Feasibility of visual signals on the construction of biosensors based on behavioral analysis of Perna perna mussels

Bivalve mollusks are well known for their sentinel characteristics being sensitive to environmental changes. Considering the capability of closing their valves to isolate their soft tissues from the aquatic medium, the analysis of bivalves behavior has been used in the construction of aquatic pollution biosensors. Hall effect sensors are a well-established and widely used valvometry method and present several advantages. However, its use still requires fixing components in the animal shells. The present study has investigated the feasibility of using visual signals for the behavioral analysis of Perna perna mussels. In this sense, it has described a computer vision algorithm for the analysis of mussels valve-activity. Moreover, the valve-activity responses obtained through the computer vision algorithm have demonstrated a very similar pattern to the signal obtained with Hall effect sensors. In conclusion, the use of visual signals in the behavioral monitoring of Perna perna mussels have proved to be feasible. Furthermore, results indicated that the study of more robust computer-vision techniques may lead to the construction of totally non-invasive biosensors.