Conservation agriculture practices are promoted to increase productivity, profitability, and sustainability across diverse cropping systems. Many studies have used these goals in decision support frameworks to identify the most effective treatment among those examined. While this approach is valuable, it lacks actionable guidance for farmers regarding maximizing return, while minimizing risk. It does not provide specific recommendations on how to allocate land across various cropping systems and tillage practices to achieve such objectives. This would require another long-term experiment exploring various combinations of treatments. To address this challenge, we propose the application of modern portfolio theory, specifically leveraging mean-variance and conditional value at risk optimization models. Using these models has enabled us to identify the optimal cropping system combinations with different tillage practices that maximized yield and net returns with minimal associated risk. The proposed approach allows for recommendations involving combinations of treatments that may not have been previously tested in a geography. In a 14-year long-term conservation agriculture study involving twelve combination of tillage and cropping systems, we showed how different combination of treatments differ in risk-return profile using mean-variance and conditional value-at-risk models that trace out a frontier of options—combinations of treatments that give highest returns at minimal risk. For example, we find that across risk neutral (most profitable) and most risk averse (lowest risk) farmers, the optimal treatments on the frontier encompass of maize-mustard-mungbean (MMuMb) under zero tillage and maize-wheat-mungbean (MWMb) under bed planting (which offer high returns and associated risk), maize-maize-Sesbania (MMS) under zero tillage (providing a balance of moderate returns and risk), and MMS under conventional tillage (yielding lower returns and risk). Additionally, risk-averse farmers stand to gain by diversifying their land allocation. For instance, they could allocate 54 % of their land to MMuMb under zero tillage and 46 % to MWMb under bed planting to target net returns of INR 1,32,000, with downside risk of INR 56,000, otherwise they can allocate 44 % and 56 % of their land to MMS under zero tillage and MWMb under bed planting, respectively, with a targeted net return of INR 1,22,000 and downside risk of INR 43,540. This highlights the nuanced trade-off between risk and return in maize based diversified cropping systems under different tillage practices. Leveraging mean-variance and conditional value at risk optimization models in the analysis of long-term experiments can yield novel treatment combinations that hold promise and can be recommended to farmers for implementation.
Read full abstract