Application of Monte Carlo simulation in evaluating sunflower productivity in Tanzania: Policy Insights

Sunflower (Helianthus annuus L.) is a key oilseed crop for Tanzania’s edible-oil self-sufficiency agenda, yet national yields remain low and highly variable. This study applies a Monte Carlo–based Stochastic Simulation Approach to nationally representative data from the 2019/20 National Sample Census of Agriculture to quantify yield distributions and production risk across six farming systems distinguished by seed type and fertilizer use: (1) local seed–no fertilizer (LS_F0), (2) local seed–organic fertilizer (LS_F1), (3) local seed–inorganic fertilizer (LS_F2), (4) improved seed–no fertilizer (IMS_F0), (5) improved seed–organic fertilizer (IMS_F1), and (6) improved seed–inorganic fertilizer (IMS_F2). For each system, 500 Latin Hypercube iterations generated empirical probability density functions and Stoplight Charts to evaluate the likelihood of yields falling below 1.0 t ha⁻¹ or exceeding the national target of 2.0 t ha⁻¹. Results show that low-input production dominates: 87% of farms remain in LS_F0, where more than half face yields below 1.0 t ha⁻¹ and only 3% surpass 2.0 t ha⁻¹. Inorganic fertilizer significantly shifts the yield distribution upward. LS_F2 increases the probability of exceeding 2.0 t ha⁻¹ to 7 %, while IMS_F2 achieves the highest mean yield (1.14 t ha⁻¹) and the greatest likelihood (16 %) of surpassing the upper target, albeit with higher variability (CV ≈ 48 %). Improved seed without adequate nutrients provides only modest gains, and improved seed with organic fertilizer underperforms. These findings demonstrate that combining improved seed with inorganic fertilizer offers the clearest pathway to Tanzania’s sunflower productivity goals, though risk-management measures and region-specific strategies are essential. The analysis provides a national baseline for future agro-ecological and policy-focused research on agricultural transformation.