Long-term effects of crop rotation and nitrogen fertilization on phosphorus cycling and balances in loess-derived Mollisols

Developing meaningful agroecosystem soil P inventories necessitates moving beyond single measures of readily extractable inorganic P (P_i) limited to surface depths. We drew on a long-term (36 year) experimental field trial in the US Maize Belt (northwestern Illinois) to evaluate how crop rotation [maize-maize (Zea mays L.) vs maize-soybean (Glycine max L. Merr.)] and N fertilization (0 vs 269 kg N ha⁻¹) impact P dynamics throughout the soil profile by using sequential fractionation and phosphatase activity assays, contextualized by soil P stocks and agronomic P balances. Distribution of P fractions by depth (0–15, 15–30, 30–60, 60–90 cm) indicate that management effects were limited to the surface soil layers (0–30 cm). Soil P fractions differed more by depth than by experimental treatments. Long-term N fertilization significantly decreased pH concurrently with labile organic P (P_o) and phosphodiesterase activity. Soil labile inorganic P (P_i) was two-fold lower under N fertilization compared to zero N fertilization, reflecting greater yield and thus P export via grain harvest. Under N fertilization, integration of soybean elevated soil phosphodiesterase activity and decreased water-extractable P_o. Higher stocks of soil P_o than labile P_i at surface depths (0–30 cm) corroborated a hypothesized appreciable pool size of soil P_o relative to the labile P_i pool to which most agronomic assessments are limited. Large negative agronomic balances over the 36-year period (−426 to −945 kg P ha⁻¹) are suggestive of legacy P from pre-experiment manure application and high native P stocks, with net P export equivalent to 11–35% of soil P stocks at 0–90 cm depth at the initiation of the experiment. These results contribute to a better understanding how N fertilization and rotation practices influence soil P cycling and stocks, thereby informing P budgets for comprehensive agroecosystem P management.