Project

S⊙N⊙SYLVA is part of the current national effort to monitor terrestrial biodiversity managed by the Office Français de la Biodiversité.

The aim of this national programme is to provide a regular and accurate overview of biodiversity throughout the country, in particular by estimating population trends and working on the links between human pressures and biodiversity responses.

The Office Français de la Biodiversité and the Muséum national d’Histoire naturelle have therefore decided to work together to monitor the biodiversity of forest environments in mainland France using ecoacoustics, a rapidly developing discipline.

Main objectives

The main objective of S⊙N⊙SYLVA consist in the development of a long-term biodiversity monitoring with sound. In particular, the project will have to:

  • test the feasibility of a large-scale ecoacoustic monitoring scheme across a large territory (France mainland)
  • initiate a baseline for the description of protected forest soundscapes
  • estimate the noise pollution inside protected areas
  • build a reference archive of forest soundscapes

Planning

S⊙N⊙SYLVA started in 2023 and will cover 3 years of recording, from 2024 to 2026.

  • 2023 : partners contact
  • 2024 : recording, data collection and curation
  • 2025 : preliminary analyses, second year of recording, first results shared with partners
  • 2026 : analyses, third year of recording, further results shared with partners, national results

Sites

S⊙N⊙SYLVA is based on a unique network of partners giving access to 103 forests covering most of the continental territory of France.

Sites were selected according to the following criteria:

  • protected forest localized in a Nature Park, Reserve or other
  • minimum ground surface of 50 ha
  • minimum distance to any edge of 400 m
  • être représentatif des grands ensembles forestiers métropolitains
  • être représentatif de la variabilité latitudinale et altitudinale du territoire métropolitaine

Recording procedure

S⊙N⊙SYLVA is a collective project based on the acquisition of sound data following a predefined sampling protocol. In each site, an autonomous recorder (Wildlife Acoustics) is settle with the following specifications:

  • at the center of the forest
  • on a tree with a diameter of approximately 25 cm and at a height of 1.5 m
  • from the beginning of March to the end of September
  • uncompressed .wav audio format (44.1 kHz, 16 bits)
  • 1′ on / 14′ off, 1 day on / 1 day off

Analysis

Different levels of analysis are currently considered:

  • acoustic indices that aim at estimating the acoustic complexity of recordings, potentially correlated with species richness
  • AI classification of large soundscape categories (biophony, geophony, technophony) or main sound components (birds, amphibians, insects, cars, aircraft, voice, music, rain wind etc.)
  • AI classification of selected species due to their ecological and/or conservation value