Speaker
Description
Insect pollinators play a vital role in maintaining biodiversity and ecosystem functions via their interactions with plants. Standardized monitoring methods capturing plant-pollinator interactions are therefore indispensable to create effective conservation strategies, in the face of accelerating global biodiversity loss. Traditional monitoring methods are often limited in their temporal and spatial scale, labor-intensive, and costly. Automated image-based monitoring systems can offer an efficient and scalable alternative.
While most research and monitoring have focused on diurnal pollinators such as bees, nocturnal pollinators, especially moths, play a crucial yet often overlooked role in pollination. Moths visit and pollinate a diverse range of plant species worldwide, with many plants benefiting from a combination of both diurnal and nocturnal visitors. However, studying moths and other nocturnal pollinators remains challenging due to the limitations of existing monitoring methods.
Here, we present a novel approach to monitoring nocturnal plant-pollinator interactions using an AI-enabled automated camera trap with an infrared camera. Unlike traditional methods that actively attract insects with UV light, our system passively detects flower-visiting insects using an object detection model trained on grayscale images. When an insect is detected, an LED is activated to provide sufficient illumination for RGB image acquisition, minimizing disturbance and active attraction. This setup can enable the documentation of natural pollination behavior and interaction analysis.
By integrating AI-based detection with minimally invasive imaging, this method complements existing diurnal datasets and helps expand our understanding of nocturnal plant-pollinator networks.
| Status Group | Master Student |
|---|---|
| Poster Presentation Option | Yes, I’m willing to present as a poster. |
