'Monts d'Azur' Biological Reserve, ThorencUniversité du Sud Toulon Var
LSIS (Systems and Sciences of Information laboratory)
Information DYNamic team
Dr Patrice Longour (director)Professor Hervé Glotin (director)
engineer in Operating and acquisition linux systems
Jean-Marc Prévot
photograph: Thanks to M Dancer Vincent for his amiable permissionthesis student:
Website of Haut Thorenc reserveProfessor Hervé Glotin website

The technological challenge is to detect, classify and measure by analyzing audio signals, interactions (such as fear, aggression, domination, flight, cries of alarm ...) between individuals of the same species and between different species.
This signal processing, with the complicity of ecologists, will allow to estimate species interactions and relationships with their environment.
Przewalski's horse, Equus przewalskii, is a species that no longer existed in the wild, until a few years ago. Because of long domestic living conditions, the species hardly establish stable social groups [Skiff Hogan, 1988 ; Winther Christensene, 2002] in natural conditions. Bioacoustics will allow to precisely study in the Biological Reserve of the Monts d'Azur relationships
-firstly, between individuals of Equus przewalskii;
-secondly, between Equus przewalskii and other present herbivores (potential competitors) : Cervus elaphus, Bison bonasus ... ;
-finally, between Equus przewalskii and carnivorous mammals such as Canis lupus italicus [Randi] or Lynx lynx.

SESSION n°1: from 01/10/2012 12:00 a.m. to 09/10/2012 12:00 a.m.
First recording session in Monts d'Azur biological reserve, private reserve directed by M Patrice Longour
Stard date: lundi 01/10/2012
Start time: 12:00 a.m.

the deer belling began September 15, 2012 this year

10am arrival on site
11:30 a.m. arrival and choice of place of installation: installation
1:00 p.m. departure from the place of installation
sharing meal with Mr Longour and the tv team in its structure.
3:30 p.m. am: departure from the reserve for return to toulon

garmin etrex GPS coordinates

Degrees-Minutes decimales: DM.d
Accuracy ± 8 meters (for 3 decimal)
Practical notions about GPS N 43°48.746' E 006°51.238'

Now I set my gps not in Degrees-Minutes decimales but in Degrees-decimales D.d.
Accuracy up to 8 centimeters (for 6 decimales).
hypothetical example that has nothing to do with the coordinates above
N 48.858282°
E 2.294624°

SABIOD Box has been installed in an oak, at 2m50 from the ground

Leave SABIOD box on the floor is not an option because
-rutting deer are dangerous, very curious, and rub their antlers against everything that is not at 2m above the ground.
-bisons also are very curious, get rid of their summer wool rubbing against trees, bushes and rocks supports
-same with boars which we do not know the reaction but whose flair is formidable.

So we hoisted the SABIOD BOX 2.50m above the ground in an oak. Boxes and microphone are strapped.

The gain was adjusted using two earphones whose jack was plugged into the out jack port of the Beagle board. Patrice made ​​noises to 10, 20, 30 meters to help me to adjust the gain.
The risk of falls are killed by simply using a strap that goes under the climber's arm and around his torso.

For a future record in another biotope in the same park, it would be easier to come with a wooden platform (by instance: a simple structure with four planks of 2 meters all that belt to a tree).

The first session during 8 days, it will be recovered Tuesday, 09/10/2012 and reactivated immediately for 15 days in another part of the park

Discharge of Session 1 batteries
batterysession 1 start voltagesession 1 end voltageremaining energy according to battery charger
F 00527 14.23 Volts 13.57 Volts70%
F 0053414.06 Volts13.63 Volts70%

---------------------------------------------------------------------------------------------------------------------------------- RESULTS OF THE SESSION 1:
total volume occupied by recordings (session 1) on the 120 GB pendrive: 60 GB
We copied the recordings on the hard drive of a Windows MiniPC : copy time = 50 minutes.


