Community structure, distribution pattern, and influencing factors of soil Archaea in the construction area of a large-scale photovoltaic power station

The photovoltaic power station in Qinghai has been built for 8 years; however, its impact on the regional soil ecological environment has not been studied in depth. To reveal the structure and distribution pattern of archaeal communities in desert soil under the influence of a large photovoltaic power station, a comparative study was carried out between the soil affected by photovoltaic panels and the bare land samples outside the photovoltaic station in Gonghe, Qinghai Province. The abundance, community structure, diversity, and distribution characteristics of archaea were analyzed by quantitative PCR and Illumina-MiSeq high-throughput sequencing, and the main environmental factors affecting the variation in soil archaeal community were identified by RDA. The contribution rate of environmental factors and human factors to microbial community diversity was quantitatively evaluated by VPA. The results showed that there was no significant difference in soil nutrients and other physicochemical factors between the photovoltaic power station and bare land. Thaumarchaeota was the dominant archaeal phylum in the area, accounting for more than 99% of archaeal phylum, while at the level of genus, Nitrososphaera was the dominant archaeal genera. There was no significant difference in archaeal community structure between and under different types of PV panels. The analysis has shown that the construction of a photovoltaic station has little effect on the community structure of soil archaea in a desert area, and it was speculated that the selection of niche played a leading role in the distribution pattern of soil archaeal community. This study provides the basis for a scientific understanding of the characteristics and distribution patterns of soil archaeal communities affected by the construction of a photovoltaic power station.


Introduction
Energy production is considered a major challenge in industrial development (Pires et al. 2012) especially in the context of exponential growth of the global population. Considering the impoverishment and depletion rate of conventional fossil energy sources, several environmentally friendly and efficient renewable energy sources have been investigated recently (Tawalbeh et al. 2021). Among the many renewable energy technologies, solar photovoltaic (PV) technology has become an increasingly important energy supply option worldwide owing to its significant advantages in mitigating climate change, reducing greenhouse gas emissions, and increasing the security of energy supply (Ebhota and Jen 2020;Nižetić et al. 2018;Turney and Fthenakis 2011;Wilberforce et al. 2019). However, based on the product life cycle theory, the development of any industry and technology is a "double-edged sword." Solar PVs are no exception, and their positive and negative feedback effects on the ecological environment are gradually emerging (Tawalbeh et al. 2021;Turney and Fthenakis 2011).
Archaea are one of the three domains of life, whose survival was once thought to be restricted to extreme environments. However, the archaeal tree has expanded rapidly with the discovery of new archaeal lineages at different taxonomic levels (Auguet et al. 2010;Bates et al. 2011;Huang et al. 2019;Shi et al. 2016). An increasing number of studies in different environmental systems have shown that archaea are widely distributed in a variety of ecosystems (Anish et al. 2015;Auguet et al. 2012;Bates et al. 2011;Brochier-Armanet et al. 2008;Hu et al. 2016;Jiang et al. 2007) and exhibit significant environmental heterogeneity. Meanwhile, archaea play an important role in mediating biogeochemical processes such as ammonia oxidation (Dai et al. 2018), sulfate reduction (Sorokin et al. 2017), organic matter decomposition (Meng et al. 2019), and greenhouse gas regulation . Despite the significant impact of archaea on the ecological restoration of soil function, previous studies have focused on bacteria and fungi, and studies on archaeal community composition and function have greatly broadened owing to extensions cultivationindependent techniques, including metagenomics and singlecell genomics techniques (Ayangbenro and Babalola 2021;Pfeifer et al. 2021;Rinke et al. 2013;Zou et al. 2020). In addition, various environmental factors including pH, salinity, soil moisture, heavy metals, nutrient levels, vegetation type, and local climatic characteristics Meng et al. 2019) have significant effects on the composition of soil archaeal communities. In contrast, the variation in soil archaeal communities caused by anthropogenic disturbances such as soil reclamation, improvement, restoration, and engineering construction has received substantially less attention (Armbruster and Strand 2020;Yang et al. 2019). Limited studies have suggested that anthropogenic disturbances might be an important factor influencing the composition (Shi et al. 2016;Wang et al. 2011) and taxon abundance of archaeal communities.
As a typical representative of the arid desert region in Northwest China, Talatan, where the Qinghai Gonghe Photovoltaic Industrial Park is located, is an ideal area for the development of solar PV power generation because of its poor soil physical structure, low soil nutrient and organic matter content, high salinity, aridity and water scarcity, extreme temperatures and dryness, strong winds, and strong UV radiation. Soil archaea is one of the most abundant organisms in the soil. However, once microbial development and activities are disturbed by engineering development in such a nutrient-limited environment, which will affect the biogeochemical cycle of nutrients (Hu et al. 2019;Shi et al. 2016;Wang et al. 2015;Wei et al. 2020). This makes the already fragile ecosystem more sensitive to interference, the rate of ecological recovery after human interference is low, and vulnerable to climate change. The increasing number of crystalline PV panels in the Qinghai Gonghe Photovoltaic Industrial Park has impacted the ecological environment of the region. Current studies mainly elucidated the ecological and environmental effects of climate, vegetation, and soil from a macroscopic perspective Li et al. 2019Li et al. , 2016, and no research on the microscopic level of soil archaeal communities has been reported.
Based on the above-mentioned background, this study used quantitative PCR and 16S rRNA high-throughput sequencing techniques to analyze the abundance and diversity of soil archaea in PV power plants. In addition, the effects of environmental and anthropogenic factors on the distribution patterns of soil archaeal communities were elucidated. This is the first exploration to investigate the soil archaeal communities under the influence of PV plant construction. This work provides some reference for the comprehensive assessment of the ecological and environmental effects of PV parks in the future.

