Abstract: 【Objective】 The objective of this study is to investigate the effects of long-term fertilization on phosphorus (P) accumulation and movement in profile of a calcareous purple paddy soil. 【Method】 The ongoing experiment, initiated in 1981, consisted of eight treatments including CK (no fertilizer), nitrogen fertilizer (N), nitrogen and phosphorus fertilizers (NP), nitrogen, phosphorus and potassium fertilizers (NPK), pig manure (M), pig manure plus nitrogen fertilizer (MN), pig manure plus nitrogen and phosphorus fertilizers (MNP) and pig manure plus nitrogen, phosphorus and potassium fertilizers (MNPK) with four replications. Status of P balance, accumulation and fate, soil total P (TP) and Olsen-P at 0-20 cm soil depths, and P accumulation and distribution at 0-100 cm soil depths were measured to get better understanding of the impacts of different treatments on these parameter changes after 33 years in the calcareous purple paddy soil in the Southwest China. 【Result】 Results showed that 613.12 kg·hm^-2 of P in P omission treatments (CK and N alone) were absorbed by crop grain and straw after 33 years in the calcareous purple paddy soil. The P nutrient from seeds, stubble, rain and irrigation waters was about 106.61 kg·hm^-2 entered into the soil. Thus soil P became deficient when P was omitted from the fertilizer program (CK and N alone), giving an average deficit value of 15.35kg·hm^-2 of P nutrient. And the situation was worsened with time, changing of 0.0011 g·kg^-1 in TP and 0.029 mg·kg^-1 in Olsen-P annually. After 33-year different fertilizer treatments, two inorganic P treatments (NP and NPK) resulted in 1 880.03 kg·hm^-2 of total P inputs and 1 275.40 kg·hm^-2 of P uptake by crops. Total inputs of P were 2 532.68 kg·hm^-2 and P absorbed by crop was 757.50 kg·hm^-2 in two P-omitted organic manure treatments (M and MN). And P inputs and uptake were 4 305.11 and 1 436.64 kg·hm^-2 in the two P-added organic manure treatments (MNP and MNPK), respectively. Thus treatments amended with P and/or organic manure yielded annual balances of total P as 18.32, 53.79 and 86.92 kg·hm^-2, fate unknown P as 4.99, 34.96 and 59.39 kg·hm^-2. TP changing as 0.015, 0.0018 and 0.018 g·kg^-1 and Olsen-P as 1.13, 0.032 and 1.17 mg·kg^-1 for the two inorganic P treatments, the two P-omitted organic manure treatments and the two P-added organic manure treatments, respectively. The P omission treatments resulted in reduction of TP with depths downward the profile, while an opposite trend was observed for Olsen-P. P treatments, however, gave P distribution patterns of highs in the top and the bottom depths but lows in the middle layers. Addition of inorganic P fertilizers could transfer P down to the 60-80 cm soil profile, while addition of organic manure alone or in combination with inorganic P fertilizers could move P further down to more than 100 cm soil depths, suggesting that organic P was more mobile than inorganic P in the soil and susceptible to loss. 【Conclusion】When soil Olsen-P reaches to an adequate level (eg. 20 mg·kg^-1 as a critical value for most of cereal crops) after consecutive applications of P fertilizers for years, reduction in P rates, organic P in particular, should be considered to maintain or effectively lower soil P pool and to minimize P losses to environment.

Key words: long-term fertilization, calcareous purple paddy soil, phosphorus, accumulation and movement