Review of an acoustic alarm strategy to minimise bycatch of humpback whales in Queensland coastal gill net fisheries

8 Pages·2002·2.7 MB·English

by  G R Mcpherson

KLV1LVV OF AN ACOUSTIC ALARM STRATEGY TO MINIMISE BYCATCH OF
humpback whales in queensland coastai gill net fisheries
gr. Mcpherson,j. lien, n a. gribble and b lane
McPhcrMm,(i.R.,LicnJ..(inbblc. N.A.& Lane, B.2001 1231: Reviewofanaeousi\<_ altera
strategy to minkiwc bycfftt h ofhumpback whales m Queensland coastal gill net fishcriefi.
Memoirs of i!w QueenslandMuseum 47(2): 499-506. Brisbane. ISSN 0079-S.X3S
Humpback whale**, MCgctpttfm navtwangHuP, in Queensland coastal waters are at ri<fe >>t
eIntanglement in a range offishing gears and obstacles. Since L99] the Queensland Shark
Contiol Programme ofthe Queensland Department ofPrimary Industries has developed an
acousticiUarmt>yca$chreductionstratcgj oUfacousticalarmtypesattachedtogtlltietshavc
I
ilised in an atlempl lo 'warn' humpback whales offhe presence oi these man-made
obstacles. Another alarm type, under development, has been distributed to cumnieivui
fiShcriC8operatinginQueenslandWfitterstoreducetherisk pfhuftlpl>ack whaleentan,-leineAl
in commercial gear. A standard acoustic warning protocol is under development for
humpback whales, intcgralmg specific alarm source levels. BCOUStk propagation and
ihmobwiedvlelrEiJtOsifhioCulledveplrs.ovHideeabeInIchaimtartOkhaguamipnsbtacwhkiwchhawlheaslethiesntsatnagnldeamredn\ts,iiotbihd oDtO'tilclmeaary.
compared OFfumpbad whale, entanglement byeatvk acou&tk:alarms;
The Queensland Shark Control Programme grounds in the Antarctic, passing southeast
tQSCP) Oi (he Oueensland Department of Oueensland m September-November, again with
Primary Industries(DPI wasinitiatedbecauseof some whales moving close to shore. QfiCP
|
a seriesoffatal shark attacks dfftbe Gold Coast. records show eight humpback whales were
SunshineCoastand otherQueenslandbeaches in napped in nets helvveen 1962-1995 off the Cold
the summers of 1958-1961 (Fig. I). The QSCP and Sunshine Coasts, with five being released
does not provide an impenetrable barrier to and three dead in Cold ( nasi netfc dribble 61 al
sharks, rather a constant fishing pressure with a (1998). Norecordswere keptofhumpback whale-
combinationofgillnetsandbailedlinesthaioperate collisions that did not tesult in entrapment (Lien
to reduce shark numbers in the immediate etaL 1998),
\geiacrin1itsytroaftmegayjoorfsnweitsmmainndgdbreuaicnhliense.sTahdeap'tmsixtehde Lien el al, i 1 990) used mechanical Mow
type ofgearto the physical characteristics ofthe frequency clangers' (50-IOOOHz), mechanical
swimming beach and allows for differences in Mow frequency beepers (3.50UH?) and
catch selectivity oflarge individuals from a wide electronic 'high frequency (ringers' (27»50fcH?)
range ofshark species. The policy has provided lo reduce bycateh o( humpback whales in
Swimmer protection, with the incidental capture Newfoundland's cod traps. The low frequency
of non-target species lower than that resulting 'clangers' did not significantly reduce the
fftmi deployment of nets alone (Dudley, |998; probability Of entrapment of humpback wh.i
Gribble etal, 1998). possibly due to logistic reasons. The Mow
frequency beepers5 did reduce the probability,
Humpback wbales. Megaptem novuc&ngliae vvbilethe 'high frequencypingers'didnot. Due to
of the eastern Australian population pass the manner in which whales were entrapped
southeast Oueensland during their northward when high frequency pingcrs* wereused. Lien el
migration to calviflg areas north ofF'raser Island al. (1990) believed that these entrapment's
lo.in June-Augusteach year Some whales ninvc occurred as the whales were manoeuvring to
close to Gold and Sunshine Coast beaches, often avoid a collision. Their suggestion was that the
between the shark nets and Che surf/one (Lien et whalesdetectedthemtoo late,eitherasthey \
aL, 1998), Alter Ihebreedingseason, whales wiih ton quietoi weredetectedal an insensitivepaii .-I
calves move southwards to summer feeding the whales hearing spectrum.
