A satellite-based mobile warning system to reduce interactions with an endangered species

Area of Science:

• Ecology
• Conservation Biology
• Remote Sensing

Background:

• Earth-observing satellites are vital for ecosystem forecasting but face challenges with data gaps and latency.
• The Atlantic sturgeon (Acipenser oxyrinchus oxyrinchus), listed under the Endangered Species Act, requires effective conservation strategies.
• Integrating real-time data into actionable products is crucial for managing protected species and minimizing human impact.

Purpose of the Study:

• To develop an Atlantic sturgeon forecast system for the Delaware Bay, USA, by integrating satellite data with an existing species occurrence model.
• To overcome satellite data limitations (clouds, latency) using data interpolating empirical orthogonal functions (DINEOF) for a 3-day ocean condition forecast.
• To create a novel application of species distribution models with gap-filled satellite data for fish forecasting.

Main Methods:

• Implemented Data Interpolating Empirical Orthogonal Functions (DINEOF) to fill gaps in daily satellite data, creating a 3-day ocean condition forecast.
• Applied an Atlantic sturgeon research occurrence model to the DINEOF-processed data to predict species telemetry occurrences.
• Collaborated with state and federal agencies to develop real-time and forecasted risk maps for managers and commercial fishers.

Main Results:

• The developed forecast system accurately predicted Atlantic sturgeon occurrences over 90% of the time within a 3-day forecast.
• This represents the first known application of a species distribution model to DINEOF gap-filled data for fish forecasting.
• Real-time and forecasted risk maps were generated for the Delaware River Estuary, accessible via mobile devices and an interactive web interface.

Conclusions:

• The Atlantic sturgeon forecast system provides timely and actionable information for endangered species conservation and management.
• The system effectively reduces potential interactions between Atlantic sturgeon and human activities, including commercial fishing.
• This framework demonstrates a successful integration of remote sensing, data imputation, and species distribution modeling for applied ecological forecasting.