Hydrologic drying drives dispersion inflation and predictability loss in alpine wetland bacterial communities
Description
Research Data and Reproducibility Statement Repository deposition Raw sequence reads have been deposited in the NCBI Sequence Read Archive (SRA) under BioProject PRJNA1397567 (Study SRP659517). Accession details are provided in the manuscript and will be publicly accessible upon release by the repository. Supporting data files included with this submission Data S1. Per-sample read-processing statistics across the amplicon pipeline. Data S2. Full ALDEx2 phylum-level trend output (all tested phyla). Data S3. Full ANCOM-BC2 phylum-level trend output (all tested phyla). Data S4. Reproducible scripts/commands and R session information used to generate the results and figures. These materials are provided to enable transparency and reproducibility of the analyses reported in the manuscript.
Files
Steps to reproduce
Soil samples were collected across an ordered alpine wetland degradation gradient from wet marsh to severely degraded sites using a stage-stratified field design. At each site, surface soils were sampled with sterile tools, homogenized, and subsampled for physicochemical measurements and DNA-based community profiling. Soil moisture and key resource-pool variables (including carbon and phosphorus pools) were quantified using standard soil analytical protocols described in the manuscript and Supplementary Information, with appropriate quality control and consistent sample handling across stages. Microbial community data were generated from 16S rRNA gene amplicon sequencing. Total DNA was extracted from soil using a commercial soil DNA extraction kit following the manufacturer’s protocol, and extraction and no-template controls were processed in parallel to monitor contamination risk. The target 16S rRNA gene region was amplified using the primer pair specified in the Methods. Amplicons were prepared for sequencing and run on an Illumina platform as paired-end reads by a professional sequencing provider. Raw sequence reads were processed using a reproducible amplicon workflow that includes demultiplexing, quality filtering, denoising to amplicon sequence variants, chimera removal, and taxonomic assignment against a curated reference database. The resulting feature table and taxonomy table were used for downstream analyses, including rarefaction to a consistent sequencing depth, computation of alpha and beta diversity, ordination, permutational tests, dispersion metrics, differential abundance trend modeling, distance-based redundancy analysis, and a hydrology-centered piecewise structural equation model. Predictability analyses were performed using repeated cross-validation, with prediction residuals and dispersion-derived distance-to-centroid metrics derived from the same processed datasets. All raw reads are deposited in a public repository, and the accompanying research data package provides (i) per-sample read-processing statistics for transparency, (ii) full differential abundance outputs for all tested taxa from two complementary methods, and (iii) complete scripts/commands and software session information to reproduce all results and figures from raw inputs to final analyses.
Institutions
- Southwest Minzu University