500 MEMOIRS OF THE QUEENSLAND MUSEUM
Lien et al. (1990) concluded that humpback databaseoperatedbyrapidresponsemarinerescue
whaleswerenotorientingusingvisualcuesduring groups(Gribbleetal., 1998). Suchoperationsare
inshore feeding activities in Newfoundland not included in the QSCP database.
waters,and itwasmorelikelythatacousticalcues QSCPnetsarenottheonlypotentialhazard for
were the primary stimuli. The observations that migrating humpback whales. A gillnet that
humpbackwhalescouldmovearoundandmostly
appeared to be from the Australian southern
avoid nets at night in extremely low light levels shark fishery was observed entangled around a
and in turbid water, without producing sounds, northward migrating whale offSydney in 2000.
suggested that acoustic cues from the net were Entanglements in anchor ropes have been
used.
reported by crews ofsmall vessels and spanner
During late 1991 Lien provided acoustic crab pot lines have also been observed trailing
alarms ofa mechanical Mow frequency beeper" from humpback whales.
type lo the QSCP and supervised positioning A
small offshore shark gillnet fishery operates
them on the Gold Coast nets. These alarms were
within Queensland continental shelf waters,
deployed during a 16 week period of the 1992
humpback whale migration season. No whales often inareaswhereadultwhalesandcalveshave
been observed but no entanglements have been
were caught in nets fitted with the alarms.
reported.
A paired comparison study of alternating
alarmed and non-alarmed nets was commenced CRITICISM OF THE ACOUSTIC BYCATCH
for a 26-week period during the 1993 humpback REDUCTION POLICY
whale migration season. C-CORE alarms were
utilised featuring a broadband signal centred on The acoustic alarm policy developed by DPI,
4kHz. Towards the end of the experimental particularly by QSCP, has been criticised from
period a whale was entrapped in a non-alarmed three major viewpoints.
net. The subsequent public pressure resulted in
all Gold Coast nets being fitted with alarms for 1) Environmental groups disagreed with the
the remainderofthe whale migration season, the potential environmental effectsofthe QSCP, and
consideredthatacousticalarmsweresuperfluous
change effectively terminating the experimental
opportunitytoexaminetheeffectivenessofalarms. to a shark control operation that should not be in
operation. Whatever the final biological results
Lien et al. (1992) demonstrated that acoustic ofanalyses ofthe QSCPdata, the outcomes will
alarms were successful in reducing humpback be considered primarily in the light of risk to
whale collisions with cod traps. Given that no human life and with regard to Government
dramatic decrease in shark catch occurredduring 'duty-of-care' legal responsibilities (McPherson
the 1992 and 1993 acoustic experiment periods et al., 1998). However, bycatch minimisation is
and that no whales had become entangled in an integral part ofthe QSCP strategy (Gribble et
alarmednets,alarmshavebeenroutinelyfittedto al., 1998).
Gold Coast nets during subsequent whale
migration periods. 2) The effectiveness of alarms, specifically the
acoustic propagation ofthe alarms in relation to
In 1994 a deliberate interaction was observed various ambient conditions, is uncertain. There
betweenalarge humpbackwhaleand an alarmed was also concernthat the alarms could affect the
netofftheGoldCoast,withthewhalecirclingfor localised migratory behaviour of humpback
some time before charging the net. Smaller whales, namely that alarmed nets offshore from
whales including calves had moved away as the specific headlands may direct close inshore
largewhaleapproachedthenet.Thematerial,and migrating whales toward waters with unfavour-
particularly the net headropes, stretched out of able navigation conditions and higher ambient
the water and disintegrated under the force. noise levels which may mask the acoustic alarm
While this behaviour has not been observed signals. While most humpback whales appearto
again, there have been three further reports of ignore alarm signals, some approach the sound
massive holes appearing in net panels and sourcewrhileotherswithdrawfromit(Toddetal.,
headropes of other alarmed nets on the Gold 1992). These concerns were well-founded and
Coast and Sunshine Coast. DPI expended research effort to assess the
From 1992-1995 a single live release of a acousticpropagationofalarmsignalsinthemain
humpbackwhale from a non-alarmednet (dueto areas where QSCP gear was deployed. These
short term logistical reasons) was recorded in a assessmentsarebeingextendedtootheroffshore
ACOUSTIC ALARMS
501
habitats where gear that poses a
potential risk for humpback whale 140--E I44"E 148°E TE 15<J°E
entanglement is deployed.
