Hydrophone recordings from three aquatic beetle species within the family Hydrophilidae, Berosus pantherinus (LeConte), Tropisternus blatchleyi (d'Orchymont), and Tropisternus collaris (F.), were used to create an automated identification program. The identification program was designed using digital signal processing techniques and is capable of identifying which species is present from hydrophone recordings alone. Using features based on the frequency content of each beetle call and background sounds in reference recordings, half-second segments of audio recordings were classified as a specific beetle species call, a general beetle distress call, or as noise. Classification accuracy ranged from 87.5% for reference recordings. Training calls had accuracies of 98% and above between beetle species and noise. The majority of beetle recordings in a mesocosm environment were classified correctly. Often, T. blatchleyi was false-positively identified in recordings with non-beetle background noises, such as frogs or traffic, suggesting similar active frequencies in the T. blatchelyi features and these noises. The use of digital signal processing to identify aquatic invertebrates by sound is a new technique that has potential uses in taxonomy, surveys, and long-term biomonitoring of aquatic systems by providing a hands-free method of detection, and eventually identification, in the field.
You have requested a machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Neither BioOne nor the owners and publishers of the content make, and they explicitly disclaim, any express or implied representations or warranties of any kind, including, without limitation, representations and warranties as to the functionality of the translation feature or the accuracy or completeness of the translations.
Translations are not retained in our system. Your use of this feature and the translations is subject to all use restrictions contained in the Terms and Conditions of Use of the BioOne website.
Vol. 69 • No. 2