Intervening into agricultural systems necessarily includes risks, uncertainties, and ultimately unknown outcomes. Decision analysis embraces uncertainty through an interdisciplinary approach that involves relevant stakeholders in evaluating complex decisions. We applied decision analysis approaches to prioritize 21 farm management interventions, which could be considered in certification schemes for banana production. We estimated their contribution to climate change adaptation and mitigation as well as ecological outcomes. We used a general model that estimated the impacts of each intervention on adaptation (benefits minus costs), mitigation (global warming potential), ecological parameters (e.g., biodiversity and water and soil quality), and farming aspects (e.g., yield, implementation costs and production risks). We used expert and documented knowledge and presented uncertainties in the form of 90% confidence intervals to feed the model and forecast the changes in system outcomes caused by each intervention compared to a baseline scenario without the measure. By iterating the model function 10,000 times, we obtained probability distributions for each of the outcomes and farm management interventions. Our results suggest that interventions associated with nutrient management (e.g., composting and nutrient management plan) positively affect climate change adaptation, mitigation, and ecological aspects. Measures with no direct yield benefits (e.g., plastic reduction) correlate negatively with adaptation but have positive impacts on ecology. Creating buffer zones and converting low-productivity farmland (incl. unused land) also have positive ecological and adaptation outcomes. Decision analysis can help in prioritizing farm management interventions, which may vary considerably in their relationship with the expected outcomes. Additional work may be required to elaborate a comprehensive assessment of the underlying aspects modulating the impacts of a given measure on the evaluated outcome. Our analysis provides insights on the most promising interventions for banana plantations and may help practitioners and researchers in focusing further studies.

干预农业系统必然包括风险、不确定性和最终未知的结果。决策分析通过涉及相关利益相关者评估复杂决策的跨学科方法来包含不确定性。我们应用决策分析方法对 21 项农场管理干预措施进行优先排序，这些措施可以在香蕉生产认证计划中加以考虑。我们估计了它们对气候变化适应和减缓以及生态结果的贡献。我们使用了一个通用模型，估计了每种干预措施对适应（收益减去成本）、减缓（全球变暖潜能值）、生态参数（例如，生物多样性和水土质量）和农业方面（例如，产量、实施成本）的影响和生产风险）。我们使用专家和书面知识，并以 90% 置信区间的形式呈现不确定性，以提供模型并预测与没有测量的基线情景相比，每种干预措施导致的系统结果变化。通过将模型函数迭代 10,000 次，我们获得了每个结果和农场管理干预措施的概率分布。我们的研究结果表明，与养分管理相关的干预措施（例如堆肥和养分管理计划）对气候变化的适应、缓解和生态方面产生积极影响。没有直接产量效益的措施（例如减少塑料）与适应呈负相关，但对生态产生积极影响。创建缓冲区和转换低生产力农田（包括 未利用的土地）也具有积极的生态和适应成果。决策分析有助于确定农场管理干预措施的优先级，这些干预措施与预期结果的关系可能存在很大差异。可能需要额外的工作来详细说明对调节给定措施对评估结果的影响的潜在方面的全面评估。我们的分析为香蕉种植园最有希望的干预措施提供了见解，并可能有助于从业者和研究人员集中进一步研究。可能需要额外的工作来详细说明对调节给定措施对评估结果的影响的潜在方面的全面评估。我们的分析为香蕉种植园最有希望的干预措施提供了见解，并可能有助于从业者和研究人员集中进一步研究。可能需要额外的工作来详细说明对调节给定措施对评估结果的影响的潜在方面的全面评估。我们的分析为香蕉种植园最有希望的干预措施提供了见解，并可能有助于从业者和研究人员集中进一步研究。