The scientific value of Amazonian protected areas

Protected areas (PAs) are under increasing pressure to demonstrate their broader value and contribution to society. Scientific research and associated knowledge production comprise one such value, which has received relatively little attention in the academic literature. Here, we use the Amazon region as a case study to quantify scientific knowledge production (as measured by scientific publications in peer-reviewed journals) in PAs and identify the main biophysical, geographical and social characteristics that influence such production. We adopt a multi-model inference approach with an innovative hurdle regression model to independently assess the factors influencing the presence of research and the number of studies in PAs. Our results indicate a highly skewed pattern of scientific production, with many PAs with few or no associated scientific articles. Larger, older and more highly protected PAs in Ecuador and Peru were most likely to have scientific production, while time since first publication was most strongly associated with the number of publications from a PA. These findings provide important insights that could be used to support and strengthen policy aimed at increasing the value of Amazonian protected areas for scientific research.