Content uploaded by Jonathan Murray Lovell
Author content
All content in this area was uploaded by Jonathan Murray Lovell on Feb 17, 2015
Content may be subject to copyright.
Content uploaded by Jonathan Murray Lovell
Author content
All content in this area was uploaded by Jonathan Murray Lovell on May 05, 2014
Content may be subject to copyright.
An investigation into the effects of underwater piling noise
on salmonids
Jeremy R. Nedwell
Subacoustech Ltd., Chase Mill, Winchester Road, Bishop’s Waltham, Hampshire SO32 1AH,
United Kingdom
Andrew W. H. Turnpenny
Jacobs Babtie Aquatic, Fawley, Southampton, Hampshire SO45 1TW, United Kingdom
Jonathan M. Lovell
University of Plymouth, Drake Circus, Plymouth PL4 8AA, United Kingdom
Bryan Edwards
Subacoustech Ltd., Chase Mill, Winchester Road, Bishop’s Waltham, Hampshire SO32 1AH,
United Kingdom
共Received 16 August 2005; revised 5 July 2006; accepted 17 July 2006兲
Underwater piling was undertaken in 2003 in Southampton Water on the South Coast of England.
Monitoring was simultaneously undertaken of the waterborne sound from impact and vibropiling
and its effects on brown trout in cages at increasing distances from the piling. Brown trout 共Salmo
trutta兲 were used as a model for salmon 共Salmo salar兲, which were the species of interest but were
not readily available. No obvious signs of trauma that could be attributed to sound exposure were
found in any fish examined, from any of the cages. No increase in activity or startle response was
seen to vibropiling. Analysis using the dB
ht
metric indicated that the noise at the nearest cages
during impact piling reached levels at which salmon were expected to react strongly. However, the
brown trout showed little reaction. An audiogram of the brown trout was measured by the Auditory
Brainstem Response method, which indicated that the hearing of the brown trout was less sensitive
than that of the salmon. Further analysis indicated that this accounted for the relative lack of
reaction, and demonstrated the importance of using the correct species of fish as a model when
assessing the effect of noise. © 2006 Acoustical Society of America. 关DOI: 10.1121/1.2335573兴
PACS number共s兲: 43.30.Nb, 43.80.Nd 关WWA兴 Pages: 2550–2554
I. INTRODUCTION
It has been documented that the driving of piles in water
causes high levels of underwater noise. Measurements made
of piling using piles of 0.7 m diameter on the River Arun in
southern England indicated a peak-to-peak Source Level of
191 dB re1
Pa@1 m 共Nedwell et al., 2002兲. Measure-
ments made of the noise from piling using piles of 4.3 m
diameter during the construction of offshore windfarms in
the UK indicated a peak-to-peak Source Level of
260 dB re1
Pa@1 m for 5 m depth, and
262 dB re1
Pa@1 m at 10 m depth, associated with a
Transmission Loss given by 22 log共R兲, where R is the range
共Nedwell et al., 2003b兲. Such levels of sound may have an
effect on marine species. A report for Caltrans on piling in
San Francisco Bay indicated mortality to several species of
fish at ranges of up to 50 m 共Abbott, 2005兲. Intense sound
from piling may also have an effect on hearing. McCauley et
al. 共2003兲 reported damage to the ears of a pink snapper
caused by the impulsive noise of an airgun having a peak-to-
peak Source Level of 222.6 dB re1
Pa@1 m and deployed
at a range of 400–800 m from the caged fish. Enger 共1981兲
reported damage to the ciliary bundles on the sensory cells of
the inner ear of cod caused by sound at frequencies of 50 to
400 Hz at a level of 180 dB.
In September 2003, piling was undertaken in Southamp-
ton Water, a shallow inlet on the South Coast of England,
adjacent to the quayside of a ferry terminal. There was con-
cern that the construction work might adversely impact local
fish populations, and, in particular, the migration of salmon
up Southampton Water into the River Test, a Special Area of
Conservation 共SAC兲. A limit of 90 dB
ht
共Salmo salar兲 at half-
channel width was set as a noise limit by the regulator 共the
UK Environment Agency兲, with the intent of leaving half of
the channel open for migration; the dB
ht
metric is discussed
in Sec. III B. Consequently, monitoring was undertaken of
the waterborne sound generated by the piling operations. Si-
multaneously, the opportunity was taken to monitor the ef-
fects of the noise on fish in cages at increasing distances
from the piling.