Watching spectrogram and listening wav file, looks like an animal is present in the surrounding (noise of trampled leaves, strong respiration)


passage of a large aircraft or helicopter
Bird sounds
between 50 and 60 seconds, the sound of broken stick ==> herbivore in the corner?
Another passage of air or helicopter

2nd species of birds: corvid?
passin airplane of helicopter
3rd species of birds
4 minutes: sequence of differents signals of different species of birds

Extract from data-19700108_224355_oiseaux
Extract from data-19700101_020154_oiseau1_3signaux_forts
Extract from data-19700101_030222_extrait_peutetrecommeoiseau1
Extract from data-19700101_030222_extrait_presence_animale
Extract from data-19700101_040249_bruit_systeme_toutes30sec
Extract from data-19700101_060344_oiseau2_trilili_5signaux_6secintervalle
Extract from data-19700101_070412_oiseau_1min_signalslash_touteslessec
Extract from data-19700101_080440_animal_inconnu_loin
Extract from data-19700101_080942_bruit_aigu_indetermine
Extract from data-19700101_090005_brame_peutetre
Extract from 19700101_090507_brame_tres_probable
Extract from 19700101_090507_indetermine_bramepeutetre
Extract from data-19700101_091010_brame_tresprobable
Extract from data-19700101_100033_brame_assez_net
Extract from data-19700101_100535_3brames_lointains
Extract from data-19700108_224355_oiseaux

SESSION n°2: from 09/10/2012 12:20 a.m. to 25/10/2012 16:30 p.m.

Departure from Six-Fours les plages: 5:45 a.m.
How to get there:
Motorway towards Nice. Take a little before Cannes exit 42 Mougins toward Grasse
Go to Grasse
Follow the Napoleon road
Follow Thorenc / Andon

Mr Olivier Dufour and Patrice Longour left reception desk of the reserve at nearly 10:10 a.m.

handling gloves (to break twigs and handling brambles without fears)
shears (to cut short the thin cords and brambles)
Ladder (lent by Mr Longour) enormously facilitate the installation of the equipment.
One strap to secure the climber
1 strap to fasten the main box
2 solid wood boards to create the foundation on which both batteries rest on

We recovered the SABIOD box located on the first location:
1-Disconnect the secondary battery connected in parallel to the main battery by pressure on the switch provided for this purpose. Remove on the ground the box containing the secondary battery.
2-Detach microphone from its branch and attach it to the main box
3-Down main box + microphone attached to it.
4-Open the main box and control the system: green leds of the motherboard and red leds of preamp were on as normal. The device was always on. No water found.
5-Connect headphones to the jack out port of the beagleboard (motherboard) to verify: the system was still correctly recording.
6-We stopped recording at 11:30 a.m. by pressing the switch.

total volume occupied by recordings on the 120 GB pendrive: 60 GB
We copied the recordings on the hard drive of a Windows MiniPC : copy time = 50 minutes.

We have taken advantage of this time to load up all the equipments in the four-by-four and to find a place for the second session.
Replace both batteries of session 1 (F 00534 and F00527) with two charged batteries (for this session 2):
F-00512: warning the battery is damaged on its top. His seal is maybe no more. Maybe it's easily reparable. Prefering avoid to lay down it, I put it in the main box (so vertically).

Gain of the session 2 is not the same than gain of session 1. I adjusted it ad hoc plugging the headphones hand playing with the potentiometer.

GPS positions
12:02:56 pm 43.81099 ° N, 006.84915 ° E
copy and paste directly into a google map or google earth or also geoportal (which has the advantages of also offer ign map for example)

Altitude: garmin etrex GPS informed 498m which is absurd. Patrice estimate altitude to 1300m.

Records of the Session 2 start at 12:20 a.m.

Recovery of SABIOD box:
At the opening of the main box, preamp LED was red (as normal).
But all green LEDs of motherboard beagle board, including three LEDS next to usb ports, were extinguished.
On the beagle board, only a red LED was lit next to the arrival of power supply port.
No sound by plugging a pair of headphones to the 3.5mm jack output port of the beagle (even by varying the gain of the preamp).
End of the acquisition
4:37 p.m. by setting off the main switch.
Within less than 10 meters, M. Longour noted the presence of fresh droppings of wild boar and deer and grasses grazed by bison.

Discharge of Session 2 batteries
batteriesession 2 start voltagesession 2 end voltageremaining energy according to battery charger
F 00519 14.23 Volts 6.81 Voltsless of 30%
F 0051214.06 Volts7.01 Voltsless of 30%


total volume occupied by recordings (session 2) on the 120 GB pendrive: 113,3 GB

Hervé Glotin (Institut Universitaire de France, team director)

Parallel study: Photo-acoustic automated surveillance of fauna in "vallon de notre Dame", Port-Cros national park

Detection and classification on audio-visual transect of North PELAGOS and Stoechades archipelago cetaceans populations (DECAV project)