現行臺灣土壤磷速測法與作物施肥推薦量之再評估

Abstract

近年來土壤與磷相關研究已由偏重磷有效性指標與磷肥推薦量評估，轉變為考量磷飽和指標與磷流失風險評估，而先前研究指出由Bray No.1萃取量和吸附實驗所計算出磷肥推薦量不盡相符；故本研究以18種臺灣農田土壤對此進行探討，並與植生試驗和其他與磷相關指標作比較，對修正臺灣現行磷肥推薦量分級標準提出初步嘗試。Bray No.1萃取量對時間的趨勢，在和基本性質比較後，初步懷疑可能與pH值以及磷、鐵、鋁含量有關。等溫吸附實驗中，由Langmuir方程式推算出來的最大吸附量和磷肥推薦量，會受到所添加磷濃度及各濃度處理組數和組距影響；故最大吸附量以存在較大磷添加濃度的組別為準，而磷肥推薦量則以較多組數、較小組距所計算出來者為準。在Neubauer生物試驗中，因發芽率和各盆磷總含量間無明顯相關，故將其扣去blank組得到的吸收量平均，換算為土壤有效磷含量(mg P/kg soil)，作為植物有效性磷的依據。由各指標間的相關係數矩陣可看出：三種萃取法(Bray、CaCl2、草酸銨)和兩種磷飽和指標(Psat、DPS)彼此有極顯著正相關(P=0.001)，而小麥吸收量所推算植物有效性磷含量與Bray No.1萃取磷、草酸銨萃取磷、Psat和DPS-1亦存在相關性(P=0.05)。另外0.01 M CaCl2萃取量對磷飽和指標(磷流失風險評估)的圖形中，可以看出當Psat>14％、DPS-1>10％、DPS-2>40％時(如二水、台南、五魁寮、將軍系)，可能具有流失風險。而由吸附實驗0.2、0.3 mg P/L計算出推薦量對各萃取量與磷飽和指標作圖，以SigmaPlot內建雙曲線衰減或指數衰減曲線做非線性迴歸，R2值分別以Bray(0.66~0.77)和DPS-1(0.87~0.93)最高；再以0.2 mg P/L對應添加量與Bray萃取量作雙曲線衰減迴歸，由方程式: y=1572.43/(0.8642+x)，可得到特定萃取範圍的磷肥推薦量；並以此作為現行(依Bray No.1萃取量來分級的)磷肥推薦量之修正。

Recently, the research of soil phosphorus has been changed from availability indices and fertilizer recommendations to saturation indices and run-off risk assessments. A report pointed out the P-fertilizer recommendations calculated by Bray-P and sorption isotherm weren't fitted well with some acid soils in Taiwan. The subject of this article is comparing the data between P availabilty and saturation indices, Neubauer bioassay, and other elementary properties of eighteen Taiwan agricultural soils, and trying to build a proper standard with different P levels to fertilizer recommendations. The trend of Bray No.1 extractable P with different extraction time and soil/solution ratios seems to be relate to pH value and contents of P, Fe, and Al of soil samples. Parameters of Langmuir sorption isotherm equation were affected by the concentration of P addition, the number of data points, and the interval within two points. Wheat germination rates and amount of P uptake had no significant correlation in the Neubauer bioassay. Three extraction methods (Bray-P, 0.01M CaCl2-P, Oxalate-P) and two saturation indices (Psat, DPS) have significant correlation (P= 0.001); P uptake by wheat and other P indices also have significant correlation (P= 0.05). The assessments of P run-off risk by P saturation indices describe that when Psat>14％, DPS-1>10％, or DPS-2>40％, there may be some risk of those soil to liberate P to surrounding environment. Nonlinear regression between recommending P fertilizer application rates, which based on sorption isotherm (0.2, 0.3 mg P/L), and P availability and saturation indices shows that Bray, DPS-1, and the fertilizer recommendations have significant correlations. Thus, from the nonlinear regression equation: y=1572.43/(0.8642+x), we can get the fertilizer recommendations to specific range of Bray extractable P levels. This is an attempt to correct the fertilizer recommendations to Taiwan agricultural soils nowaday.