长期不同施肥下南方黄泥田有效磷对磷盈亏的响应特征

【目的】黄泥田为南方主要中低产田类型之一。通过研究长期施肥条件下南方黄泥田土壤磷素累积盈亏与有效磷的关系，为黄泥田科学施用磷肥提供理论依据。【方法】基于连续33年水稻长期定位试验，研究了不施肥 (CK)、单施化肥 (NPK)、化肥配施牛粪 (NPKM)、化肥配施秸秆 (NPKS) 四个处理土壤有效磷的变化规律及土壤磷累积盈亏状况，计算有效磷–磷盈亏响应系数。【结果】各施肥处理双季稻年份 (1983～2004年) 土壤有效磷与全磷含量呈年际上升趋势，以NPKM增幅最为明显，改为单季稻后 (2005～2015年) 则呈下降趋势，也以NPKM降幅最为明显。试验至2015年，NPK、NPKS处理有效磷含量分别为9.7 mg/kg、8.7 mg/kg，较试验初期分别下降8.3 mg/kg、9.3 mg/kg，NPKM处理与试验初期持平。CK、NPK、NPKM、NPKS处理的磷素活化系数 (PAC) 总体呈年际下降趋势，其多年平均磷素活化系数值 (PAC) 分别为2.9%、3.5%、4.7%、4.1%，其中NPKM与NPKS处理显著高于NPK与CK处理。除双季稻NPKS处理外，不同施肥模式下双季稻与单季稻年份的土壤有效磷增减与土壤累积磷盈亏均呈显著正相关，其中双季稻年份土壤磷素 (P) 每盈余100 kg/hm2，NPK、NPKM处理有效磷分别增加4.5 mg/kg与11.2 mg/kg，而单季稻年份土壤磷素每亏缺100 kg/hm2，NPK、NPKM、NPKS处理有效磷分别减少14.6 mg/kg、23.9 mg/kg、25.9 mg/kg。双季稻磷肥年施用量 (P) 为26.2 kg/hm2时，土壤磷素呈盈亏持平状态。【结论】有机无机肥配施比单施化肥能够显著提高黄泥田土壤有效磷、全磷含量和磷素活化系数，有效磷含量与磷素累积盈亏密切相关，等磷素盈亏量下，有机无机肥配施的有效磷响应系数要高于单施化肥，而磷累积亏缺下，有效磷降幅响应比磷累积盈余下有效磷增幅响应大。

Response characteristics of Olsen-P to P balance in yellow paddy fields of southern China

Objectives]In order to provide theory for phosphate (P) fertilizer application in the yellow paddy, a study was conducted to research the relationship between P accumulation and its efficiency.Methods]The variation and accumulate of P in paddy fields were studied in a 33 years' long-term experiment. Four fertilizer treatments were setup: no fertilizer (CK), chemical fertilizer (NPK), chemical fertilizer plus cattle manure (NPKM), and chemical fertilizer plus straw (NPKS) and available P-P response coefficient was calculated.Results]The contents of soil available and total P in the CK treatment decreased with the treatment years in both the double-rice and the single-rice systems, and those in the other three treatments increased with the fertilizer treatment years under double-rice system (1983-2004) and decreased in single rice system (2005-2015). Until 2015, the contents of soil available P in the NPK and NPKS treatments were 9.7 mg/kg and 8.7 mg/kg, which were significantly lower than that of initial soil sample in 1983, and decreased by 8.3 mg/kg and 9.3 mg/kg, respectively. The available P content in the NPKM treatment was 18.0 mg/kg in 2005, which was in line with initial soil sample. The phosphorous activity coefficients of the CK, NPK, NPKM and NPKS treatments were 2.9%, 3.5%, 4.7% and 4.1% in turn, those in the NPKM and NPKS treatments were significantly higher than that in the NPK treatment. There was a linear relationship between the balance of P and soil available P contents in all the treatments except that in NPKS under double-rice system. Under double-rice system, annual soil P surplus of 100 kg/hm2 would lead to the increment of soil available P contents of 4.5 and 11.2 mg/kg in the NPK and NPKM treatments, and under single-rice system, annual soil P defect of 100 kg/hm2 would lead to the decrement of available P content by 14.6, 23.9 and 25.9 mg/kg in the NPK, NPKM and NKPS treatments, respectively. The soil could keep in balance when the amount of P fertilizers was 26.2 kg/hm2.Conclusions]Available P content is closely related to the balance of soil P. The available P response coefficient of the NPK combined with manure or straw fertilization is higher than that of NPK fertilization under the equivalent phosphorus balance. Acceleration response of available P is higher under phosphorus-deficient conditions than phosphorus-sufficient conditions.