Site description
The experimental site was located in Haibei Autonomous Prefecture, Qinghai Province, China (36°92′ N, 100°93′ E, 3029 m a.s.l). The area is a typical plateau continental climate with an average annual precipitation of about 410 mm. The year is divided into cold and warm seasons with an annual average temperature of 1.4°C. The highest temperature in the warm season (June to August) is about 15℃, and the lowest temperature in the cold season (December to February) is about − 13℃. Annual mean temperature and annual mean precipitation from 1976 to 2016 are shown in Fig. S1. Elymus nutans, Poa pratensis L., Kobresia humilis and Potentilla bifurca are the main plant species. The soil is classified as Mat-Gryic Cambisol with a clay loam texture under the Chinese soil classification system (Ma et al. 2017) with 3.5 g kg− 1 soil total nitrogen, 40.9 g kg− 1 soil total carbon and 2.32 mg kg− 1 available phosphorus.
Experimental Design
In 2012, a flat alpine meadow was fenced to prevent livestock from disturbance and the dominant species in this meadow was Elymus nutans. In order to simulate the global warming environment in 2014, open-top chambers (OTC) (top diameter 1.5 m, height 0.7 m, bottom diameter 2.2 m) were permanently installed in the meadow. The meadow was divided into Control (CK) and Warming (W; +~2°C), with 4 replications of each treatment. Each experimental plot had a diameter of 2.2 m with a 3 m buffer zone in adjacent plots. Three extractor fans were installed at the top of each chamber to maintain a ~ 2 ◦C difference between the warmed and control plots. A multipoint soil (CS655-L30 decagon devices, USA) temperature and humidity meter was used to automatically record soil temperature and humidity, and data were exported by an automatic data collector (CR800-XT).
Sample Collection In The Field
The plant community characteristics were investigated using the quadrat sampling method (1 × 1 m) in August 2021. Plant species richness in each experimental plot was recorded, with the frequency and coverage of each plant species being assessed. Then five soil cores (3.5 cm in diameter and 0–10 cm in depth) were randomly selected from each plot, and after removing rock from the cores, they were uniformly mixed and combined into one soil sample. The soil samples were divided into two parts: one part was frozen in liquid nitrogen and transported to the laboratory within 24 hours and stored at -80℃ for AMF community analysis; the other part was sieved through a 2 mm sieve and air-dried for soil physical and chemical property analysis. To explore the effect of long-term warming on the AMF community at the individual-plant-species level, we sampled a dominant plant species, E. nutans. Five individuals of E. nutans were selected randomly from each plot, excavated and collected as one sample (Jiang et al. 2018). Plant samples for the analysis of AMF community and mycorrhizal colonization were stored at − 80℃, while those for plant nutrient contents were dried at 65°C oven.
Determination Of Mycorrhizal Colonization Rate
Mycorrhizal colonization was examined by following the procedure of Koske and Gemma (1989). When the plants are excavated, select the fine roots and washed them to remove soil particles, weighed 0.5-1.0 g and cut them into 1cm. Soaked them in KOH (10%, w/v) boiled until the roots were transparent. Then acidified in HCl (2%, v/v), and stained with trypan blue (0.05%, w/v). After 30 min, these roots were immersed in destaining solution (glycerol: lactic = 1:1). Finally, we mounted stained roots on slides and calculated AMF colonization using the magnified gridline intersect method (McGonigle et al. 1990) with a compound microscope under 100–400 magnification. A root intersection was considered colonized if hyphae, arbuscules or vesicles were present.
Analysis Of Soil Physicochemical And Plant Properties
The soil moisture content (SMC) was measured gravimetrically (dried at 105℃ for 12 h) and soil pH was measured in 1 M KCl (1:5 w/v) by using a pH meter (Orion Star A215, Thermo Fisher Scientific, USA). NO3−-N and NH4+-N concentrations were determined by a flow-through solution analyzer (Flowsys, Ecotech, Germany). Soil total C (TC) and total N (TN) were measured by a carbon-nitrogen element analyzer (Elementar, Hanau, Germany), as well as plant TC and TN. Soil active phosphorus (AP) was extracted from 2.5 g of dry soil with 0.5 M NaHCO3 solution (pH 8.5) for 30 min and determined by colorimetric method (Bao, 2008). Soil organic carbon (SOC) was measured with an Organic Carbon Analyzer (Vario Macro, elementary, Germany) after 8 hours of fumigation with concentrated hydrochloric acid. HClO4–H2SO4 colorimetry was used to measure plant total phosphorus (TP) which was extracted by 0.5 mol/L NaHCO3.