II. FIELD WORK
A. Location of measurements
The piling was carried out adjacent to a quay, in other-
wise open water conditions. A total of ten piles, four 914 mm
in diameter and six 508 mm in diameter, were driven. Vibro-
driving was used for driving the piles, the duration being
approximately 20 min per pile. At the end of the operation 3
of the 10 piles were driven to final depth by impact driving
2550 J. Acoust. Soc. Am. 120 共5兲, November 2006 © 2006 Acoustical Society of America0001-4966/2006/120共5兲/2550/5/$22.50
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 192.129.2.17 On: Tue, 17 Feb 2015 13:02:46
for dynamic test purposes, the duration of each test being
between one and three minutes, with approximately 20 blows
during the period.
B. Fish response monitoring
Hydrophones and fish cages for reaction monitoring
were located at 30, 54, 96, 234, and 417 m from the piling.
The cages were 1 m cubes, made from a mild steel angle and
covered with plastic mesh of 25 mm square aperture and
suspended 2.5 m below the water surface. A closed circuit
television camera was placed in each cage. A control cage
was suspended in a dock 6 km from the site of the piling,
where the piling noise had fallen below background levels.
Although the species of interest was salmon 共Salmo
salar兲, suitable fish were not available and instead farmed
brown trout 共Salmo trutta兲 were substituted as a close rela-
tive, which it was thought would react similarly. As is noted
in Sec. III B, this assumption proved to be invalid.
C. Fish behavioral reactions
The video recordings were reviewed to identify any
changes in behavior that might have resulted from the piling
noise. These data were reviewed “blind,” with the reviewer
unaware of the condition of exposure. Two types of behavior
were investigated.
共i兲 Startle reactions, which are here defined as sudden
C-shaped flexure of the fish’s body, described by Blaxter and
Hoss 共1981兲. The analysis of the startle reactions was based
on a frame-by-frame inspection of the video at the start-up
instant of each vibropiling session and for the next 5 s;
共ii兲 Fish activity level; captive fish that are exposed to
irritating stimuli commonly show “milling behavior,” in
which the fish swim faster and make random turns. These
were measured by counting the number of times fish entered
the camera’s field of view within a two-minute observation
period.
Due to the fish being caged avoidance reaction 共i.e.,
fleeing the noise兲 could not be directly observed, and the
activity level was used as a measure of the behavioral effect.
It may be commented, however, that human experience
would indicate that the avoidance of noise may well occur at
lower levels of noise than an increase in activity, and startle
probably occurs at relatively high levels of noise. Since the
caged fish could not flee, the experiment was imperfect in
that it probably did not demonstrate the lowest level of sound
at which avoidance of the piling noise occurred.
1. Reactions during vibropiling
No startle response was seen for any of the piles driven
by vibropiling.
Table I gives the activity level observations for the
2 min period prior to piling and then for the first 2 min dur-
ing vibropiling; they are seen to have remained similar be-
fore and after the start of piling. The control and event ac-
tivity levels were compared using the nonparametric Mann-
Whitney U-test 共Campbell, 1974兲 with the null hypothesis
that activity levels were not significantly different at the
P= 0.05 level; it was found that there was no significant dif-
ference in activity level following the commencement of vi-
bropiling 共P = 0.001兲.
2. Reactions during impact piling
One pile of 914 mm diameter and two of 508 mm were
impact driven. No startle reactions were observed at any of
the five locations.
Activity levels recorded before and during the impact
piling sessions are given in Table II. Fish activity is seen to
have remained similar in most cases before and after the start
of impact piling. The Mann-Whitney U test shows that there
was no significant difference in activity level following the
commencement of pile driving 共P = 0.001兲 in cages at 30, 96,
234, and 417 m; a 36% increase in fish activity level was,
however, observed in the cage at 54 m, which was significant
at the P = 0.05 level.
III. LABORATORY STUDIES
A. Electron microscope examination of the inner ear
of the trout „Salmo trutta…
The fish were initially examined under an optical micro-
scope after the end of the piling operation for any evidence
of swim bladder rupture, eye haemorrhage, or eye embolism
as a result of exposure to the sound from the piling; none
was found. The saccule, the primary auditory region of the
fish ear 共Popper and Fay, 1993兲, was investigated for evi-
dence of trauma to the inner ear ultrastructure. Five fish from
each of the cages were examined for ultrastructural damage
to the hearing organs; the examiner was unaware of which
fish came from which cage. A typical Scanning Electron Mi-
crograph 共SEM兲 of saccular hair cells from S. trutta is pre-
sented in Fig. 1, which is from a central region of the
macula. The image shows apical hair cell bundles with an
anterior positioned kinocilia, surrounded by approximately
40 cilia arranged in 4 to 5 consecutively shorter rows in a
format common to many fish species 共e.g., Platt and Popper,
1981; Lovell et al., 2005兲. Each hair cell is separated from
TABLE I. Fish activity statistics 共No. of movements per 2 min period兲 for
vibropiling.