Study area and sample collection
T h e G o n g h e P h o t ovo l t a i c I n d u s t r i a l P a r k (98°54′E-101°22′E, 35°46′N-37°10′N) is in Talatan, Gonghe County, Qinghai Province ( Fig. 1), with an altitude of 2900-3100 m. Gonghe County has a plateau subtropical climate, with average annual evaporation and precipitation of 1800 mm and 303 mm, respectively, the sunlight of 2772 h, the temperature of 4.1 ºC (extreme low temperature of -26.9 ºC), and solar radiation of 6564.26 MJ/m 2 . The number of days in which the daily minimum temperature is less than 0 ℃ can reach 200 days. The soil types in the study area are mainly grassland soil, aeolian sandy soil, and saline alkali soil. The soil layer is thin, but the soil permeability and permeability are good. The performance of soil water retention and fertility is poor, and the content of organic matter is low. At present, the park is divided into new energy generation (PV power park and wind power park) and industrial processing and manufacturing areas. The total area is 3,028.5 km 2 , including a 609.6 km 2 PV power park, 2,400 km 2 wind power park, and 18.9 km 2 industrial processing and manufacturing area.
In mid-August 2020, a total of 27 representative soil samples were collected from the PV park, as shown in.
Table. S1. GPS (UG802, Tibbo, China) was used to locate the sites throughout the sampling process, and the latitude, longitude, and elevation information of the sites were recorded. The sampling points were divided into three groups: flat single axis (FSA, sampling number 1, 10), oblique single axis (OSA, sampling number 2, 3, 9), and fixed axis (FIX,sampling number 6,11,12,17,18,19,21) according to the orientation of the PV tracking brackets. For each group, samples were collected at the sampling points under the PV stand panels (SX) and between the panels (SJ). To distinguish between the characteristics of soil archaea at sample sites with and without PV stand panels, three control sampling sites (CK) were set up on the bare ground outside the PV power plant. Five independent surface soil samples (sampling depth: 0-10 cm) were collected separately from each sampling site (2 m × 2 m sample square) and mixed into one composite sample. Samples for characterizing the physical and chemical properties of the soil were packed in sterile polyethylene bags, transported to the laboratory, air-dried, sieved through a 1-mm screen to remove debris and plant roots, and stored. Samples for microbiological analysis were packed in 50-mL sterile polyethylene conical centrifuge tubes, stored in liquid nitrogen, rapidly transported back to the laboratory, and stored at -80 ºC in a freezer.

Soil physicochemical parameters and vegetation diversity analysis
Soil samples were dried at 65 ºC and analyzed for chemical properties according to standard methods. Soil water content was determined using the weight method, soil pH, and conductivity (FE30, Mettler-Toledo, Switzerland) at a soil-to-water ratio of 1:2.5. Soil organic matter (SOM) and soil organic carbon (SOC) were determined by the dichromate oxidation method (Nelson et al. 1982). Total nitrogen (TN) was determined by the dry combustion method in an elemental analyzer (Elementar Vario Macro cube, Hanau, Germany). The soil carbon-to-nitrogen ratio was calculated from OC and TN. Alkaline nitrogen was determined using the right show the sampling sites under the PV panels, between PV panels and outside the PV park from top to bottom a Kjeldahl distillation system (C. Gerhardt, Germany). The soil total phosphorus (TP) and total potassium content was determined using the perchloric acid digestion method Nelson and Sommers (1973). Soil alkaline nitrogen, fastacting potassium, and fast-acting phosphorus content was determined by visible-NIR spectroscopy ). The vegetation survey covered the entire study area, and the plant community species diversity index was calculated (unpublished data).

