Soil chemical fertility has steadily declined in tropical and subtropical agriculture with depleted stocks of phosphorus (P), nitrogen (N), and carbon (C). Assessing the dynamics of these elements and their interactions on crop productivity in dryland cropping are complex because climate often dictates crop nutrient response. This results in under- or over- fertilising crops, suboptimal crop yield, and fertiliser inefficiency. The Agricultural Productions Systems sIMulator (APSIM) model accounts for C x N x climate interactions, but simulation of P dynamics is constrained by a dearth of suitable data. To address this problem, we used a novel approach to simulate P, N, and C dynamics at a 35-year long-term field trial, where a broad range of N (0, 40, 80, 120 kg ha⁻¹) and P (0, 10, 20 kg ha⁻¹) fertiliser rates were consistently applied. We parameterised the soil P model with quantified adsorption isotherms and by assuming correspondence between conceptual soil P pools and Hedley fractionation pools. Soil N and C dynamics were parameterised with measured organic N, C, and charcoal content to estimate organic matter decay coefficients, pool sizes, and C:N ratios. APSIM accounted for variation in mean N export (94%), crop yield (88%), and P export (62%) across the 12 treatments, and reproduced interannual variation in N × P effects for crop yield and N export, where crop response was strongly mediated by N supply and water availability. APSIM also identified the long-term depletion or accumulation of soil P, N, and C in most treatments. P fractionation and isotherm measurements are labour intensive but worthwhile, and future efforts should work to consolidate a database for different soil types. Better informed P modelling will provide insights into the effects of climate variability on soil fertility and crop productivity, and guide management practices to deliver better fertiliser efficiency and maintain soil organic C.

热带和亚热带农业的土壤化学肥力稳步下降，磷 (P)、氮 (N) 和碳 (C) 储量枯竭。评估这些元素的动态及其对旱地种植作物生产力的相互作用是复杂的，因为气候通常决定了作物的养分反应。这会导致作物施肥不足或过度施肥、作物产量欠佳和肥料效率低下。农业生产系统模拟器 (APSIM) 模型考虑了 C x N x 气候相互作用，但 P 动力学的模拟受到缺乏合适数据的限制。为了解决这个问题，我们在一项为期 35 年的长期现场试验中使用了一种新方法来模拟 P、N 和 C 动力学，其中 N（0、40、80、120 kg ha ^-1）和P (0, 10, 20 公斤公顷^-1) 施肥量始终如一。我们用量化的吸附等温线参数化土壤 P 模型，并假设概念土壤 P 池和 Hedley 分馏池之间的对应关系。土壤 N 和 C 动态通过测量的有机 N、C 和木炭含量进行参数化，以估计有机质衰减系数、池大小和 C:N 比率。APSIM 解释了 12 种处理的平均 N 输出 (94%)、作物产量 (88%) 和 P 输出 (62%) 的变化，并再现了作物产量和 N 输出的 N × P 效应的年际变化，其中作物氮的供应和水的可用性强烈地介导了响应。APSIM 还确定了大多数处理中土壤 P、N 和 C 的长期消耗或积累。P 分馏和等温线测量是劳动密集型的，但值得，未来应该努力整合不同土壤类型的数据库。更明智的 P 模型将深入了解气候变化对土壤肥力和作物生产力的影响，并指导管理实践以提供更好的肥料效率和保持土壤有机 C。