
Sustainable management of hardsetting soils requires coordinated improvement of chemical, hydrological and mechanical functioning. Whether biochar can deliver these benefits simultaneously remains uncertain. We tested a 700 °C wood biochar at soil-specific rates of 3%–4% w/w in 18 sloping lysimeters paired with 18 pots used for destructive measurements. Three Portuguese pasture soils were monitored during one growing season and nine natural rainfall events. Descriptive composite measurements indicated higher soil pH, 70%–100% lower exchangeable Al and 57%–199% higher measured organic matter following biochar addition, although the latter included incorporated biochar. Repeated measurements showed that soil moisture increased by 39%–76%. However, bulk density did not decrease, and penetration resistance at a common soil-moisture level increased by 56%–121% across the three soils (all p < 0.001). Aboveground biomass declined by 37%–61% (all p ≤ 0.006), whereas final vegetation cover, runoff, estimated infiltration, erosion and erosion-associated organic-matter export showed no significant response. Rainfall amount received stronger model support than maximum 30-min intensity as a driver of runoff and erosion. Greater soil water content therefore reflected increased water storage, but not improved hydrological or mechanical functioning. These findings identify an important constraint for biochar-based management of hardsetting soils: incorporation conditions should be considered alongside amendment properties and application rate when multiple soil functions are targeted simultaneously.
Recurrent daytime heat can modify plant–soil–microbe relationships, but cultivar-associated rhizosphere responses remain poorly resolved in perennial tea. We compared three Wuyi rock tea cultivars — Shuixian (SX), Rougui (RG) and Qilan (QL) — under two 30-d chamber programs that differed during a 5-h midday period (21 °C in the Reference chamber and 40 °C in the Heat-Pulse chamber). Because each program was implemented in one dedicated chamber, replicated inference was restricted to cultivar differences within chambers; between-program contrasts were descriptive. Rhizosphere soil chemistry, bacterial 16S rRNA amplicon profiles and CNPS functional-gene abundance were quantified from 18 pot-level samples. Six of eight soil variables showed FDR-significant cultivar effects in each chamber. Bacterial composition was strongly cultivar-differentiated (PERMANOVA R2 = 0.640, P = 0.004 in Reference; R2 = 0.712, P = 0.005 in Heat-Pulse). In contrast, finalized absolute CNPS profiles were not significantly differentiated among cultivars in either chamber, although a 16S-normalized sensitivity representation was significant in Reference only. Across all samples, soil and bacterial dissimilarities covaried (Mantel r = 0.411, FDR-adjusted P = 0.0003), whereas global associations involving CNPS were not significant after FDR correction. Partial db-RDA identified TN and AP as conditional correlates of bacterial composition, with no FDR-significant soil variable for CNPS. Thus, cultivar-associated differentiation was robust for rhizosphere soil and bacterial communities but weaker and abundance-representation-sensitive at the DNA-level functional-gene layer.