High-throughput sequencing and fluorescent quantitative polymerase chain reaction
Total DNA was extracted from soil archaea using a Fast DNA Spin Kit (MP Biomedicals, USA) and stored at -80 ºC. The quality of extracted DNA was determined by 1.0% agarose gel electrophoresis. Polymerase chain reaction (PCR) amplification was performed on the V4-V5 region of the archaeal 16S r RNA gene, and the primers used for PCR amplification were Arch524F (5′-TGY CAG CCG CCG CGG TAA -3′) and Arch958R (5′-YCC GGC GTTGAVTCC AAT T-3′) (Pires et al. 2012). The PCR amplification system consisted of a 2.5 μL 10 × PCR buffer, 3 mM MgCl 2 , 160 μM dNTPs, 100 pmol of upstream primer, 100 pmol of downstream primer, 0.625 U Taq polymerase, and 2 μL DNA template (1:20 dilution of total DNA). The PCR amplification reaction protocol consisted of initial denaturation at 95 °C for 5 min, 35 cycles of denaturation at 95 °C for 1 min, annealing at 56 °C for 30 s, and extension at 72 °C for 1.5 min, followed by final extension at 72 °C for 10 min. The PCR products were separated by electrophoresis on 2% agarose gels, and gel bands containing target fragments were cut and recovered with a gel extraction kit (AXYGEN, USA). The PCR products were quantified using a fluorescence microplate reader (BioTek, FLx800, USA) with the Quant-iT PicoGreen dsDNA Assay Kit (ThermoFisher, USA). Based on the quantification results, samples were pooled in the appropriate ratio according to the required read depth. Sequencing libraries were prepared using an Illumina TruSeq Nano DNA LT Library Prep Kit. After the samples passed quality check, paired-end sequencing with readlength of 250 bps was performed on an Illumina sequencer by Shanghai Personal Biotechnology Co., Ltd.
The archaeal 16S rRNA gene was quantified using an AceQ qPCR SYBR Green Master Mix on a CFX96 Touch model real-time fluorescent quantitative PCR instrument (Bio-Rad, USA). Amplification primers were consistent with high-throughput sequencing primers. The total reaction volume was 15 μL, consisting of 12.5 μL of AceQ qPCR SYBR Green Master Mix (2X), 0.7 μL of upstream and downstream primers, 1 μL of DNA template, and 15 μL of ddH 2 O. The reaction protocol consisted of initial denaturation at 95 °C for 5 min, followed by 40 cycles of denaturation at 95 °C for 10 s, and annealing and extension at 60 °C for 60 s.
All sequence data from Illumina MiSeq sequencing in this study have been submitted to the public NCBI database (http:// www. ncbi. nlm. nih. gov/) under the accession number PRJNA690680.

Bioinformatics data processing and analysis
The raw data from Illumina high-throughput sequencing were first demultiplexed using the barcodes of all libraries. Quality control filtering was performed based on read quality. Paired reads were then merged into one sequence based on the overlap of the reads. Finally, high-quality sequences for each sample were obtained by demultiplexing with barcode and primer sequences, and sequence orientation was corrected according to the forward and reverse barcodes and primer orientation. Chimeras were removed using a combination of de novo and reference-based methods (Usearch software and Gold database). QIIME (version 1.8.0, http:// drive5. com/ uparse/) was used. Operational Taxonomic Units (OTUs) were grouped and sorted according to a 97% similarity level. OTUs with abundance < 0.01% of the total sequencing depth were removed. The sequence with the highest abundance in each OTU was selected as the OTU representative sequence. Each OTU representative sequence was then compared with the Greengenes database (Luvizott et al. 2019) to obtain the annotation proportion and relative abundance of species of each taxonomic rank of OTUs. The RDP classifier Bayesian algorithm was used to taxonomically analyze the OTU representative sequences at 97% similarity level, and QIIME software (version 1.8.0) was used to obtain the abundance distribution table for each sample at five taxonomic levels including Phylum, Class, Order, Family, and Genus.
Alpha diversity was calculated using the vegan package and OTU table package in R software (version 3.5.1), and the Kruskal-Wallis test was used to compare the differences in diversity between different subgroups. Non-metric multidimensional scaling analysis (NMDS), analysis of similarity (ANOSIM), and multivariate analysis of variance (ADONIS) were performed using different functions in the vegan package to test the significance of differences in the structure of archaeal communities in soils. Hierarchical cluster analysis (beta diversity) based on Bray-Curtis distance was performed on soil samples. To estimate the relative importance of environmental and anthropogenic factors in the formation of community composition, variance decomposition analysis (VPA) was performed using (partially) constrained principal coordinate analysis (CAP) using the varPart function in the vegan package in R software based on Bray-Curtis dissimilarity (Rui et al. 2015). Spearman's rank correlation analysis was performed using SPSS 22.0 to test the relationship between archaeal abundance and diversity and different environmental variables. In addition, redundancy analysis (RDA) was performed using CANOCO 5.0 to reveal the correlation between archaeal community composition and environmental variables.

