Data or statistical facts on the situation and perspectives of agri-food systems and the impact of policies
CAF and Oxford presented a report at COP30 proposing the use of science and bio-inspired innovation to turn biodiversity into an engine of development (CAF, 2025).
The study assessed how precision technologies—such as sensors, GIS, and drones—relate to agricultural input management in the San Martín region of Peru. Based on surveys of 120 producers, it found that productivity and sustainability positively influence efficient input management, while costs, limited infrastructure, and resistance to change are the main barriers to their adoption
0.006 was the average DID effect on mean NDVI for citrus, indicating productivity improvement.
The combined strategy of using surveys and satellite images allows for measuring mechanisms (adoption, credit) and returns, offering consistent evidence.
The monthly NDVI signal shows greater variation in annual crops such as cotton and greater stability in perennial crops such as citrus.
Annual NDVI (average and maximum) is used to observe pre- and post-intervention trends for treated and control units. The longitudinal approach captures the "before and after" dynamics of the program.
NDVI-based remote sensing is a valuable, cost-effective proxy for yield to assess temporal dynamics of impacts.
262 (cotton) and 190 (citrus) plots were georeferenced to build NDVI series.
The study conducts a comprehensive review of the use of predictive models to strengthen food security in Latin America, with an emphasis on Colombia. It analyzes 65 studies published between 2014 and 2024 that apply algorithms such as Random Forests and Neural Networks to anticipate agricultural yields, manage climate risks, and optimize resources. The results show that agricultural, climatic, and...
The study conducts a systematic review of the methods applied in prospective studies for strategic planning of agri-food chains in Latin America and the Caribbean between 2010 and 2021. It examines the most widely used methodologies (such as MACTOR, MICMAC, and Delphi), their contributions to decision-making, and the challenges that remain in institutionalizing regional foresight. It also highligh...
