG-3^’) and 533R (5^’-TTACCGCGGCTGCTGGCAC-3^’) (Franke-Whittle et al., 2015). The 20 μL reaction mixture includes 0.25 μL of Q5 high-fidelity DNA polymerase, 5 μL of Q5 reaction buffer, 5 μL of Q5 high GC buffer, 10 μL of dNTP (10mM), 2 μL of DNA template, 1 μL of forward and reverse primers (10mM), and 8.75 μL of ddH₂O. The thermal cycling reaction is as follow: After preparing the necessary components for the PCR reaction, the PCR machine performs a 30-second Pre-denaturation at 98°C to thoroughly denature the template DNA before beginning the amplification cycle. In each cycle, the template was denatured by holding it at 98°C for 15 seconds. At 50°C for 30 s, the primer was completely annealed to the template; at 72°C for30 s, the primer was extended to synthesize DNA, thus completing a cycle. This cycle is repeated 25–27 times, resulting in a substantial collection of DNA fragments that have been amplified. The reaction mixture was maintain it at 72°C for 5 minutes to finish the product’s extension. The products were stored at 4°C. The amplification products were electrophoresed on a 2% agarose gel. The fragments of the target were excised and then recovered using an Axygen gel recovery kit. PCR results were then measured with the Quant-iT PicoGreen dsDNA Assay Kit on a microplate reader (BioTek, FLx800). Shanghai Pesennuo Technology Co., Ltd. (Shanghai, China) was tasked with performing two 250 bp paired-end sequencing using the Illumina NovasSeq-PE300 platform, after individual quantification and pooling of equal numbers of amplicons.

Vsearch (v2.13.4 Linux x86 64) was used for microbiome bioinformatics analysis (Edgar et al., 2011). Firstly, the original sequence data were preprocessed by using Cutadapt to remove primer fragments and eliminate sequences that do not match the primers. Then, the fastq_ series module is used to perform concatenation, quality filtering, de-duplication, de-chimerism, and clustering on the sequence, resulting in singletons OTUs and their representative sequences. Based on the Geobacter abundance data and soil physicochemical properties of the samples in the Perl script, redundancy analysis (RDA) was performed using Canoco 5.1 software. We identify the factors that have the greatest influence on Geobacter based on the RDA result. We sort soil samples according to the most significant factors (different arable soils, different times). We utilize the R programming language to calculate the alpha-diversity of microorganisms, such as Chao1, Faith-PD, Observed-species, and Shannon index. The Derived Gene Cloud Platform (https://www.genescloud.cn/home) is used to integrate data for species random forest analysis, MetagenomeSeq analysis, and species composition analysis. Using the GS+9.0 tool, a Semi-Variance Analysis should be conducted to find the semi-variance model and parameters for the most important Geobacter component. ArcGIS 10.8 was used to draw a soil pH transformation map of study area, SPSS 25.0 was used to do other routine statistical analyses.

Soil pH of arable soil in this study ranges from 4.1 to 8.2. They were divided into eight sections (<5.0, 5.0–5.5, 5.5-6.0, 6.0–6.5, 6.5-7.0, 7.0–7.5, 7.5-8.0, >8.0), with ratios of 11.1%, 5.6%, 5.5%, 9.3%, 12.9%, 20.4%, 24.1%, and 11.1%, respectively. TN content in soil was ranges from 0.09 to 0.33%. SOM content varies from 1.39 to 4.13%. TP content ranges from 0.02 to 0.13%. According to the results of XRD analysis, the iron minerals in paddy soil are mainly Hematite, Goethite, Ferrihydrite, Maghemite and Magnetite. These minerals correspond to the standard PDF cards in Jade software, and they correspond to the cards as follows: PDF#97-015-7689, PDF#97-007-7327, PDF#97-024-9048, PDF#97-009-6073, and PDF#97-016-6135.IBM content ranges from 2.13 to 16.53%. Geobacter abundances range from 0.11 to 1.17%. Through RDA analysis, it was found that TN, SOM, IBM, and C/N showed positive correlations with Geobacter and soil pH and TP showed negative relationship with Geobacter ( Figure 2 ). The relative length of the vertical projection between pH and Geobacter is relatively long, indicating that pH is a key factor affecting the abundance of Geobacter. According to the ranking of the influence of Geobacter, the importance of physicochemical factors in paddy soil is as follows: pH>TN>TP>OM>IBM>C/N. The proportions of pH and TN to all soil physicochemical factors are 72.5% and 16.9%, respectively, and pH and TN reach a very significant level (p<0.01), while TN also reaches a significant level (p<0.05) ( Table 1 ).