Spatial variation characteristics of soil physicochemical factors
Soil physicochemical characteristics were analyzed for three types of sample plots and three different PV stand sampling sites, respectively (Fig. 2). The results of sampling analysis were analyzed by one-way ANOVA, and the significant differences between samples were tested by Student-Newman-Keuls (SNK). The soils in the study area were semiarid alkaline (pH = 8.41-9.60) with water content ranging from 4.5% to 26.25%. Soil pH and water content were both higher inside the PV stand sites than outside but did not show significant differences. Soil temperature was significantly lower in the field than in the bare ground outside the field during the sampling period due to shading by PV panels. TN and TC in the soil samples did not show statistically significant differences (P > 0.05), were highest in the samples under the PV stand panels, and did not show a significant pattern between different PV tracking stands. Fig. 2 Physicochemical parameters of the soil samples in photovoltaic industrial park of Gonghe, Qinghai. FSA, flat single-axis; OSA, oblique single-axis; FIX, fixed trackers; Temp., water temperature; EC, electrical conductivity; TN, total nitrogen; TC, total carbon; TP, total phosphorus; TK, total potassium; AN, alkaline-hydrolysable nitrogen; FP, fast-acting phosphorus; FK, fast-acting potassium; SOC, soil organic carbon; SOM, soil organic matter. Different letters after the columns indicate significant differences (P < 0.05) Soil OM and OC content was the highest in the inter-panels sampling sites and the lowest in the off-site sampling sites; this was also the highest in the flat uniaxial sampling sites and the lowest in the oblique uniaxial sampling sites, with no significant differences among sampling sites (P > 0.05). TP and potassium content did not differ significantly (P > 0.05) among sampling sites. Soil alkaline nitrogen and fast-acting phosphorus did not differ significantly (P > 0.05) among sampling sites, and fast-acting potassium was the highest under the flat single-axis PV panels and the lowest on the bare ground outside the field.

Abundance of soil archaeal
A total of 1,722,647 valid reads were obtained by highthroughput sequencing of different soil archaea 16S rRNAs in the PV park, and the number of valid high-quality reads of archaea ranged from 50,730 to 91,564 (Table S2). The corresponding library coverage rate ranged from 80.9% to 89.8%, indicating that the libraries of samples in this study contained most archaea taxa and basically reflected the structural composition of the endophytic Archaea community in the samples. As shown in Fig. 3, the number of archaeal 16S rRNA genes per gram of soil (dry weight) under the PV tracker panels ranged from 2.30 × 10 9 to 5.01 × 10 10 copies, with a mean value of 2.23 × 10 10 copies. The number of archaeal 16S rRNA genes per gram of soil (dry weight) between the panels ranged from 3.84 × 10 9 to 6.37 × 10 10 copies, with a mean value of 2.23 × 10 10 copies. In the bare ground outside the PV park, the number of archaeal 16S rRNA genes per gram of soil (dry weight) ranged from 2.25 × 10 10 to 8.37 × 10 10 copies, with a mean value of 4.39 × 10 10 copies. In the PV park area, the number of archaeal 16S rRNA genes per gram of soil (dry weight) ranged from 2.25 × 10 10 to 8.37 × 10 10 copies, with a mean value of 4.39 × 10 10 copies. The number of soil archaea was higher under the panels than between the panels except sampling points S1, S2, S6, S10, S17, and S21 in the PV park, where the number of soil archaea between the panels was higher. The number of archaea was higher between the plates than under the plates at all the sampling points of flat singleaxis tracking bracket PV panels (S1 and S10). The number of archaea between the plates and under the plates did not have clear pattern at the sampling points of oblique singleaxis tracking bracket PV panels (S2, S3, and S9) and fixed tracking bracket sampling points (S6, S11, S12, S17, S18, S19, and S21). The number of archaea per gram of soil (dry weight) was not spatially significantly different (P > 0.05) among the sampling sites within the PV plant.

