Audiomonitoring

Biodiversity - the variety of genes, species and ecosystems - is in crisis. Monitoring is key to better understand which biodiversity facets are particularly vulnerable to ongoing global change or how they benefit from conservation and restoration measures. How can we improve or even automate monitoring systems? An example with great potential are audio detectors that record biological sounds, which can then be compared with known sounds on the computer for species identification (with the help of artificial intelligence).

In our pilot project, we are testing the benefits and quality of automatic monitoring systems for recording different groups of species (birds, bats, insects). Also, we use the systems to answer scientific questions related to spatial and temporal dynamics of biodiversity, and the influence of various environmental conditions and disturbances. 

Project team: 

• Ecology & Macroecology lab: Damaris Zurell, Jette Reeg 
• Landscape ecology lab: Bertrand Fournier

Current/Planned pilot studies

• Spatial distribution of species and communities along habitat gradients (e.g. open land to forest, grassland to cropland) 
• Temporal changes in the annual cycle of species (e.g. timing of spring and autumn migration in migratory birds, start of breeding, activity peaks in insects) 
• Changes in species compositions over time (e.g. after habitat changes/improvements) 
• Comparison of different facets of biodiversity: species diversity vs. sound diversity

Further links:

• Related project: COMECO - Combining machine learning, ecological modelling and plant morphological characteristics to develop a tool for identifying plants of Switzerland