Perspective Chapter: Insect Conservation, Machine Learning, and Citizen Science Take Flight

Abstract

Insect species are likely declining, resulting in an urgent need for more conservation and management action to maintain ecosystem function and human health. Inexpensive community scientists and mechanical sensors are accelerating data acquisition in insect ecology. These data have a great potential to help inform insect conservation and management decision making, but current approaches and training limit the utility and impact of this potential. Careful application of machine learning will likely improve the speed, efficacy, and reproducibility of insect ecology workflow and hopefully conservation efforts, specifically in insect monitoring, species identification and validation, and ecological modeling. Of course, machine learning will not be a panacea for all things that ail us and continued work on taxonomy, species identification, and sampling will continue. Regardless, the addition of machine learning to the insect ecologist tool kit is critical to help conserve and manage various insect species in a quickly changing world.