Soil archaea alpha diversity
The alpha diversity (Chao1, observed species, Shannon, Simpson, Faith's PD, Pielou's evenness, and Good's coverage index) of soil archaea was not significantly different (P > 0.05) between the under-panels (INX), inter-panels  Fig. 4. The Chao1 index and observed species index indicated that soil microbial diversity and richness were slightly higher in the INX and INJ samples than in the OUT samples but not significantly different (P > 0.05). The overall trend of the Shannon and Simpson index was inconsistent among the 27 samples. The Shannon index showed slightly lower species diversity and evenness at the INJ) and INX sites than at the OUT sites, again not significantly different (P > 0.05) and with little fluctuation at each site. The Faith's PD index showed that the evolutionary diversity at the OUT sites was significantly lower than at INX and INJ sites. The Pielou's evenness index indicated that the species evenness of the INJ sample site was slightly lower than those of the INX and OUT sample sites. The Good's coverage index of the three types of sampling sites indicated that the species diversity index of the OUT sampling sites was relatively lower.

Structural characteristics and beta-diversity of archaeal communities
The relative abundance of soil archaea species at the taxonomic level of Phylum and Genus in the PV park of Qinghai Gonghe Photovoltaic Industrial Park is shown in Fig. 5. The soil archaeal community in the study area was structurally homogeneous. Analysis of archaeal phylum or candidate phylum with a relative proportion > 0.1% showed a total of two phyla, among which Thaumarchaeota (99.84%) dominated and was widely distributed in all sampling sites, As shown in Fig. 6(a, b, c), the stress values of NMDS analysis for soil bacterial communities in different groups of sampling sites were 0.125 and 0.045, respectively, which indicated that the ranking results in this study reflected the differences of soil archaeal communities in different sampling sites. Based on the Bray-Curtis distance algorithm (Table S3), no significant variation (P > 0.05) in soil archaeal communities was observed in the three types of fields, which was confirmed by ANOSIM (R = 0.03, P = 0.26) and ADONIS (R = 0.25, P = 0.53). The statistical results from the weighted UniFrac (ANOSIM: R = 0.01, P = 0.35; ADONIS: R = 0.24, P = 0.59) and unweighted UniFrac distance (ANOSIM: R = 0.02, P = 0.34; ADONIS: R = 0.32, P = 0.12) were consistent with the results from the Bray-Curtis distance. The Mantel tests did not show significant correlation between Bray-Curtis and unweighted UniFrac distances (Mantel R = 0.14, P > 0.05 with 5,000 permutations), Bray-Curtis and weighted UniFrac distances (Mantel R = 0.17, P > 0.05 with 5,000 permutations), and unweighted and weighted UniFrac distances (Mantel R = 0.128, P > 0.05 with 5,000 permutations).
To investigate the microbial community characteristics and functions in the soil archaeal communities of the PV field, the archaeal genera whose relative proportions reached 0.1% in each sample site were screened. Cluster analysis was performed on the genera that were picked up from the screening using the Bray-Curtis algorithm. As shown in Fig. 6(c), the archaeal community composition of the sampling sites formed two major branches according to the regional soil geographic environmental conditions and the deployment of PV tracking support panels. The sampling sites S2, S3, S6, S12 (containing INX and INJ), and CK18 clustered into one group. S1, S9, S17 (containing INX and INJ), CK8, and CK17 clustered into another group, and the remaining sampling sites were randomly distributed. Similar to the UPGMA cluster analysis of bacteria, the analysis of archaeal communities showed that rather than clustering into groups based on sampling sites or geographical locations, the INX and INJ soil samples showed a random distribution.