3) QSCPstudies did not demonstrate
sufficientstatistical rigourtoprovide
clear cut conclusions to assess the
effectiveness ofalarms. These critic-
isms were based on a premise that if
somethingcouldnotbedemonstrated AUSTRALIA ^lj
tobeeffectivewith>95%probability
then there was no effectiveness and
noconclusionsshouldbe drawn. The
Acoustics Deterrents Workshop
hosted by the U.S. National Marine
Fisheries Service (Reeves et al.,
1996) recognised that rigorous
experimental procedures should be
incorporated into any fishery study
using acoustic alarms. However, the
report recognisedthat some fisheries
would never have sufficient fishing
power to demonstrate statistically Sujishuu-Const
whether acoustic alarms could PointLookoul
reduce marine mammal bycatch. Kilometres ColdCoast
Reeves et al. (1996) indicated that
experiments that could not provide
statistical probabilities beyond the FIG. 1. Map of Queensland showing selected Queensland Shark
most rigorous standards were still Control Programme contract locations.
relevant provided the observations
were taken in context ofother observations that Withthesteady increaseinnumbershumpback
demonstrated the same trend. The report whales have appeared in waters where they had
suggested that behavioural studies monitoring not been observed, at least over the past 35-40
responses of mammals to dummy or 'pseudo" years.ThereisanecdotalinformationfromQSCP
nets with active and non-active alarms contractors (e.g. J. Backmann, pers. comm.)
(Koschinski & Culik, 1996; Stone et al., 1997) indicating that humpback whales hadpreviously
could provide larger sample sizes to determine visited those areas, but not since themid 1960\
effectiveness ofalarms. prior to when the eastern Australian population
was reportedto havebeen at its lowest(Paterson
CHANGES IN RISKTO WHALE etal., 1994). In 1996 ahumpback whalecalfwas
ENTANGLEMENT SINCE 1991 entangled in a QSCP gillnet off the Sunshine
Coast (NW of the Gold Coast) during the
In 1991 the only gear that appeared to pose a southward migration and, as a result, was
threatto humpback whales inQueenslandwaters temporarily beached in the surf zone. In 1997
were eleven 186m gillnets anchored offthe surf near entanglements occurred off the harbour
zone on Gold Coast beaches. Since that time mouth at Mackay (Fig. 1). Acoustic alarms have
Paterson et al. (1994) have reported increases in now been attached to QSCP gillnets at Mackay
whalenumbersof .7%perannum. Theobserv-
1 1 (5) and Sunshine Coast (11).
ations of Paterson el al. (1994) were conducted
off Stradbroke Island immediately north of the FIELD AND ANALYTICAL METHODS
GoldCoast. It is not clearwhat proportionofthe
humpback whale population observed from Acoustic signals from alarms were recorded
Stradbroke Islandpassedwithin close proximity with aGEC-Marconi SH101X calibrated 100kHz
of Gold Coast QSCP nets, although it is hydrophone, a low noise Royal Australian Navy
reasonabletoassumethatthenumberpassingthe Research Laboratory pre-amplifier and a Sony
Gold Coast has increased in proportion to the TCD-D8 DAT recorder. The system had a
population increase. frequency response of 15-22,000Hz. Tapes were
502 MEMOIRS OF THE QUEENSLAND MUSEUM
Left_flannel
15.0k'
140k.
130k.
12Oki
11Ok.
100k. -6O0
s.dk
8Oh.
01
aih
50k' .80.0
40k'
3CHc
in.'
