Organization Across a Caragana korshinskii Plantation Chronosequence: Edaphic Gradients, Hierarchical Filtering and Domain-Specific Assembly

Dryland shrub restoration is often assessed by vegetation recovery, yet belowground microbial organization remains less resolved. We analyzed bacterial 16S rRNA and fungal ITS amplicon sequence variants (ASVs) from four Caragana korshinskii plantations established 5, 10, 20 and 30 years before sampling. Within each plantation, five shrubs were sampled as within-plantation subsamples. Fine roots and rhizosphere samples were collected at 25–30 cm soil depth, while non-rhizosphere soils were collected at the same depth but 10 or 30 cm horizontally from the sampled root system. A soil-chemistry-matched subset defined a PCA-based edaphic axis (PC1 = 48.4%) that ordered the four sampled plantations and was associated with higher soil water content, lower electrical conductivity and lower total nitrogen and soil organic matter in the older plantations. ASV pools contracted from the soil to the rhizosphere and root compartments, indicating hierarchical filtering. Sample-level Bray–Curtis analyses showed strong microhabitat structuring, particularly for bacteria, while null-model analysis assigned most bacterial turnover to homogeneous selection (91.9%). Fungal turnover more often fell into the dispersal-limitation category (79.4%), but this ITS-based result should be interpreted cautiously. Rhizosphere candidate indicators and exploratory association-network summaries suggested differences among plantation age/site classes, including an apparent increase in retained associations in the 10-year plantation; however, these network comparisons are treated as hypothesis-generating because the network for each plantation age/site class was based on only five samples. Because each plantation age/site class was represented by a single plantation, plantation age is confounded with site, and all age-related patterns are interpreted descriptively within this chronosequence rather than as replicated temporal effects.