Environmental drivers of soil archaeal communities and VPA analysis
As shown in Table 1, Spearman correlation analysis indicated that the abundance of archaea in the soil had significant positive correlations with soil OM and total OC content (P < 0.01). The Chao1 index of archaea had a significant negative correlation with soil alkaline nitrogen content (P < 0.01). The Faith's phylogenetic diversity index had   Temp., water temperature; EC, electrical conductivity; TN, total nitrogen; TC, total carbon; TP, total phosphorus; TK, total potassium; AN, alkaline-hydrolysable nitrogen; FP, fast-acting phosphorus; FK, fast-acting potassium; SOC, soil organic carbon; SOM, soil organic matter * indicate significant differences (P < 0.05); ** indicate significant differences (P < 0.01) a highly significant positive correlation with total carbon content (P < 0.01). The Shannon index had a highly significant negative correlation with alkaline nitrogen content (P < 0.01) and a highly significant positive correlation with total carbon (P < 0.01). The Simpson's index had significant positive correlations with total carbon and OM content (P < 0.01) and a highly significant negative correlation with alkaline soluble nitrogen (P < 0.01). Thaumarchaeota and Nitrososphaera had a highly significant negative correlation with water content (P < 0.01), respectively. To further investigate the contribution of soil physicochemical factors to soil archaea in the PV park, a ranking analysis was performed using CANOCO 5.0, and the results are shown in Fig. 7. For the archaeal community, the first two axes of the RDA analysis explained 64.82% and 0.01% of the archaeal community composition, respectively. For the ranking plot based on the first two RDA ranking axes, soil electrical conductivity (EC) (F = 11.5, P = 0.028, 499 permutations), alkaline nitrogen (Pulipaka et al. 2016) (F = 4.4, P = 0.03, 499 permutations), and water content (WC) (F = 5.4, P = 0.028, 499 permutations) passed the Monte Carlo significance test, indicating that these three environmental factors can significantly affect the entire archaeal community. VPA was performed to quantify the relative contribution of different variables that may change the structure of the archaeal community (Fig. 8). Based on the available research data, the variables were divided into three groups: soil, vegetation, and anthropogenic influences (PV panel installation). Among them, soil physicochemical characteristics, vegetation growth condition, and PV panel installation parameters explained 33.67%, 25.77%, and 7.02% of the observed variations, respectively, leaving 48.59% of the variations unexplained (ANOVA, p = 0.001, Fig. 8). The PV panel installation parameters alone explained only 0.09% of the variations, and its contribution was mainly related to plant growth conditions (6.91%) and soil physicochemical properties (2.56%). Soil physicochemical properties explained 25.53% of the variations, among which soil moisture content and electrical conductivity passed the substitution test (P < 0.05).

Discussion
The Gonghe Photovoltaic Industrial Park is located in the semi-arid desert region of Northwest China, and the soil in the park exhibits characteristics of semi-arid conditions, high salinity, and low OM concentrations, providing a unique ecosystem for archaeal communities. In this study, we used 16S rRNA amplicon sequencing to identify archaeal populations in the surface soil collected from specific areas.

Soil archaea abundance
Information on the abundance of archaeal communities in desert soil is still scarce. Spatiotemporally, soil archaebacterial abundance varies considerably among studies. A study of archaea in the Tengger Desert during the vegetation recovery period showed that the 16S rRNA gene copy number of archaeal communities increased significantly  (Zhao et al. 2020). However, another study on Tengger Desert soil showed that the abundance of archaeal genes generally increased throughout the succession after 61 annual of vegetation restoration (Hu et al. 2019). The results of several previous studies have also shown that soil microbial communities can increase rapidly in the early stages of desert vegetation restoration . The unique desert soil recovery habitat in which the park is located may also affect archaeal abundance. However, our study was unable to track the abundance of archaea in the same areas prior to construction of the PV park, and future studies should focus on historical data and spatiotemporal relationships.
It is commonly assumed that the placement of solar PV panels plays an important role in reducing external environmental stresses such as wind and heat, enhancing the level of development of shrubs and herbaceous plants under and between the panels and promoting nutrient and energy acquisition by microorganisms; therefore, the abundance of soil microorganisms will increase significantly within the PV park area. However, our study on soil archaea showed the opposite results, with a decreasing trend in archaeal abundance within the PV park (2.20 × 10 10 ) compared with the bare ground outside the PV park (mean value: 4.39 × 10 10 ). The most likely reason is that archaea have formed a specific ecological succession pattern and ecological niche under the extreme environment. They have strong adaptability to the poor soil texture and nutrient environment and can effectively use the limited nutrients and low moisture in the desert ecosystem for growth and reproduction. The construction of the PV site broke the original ecological succession pattern and resource utilization sequence. While there was competition among bacterial, fungal, and archaeal communities (Feng et al. 2017;Rothrock and Garcia-Pichel 2005), it may be at a disadvantage in competition with bacteria or fungi (Bates et al. 2011;Zhao et al. 2020). To the best of the authors' knowledge, this study is the first to reveal differences in the abundance of soil archaea under the influence of different types of PV stands.
Soil archaeal abundance is closely related to OC and OM, which suggests that OM and the most basic metabolic source of archaea can significantly affect soil archaeal abundance (Yang et al. 2014). Moisture, pH, and soil EC were The pure effects of spatial and environmental factors were tested for significance with the permutation test considered as possible determinants limiting the abundance of soil (Hu et al. 2019;Liu et al. 2017). However, in this study, Spearman correlation analysis showed that moisture was not a determinant of soil archaebacterial abundance in desert soil of PV parks. The insignificant role of pH in regulating archaeal community abundance in this study may be attributed to the small variation in soil pH across sites. The lack of correlation between archaeal abundance and EC may be due to high soil salinity levels that exceeded the migration threshold of the archaeal community in response to salinity changes (Hu et al. 2019). The latest research also shows that temperature and water availability have decisive effects on soil microbial extracellular enzyme activity (EEA) and its persistence in soil. EEA allows microorganisms to respond more quickly to environmental changes (i.e., temperature and water availability) in mesophilic and thermophilic growth temperature range and produce a significant feedback mechanisms to self-maintain particular microbial communities in the soil (Gómez et al. 2020). At the same time, the optimal water availability conditions for the maximum EEA of soil microorganisms were mainly affected by soil texture. Soil properties and climatic parameters are not only the main environmental factors controlling soil water availability and temperature, but also the decisive factors regulating soil microbial EEA (Gomez et al. 2020).

