Soil available P for crops, derived from P inputs and regulated by soil chemical, physical and biological factors, represents a small fraction of the total P content. Soil P balance (P input minus P output) is potentially a good descriptor of fluctuations in the available P pool, making it possible to compare experiments that differ in management and duration. Using a data set compiled from three long-term experiments in NW Italy that tested different P fertilization types (nil, mineral, organic and manures), cropping systems and P balances in different soils, this paper focused on (a) proposing an empirical exponential model to describe the relationship between variables and (b) providing an agronomical interpretation of optimized parameter values, to set the basis for a unified interpretation of long-term dynamics of soil plant-available P. The pooled data fit well in an exponential model that identified three factors supportive of an agronomic interpretation, two linked to soil attributes and one that differentiated fertilizer types. Soil characteristics were found to affect minimum soil Olsen P values in depletion and response rates with additions, while fertilizer type had a multiplier effect on cumulative P balance. Farmyard manure was shown to build the available P pool better than other organic sources (bovine slurry, commercial organic fertilizers, green manure) and inorganic fertilizers, owing to a high C/P. The exponential model provided a good theoretical basis, but further studies are needed to correlate parameter values to soil (e.g., carbonates, Al and Fe oxide content) and fertilizer (e.g., C:P, lignin:P) characteristics and to verify the model applicability to different conditions.

中文翻译：

---

意大利西北部长期试验中土壤奥尔森磷对不同磷肥策略的反应

作物的土壤有效磷来源于磷的投入并受土壤化学、物理和生物因素的调节，仅占总磷含量的一小部分。土壤 P 平衡（P 输入减去 P 输出）可能是可用 P 池波动的一个很好的描述，从而可以比较管理和持续时间不同的实验。使用从意大利西北部的三个长期实验中收集的数据集，这些实验测试了不同土壤中的不同磷肥类型（零、矿物、有机和粪肥）、种植系统和磷平衡，本文重点关注 (a) 提出经验指数模型来描述变量之间的关系和（b）提供优化参数值的农艺解释，为土壤植物有效磷的长期动态的统一解释奠定基础。汇总数据非常适合指数模型，该模型确定了支持农艺解释的三个因素，两个与土壤属性相关，一个区分肥料类型。发现土壤特性会影响消耗的最小土壤 Olsen P 值和添加的响应率，而肥料类型对累积 P 平衡具有乘数效应。由于 C/P 高，农家肥显示出比其他有机来源（牛粪、商业有机肥料、绿肥）和无机肥料更好地构建可用 P 池。指数模型提供了良好的理论基础，但需要进一步研究将参数值与土壤（例如碳酸盐、Al 和 Fe 氧化物含量）和肥料（例如 C:P、木质素）关联起来：