Overall, the correlation between IBM and Geobacter has no significance (p>0.05). Therefore, based on the IBM content, the samples were divided into five groups: IBM1 (5%), IBM2 (7.5%), IBM3 (9%), IBM4 (10.5%), and IBM5 (12%). We found that the greater the difference in IBM content in soil, the greater the difference in microbial abundance (p>0.05) ( Table 2 ). From the results of hierarchical clustering analysis, IBM1 and IBM2 groups which have lower IBM content were relatively close in terms of species evolutionary composition, while the IBM3, IBM4 and IBM5 groups which have a higher IBM content exhibit an interlocking state ( Figure 3 ). For the abundance of Geobacter, the lower content of IBM1 and IBM2 groups showed a negative correlation with Geobacter abundance, while the higher content of IBM3, IBM4 and IBM5 groups showed a significant positive correlation with Geobacter abundance ( Figure 4 ). These data implied that arable soil with higher IBM content has a significant impact on the relative abundance of Geobacter, especially when the IBM content exceeds 9%, the effect of IBM content on Geobacter abundance reaches a significant level.

Soil samples were divided into different categories by soil types (purple soil and paddy soil), soil pH (<6.5:acidic, 6.5-7.5:neutral, and >7.5:alkaline soils) and sample time (May and August). It was found that microbial diversity in arable soil increased at the genus level (p<0.01) with increasing pH. Soil microbial diversity shows significant differences in different sample months and soil types. The diversity of soil microorganisms is significantly higher in August than in May, and the diversity of microorganisms in purple soil is significantly lower than that in paddy soil ( Figure 5 ).

The results of random forest analysis indicate that Geobacter ranked 30th in importance of our sampled soil ( Figure 6 ). Paddy soil sampled in May revealed a positive link with Geobacter in acidic environment, a weak correlation with a neutral environment, and a negative correlation with an alkaline environment. Paddy soils which were collected in August demonstrated a negative correlation in acidic, neutral, and alkaline environments. Geobacter in the purple soil collected in May and August demonstrated a weaker correlation in the neutral environment, which is negatively correlated in acidic and alkaline environments. These results indicate that Geobacter is more adaptable to acidic environmental in paddy soil, but in purple soil, Geobacter is more adaptable to neutral environment. Geostatistical analysis revealed that the pH of soil in the Sichuan Basin has decreased by an average of 0.7 over the past 40 years ( Figure 7 ). Soil acidification phenomenon was serious especially in soil that long term affected by drainage and irrigation.

Abundance of microorganisms was greater in August than in May whether in paddy soil or purple soil. Abundance of Geobacter is also higher in August than in May. In paddy soil, the abundance of Geobacter reaches 0.83% and decreased with increasing pH. In purple soil, the abundance of Geobacter first increases and then decreases with the increase of pH, and Geobacter prefer neutral purple soil, its maximum abundance reaches 0.68% ( Figure 8 ). Further analysis through metagenomeSeq revealed that there was a significant upregulation of Geobacter in paddy soil in August compared to May, and its abundance ranked among the top 10 at the genus level ( Figure 9 ). However, there was no significant difference in Geobacter in purple soil. Through the analysis of the content changes of non-stable crystal Ferrihydrite in different soil types and months, it was found that the content of ferrihydrite in paddy soil in August was lower than that in May, while this trend was not observed in purple soil ( Table 3 ).

pH had a negative correlation with Geobacter abundance and its contribution reached 72.5% (p<0.01). This result mutually agree with another result which demonstrated that Geobacter abundance declined from 2.89 percent to 0.9 percent in paddy soils as pH increased (Li et al., 2019). Geobacter can utilize a variety of substrates for respiratory metabolism that the best growth pH range for Geobacter is between 5.5 and 7.0, which corresponds to a neutral and acidic environment (Straub and Buchholz-Cleven, 2001; Wang et al., 2021; Xu et al., 2019).