Mann-Whitney U test
Range 共m兲 Before During UZ P
30 147.5 152.5 69.5 0.1443 0.8852
54 150.5 149.5 71.5 −0.028 0.9769
TABLE II. Fish activity statistics 共No. of movements per 2 min period兲 for
impact piling.
Mann-Whitney U test
Distance Before After UZ P
30 m 9 12 3 −0.654 0.513
54 m 6 15 0 −1.96 0.0495
96 m 12 9 3 0.654 0.512
234 m 6.5 3.5 0.5 1.16 0.245
417 m 12 9 3 0.654 0.512
J. Acoust. Soc. Am., Vol. 120, No. 5, November 2006 Nedwell et al.: Effects on fish of piling noise 2551
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 192.129.2.17 On: Tue, 17 Feb 2015 13:02:46
neighboring cells by an area of epithelium bearing short
共⬍1
m兲 microvilli. No obvious signs of trauma that could
be attributed to sound exposure were found in any fish ex-
amined, from any of the cages.
B. Fish audiograms
1. Quantifying the effects of noise on species
The response of marine animals to a given sound is de-
pendent on the species since these vary greatly in hearing
sensitivity and frequencies range. A metric has been de-
scribed 共Nedwell et al., 2003a兲, 共Nedwell et al., 2004兲
termed the dB
ht
共Species兲 scale, which may be regarded as a
generalization of the dB共A兲 metric used for the rating of the
behavioral effects of noise on humans and hence that is in-
dicative of how much a species will be affected by that
sound. It is an estimate of the level that the sound is above
the hearing threshold of the species and hence is an indica-
tion of the likely perception by, or “loudness” of the sound
to, that species. Initial research with fish 共Nedwell, in prepa-
ration兲 indicates that at levels of 90 dB or more above the
species’ threshold of hearing 关i.e., of 90 dB
ht
共Species兲 or
more兴 strong avoidance of sound occurs.
2. Inadequacies of brown trout as a model for salmon
At the time of the experimental measurements, it was
noted that, despite the piling reaching levels of noise at
which salmon, S. salar, were expected to react 关i.e., in excess
of 90 dB
ht
共S. salar兲兴, the brown trout, S. trutta, were show-
ing no obvious signs of reaction. It was suspected that this
was a result of the hearing of the latter being less sensitive.
Since there was no audiogram available for the brown trout,
this was determined by the Auditory Brainstem Response
共ABR兲 method, and compared with previous published au-
diograms of salmon 共Hawkins and Johnstone, 1978兲, in order
to attempt to explain this unexpected observation. It should
be noted, however, that the latter audiogram was obtained by
the behavioral method, which in humans has been noted to
yield slightly different results from ABR audiograms.
3. The ABR method
The ABR method was similar to that described by
Kenyon et al. 共1998兲. Fish were held in a cradle in a water
tank and a pair of electrodes placed cutaneously on the cra-
nium such that they spanned the VIIIth nerve, recording the
evoked auditory response to sound. The sound stimulus of
several periods of a sine wave at a given frequency was
generated by two underwater projectors within a water tank.
Responses were averaged over 2000 presentations. The level
of the stimulus was reduced until the stimulus trace was no
longer discernible in the response trace. The sound level at
which the stimulus trace was just discernible was taken as
the threshold level for that frequency. The procedure was
carried out using four fish at all of the frequencies tested, and
24 fish at frequencies between 300 and 1000 Hz.
Figure 2 shows the resulting S. trutta audiogram, as well
as that for S. salar from Hawkins and Johnstone 共1978兲.
First, it should be noted that the latter audiogram was ob-
tained by the behavioral method, which may give slightly
different results from the ABR method. Nevertheless, it may
be seen that, despite being a close relative with morphologi-
cally similar hearing, the audiogram of the brown trout is
significantly different from that of the salmon. It is less sen-
sitive, and broader and flatter in frequency response, with
effective hearing from 30 Hz to above 1 kHz. It was con-
cluded that the common assumption that closely related spe-
cies will have similar hearing abilities is not reliable. Conse-
quently, the data recorded during the piling were reprocessed
using the brown trout audiogram to yield dB
ht
共S. trutta兲 val-
ues in addition to the dB
ht
共S. salar兲 values recorded on-site at
the time of the noise measurements; these data are presented
in Sec. IV.
