In the forest (photo Vlad Barza)
Artificial intelligence techniques have been used in a lot of fields and a lot of interesting things have been done with their help. Here, the AI has become very useful in the woods, to discover when there are the highest chances to hear the song of dozens of bird species.
A team from the University of Göttingen and the German Forest Research Institute in the same city analyzed with the help of artificial intelligence the songs and sounds issued by 53 European species of forest birds, during an entire breeding season. The results were published in the Journal of Ornithology, and the study wrote Phys.org.
The researchers collected data from more than 250 dwellings from the forests of the German Land Saxony and recorded the sounds of birds with small reports, for 30 seconds, every 10 minutes, 24 hours a day, from March to May.
They used to identify the species of birds, based on the songs and sounds emitted by them.
The high “resolution” analysis of the 6.4 million registered sounds showed that the species of forest birds have individual activity patterns.
In addition to a large group of active species during the day, such as honey, others were active at sunrise and dusk, and could be heard twice a day. The honey could be heard much more often in the evening, than in the morning, a result that had not appeared in previous statistics.
The study showed that the recommended hours in various guides for observing many species of birds do not overlap with their periods of maximum vocal activity.
So far, the detailed information about the daily and seasonal patterns of the bird song was only available for a few species, because all the necessary observations last long.
The study has the title: Diel and Seasonal Vocal Activity Patterns Revealed by Passive ACOUTIC MONITORING SUGGEST Expert Recommendations for Breeding Bird Survey Need Adjustment
Photo source: https://www.dreamstime.com/dreamstime.com
