HerbSpectra-Amazon
Collaborative network for the construction of spectral herbaria of the Amazon. Image credit © 2025 Biodivspec group.
Amazon Spectral Herbarium Network: Spectroscopy and artificial intelligence for the automated identification of biodiversity in Amazonian herbaria - HerbSpectra-Amazon
Description
Innovative technologies applied to botany are essential for the conservation and sustainable use of biodiversity, especially in megadiverse regions such as the Amazon. The correct identification of species is a basic requirement for initiatives in biotechnology, restoration, and environmental monitoring. In the Amazon, this process depends on herbariums and traditional taxonomy, which face limitations given the high diversity and morphological complexity of species. Visible and near-infrared spectroscopy has proven to be a promising tool for overcoming these challenges, allowing species to be identified quickly and non-destructively. This proposal aims to modernize botanical taxonomy in the Amazon region, making it more efficient, accessible, and integrated with digital technologies. We hypothesize that it is possible to create a network of standardized spectral herbaria in the Amazon and apply them in Artificial Intelligence (AI) systems for species identification. We count on the collaboration of sixteen national and international institutions. There are ten executing herbaria, five of which are Brazilian (AM, AC, RR, and two in PA) and five from ACTO countries (Bolivia, Colombia, Peru, and two in Ecuador), all with a high level of interest in developing this proposal: spectrally digitizing their collections, building spectral herbaria, and applying AI to biodiversity recognition. The methodology involves consolidating a network of Amazonian herbaria trained in standardized spectral reading of herbarium specimens with a MicroNIR (900–1700 nm). This data will be used to train AI algorithms to automatically identify species in different regions of the Amazon. The other six collaborating institutions will provide support to advance the spectral network’s activities technologically. The proposal also includes training, strengthening regional science, and modernizing the Amazon’s biological collections.
Find out more on the HerbSpectra-Amazon website.
Related link(s)
Caroline C. Vasconcelos
Research Fellow
My research interests include taxonomy and systematics (especially neotropical Sapotaceae), spectroscopy as a integrative tools, Amazonian flora, species distribution modeling, floristic studies, and tropical forest ecology.