Diversity and Community Composition of Archaea in Desert Soils
Despite some variation within archaeal communities, they had low diversity in the desert soil of the PV park in this study (Fig. 4). This is consistent with previous studies (Hu et al. 2019;Nan et al. 2020). Low water usage, high evapotranspiration, high temperature variability, high salinity, lack of vegetation, low decomposition rates, and high soil pH are general trends in desert biomes, all of which may affect the proliferation of archaea, resulting in less diversity than bacterial and fungal communities. We did not find significant differences in archaeal alpha diversity when compared to bare ground soil outside the PV park, likely because the PV park was built recently. Furthermore, as a vehicle for archaeal activity, it may take a long time for the soil environment and vegetation to evolve in the study area; thus, it is unlikely that the diversity of archaea in the comparison sample sites would show significant differences in a short period of time. As for the PV panels themselves, the installation height and rotation angle of the panels may impact the richness of the vegetation in the field area, which in turn impacts the development of soil OM, which is a carrier of microbial growth.
To date, there is no consensus on the dominant archaeal phylum in different environments (Anish et al. 2015;Auguet et al. 2010;Huang et al. 2019;Li et al. 2021;Liu et al. 2009;Zhao et al. 2020). Several previous studies have documented changes in soil archaeal community structure with elevation, latitude, and sampling time Shi et al. 2016). The community structure of Archaea is obviously affected by a variety of geographical environment features, and our study also shows similar results. In our study, the archaeal communities in all samples from areas > 50 km 2 consisted mainly of Thaumarchaeota and Euryarchaeota, with Thaumarchaeota being much more dominant than Euryarchaeota, It is similar to the pattern observed in the Tibetan Plateau and other soils (Brochier-Armanet et al. 2008;Shi et al. 2016). Thaumarchaeota is a mesophilic group of archaea that is widespread across terrestrial and aquatic environments, and as the third phylum of archaea recognized in 2008 (Brochier-Armanet et al. 2008), most soil Thaumarchaea have ammonia oxidation capacity (Auguet et al. 2012;Pedneault et al. 2014) and play an important role in the global nitrogen and carbon cycles. It was found that the ammonia oxidation activity of Thaumarchaea even exceeds that of bacteria (Ke et al. 2014). The relative abundance of Thaumarchaea in the Gonghe PV Park was significant compared to other desert soil, and this increased abundance may be determined by the microbial ecological niche under extreme environmental stress. Meanwhile, the genus of Nitrososphaera under Thaumarchaea acts as an important ammonia oxidizer, using CO 2 as the only carbon source and exhibiting strong autotrophic activity in the soil (Tourna et al. 2011). Nitrososphaera has been shown to be pro-alkaline (Tripathi et al. 2017). The dominant phylum, class, order, and genus of soil archaeal communities are consistent under different PV panel types in the desert zone. The taxonomic structure is simple and not significantly different. The similarity in structure may be mainly due to the similarity in relative abundance of the constituent taxa. However, some sub-taxa were found in specific PV panel types; therefore, it can be inferred that soil microbial communities differed among PV panel types due to the variation in PV panels and their surroundings. In addition, since higher vegetation coverage due to PV panel construction may improve the harsh environmental conditions of the original desert area, plants may indirectly affect the archaeal community through their impact on the physical environment of their root area. Various organic compounds released by various plant roots may also cause changes in soil archaeal communities (Ayangbenro and Babalola 2021).