IV. ANALYSIS OF NOISE MEASUREMENTS
A. Introduction
The piling was recorded using Brüel & Kjær hydro-
phones calibrated to International Standards, and conditioned
by Brüel & Kjær charge amplifiers, before being digitized,
FIG. 1. Scanning Electron Micrograph 共SEM兲 of saccular hair cells from S.
trutta, from a central region of the macula. Fish from cage number 1. k.
kinocilia, c. cilia, mv. microvilli.
FIG. 2. Audiograms for the brown trout 共Salmo trutta兲 and the salmon
共Salmo salar兲. Salmon audiogram from Hawkins and Johnstone 共1978兲.
2552 J. Acoust. Soc. Am., Vol. 120, No. 5, November 2006 Nedwell et al.: Effects on fish of piling noise
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 192.129.2.17 On: Tue, 17 Feb 2015 13:02:46
archived, and analyzed using a program written in the Na-
tional Instruments LabVIEW environment that calculated
one second dB
ht
共Species兲 values.
B. Results
It was found that the unweighted Source Level of the
impact piling of the 508 mm diameter pile was
193 dB re1
Pa@1 m, with a linear Transmission Loss rate
of 0.13 dB per meter, and for the 914 mm diameter pile the
Source Level was 201 dB re1
Pa@1 m and the Transmis-
sion Loss 0.13 dB per meter.
Figure 3 presents the dB
ht
levels of the impact piling as
a function of range. The levels recorded at each of the mea-
surement positions are indicated on the figure; for each of the
two pile diameters driven the values of the noise in
dB
ht
共Species兲 units for both species have been plotted. It will
be noted that the levels are different for the same data for the
two different species since they have different hearing ability
and frequency range of hearing. The figure may therefore be
interpreted as indicating “loudness” for the two species as a
function of range. Also illustrated in the figure is the
90 dB
ht
共Species兲 limit at which it is believed noise becomes
intolerable.
For each set of results a least squares best fit of the data
has been calculated, the results being
SPL = 85 − 0.12R dB
ht
共S . salar兲 and SPL = 62
− 0.09R dB
ht
共S . trutta兲, 共1兲
for the 914 mm diameter pile, and
SPL = 85.4 − 0.16R dB
ht
共S . salar兲 and SPL = 70.7
− 0.17R dB
ht
共S . trutta兲, 共2兲
for the 508 mm diameter pile.
In general, the dB
ht
共S. salar兲 levels are higher than the
dB
ht
共S. trutta兲 levels. This is to be expected, as the hearing of
the salmon is more sensitive than that of the brown trout. It is
noticeable that there is a significant difference in Transmis-
sion Loss for the case of S. trutta for the two pile diameter
cases. It is thought likely that this might have been due to the
fit of Source Level and Transmission Loss to the data being
poor in the case of the 914 mm diameter pile.
In respect of the results for S. trutta, it may be seen that
both the measured and the estimated levels near to the piling
are well below the level at which a mild reaction would be
expected to occur. In other words, no reaction would be ex-
pected by S. trutta at any range from the piling. By compari-
son, the results for S. salar would indicate that a mild reac-
tion would be expected at a range of about 60–80 m, and an
increasingly stronger reaction as this range was reduced.
These results may explain the relative lack of a reaction to
the piling from the brown trout S. trutta, and indicate the
importance of using the correct species of fish as a model
when assessing the effect of noise.
V. CONCLUSION
In summary we have the following.
共1兲 Observations were made of caged brown trout 共S.
trutta兲 during vibropiling of four 914 mm and six 508 mm
diameter piles. The brown trout was used as a model for
salmon 共S. salar兲, which were not readily available.
共2兲 No reaction to vibropiling was noted at any of the
cages. No startle reactions were observed for the impact driv-
ing. However, there was a 36% increase in the fish activity
level observed in the cage at 54 m, which was significant at
the P =0.05 level, although in the two cages nearest to the
piling 共at 30 and 54 m兲, no significant difference in activity
level was detected.
共3兲 Despite the noise from the piling at the nearest cages
reaching levels at which salmon, S. salar, were expected to
react strongly, the brown trout, S. trutta, showed little reac-
tion. An audiogram of the brown trout measured by the Au-
ditory Brainstem Response method indicated that the hearing
of the brown trout was less sensitive than that of the salmon.
