.png){kind=icon}
返回
Understanding differences between static and dynamic nitrogen fertilizer tools using simulation modeling
.png){kind=icon}
PRE
AI
DOI：10.1016/j.agsy.2021.103275.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
CONTEXT: Improving nitrogen (N) fertilizer recommendations for maize (Zea mays L.) in the US Midwest has been the focus of much research, yet there is no agreement for which methodology is the best to balance trade-offs between production and environmental outcomes. This study investigated the strengths and limitations of two broad approaches: dynamic and static recommendation tools. Dynamic tools use advanced technology to predict the Economically Optimum N Rate (EONR) using year-specific soil, weather, and crop growth charac-teristics to detect conditions that need lower or higher N rates. Static tools provide regional N recommendations that are static over time, maximizing long-term profits rather than predicting the best EONR for each field and season. OBJECTIVE: The objective of this work was to explain the interactions between the accuracy, profitability, and environmental losses for different N recommendation tools under a wide range of production scenarios. METHODS: For this, we used a calibrated synthetic dataset of 4200 fields over 30 years. In the first part, we compared multiple N recommendations tools belonging to the static and dynamic groups. In the second part, we selected each group's best tools and compared them in detail. RESULTS AND CONCLUSION: From an economic view, results indicate that increasing profitability by increasing the accuracy in EONR predictions with dynamic tools is challenging. The reason is that these more accurate tools are not perfect, and around half of the time, they under predict. In that situation, the yield penalty is higher, and the economic loss is usually not compensated by savings in N fertilizer costs associated with other more accurate recommendations. The static recommendations avoid this penalty by recommending slightly higher N rates, providing similar profits. From an environmental view, both tools can reduce N leaching by 15%, dynamic tool by being more accurate overall, but the static tool could achieve it by recommending on the lower end of their current recommended N profitable range. SIGNIFICANCE: Our analysis suggests that we need to re-think the goals of N management tools. Higher complexity in N management may not necessarily increase profits and reduce N leaching. In fact, there is current potential to reduce N leaching by simply reducing static recommendations without hurting profits. For either approach, this study highlights the need to develop other ways (education, environmental awareness, policies) to account for environmental benefits and provide clear incentives for farmers to adopt these tools and increase the eco-efficiency of agriculture.
Keyword:
Crop modeling
Machine learning
Environmental indicators
Economic analysis
Maize
Nitrogen fertilizer
期刊
IF:
6.1
论文数:
4.0K
被引数:
1.4W
机构
引用论文
.png){kind=icon}
Monitoring US agriculture: the US Department of Agriculture, National Agricultural Statistics Service, Cropland Data Layer Program监测美国农业: 美国农业部，国家农业统计局，耕地数据层计划
PREAI
分享收藏
查看更多内容