Factors controlling the structure of soil archaeal communities
The factors affecting the distribution of soil archaea remain unknown Meng et al. 2019;Shi et al. 2016). Previous studies have suggested that the structure of the archaeal community in desert soil may be shaped by a combination of soil physical properties, nutrient content, heavy metal disturbance, plant root action, and pollution (Nan et al. 2020;Wang et al. 2019Wang et al. , 2011Zhao et al. 2020). In this study, the combined results of RDA analysis and VPA suggest that a combination of many environmental variables and anthropogenic disturbances determine the distribution patterns of archaeal community structure in desert soil. Of all the factors we studied, soil conductivity received more attention because of its significant influence on archaeal community composition and diversity patterns. Soil conductivity serves as an equivalent indicator of salinity, which previous studies have found to be the most important environmental factor controlling the global distribution of archaea, and increased conductivity may reduce soil respiration and strongly affect archaeal community composition (Nan et al. 2020;Zhao et al. 2020). The strong explanatory power of several measurable environmental variables in predicting community composition in the current small-scale archaeal studies may reflect the importance of many soil and vegetation factors in influencing archaeal ecology and the dominant influence of only a few key factors. In addition, this study also found that there was a significant correlation between total carbon in soil and microbial community. Soil total carbon contains soil active organic carbon, which is unstable, fast-moving, easy to oxidize, and mineralize in soil. The part of organic carbon with high activity to soil microorganisms, although it only accounts for a small part of the total soil organic carbon, is the energy of soil nutrients and the driving force of soil microbial activities. Previous studies have reported the relationship between soil organic carbon and microbial community structure (Cui et al. 2021;Kang et al. 2021). However, the impact of soil total carbon is also indispensable. Soil microorganisms play a central role in decomposing organic matter, determining mineral nutrient release and nutrient cycling. Our soil samples show that there is a positive correlation between the diversity of archaea and total carbon, indicating that the high content of soil total carbon is conducive to the diversity of these microorganisms. The high growth of diversity and richness of the archaeal community may be consistent with the multiple adaptation strategies to resist the desert oligotrophic ecological environment. From the perspective of genetics, these species communities show unique genes, metabolites, and eco-physiological characteristics, which is an important mechanism for microorganisms to resist the harsh desert environment (Nan et al. 2020). Desertification seems likely to lead to the enrichment of dominant communities and increase tolerance to harsh soil environmental composition. Soil conductivity serves as an equivalent indicator of salinity, which previous studies have found to be the most important environmental factor controlling the global distribution of archaea, and increased conductivity may reduce soil respiration and strongly affect archaeal community composition (Nan et al. 2020;Zhao et al. 2020). The strong explanatory power of several measurable environmental variables in predicting community composition in the current smallscale archaeal studies may reflect the importance of many soil and vegetation factors in influencing archaeal ecology and the dominant influence of only a few key factors. In the case of archaea, the residual spatial variation unexplained by measured environmental factors is often greater in other studies, leaving more room for dispersal-limiting effects in archaeal microbial geography. However, this unexplained variation simply reflects the spatial variability of unknown environmental factors (Shi et al. 2016;Wei et al. 2020). In future studies, in addition to soil factors, the climate of the Tibetan Plateau, where the study area is located, should also be considered. The influence of climate can be explained at regional and district scales. Regionally, the harsh climatic conditions make the Tibetan Plateau a unique microbial geographic distribution area.

Conclusions
In this study, we reported for the first time the characteristics of soil archaeal communities in a PV park in Qinghai Gonghe and explored the effects of PV park construction on archaea. The quantity and diversity of soil Archaea in photovoltaic power stations were analyzed. The structure, quantity, diversity, and distribution characteristics of soil bacterial community in large-scale photovoltaic power stations were analyzed. At the same time, the influence of environmental factors on the distribution pattern of the soil archaeal community was revealed. The abundance of soil archaeal communities was lower in the PV park than outside the park, and soil archaeal abundance was closely related to OC and OM. At the taxonomic level, Thaumarchaeota (99.84%) was dominant in the PV park, while Nitrososphaera was dominant at the genus level. The structure of soil archaeal communities was similar and did not show a significant difference between inside and outside of the park, or among different types of PV panels. The analysis showed that total carbon, alkaline decomposition nitrogen, and OM were the main factors affecting the alpha diversity of soil archaea, and soil water content was the main environmental factor affecting the phylum of Thaumarchaeota and the genus of Nitrososphaera. Quantitative VPA showed that soil physicochemical characteristics, vegetation growth condition, and PV panel installation parameters explained 33.67%, 25.77%, and 7.02% of the observed variation.