Dataset of the effects of forest recovery on soil microbial carbon use efficiency and its drivers in a subtropical karst region of Southwest China

Soil moisture was adjusted to approximately 60% of water-holding capacity, and the soils were incubated at 20 °C for 24 h. After incubation, soil DNA was extracted, and microbial growth and CUE were calculated based on 18O incorporation into DNA. qGrowth and qCO2 were further calculated using microbial biomass carbon and soil respiration-related measurements. Soil physicochemical properties, nutrients, microbial biomass, and soil carbon fractions were determined using corresponding standard analytical methods.

Microbial community data were obtained from soil microbial sequencing and related analyses. Data processing and statistical analyses were primarily conducted using R, including descriptive statistics, paired-sample comparisons, correlation analyses, random forest analysis, and structural equation modeling. The original measurements were organized, harmonized in terms of units, and subjected to quality checks before being compiled into the dataset.

Derived variables were calculated from the corresponding measurements according to the relevant calculation procedures, while the original measurements or variables used for calculation were retained where applicable to ensure data traceability.The spatial information is represented by the geographic coordinates of individual sampling plots, with the plot level serving as the basic spatial unit. The dataset does not contain continuous spatial raster data.

The temporal information mainly corresponds to field surveys and soil sampling conducted during September–October 2023, while the 18O-H2O DNA tracing experiment involved a 24 h incubation period. The dataset uses individual sampling plots and paired sites as the basic observational units and contains 46 sampling plot records from 23 paired sites, including 23 secondary forest plots and 23 maize cropland plots. Each row in the tabular dataset represents one sampling plot.

The land-use type field distinguishes between secondary forest and maize cropland, while the paired-site identifier is used to link the two land-use types within each pair. The columns represent site information, soil physicochemical properties, soil carbon, nitrogen and phosphorus fractions, microbial biomass, microbial CUE, and related variables. The specific measurement units of individual variables are provided in the corresponding data tables and variable descriptions.Missing data were identified based on the original experimental records and the final data tables.

No missing values were artificially imputed or interpolated. Where missing values occur, the corresponding cells are left blank or recorded using the missing-value codes specified in the relevant data files. The dataset is primarily derived from field sampling and laboratory measurements, and potential measurement errors may arise from spatial heterogeneity of soil samples, field sampling variability, experimental procedures, instrument precision, and differences among analytical methods.

For variables involving replicate or technical replicate measurements, quality control was conducted according to the corresponding experimental procedures, and the final values were calculated following the specified analytical protocols.The data files mainly contain information on sampling sites, soil microbial characteristics, basic soil physicochemical properties, soil microbial carbon use efficiency, and related variables.

All data files are provided in commonly used tabular formats and can be linked through variable names, sampling-site identifiers, and paired-site identifiers. The numerical values represent organized original measurements or variables calculated from the original measurements. The data files can be opened and viewed using commonly used spreadsheet software such as Microsoft Excel.

Further statistical analyses can be conducted using R and relevant packages.