Opinion

And the forests have … ears: What special sensors recorded

by

What does the forest of Rodopi “hear”? Is he “decapitated” by illegal shootings? In general, are the “ears” of the forests full of animal calls or can only the sound of chainsaws be heard?

Acoustic sensors in forests were installed by scientists from the Forest Research Institute of ELGO-DIMITRA in the framework of the project “Research-Create-Innovate” (EPANEK – NSRF), which includes the action of ” monitoring of fauna and human activities “. The project has been undertaken by a partnership of companies and research organizations, including their Department of Zoology, Department of Biology, Aristotle University of Thessaloniki and the Foundation.

The results that emerged from them 18 acoustic sensors which were placed in an area of ​​about 1700 square kilometers in the Rhodopes presented, in the morning, in an online event, the researcher of the Forest Research Institute Dr. Christos Astaras.

“The sensors were placed on a sample basis in NATURA areas and Wildlife Refuges (KAZ) and from July 2019 to January 2020 they recorded 24 hours a day. “When we got the data and ‘run’ the relevant algorithms, we detected more than 2,500 shots on 12 of the 18 sensors,” Astaras said.

Shots inside Wildlife Refuges

As it turned out, a very small percentage of the shots (0.5%) were out of the hunting season and are not necessarily attributed to poaching, according to the researcher, but because they were “heard” near herds, they may be people who intimidate wolves from approaching. their herds.

“Of the total number of shootings, however, 3.3% were out of hours, ie half an hour before sunrise and half an hour after sunset, while about 17% took place inside Wildlife Refuges,” Mr. Astaras said. At the same time, in the virgin forest of Fraktos, 13 shots were recorded near the border with Bulgaria, which according to the researchers, target either the wild goat or the red deer.

Three acoustic sensors were also installed in the Evros Delta National Park during the entire hunting season 2019-2020 and 43,000 shots were detected. “But the worrying thing is,” Mr. Astaras said, “that the shootings were 40% off-schedule, meaning that there were legal hunters in legal areas but they hunted before dawn and long after sunset, so you can not distinguish the protected species.” ».

The “ears” of the forest and the algorithms

The sensors for the passive acoustic monitoring are like ears in the forest since, as Mr. Astaras explains, they record all the sounds of the area, which consist of winds, the flow of a river, rain, lightning, the calls of the animals and their man-made sounds, such as a motorcycle march or shots. “Acoustic sensors hear up and down what a human ear hears. “They record the sound landscape on memory cards, but especially when they record continuously, the problem is that you are drowning in data”, points out Mr. Astaras.

Automatic incident detection algorithms are a valuable aid in understanding the data, without which – the researchers admit – the ability to use passive acoustic monitoring would be greatly reduced. “For example, especially for shots, they quickly scan your audio data – within a few hours they can scan 3-4 months of data – and give you some potential results. Of course, there is no perfect algorithm, but we want efficient algorithms and in the framework of the project we had relevant for shooting, for sheep and goats and in collaboration with the Institute of Technology and Research in Crete, we developed our own algorithm for chainsaws “, says Mr. Astaras.

Passive acoustic monitoring comes to fill the gap in terms of pressures and threats to ecosystems, having as main advantages the recording of events with high accuracy in a specific space and time, with the release of staff, transparency and repetition of analyzes. However, no threats can be recorded, such as bird traps or poisoned baits.

See all the news

Follow Skai.gr on Google News
and be the first to know all the news

ELGO-DIMITRAenvironmentforestsnewsSkai.grsounds

You May Also Like

Recommended for you