FIG.2.SpectrogramofrepeatedsignalsfromatleastsixC-COREmechanicalalarms(verticalbroadbandsignals
between 2-12kHz), three Dukane 'Netmark' alarms {horizontal tone burst at around 11 kHz) and humpback
songcomponentsofftheGold Coast. C-CORE andDukanealarmswereonanet 100mfromthe hydrophone,
and possibly another furtheraway. Location ofthecallingwhale was not known.
analysed using 'Spectra Plus' acoustics software to draw the attention of whales to the sound
with anAWE-64soundcardatasamplingrateof source,whichuponcloserinspectionwasavoided
44,100Hz, with a Fast Fourier Transform (FFT) along with the gillnet to which itwas attached.
woifd1t,h0)2o4fp4o3i.n0t7sHazn.dWahefinltemreabsaunrdiwnigdtthhe(lFeFveTlsbionf Corrosion anddamage incurredby nethauling
operations rapidly reduced the number ofwork-
the fundamental frequencies of the alarms, no
ing alarms. These were replaced during the
correction was made for the filter bandwidth 1994-1996 migrations by 'C-CORE* alarms
becauseofthesinusoidalcharacterofthesignals.
(Centre for Cold Ocean Research Engineering,
Sound pressure levels (SPL) were expressed as
Memorial University ofNewfoundland, Canada).
dB re luPa. The analysis system was calibrated The acoustic signature of these mechanical
wwiitthh aanTFeFktTrosnpiexctTrDumS-a2n1al0ysdeirgimtoalduolsec.illoscope 2a-l1ar2mksHz.feAatsupreecdtroagrbarmooafdCb-aCnOdRrEaanlgaermsfraonmd
Background noise spectrum levels (in 1Hz 'Dukane' high frequency alarms (Dukane
bands) were calculated from the FFT results by Corporation, Seacom Division, IL, USA) is
correcting forthe filterbandwidth from the level giveninFig.2.Assomeacousticenergyoccurred
in the FFT bin (values given are in dB re <2.0kHz, which approaches the known audible
luPaVHz). One-third octave bandwidth levels capacity of most shark species investigated
were estimated by adding the bandwidth (Corwin, 1981), there was concern that sharks,
correctionforthe2,810-3,540Hz 1/3octaveband the target species of the gear, would detect the
to the spectrum level. acoustic signal. Given the short duration that the
ACOUSTIC ALARM VARIATIONS alarms were deployed on QSCP gillnets, no
consistent trend in shark catch was detected.
Concerns were also expressed that the electro-
Since 1991 fouracoustic alarms types have been magnetic nature of the C-CORE alarm signal
used to 'warn' humpback whales ofthepresence mayaffectcatchesalthoughnodataareavailable
of QSCP gillnets. Original alarm deployments on this aspect ofperformance.
were courtesy ofJon Lien who provided mech-
anical type alarms centred around a fundamental On Lien's second visit to Queensland he
frequency of 4.0kHz that had been used supervised the development of a piezo buzzer
effectively to enhance the acoustic signature of type alarm, similar to his earlier design and
cod traps (Lien et al., 1992). Source levels were described by Lien et al. (1995). At that time the
upto 145dB re 1uPaat lmetre. Thesehad shown 50mm diameter plastic sewer pipe and
ACOUSTIC ALARMS
503
_
appropriateendcapsandthreadedfittingsused in 150
Canada and USA were not available in Cairns, £
Australia. The nearest equivalent pipe was I«S 130
100mm diameter. To minimise damage due to a.
o 120
water intrusion, the piezo buzzer (a truck
reversing alarm with a fundamental frequency
centredaround2.9-3.0kHz)wassetinresin inthe
base ofthe unit with only the terminals exposed.
Acoustic output ofthe alarms were not as high i 9 12
(source levels-125-130dB re luPaat lm) as the Frequency(kHz)
original alarm described by Lien et al. (1995). FIG. 3. SpectrumofBASATechnical Services 'whale'
The new alarm was -3 times heavier due to the a ami.
