Research Report
Gregory Dawe and Miriam Goosem
School of Earth and Environmental Sciences, James Cook
University
ISBN 9781921359132
Published October 2008
Rainforest CRC / MTSRF Transition
Report
Executive Summary
Section 1: Literature Review – The
effect of traffic noise on wildlife
This review of scientific literature was undertaken in order to
identify known effects to wildlife caused by exposure to traffic
noise. Such effects include:
-
Physiological impacts such as ear trauma and raised hormone
levels; and
-
Behavioural responses such as elevated stress levels, acoustic
adjustment and road avoidance.
Due to paucity of data on animal species, the review also
includes known noise effects on human subjects
Significant findings include:
-
Most literature discussed animal avoidance responses to roads,
and mostly in the temperate climes of the northern hemisphere.
-
Information lost from calls and noise attenuation is determined
by the loudness of sound, frequency range, the receiver’s
ability to detect the sound and the attenuation characteristics of
the habitat – losses in forests are raised by five to ten
decibels over areas with no obstruction. Animals adapt their calls
to compensate for the type of habitat so that individuals can be
recognised and distances estimated.
-
Researchers link traffic noise with reduced bird diversity and
species abundance adjacent to roads to distances of up to 1,750
metres from highways through forests and further through other
habitats. Edge effects in the absence of noise may contribute to
the reduced bird density. Frogs also exhibit reduced populations
although it is unclear whether this is due to noise, pollution or
mortality from traffic. Avoidance in coniferous forests has been
measured and varies between forty and 1,750 metres depending on
species and traffic levels, with even greater distances observed in
less dense habitats such as grasslands.
-
Traffic noise impedes movement by mammals through culverts under
highways constructed in Canada for habitat connectivity and causes
deer to flee in other areas. Noise also increases stress hormone
levels in mammals. Some mammals also suffer raised stress hormone
levels.
-
Studies on acoustic responses to noise by fauna (mostly birds)
have been predominantly laboratory-based, finding traffic noise to
impede the recognition of mating calls in five North American frog
and toad species, and to induce raised amplitude levels in songs or
calls of tree swallow nestlings, zebra finches, lovebirds, African
bush shrikes, nightingales, canaries and budgerigars.
-
Field experiments have found some temperate birds overcome
traffic noise blanketing by singing louder or by making adjustment
to the pitch of their songs. This may impact their general fitness
by requiring expenditure of greater amounts of energy. Birds
singing songs with higher dominant frequencies appear, in some
cases, to be less affected.
-
Anthropogenic noise in the range of 65-85 dB(A) has caused
flight and alert responses in birds and behavioural changes.
-
Mating behaviour of North American frog species is altered by
playback of traffic noise in the field. Similar responses might be
expected in rainforest frogs.
-
High traffic noise levels induce a range of severe, often
chronic clinical responses in humans, with limited tests on
selected fauna indicating similar adverse reactions.
-
No studies have examined the effect of traffic noise on tropical
rainforest fauna including those communicating by sound such as
birds and frogs.
Section 2: Traffic noise propagation through montane
rainforest
The project was designed to collect daytime noise-level data
along the Kuranda Range Road, north of Cairns, with the primary aim
of accurately representing third-octave noise conditions through
the rainforest adjacent to the road at heights of one and fifteen
metres above the forest floor under varying microtopography.
A secondary objective was to delineate any acoustic refugia
(areas of relative quiet shielded from traffic noise) along the
current road and examine potential impacts on these from the
proposed road upgrade.
Data covered a suite of acoustic descriptors including
third-octave dissemination. Noise data was collected at
heights of one metre (representing ground level) and fifteen metres
above the forest floor (representing lower canopy level) for a
period of twelve hours from each site. Sample sites along
eight transects perpendicular to the Kuranda Range Road were
located at the highway edge, one hundred metres from the edge, and
at two hundred metres into the rainforest. Data was used for
comparison with noise modelling generated for the current and
upgraded road.
Significant findings include:
-
Traffic noise at the edge of the forest was louder at ground
level than in the canopy, whereas traffic noise levels in the
forest interior were greater at canopy level than near the
ground.
-
Traffic noise levels decreased with distance into the forest;
being louder at the edge than at one hundred metres, and louder at
one hundred metres than at two hundred metres.
-
Traffic noise was still a significant component of the acoustic
environment at two hundred metres inside the forest away from the
rainforest edge.
-
There was no significant difference in edge noise levels at one
metre above the ground between dawn, noon and evening during
weekdays. However, edge canopy noise levels did vary significantly
when tested across the three one-hour time periods. Traffic noise
is relatively continuous during the day when examined over
ten-minute intervals.
-
Noise levels were strongly correlated to traffic flows at the
edge of the rainforest at both height treatments, however further
into the forest, biotic sounds and noise from other natural sources
such as wind were more significant.
