2.2. Acoustic recordings and data analysis
Sensor arrays comprising six Song Meter Digital Field Recorders
(SM2) (Wildlife Acoustics, Inc., Massachusetts), distributed in two triangles,
were installed at two sites andwere programmed to record continuously
during seven days every two months from October 2012 to
August 2013 (six recording sessions). Both sites were matched by habitat
and were located in the same Atlantic forest fragment. The 6-SM2
array close to the active open-cast mine was installed at a distance of
500 m from the mine and 25 m from the closest mining road. The 6-
SM2 array located at the site that was far from the mine was installed
at a distance of approximately 2500 m from the mine and 25 m from a
rarely used road in order to control for a potential border effect due to
the physical structure of the road (Fig. 1).
In order to avoid overlap of the sounds recorded, each SM2 within
each sensor triangle,was placed 80mfromeach other. This distance between
recorders was established during a pilot study conducted in the
area. The distance between the two SM2 triangles was at least 100 m
in order to have two independent recording samples at each site
(close and far from the mine). The distance between the arrays (far
and close sites) was approximately 2300 m (Fig. 1). The triangular
array geometry was chosen to have one SM2 at the forest border and
two located 80 m toward the interior of the forest.
Each SM2 was fixed on a tree at 1.5 m above the ground and was
placed to have the two lateral microphones clear of any surface that
could be an obstacle to incoming sound waves. They were configured
to record in wave format at a sampling rate of 44.1 kHz, at 16 bits. No
high-pass or low-pass filters were applied. One SM2 disappeared during
the fifth session (at the site close to the mine), and the second session
was not considered for one SM2 installed at the site distant from the
mine because the noise produced by a flooded rivermasked all incoming
sounds.