Global yield responses of cereals to biostimulants under field conditions: Insights from a random forest meta-analysis

Context: Biostimulants represent a promising strategy for improving crop productivity in a sustainable manner, yet their effectiveness is often inconsistent due to heterogeneous field conditions. Research question: Understanding how environmental and management factors influence biostimulant performance is crucial for guiding their practical use. Method: A systematic literature review was conducted, focusing on the three major cereals (rice, wheat, and maize) and including only experiments carried out under open-field conditions. Results: The analysis revealed that many published studies lacked key information on yield variability measurement. Nevertheless, the metaforest package identified several agro-environmental and treatment descriptors parameters that had a consistent impact on crop responses. The most influential factors included mineral and organic fertilization doses, Köppen–Geiger climate class, soil type, soil pH, crop species, soil organic matter, and the physical state of the applied formulation. In contrast, specific formulation related parameters were not relevant in modelling yield outcomes. The model output highlighted different patterns. Biostimulants, generally, were found to offer the most significant yield benefits in suboptimal growing conditions. Reduced fertilization enhanced the positive effects of biostimulants, suggesting that their contribution is enhanced when nutrient inputs are limited. Similarly, there was a more noticeable increase in yields under extreme climate conditions, compared to suitable climate classes for cereal production. Soil conditions also played a role: alkaline pH and organic matter levels of around 2% were associated with stronger responses to biostimulant application. Among the analysed crops, the model output estimated a median yield gain of 13.3% for rice, 13.2% for wheat and 11.6% for maize. On the other hand, type, and formulation state were less important in defining plant biostimulants efficacy. Conclusions: Overall, these results show that the success of biostimulant applications is driven more by agro-environmental context than by product formulation. Biostimulants appear particularly effective when crops are exposed to stress or non-optimal growing conditions, highlighting their potential as tools to stabilize yields in challenging environments. Implications: This meta-analysis provides new evidence to support the targeted use of biostimulants and offers practical guidance for farmers, agronomists, and policymakers aiming to integrate biostimulants into sustainable cereal production systems.