Long-term manure compost application induces a plow pan–like resistant layer in the subsoil of an extremely sandy tropical Ultisol

Abstract

Subsoil hardening is generally regarded as detrimental in arable systems, but its occurrence and functional significance in extremely sandy soils remain poorly understood. This study examined whether long-term manure compost application induced a plow pan–like mechanically resistant layer in an extremely sandy tropical Ultisol in Khon Kaen, Thailand, and whether this layer affected cassava ( Manihot esculenta Crantz) yield and yield stability. A 47-year field experiment included five treatments with four replicates: control, NPK fertilizer, cassava residue, NPK + residue, and NPK + manure compost. Soil penetration resistance (PR) was measured at 1-cm intervals to 0.90 m depth, surface volumetric water content was measured concurrently, and soil chemical properties to 1.0 m depth and cassava yield records from 2010 to 2020 were analyzed. PR was sharply elevated at 0.20–0.30 m depth only under NPK + manure compost, with peak values exceeding 2500 kPa, whereas no comparable layer was detected under the other treatments. Surface water content was higher in fertilized treatments, including NPK + manure compost. The resistant layer coincided with higher soil pH and Ca saturation, and depth-specific regression analysis indicated that Ca saturation was positively associated with PR in the 0.2–0.4 m layer. Despite this mechanically meaningful resistance, cassava yield and interannual yield variability did not clearly differ between NPK + manure compost and NPK + residue. Overall, this study shows that long-term manure compost application was associated with an unexpected, localized mechanical response in extremely sandy tropical soils, likely associated with amendment quality and Ca-mediated stabilization, without causing an apparent yield penalty under the present conditions. The significance of this study lies in documenting a field-scale phenomenon detected after 47 years of continuous treatment in a rare long-term experiment, although the temporal development of the layer cannot be determined from this cross-sectional dataset. The mechanism remains inferential because pore structure, particle bonding, root distribution, and depth-resolved soil water dynamics were not directly measured.