Machine-learning to enhance RBG Kew’s plant conservation work

Where can the use of machine learning most enhance our plant conservation work? RBG Kew is exploring new approaches to prioritise conservation actions and speed up time-consuming analytical stages of the seed conservation process. Building on work to automate the Least Concern preliminary conservation assessments under the IUCN red list, RBG Kew authors have produced species-level threat estimates for angiosperms helping prioritise full Red List assessment and fast-tracking species for Least Concern assessments. At a regional level, an ex situ conservation gap analysis of the Sonoran Desert ecoregion has identified taxonomic and geographical priorities for collecting, and has automated creation of potential distribution maps that could support field research. At the Millennium Seed Bank, RBG Kew is seeking to (1) improve speed and consistency of assessment of X-ray images for purity analysis and (2) enhance analysis of Tetrazolium staining using machine learning approaches.