Invasive alien species respond to biologically-inspired robotic predators

摘要

Invasive alien species threaten natural ecosystems worldwide, prey on native species, and deplete their food sources. Mosquitofish is one of the most invasive freshwater fish worldwide and its negative impacts on the native fauna are alarming. Despite the urgency of contrasting the mosquitofish invasion, we have access to very few methods to combat them. Even when successful, these methods can be excessively labor-intensive or dangerous to native species. Robotic predators may constitute a promising tool in combating mosquitofish. Our group has recently proposed the use of a robotic predator that can perform targeted attacks against mosquitofish. The robotic predator consists of three operational parts: a two-dimensional robotic platform, a magnetically connected replica of a native mosquitofish predator, and an in-house developed live tracking software. The robotic replica was programmed to swim along a predetermined trajectory and randomly target mosquitofish in real time through a dedicated tracking software. Building on available experimental results, we put forward a comprehensive mathematical toolbox based on symbolic dynamics, recurrence quantification, and information theory to detail the behavioral interaction between the robotic predator and mosquitofish.