A dataset of yield, quality, soil A dataset of yield, quality, soil A dataset of yield, quality, soil physical and chemical properties, enzyme activities, and microbial community characteristics of continuously cropped Ligusticum chuanxiong under rotation and microbial fertilizer treatmentsl properties, enzyme activities, and microbial community characteristics of continuously cropped Ligusticum chuanxiong under rotation and microbial fertilizer treatments properties, enzyme activities, and microbial community characteristics of continuously cropped Ligusticum chuanxiong under rotation and microbial fertilizer treatments

Simultaneously, a microbial fertilizer treatment (JF) and a continuous cropping treatment (LZ) were established, and these two plots were left fallow during the same period. The second crop of L. chuanxiong was grown from October 2024 to April 2025. Before planting, compound fertilizer (N : P₂O₅ : K₂O = 15 : 15 : 15) at 600 kg ha⁻¹ and calcium magnesium phosphate fertilizer (P₂O₅ ≥ 12%) at 600 kg ha⁻¹ were applied to the plough layer (0–20 cm).

In the JF treatment, “Shenchu” compound microbial fertilizer was additionally applied at 5 t ha⁻¹ one month before planting and incorporated into the same soil layer. All treatments received topdressing with compound fertilizer at 150 kg ha⁻¹ in December 2024, and again with compound fertilizer at 150 kg ha⁻¹ plus superphosphate at 75 kg ha⁻¹ in March 2025, broadcast on the soil surface.Rhizosphere soil samples were collected at four key growth stages of L. chuanxiong: the seedling stage (December 2024, SⅠ), the stem initiation stage (January 2025, SⅡ), the secondary stem initiation stage (March 2025, SⅢ) and the harvest stage (April 2025, SⅣ).

At each stage, soil samples were taken from each treatment plot using a five‑point sampling method. After removing debris, the soil was thoroughly mixed, placed in self‑sealing bags, air‑dried naturally in a shaded, well‑ventilated place, and sieved for the determination of soil chemical properties and enzyme activities. A portion of the harvest‑stage soil was kept at low temperature for subsequent soil microbiome sequencing.

At harvest, the yield and quality of L. chuanxiong were also measured.2. Determination of soil chemical properties and enzyme activitiesAfter air‑drying and sieving, soil pH was measured by the potentiometric method. Total nitrogen (TN) and total phosphorus (TP) were determined by the alkaline hydrolysis diffusion method and the molybdenum‑antimony anti‑colorimetric method, respectively, following digestion.

Total potassium (TK) was determined by flame photometry. Organic carbon (OC) content was measured by the potassium dichromate external heating method. Soil alkaline‑hydrolyzable nitrogen (AN), available phosphorus (AP) and available potassium (AK) were determined by the alkaline hydrolysis diffusion method, the molybdenum‑antimony anti‑colorimetric method and flame photometry, respectively.The activities of soil phosphatase (ALP), sucrase (SUC), urease (URE) and catalase (CAT) were determined using the disodium phenyl phosphate colorimetric method, the 3,5‑dinitrosalicylic acid colorimetric method, the sodium phenate colorimetric method and the potassium permanganate titration method, respectively.3.

Soil microbial DNA sequencingAt harvest, a portion of the soil samples was placed in sterile bags, transported to the laboratory at 4 °C, and stored at –80 °C. DNA extraction, PCR amplification, library construction and high‑throughput sequencing were performed by Fujian Manxiu Biotechnology Co., Ltd. (Fuzhou, Fujian Province). The V3–V4 region of the bacterial 16S rRNA gene was amplified with the primer pair 338F/806R, and the fungal ITS region was amplified with ITS1/ITS2.

Sequencing was carried out on the Illumina HiSeq platform using the PE300 mode.Raw paired‑end Fastq data were quality‑controlled with Trimmomatic (v0.36) and Pear (v0.9.6). Paired‑end reads were then merged based on overlapping regions using Flash (v1.20) and Pear to obtain Fasta sequences. Chimeric sequences were removed by the uchime method against a reference database and by the de novo method for unknown databases; sequences with inappropriate lengths were also discarded.

The resulting clean tags were clustered into operational taxonomic units (OTUs) at 97% sequence similarity using uclust and uparse. Taxonomic assignment of OTU representative sequences was performed with the RDP Classifier, using the Silva 128 database for bacteria and the Unite 7.2 database for fungi.4. Yield measurement of Ligusticum chuanxiongAt harvest, three 1 m² quadrats (1 m × 1 m) were randomly selected in each treatment plot.

All L. chuanxiong plants within each quadrat were excavated, the aboveground parts were removed, and the remaining underground parts (rhizomes) were washed, cleaned of impurities, and air‑dried. The fresh weight of the rhizomes from each quadrat was recorded, and the yield per hectare was calculated accordingly.5. Quality determination of Ligusticum chuanxiongQuality components were determined with reference to the method of Yan Yufen et al. Briefly, dried L. chuanxiong rhizomes were ground and sieved.

A sample was mixed with 75% ethanol at a ratio of 1:50 (w/v), ultrasonicated for 30 min, cooled to room temperature, and weighed; the weight loss was replenished with 75% ethanol. Separation was performed on a C₁₈ silica gel column (Shimadzu Shim‑pack GWS C₁₈) at a column temperature of 30 °C and a flow rate of 1.0 mL min⁻¹, with detection at 285 nm. The mobile phases were 0.5% aqueous acetic acid (A) and acetonitrile (B).

The gradient elution program, expressed as the proportion of acetonitrile (B), was: 0–15 min, 10% → 16% B; 15–30 min, 16% → 30% B; 30–35 min, 30% → 40% B; 35–40 min, 40% → 55% B; hold at 55% B until 70 min, then stop. The contents of chlorogenic acid, ligustrazine, ferulic acid, senkyunolide I and senkyunolide A were quantified.6. Data processingData were organized using Excel.

Significance analysis was conducted with SPSS 26. Graphs of soil chemical properties, soil enzyme activities, microbial relative abundances, co‑occurrence network indices, microbial alpha diversity, and L. chuanxiong yield and quality were created with Origin 2024. Redundancy analysis (RDA) and associated plots were generated with Canoco 5.

Non‑metric multidimensional scaling (NMDS) was performed in R (v4.5.1) using the packages vegan, ggplot2, dplyr and ggrepel. The Mantel test was carried out in R using the packages linkET, tidyverse and RColorBrewer. For network analysis, bacterial and fungal genus‑level abundance data were first filtered (relative abundance < 0.1%) in R (v4.5.1).

Spearman correlations were computed using the igraph, psych, tidyverse and stringr packages (p < 0.05; |correlation| > 0.7), and co‑occurrence networks were visualized with Gephi and Adobe Illustrator 2025.