IBM is a receptor for Geobacter to directly transfer electrons through conductive pili. When IBM is Added under ideal cultivation conditions, an increase in the abundance of Geobacter can be observed, which is mainly caused by Magnetite that is the main content of IBM (Kato et al., 2010; Zhang et al., 2020). Hematite, Goethite, and Maghemite have lower transfer efficiency than direct conductors when accepting electrons and transferring them to end acceptors. At the same time, these iron containing semiconductor minerals accept photoelectrons catalyzed by sunlight under natural conditions. They accept electrons produced by Geobacter while reacting with Geobacter (Lu et al., 2012). When the content of IBM is low, it often cannot reach the level that affects Geobacter abundance (Qiu et al., 2019). This may be the reason why IBM content did not have a widespread impact on the abundance of Geobacter in the topsoil. Therefore, we divided the IBM content in our study and found that when the IBM content exceeded 9%, it had a positive correlation with the abundance of Geobacter. Due to the ubiquitous presence of IBM in the topsoil, its involvement in the study of Geobacter mediated energy cycling of other substances cannot be ignored.

Under the dual effects of human activities and natural soil formation, paddy soils, especially winter paddy fields in the Sichuan Basin, have long been in a state of flooding and hypoxia (Xu et al., 2020). Due to differences in the soil formation status and utilization methods of the plow layer, even with the same tillage measures, differences in soil microbial abundance can still be observed (Yan et al., 2019). The abundance of Geobacter in paddy soil under long-term flooding conditions is significantly higher than that in purple soil. The abundance of Geobacter in acidic paddy soil and neutral purple soil is higher than that in cultivated soil during the same period, which may be related to the acidification of paddy soil. In the past 40 years, the average pH decrease in cultivated soil in the Sichuan Basin has reached 0.7 units, which has led to Geobacter tending towards neutral and acidic environments. There is research confirming that small changes in water balance can lead to a steep transition from alkaline to acidic soil. During the continuous drainage and irrigation process, alkaline cations and Si in the soil are completely lost, while Al³⁺ loss is relatively small, resulting in a decrease in soil pH. Finally, Geobacter in paddy soil tends to adapt more to acidic environments (Chadwick et al., 2003; Slessarev et al., 2016). As a primary fertile purple soil, although short-term irrigation is carried out during the rice planting period, the linear relationship between soil leaching coefficient and soil renewal always maintains the relative stability of soil pH in the plow layer (Li et al., 2010; Wang and Zhu, 2011).

Overall, the abundance of Geobacter in paddy soil and purple soil showed a higher occurrence in August than in May, which can be attributed to the fact that the temperature in August was higher than that in May in the Sichuan Basin. This is because the optimal growth temperature for Geobacter is 30-35 °C, which is conducive to improving its electron transfer efficiency (Bannister et al., 2017; Heidrich et al., 2018; Xiao et al., 2021). On the other hand, in May, rice entered the greening period, and the flooded environment caused by a large amount of irrigation led to an upward trend in soil pH. In August, rice entered the maturity period, and the drainage process caused a linear decrease in soil pH (Ding et al., 2019). In addition, after the drainage of paddy soil, the soil redox potential continues to increase, and the presence of the differentiated reduction product Fe²⁺ is very conducive to the transformation of oxides from amorphous to crystalline form (Minamikawa and Sakai, 2006; Zachara et al., 2002). Some studies has found that Fe²⁺oxidation can lead to a decrease in soil pH (Hall and Silver, 2013). At the same time, the IBM content will increase because under stable oxygen-containing conditions, Fe²⁺ rapidly re oxidizes to form Ferrihydrate, which then slowly transforms into thermodynamically more stable hematite or goethite (Vogelsang et al., 2016). Overall, these factors better explain the significantly higher abundance of Geobacter in paddy soil in August compared to May.