2026-06-01 GEODERMA REGIONAL 2026 45(卷), null(期), (null页)
Phosphorus (P) is essential for plant growth but is frequently limiting in tropical agroecosystems due to strong sorption and low availability. The adoption of cover crops (CCs) with high P uptake efficiency can enhance P cycling for cash crops, reducing fertilizer use and improving P-use efficiency. This study aimed to assess changes in soil P dynamics after adopting CCs under contrasting P sources in a sandy soil in dryland conditions in Brazil. The experiment was carried out in sandy loam soil over two maize-growing seasons under no-tillage. Treatments consisted of two P sources (single superphosphate - SSP, and rock phosphate - RP) and a control without P (NilP), combined with six summer CC species (sunn hemp, spectabilis, pigeon pea, lablab, jack bean, and pearl millet) plus fallow. P uptake, P offtake, and soil P fractions were determined in both seasons. Cover crops increased PUE under both P sources. Under SSP, PUE increased significantly after lablab cultivation, reaching 58% compared to fallow. Under RP, all CCs promoted a significant increase in PUE, except for spectabilis. Most cover crops reduced system P balance relative to fallow, indicating greater P export by maize harvest. Sunn hemp was the most efficient CCs in P uptake, similar to 36 kg ha-1 of P from the soil on average across both growing seasons (36% higher than the average of other CCs). In the first season, CCs increased the labile P pool under Nil-P at 0-10 cm depth, especially with spectabilis, pigeon pea, sunn hemp, and pearl millet (similar to 38% higher than fallow). However, this pool was depleted in the second season. Under P fertilization, labile and moderately labile P increased substantially, particularly with pearl millet under RP at 0-5 cm. Across seasons, CCs depleted non-labile P at 10-20 cm and accumulated it in the surface depth. Overall, CCs efficiently cycled P and improved maize PUE most consistently under RP, highlighting that CC selection can be used to increase the effectiveness of sparingly soluble P inputs in sandy, low-buffering dryland soils.