For their work in rethinking sustainable farming in Africa and understanding what drives behavioral change in addition, Willy Mulimbi and Peter Soyster, respectively, are each a winner of the 2026 NOMIS & Science Young Explorer Award.
Conventional tillage remains deeply embedded in African smallholder agriculture, despite its association with soil degradation, lowered productivity, and biodiversity loss. Willy Mulimbi's prize-winning essay examines how conservation agriculture, which emphasizes soil preservation and crop diversity, could improve agricultural sustainability and food security among African smallholder farmers. His research in the Democratic Republic of the Congo (DRC) shows that adopting these practices depends not only on economic and environmental conditions but also on access to training, credit, farmer networks, cultural traditions, and gender-related risks. Field experiments found that CA improved soil health and cowpea yields without requiring fertilizers or herbicides, while also reducing pressure on forests and potentially protecting women from dangers associated with traveling long distances to find farmland. However, urban consumers showed no significant willingness to pay more for sustainably produced crops, suggesting that wider adoption will depend on government support and education rather than consumer demand. Ultimately, Mulimbi argues that expanding CA requires addressing deeply rooted farming traditions and socioeconomic barriers to help farmers build more resilient and productive agricultural systems. "Resilience emerges not from forcefully shaping the land, but from working with natural processes," writes Mulimbi. "I seek to map how African farmers can transition from tradition-bound tillage to an ecological stewardship mindset to enable them to adapt to climate change, strengthen food security, and reclaim autonomy over their landscapes and livelihoods."
In the other prize-winning essay, Peter Soyster argues that traditional addiction research, which relies on group averages to develop treatments, often fails to account for the differences in individual behavior and real-world experiences. To address this, Soyster's research uses a precision-medicine approach to addiction psychology, combining frequent smartphone-based surveys with machine learning to develop predictive behavioral models tailored to each person. In a study of cannabis use, these individualized models predicted future use with substantially greater accuracy than conventional group-based models (average accuracy score of 0.87 versus 0.68), while revealing that the factors influencing substance use can vary considerably between individuals. Soyster suggests that personalized models could help clinicians identify specific behavioral triggers and develop more targeted interventions that move addiction treatment beyond one-size-fits-all strategies. "Decades of group-based research have given us a vital, generalized understanding of the possibility space of human psychology," Soyster writes. "By bridging the humanistic methods of community-partnered social science with the mathematical precision and personalized modeling of the life sciences, we can take the necessary next step."
