Resolving the identification of weak-flying insects during flight: a coupling between rigorous data processing and...

Area of Science:

• Agricultural Entomology
• Bioacoustics
• Insect Pest Monitoring

Background:

• Bioacoustic methods are vital for insect surveillance but struggle with weak-flying species.
• These insects, including aphids and beetles, pose significant threats to global food security as agricultural pests.
• Automatic detection is essential for effective crop protection strategies.

Purpose of the Study:

• To develop a novel bioacoustic approach for detecting and identifying weak-flying insect pests.
• To extract and analyze audio features for species-level discrimination.
• To improve early pest risk assessment for agricultural crops.

Main Methods:

• An experimental setup utilizing an LED array to derive audio recordings from weak-flying insects.
• Development of a data processing pipeline to extract bioacoustic features from audio and harmonic series.
• Utilizing over 50 bioacoustic parameters, including dominant and fundamental frequencies, for species discrimination.

Main Results:

• Successful species discrimination with an 80% success rate for six aphid and two beetle species.
• Improved differentiation between beetles and aphids by incorporating wingbeat frequencies.
• Minimized error rates at the species level by combining harmonic features with flight energy indicators.

Conclusions:

• The described bioacoustic method effectively detects and discriminates between weak-flying insect species.
• This technology offers a promising tool for automated pest monitoring and crop protection.
• Enhanced accuracy in pest identification contributes to safeguarding agricultural production and food security.