Analysis using the dB
ht
metric indicated that this could ac-
count for the relative lack of reaction, and indicated the im-
portance of using the correct species of fish as a model when
assessing the effect of noise.
Abbott, R. 共2005兲. “Fisheries and hydroacoustic monitoring program com-
pliance report addendum.” San Francisco – Oakland Bay Bridge East Span
Seismic Safety Project. May 2005, Office of Biological Sciences and Per-
mits, Caltrans District 4.
Blaxter, J. H. S., and Hoss, D. E. 共1981兲. “Startle responses in herring: the
effect of sound stimulus frequency, size of fish and selective interference
with the acousto-lateralis system,” J. Mar. Biol. Assoc. 61, 871–879.
Campbell, R. C. 共1974兲. Statistics for Biologists 共Cambridge University
Press, Cambridge兲.
Enger, P. S. 共1981兲. “Frequency discrimination in teleosts—central or pe-
ripheral?,” in Hearing and Sound Communication in Fishes, edited by W.
N. Tavolga, A. N. Popper, and R. R. Fay. 共Springer, New York兲,pp.
FIG. 3. The average peak-to-peak sound pressure level as a function of
range, for the impact piling case, in dB
ht
共Species兲 units, for two diameters of
pile.
J. Acoust. Soc. Am., Vol. 120, No. 5, November 2006 Nedwell et al.: Effects on fish of piling noise 2553
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 192.129.2.17 On: Tue, 17 Feb 2015 13:02:46
243–255.
Hawkins, A. D., and Johnstone, A. D. F. 共1978兲. “The hearing of the Atlantic
salmon, Salmo salar,” J. Fish Biol. 13, 655–673.
Kenyon, T. N., Ladich, F., and Yan, H. Y. 共1998兲. “A comparative study of
hearing ability in fishes: the auditory brainstem response approach,” J.
Comp. Physiol., A 182, 307–318.
Lovell, J. M., Findlay, M. M., Moate, R. M., and Pilgrim, D. A. 共2005兲.
“The polarization of inner ear ciliary bundles from a scorpaeniform fish,”
J. Fish Biol. 66, 836–846.
McCauley, R. D., Fewtrell, J., and Popper, A. N. 共2003兲. “High intensity
anthropogenic sound damages fish ears,” J. Acoust. Soc. Am. 113, 638–
642.
Nedwell, J. R. 共in preparation兲.
Nedwell, J. R., Turnpenny, A. W. H., and Edwards, B. 共2002兲. “Piling on the
River Arun; implications for salmon migration,” American Fisheries So-
ciety 132nd Annual Meeting, 18–22 August 2002, Hyatt Regency Hotel,
Baltimore, MD.
Nedwell, J. R., Turnpenny, A. W. H., and Lambert, D. R. 共2003a兲. “Objec-
tive design of acoustic fish deflection systems,” Proceedings of the US
EPA Symposium on Cooling Water Intake Technologies to Protect Aquatic
Organisms, 6–7 May, 2003, Hilton Crystal City at National Airport, Ar-
lington, VA.
Nedwell, J. R., Langworthy, J., and Howell, D. 共2003b兲. “Assessment of
sub-sea acoustic noise and vibration from offshore wind turbines and its
impact on marine wildlife; initial measurements of underwater noise dur-
ing construction of offshore windfarms, and comparison with background
noise,” Subacoustech Report No. 544 R 0424 共for the Crown Estate兲,May
2003.
Nedwell, J. R., Langworthy, J., and Howell, D. 共2004兲.“ThedB
ht
; a meth-
odology for evaluating the behavioral effects of underwater noise and
some examples of its use,” Proceedings of the Symposium on Bio-Sonar
Systems and Bio-Acoustics, Institute of Acoustics, 16 September 2004,
Loughborough University, UK.
Platt, C., and Popper, A. N. 共1981兲. “Fine structure and function of the ear,”
in Hearing and Sound Communication in Fishes, edited by W. N. Tavolga,
A. A. Popper, and R. R. Fay 共Springer-Verlag, New York兲, pp. 3–36.
Popper, A. N., and Fay, R. R. 共1993兲. “Sound detection and processing by
fish: critical review and major research questions,” Brain Behav. Evol. 41,
14–38.
2554 J. Acoust. Soc. Am., Vol. 120, No. 5, November 2006 Nedwell et al.: Effects on fish of piling noise
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 192.129.2.17 On: Tue, 17 Feb 2015 13:02:46
