Although the variable-rate manure application (VRMA) offers a promising solution to overcome the shortcomings of the conventional uniform-rate (UR) application, effectively incorporating this technique into precision agriculture solutions presents a substantial challenge. This study entailed developing a quantitative analysis tool to calculate manure consumption, evaluate environmental risks, and compare different treatment schemes under different implementation scenarios. Firstly, a simulator was designed and developed in the LabVIEW development environment with MATLAB. Next, this tool was used to compare traditional UR application with VRMA, by which manure application rate was varied according to soil phosphorus (P) maps measured with an on-line visible and near infrared soil sensor, or according to data-fusion (DF)-based management zone (MZ) maps, developed by fusion of on-line measured soil properties and crop normalized difference vegetation index (NDVI). In the DF case, simulations included Robin Hood (RH) and Kings approaches, which meant to apply the largest manure application rate at the poorest fertility zones, or in the richest fertility zones, respectively. The implementation scenarios included both map-based (MB) without accounting for on-line manure sensing and map-sensor-based (MSB) with on-line manure sensing in three commercial fields in Flanders, Belgium. The results revealed that the P-based-VRMA scheme consumed 1–13% less nitrogen (N) and P₂O₅ in the case of ignoring legislative limits, compared to the UR and this was true for the RH-VRMA scheme in all three fields and for the Kings-VRMA scheme in two out of three fields. This was because the P-based maps have larger rich fertility areas (e.g., large P concentration) than the MZ maps that both the RH and Kings approaches used. Furthermore, imposing the “MAP6” legislation limits caused all VRMA schemes to consume less fertilizer than would be the case without restrictions, reducing environmental risks due to decreasing the amount of applied N and P₂O₅. According to the simulation results, the MSB scenario saved 6–9% manure compared to the MB scenario in the P-based-VRMA scheme when the mean value of real-time sensed P₂O₅ in the applied manure was greater than the nominal values measured in the laboratory. Opposite results were observed in the UR, RH-, and Kings-VRMA schemes. The MSB was more expensive than the MB by 5–7%, when the mean N content of manure used for MSB was lower than that for MB. We concluded that the manure consumed in VRMA depends on treatment schemes, measured manure quality, imposing legislative limits, and the proportion of rich, medium, and poor fertility areas in a field. It is essential to impose the legislative limits to prevent over-applications of N and P₂O₅, while further agronomy study should evaluate the impact of these limits on crop responses.

尽管可变施肥 (VRMA) 提供了一种有前途的解决方案，可以克服传统均匀施肥 (UR) 施肥的缺点，但将这种技术有效地融入精准农业解决方案是一个巨大的挑战。本研究需要开发一种定量分析工具来计算粪便消耗量、评估环境风险并比较不同实施方案下的不同处理方案。首先，在LabVIEW开发环境下，利用MATLAB设计开发了一个模拟器。接下来，该工具用于将传统的 UR 应用与 VRMA 进行比较，其中肥料施用率根据使用在线可见光和近红外土壤传感器测量的土壤磷 (P) 图而变化，或根据基于数据融合（DF）的管理区（MZ）地图，通过融合在线测量的土壤特性和作物归一化差异植被指数（NDVI）开发。在 DF 案例中，模拟包括 Robin Hood (RH) 和 Kings 方法，这意味着分别在最贫瘠的肥沃区或最肥沃的肥沃区应用最大的肥料施用率。实施方案包括在比利时法兰德斯的三个商业领域中，不考虑在线粪便传感的基于地图（MB）和基于地图传感器（MSB）的在线粪便传感。结果表明，基于 P 的 VRMA 方案消耗的氮 (N) 和 P 模拟包括 Robin Hood (RH) 和 Kings 方法，这意味着分别在最贫穷的肥力区或最富有的肥力区应用最大的肥料施用率。实施方案包括在比利时法兰德斯的三个商业领域中，不考虑在线粪便传感的基于地图（MB）和基于地图传感器（MSB）的在线粪便传感。结果表明，基于 P 的 VRMA 方案消耗的氮 (N) 和 P 模拟包括 Robin Hood (RH) 和 Kings 方法，这意味着分别在最贫穷的肥力区或最富有的肥力区应用最大的肥料施用率。实施方案包括在比利时法兰德斯的三个商业领域中，不考虑在线粪便传感的基于地图（MB）和基于地图传感器（MSB）的在线粪便传感。结果表明，基于 P 的 VRMA 方案消耗的氮 (N) 和 P₂ O ₅在忽略立法限制的情况下，与 UR 相比，这对于所有三个领域的 RH-VRMA 计划和三分之二的 Kings-VRMA 计划都是如此。这是因为基于 P 的地图比 RH 和 Kings 方法都使用的 MZ 地图具有更大的富肥区（例如，大的 P 浓度）。此外，实施“MAP6”立法限制导致所有 VRMA 计划消耗的肥料比没有限制的情况下消耗更少，从而降低了由于 N 和 P ₂ O ₅的施用量而导致的环境风险。根据模拟结果，当实时感测的 P ₂施肥中的O ₅大于实验室测量的标称值。在 UR、RH- 和 Kings-VRMA 方案中观察到相反的结果。当用于 MSB 的粪肥的平均 N 含量低于 MB 时，MSB 比 MB 贵 5-7%。我们得出的结论是，在 VRMA 中消耗的粪便取决于处理方案、测量的粪便质量、施加的立法限制以及田地中肥沃、中等和贫瘠肥沃区域的比例。必须施加立法限制以防止过度施用 N 和 P ₂ O ₅，而进一步的农学研究应评估这些限制对作物反应的影响。