Soil physical responses to novel rice cultural practices in the rice–wheat system: Comparative evidence from a swelling soil in Nepal

Soil puddling in advance of rice (Oryza sativa L.) transplanting disperses surface aggregates and generates compaction at depth. As a management scheme for rice, puddling is typically considered advantageous for maximizing resource availability and yield. However, some experimental findings suggest a conflict between edaphic conditions created by this establishment technique and the performance of subsequent non-rice crops like wheat (Triticum aestivum L.). At a site in the mid-hills region of Nepal on a silt loam soil with vertic characteristics, we compared the impact of six rice tillage (surface tillage—T₁, shank subsoiler—T₂, shank subsoiler + moldboard plough—T₃) and establishment (soil puddling + transplanting—TPR, direct seeding—DSR) combinations on soil physical properties over two cycles of the rice–wheat rotation. For the rice season, 0–20 cm saturated hydraulic conductivity (K_sat) in the DSR plots was 2.6 and 4.3 times higher than their TPR counterparts in the first (Y₁) and second (Y₂) years, respectively (TPR-Y₁ = 93 mm day^?1, DSR-Y₁ = 241 mm day^?1, TPR-Y₂ = 133 mm day^?1, DSR-Y₂ = 582 mm day^?1), whereas tillage method did not significantly influence K_sat in this soil layer. The impact of rice establishment method was reflected in higher TPR bulk densities in the 5–10 (DSR = 1.19 g cm^?3, TPR = 1.24 g cm^?3) and 10–15 cm (DSR = 1.24 g cm^?3, TPR = 1.29 g cm^?3) depth increments in the wet season. Although none of the treatments significantly influenced the position or thickness of the plough sole, penetration resistance profiles suggest that vertical fractures with reduced soil strength were created within the pan region by deep tillage (T₂ and T₃), although these features were not associated with higher hydraulic conductivities from 20 to 50 cm. As the soils dried at the end of the rice season, crack propagation in the deep tilled plots (T₂ and T₃) was more pervasive. During the wheat season, comparable bulk density profiles and soil moisture retention characteristics across the treatments suggest that many of the edaphic changes induced by contrasting rice tillage and establishment practices did not persist in the self-mulching, vertic soils at our site. Conversely, significant increases in K_sat among the DSR plots from Y₁ to Y₂ (Y₁ = 241 mm day^?1, Y₂ = 582 mm day^?1) imply a temporal element to soil structural regeneration with adoption of direct seeding.