The deer rutting season is in full swing! Although the purpose of the red deer (Cervis elaphus) stag rutting call is to attract females, deer groans help us at the Institute for Environmental Solutions to understand the count of deer in the area and to analyze their habits and behavior in the area. The number of deer stags is an important factor in population dynamics – because one stag can impregnate several hinds during the mating season. We use two types of microphones to collect deer rutting call sound data.
Localization microphones – these microphones are synchronized with each other by a GPS clock that records time with an error of one millisecond. The distance between the microphones is 300-400 meters. Initially, the study was carried out using 3 localization microphones, but recently the fourth microphone was added to the system. Data is collected automatically. Afterwards our machine learning algorithm identifies stag grunts in the sound recording. With the help of the triangulation algorithm, the exact locations of the animals are calculated using the sound deviation in time. The resulting deer locations can then be displayed in various data visualizations and geographic data analysis can be performed. The tests performed confirm that with this method it is possible to determine the location of forest animals very accurately. The error is ±10-12 meters inside the microphone network and about ±30 meters outside the network.
Microphone grid – a wide area covered by a network of 33 microphones. Distance between the microphones – 400 meters. The microphones are not synchronized with each other, but by analyzing the number of calls recorded in each microphone (for example, per hour or per night), a deer activity map can be created. From it, it's possible to draw conclusions on which places the deer prefer. Also by analyzing the sound intensity, the approximate location of the deer in the grid can be determined. The main advantage of the grid method is the relatively low cost and the ability to cover a large area. The limitation of this technology is that it is not possible to determine the exact location of the animals.
That is why the project combines microphones with several other technologies. The sound data is complemented by drone and forest camera observations, as well as data from GPS collars. By performing different types of observations at the same time, it is possible to more objectively analyze the behavior of animals and reveal different legitimacies.
The census methods developed during the project will be useful both for nature protection planning and for forest managers to make decisions based on real data. By realizing the actual size of the ungulate population, it would be possible to determine objective quotas of wild animals or, on the contrary, promote the development of habitats favored by certain ungulate species.
Ner technologies for wildlife census are developed within the project “ICT-based wild animal census approach for sustainable wildlife management No. 220.127.116.11/18/A/146” (No.18.104.22.168 / 18 / A / 146). European Regional Development Fund, program 1.1.1 "Improve research and innovation capacity and the ability of Latvian research institutions to attract external funding, by investing in human capital and infrastructure" 22.214.171.124. measure “Support for applied research”.
More about the project here.