Opportunities and challenges of crowdsourcing in invasion science

My research leverages statistical models and massive crowdsourcing data to answer ecological questions with a focus on invasive plant species, from early detection to future dynamics forecast. Presence-only records arising from crowdsourcing projects (e.g. Pl@ntNet, iNaturalist) have a great potential to complete the existing biodiversity data. These open access and perennially stored records may be used to produce distribution assessments which can be updated in the future, even though it requires to deal with noises and biases. The spatio-temporal distributions of records can be used to inform bayesian dynamic models of invasions, which must disentangle ecological and observation processes. I’ll introduce several projects of my current PostDoc at Stellenbosch University which explore these opportunities and their challenges.