Molecular Identification Of Amf
Microbial DNA was extracted from root and soil samples using the E.Z.N.A.® Soil DNA Kit (Omega Biotek, Norcross, GA, US) according to the manufacturer's statement. Final genomic DNA purification and concentration were determined by a NanoDrop NC2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and DNA quality was checked by 1% agarose gel electrophoresis. Afterwards, the fungal 18S rRNA gene was amplified by PCR (BioRad S1000, Bio- Rad Laboratory, CA) with primer set AMV4.5NF/AMDGR (5′-AAGCTCGTAGTTGAATTTCG-3’/5′-CCCAACTATCCCTATTAATCA T -3′) according to the study of Lumini et al. 2009. PCRs were performed in a 20 µL mixtures containing 5 µl of buffer (5×), 0.25 µl of Fast pfu DNA Polymerase (NEB, USA), 2 µl (2.5 mM) of dNTPs, 1 µl (10 uM) of each Forward and Reverse primer, 2µl of DNA Template, and 8.75 µl of ddH2O. The following PCR conditions were: initial denaturation at 98℃ for 2min, denaturation with 30 cycles at 98℃ for 15s, annealing at 55℃ for 30s and extension at 72℃ for 30s, with a final extension of 5 min at 72°C. PCR reaction mixtures were purified using the TruSeq Nano DNA Gel Extraction Kit (Illumina, USA) and quantified using the Agilent Bioanalyzer 2100 system (Agilent, USA). Sequencing libraries were generated according to the standard procedures of N (New England Biolabs, USA). Finally, the library was sequenced on the Illumina Hiseq 2500 platform.
Microbiome bioinformatics were performed with QIIME2 2019.4 (Bolyen et al.2018), with slight modification according to the official tutorials (https://docs.qiime2.org/2019.4/tutorials/). Multiplex the raw sequence data using the demux plugin and then cut the primers using the cutadapt plugin (Martin et al.2011). The sequences were then quality denoised, filtered, merged and chimeras removed using the DADA2 plugin (Callahan et al. 2016) The mafft (Katoh et al. 2002) was used for comparison with non-single amplicon sequence variants (ASV). Fasttree2 (Price et al. 2009) was used to construct the phylogeny. For the MAARJAM database (https://www.maarjam.botany.ut.ee/) (Bokulich et al. 2018), classifications were assigned to ASVs using the classify-sklearn naïve Bayes taxonomy classifier in the Feature Classifier plugin (Bokulich et al.2018).
Statistical analysis
To test the significance of the differences between warming and control treatment for various variables, two-tailed paired t-tests, Wilcoxon-test were employed by SPSS19.0. For Illumina MiSeq sequencing data, alpha diversity indices (Chao1, Faith's PD, species and Simpson diversity indices) were generated using QIIME2 (2019.4). Using the “qiime feature- table rarefy” function in QIIME2, all samples are thinned to the same sequence for downstream analysis based on 95% of the lowest sequence number to ensure that the samples are sequenced to the same depth. Sequencing reads are considered sufficient if the number of species no longer increases with the number of sequencing reads. Differences between treatments were compared using Kruskal-Wallis rank sum test and Dunn's test was used for pairwise comparison. For beta diversity analysis, Bray-Curtis distances were calculated and principal coordinate analysis (PCoA) was performed using the ‘vegan’ package in ‘R’ (Version 4.1.3) to visualize community similarity. The significant difference of AMF community among treatments was detected by permutational multivariate analysis of variance (PERMANOVA). Canonical correspondence analysis (CCA) was used to identify soil properties that predict changes in AMF communities. The Mantel test was used to assess the relationships between the AMF community and the soil characteristics. CCA and Mantel tests were analyzed using the cca function in the ‘vegan’ package and the mantel.rtest function in in the ‘ade4’ package in ‘R’ (version 4.1.3), respectively. Origin2018 (OriginLab, US) and Adobe Illustrator (2020) were used for plotting.