ABSTRACT The practice, geography and nature of tropical ecology have shifted substantially over past decades, but existing syntheses rarely resolve long‐term, pantropical patterns across biomes. This gap is partly due to the challenge of analyzing vast and rapidly growing literature. The advance of large language models enables such analysis for the first time. We analyzed over 170,000 tropical ecology publications (1960–2023) to quantify long‐term research trends across five major biomes, identify the most studied countries and leading contributors (based on first‐author affiliations), and evaluate their relationships with national socio‐economic indicators. Using GPT‐4o mini to extract contextual location and biome data, we found that research output has increased across all biomes, but the rate of increase has been highest for tropical forests and lowest for grassy biomes, resulting in a growing dominance of forests and a relative decline of grassy ecosystems. Tropical coastal and freshwater biomes have seen modest relative increases, while oceanic biomes remain stable in relative effort. This relative decline is particularly concerning for Africa, where savannas cover ~70% of tropical land. Since 2010, 73% of tropical ecology papers have come from research hosted in ten tropical countries, led by Brazil (25%). China, India, Mexico, Indonesia and Malaysia are slowly rising to be powerhouses of tropical ecology. Conversely, early research featured broader African representation. This geographical disparity is more closely tied to socio‐economic capacity than biome extent. Our analysis highlights both this encouraging shift and the remaining disparities, underscoring the need to strengthen research capacity in underrepresented tropical regions.
Moorthy et al. (Fri,) studied this question.