-
Total noise levels at the edge of the forest during the daytime
are relatively continuous at a loudness recognised to have serious
implications for human health. Without further information
concerning the impact of continuous noise on the variety of
wildlife present on the Kuranda Range, a precautionary approach
would suggest that such a level might also cause problems in at
least some wildlife species.
-
The dominant frequency of traffic noise on the Kuranda Range was
1 kHz however traffic noise caused changes to the forest sound
frequency spectrum from 31.5 Hz to 2 kHz, which has the potential
to blanket areas in which some bird and frog species communicate,
particularly at the edge of the forest.
-
Modelling prepared for the Kuranda Range Road Upgrade Impact
Assessment Study by acoustic engineers failed to predict the high
noise levels at the forest edge of the existing highway, although
the same model was relatively accurate for most forest interior
sample sites. At the edge, model underestimates of noise levels
varied from 17 to 31 dB less than measured. This means that in some
cases, the edge of the road was approximately four times as noisy
as had been modelled.
-
It is suggested that modelling for the road upgrade would be
similarly inaccurate at the forest edge.
-
Acoustic refugia that exist under the current alignment of the
road were identified by extrapolating results from transects
sampled, together with consideration of topographic data.
-
Two quiet refuges in particular are likely to become far noisier
if the upgrade proceeds, and several other refuges are also likely
to be affected to some extent.
-
New acoustic refuges may be created by the new alignment;
however re-calibration of the noise models should be undertaken
prior to delineation of these areas.
-
Individual noisy vehicles do much to elevate noise conditions at
both edge and interior sites. These peaks are missed from models
using L10 descriptors, with L1 noise levels at the edge being
typically about ten decibels higher (double the perceived loudness)
than L10 levels. Peak levels along the edge are usually more than
ten decibels above the L1 levels, with occasional A-weighted peaks
above 110 decibels.
Section 3: Effect of traffic noise on avian
vocalisation
The objective of this section of the project was to analyse any
frequency shifts in birdsong recorded at the edge of the Kuranda
Range Road. This was achieved through comparison of dominant
frequency of birdsongs recorded in the forest interior and at the
edge of a road with minimal traffic with dominant frequencies of
songs recorded at the edge of the Kuranda Range Road.
Significant findings include:
-
Songs of fifty-nine bird species were recorded along transects
adjacent to the Kuranda Range Road and at control sites adjacent to
Black Mountain Road. The dominant frequencies of songs from
eighteen of these species recorded at locations adjacent to the
highway and at two hundred metres into the interior were analysed
for evidence of any acoustic modification over distance.
-
Nine of the eighteen species showed significant differences in
dominant song frequencies between individuals recorded at the edge
of the forest closest to traffic noise and individuals recorded in
the forest interior. Of these, five species were considered to have
sufficient replication of the effect between individuals to be
attributable to traffic noise and not to other potential
confounding factors.
-
At least three of the tested species appear to adjust their
songs’ dominant frequency in order to overcome traffic noise
masking,
-
It appears that, at least in some species, traffic noise can
have deleterious impacts on rainforest bird species through
adjustment of song frequency, which has the potential to alter
energy budgets, increase predation risk and reduce success in
reproduction.
Section 4: Effect of distance from edge on
avian populations and biodiversity
The objectives of this section were to examine potential bird
avoidance of habitat adjacent to the Kuranda Range Road in
comparison with the forest interior and with a control road with
minimal traffic.
Signficant findings include:
-
There was evidence of a pronounced edge effect producing lower
bird densities and reduced species richness immediately adjacent to
the highway. Bird abundance increased with distance from the road
edge and significantly more birds were observed at two hundred
metres inside the forest than at the road edge.
-
Abundance of bird species most dependent on rainforest increased
significantly with distance into the forest with greatest
abundances found in the forest interior (one hundred and two
hundred metres from the edge). Species richness of
rainforest-dependent birds was also greatest at these interior
zones.
-
No rainforest obligates were recorded at the edge zone.
-
Opportunist species not normally associated with rainforest were
found only at the edge zone.
-
Increased bird abundance and species richness were observed at
thirty metres from the highway at Streets Creek; most evident in
the frugivore and omnivore feeding guilds. This suggests that the
Streets Creek area is a special habitat that requires
protection.
-
Poor weather conditions during many surveys resulted in fewer
bird observations.
-
Although the observed edge effects cannot be attributed entirely
to traffic noise, given the variety of other impacts on vegetation
occurring in the vicinity of road edges, there is circumstantial
evidence to suggest that road noise was a contributor to the edge
effect. This evidence results from the presence of species that
otherwise avoided the road edge in a zone thirty metres in from the
road where an acoustic refuge occurred caused by the protection of
a topographic ridge. Similarly, a group which appeared to avoid the
road edge was also one in which members reacted to traffic noise by
altering the frequency of their song.
