Stony coral tissue loss disease (SCTLD) is a lethal disease affecting over 20 coral species throughout the Caribbean. To enable future coral restoration, the Florida Fish and Wildlife Conservation Commission (FWC) and research partners began removing susceptible corals from the Florida Reef Tract and placing them in aquaria before SCTLD spread throughout the region. However, basic coral spawning information that is critical to restoration success, such as timing and duration, is not known for many of the coral species susceptible to SCTLD. For example, an observer from the Florida Aquarium unexpectedly identified Mycetophyllia lamarckiana spawning for the first time in March 2020, even though many other coral species spawn in late summer.

The process of monitoring and observing potential spawning is labor intensive given the uncertainty in timing and duration, and staff-time is valuable given the magnitude of the SCTLD response. New technologies, such as artificial intelligence algorithms, can identify and inform users of coral spawning from video feeds. Artificial intelligence (AI) refers to computer algorithms that perform complex tasks that previously required human expertise and are designed to address many problems conservationists experience.

The objective of this project was to improve an existing camera-based monitoring system and reduce partner staff time by developing an automated coral spawning alert system. An anomaly algorithm was created to detect coral spawning from real-time underwater video in aquaria, and a pipeline that alerts users to coral spawning and maintains summary information on spawning events (e.g., timing and duration).