I
volumeofmaterialsusedandtrials indicatedthat
alarm source levels declined as alarm weight CURRENT STATUS OF ACOUSTIC ALARM
increased. In many alarms the sound pressure
STRATEGY
level ofthe secondharmonic frequencywashigher
than the fundamental frequency. Nonetheless,
this inexpensive alarm (~AUD$20), was utilised Research is continuing on the acoustic
during the 1997-1998 humpback whale propagation of alarm signals of the lower fre-
migration seasons with no entanglements on quency alarms (-3kHz fundamental frequency,
alarmed nets resulting. considered to be most effective for humpback
whales) within different environments. QSCP
Overall size ofthese 100mm diameter alarms areas include close proximity to high wave
aisnstorcoidautecdedwiathradnegpeloyomfenltogiosnticgailllnpetrsobwlheimcsh aennedrgSyusnasnhdinbeeacChoeasstisn,5-a1n0dmbwoattherdoeffeptheerGoalndd
TrehseulQteSdCiPn acaslulbesdtafnotriaelxplroessssiroatnesforfoimnttehreesgteafro.r snhoarltlhoewrenrwwaatteerrss. wCiotmhmemrorceialmufdisbhoertytomasreaisn
ttheendceornsftorrucstuipopnlyofwaasrelpetlatcoemBenAtSAalaTremchannidcala oifnfcslhuodreeshwaaltleorws nbeeatrw/seheonretehenvcioraosntmeanntdsQtuoemeonrse-
Services (BASA Technical Services, Brisbane,
Australia). BASAproduced apiezobuzzeralarm land's coral reefs in 20-30m.
with afundamental outputat-3.4kHz. Thealarm Alarm performance attributes such as source
was relatively small andused four ,5Vbatteries levels,total acoustic intensityofshort tonebursts
1
which proved to be light and cost effective. The relative to ambient sound levels, and alarm
SpectrumisgiveninFig.3;sourcelevelexceeded longevity are being developed and assessed.
140dB re 1uPaat 1m. Longevityofthesignalhas Until the BASA and Lien (Cairns) alarms
yet to be determined although it is anticipated to currently in use have attained their fillI develop-
be -21 days continuous operation. ment potential, specific recommendations on
McPherson etal. (1999)describedtheacoustic waaltaerrmsdceapnlnooytmbeentmaodne.obstacles in Queensland
features and construction of the Lien (Cairns)
piezo alarm, adevelopment ofthe original piezo Thethresholdforauditorydetectionofasignal
alarm described by Lien et al. (1995). Further is considered to occur when the signal level
work has increased the longevityofthesealarms equals the background noise level in a certain
to 40 days continuous operation and the alarm is bandwidth, known as the masking band
seen as a cheaper variation suitable for deploy- (Richardson etal., 1995).Noiseoutsidethisband
mentwithin Queenslandcommercial fisheries,at would have little effect on the detection of
least until a full production commercial model is signals. Research on hearingin marinemammals
available. EnvironmentAustraliahasfundedDPI has shown that a range ofvalues for the width of
tocontinuedevelopmentand construction ofthis the masking band exists for tonal signals. Most
alarmtype for immediateuse within commercial results vary between 1/6 and 1/3 of an octave,
fisheries that may take marine mammals. One although some are less (Richardson et al., 1995);
hundred alarms have been constructed with a the most conservative approach is to assume a
numberhavingbeenprovidedtogillnetoperators masking band of1/3 octave. As the fundamental
lo conduct logistical gear deployment trials frequencyofthepresentBASAwhalealarmsand
includingattachmenttonets,operatingdepthand Lien (Cairns) alarms fall within the 1/3 octave
vessel storage. band of 2,810-3,540Hz, the signal-noise-ratio
504 MEMOIRS OF THE QUEENSLAND MUSEUM
(SNR) ofalarm tone bursts are compared to the instancesthewhaleswereattemptingtoavoidthe
background noise within this 1/3 octave band. gear, but probably detecied it too late to avoid
Background ambient noise levels include collision. No SNR data were available for these
experiments.
biological noise such as snapping shrimp, wave
motionandbreakingsurfwithin20-S()m fromthe For a particular background noise level, the
nets, depending on tide state. Considerable spacing of alarms required to give a minimum
variability has been detected between different SNR of a chosen value of +10dB (or the more
beaches within QSCP contract areas. Ambient conservative +15dB) within 15m ofthe net can
levels may change with sea state and wind be determined using the method given by
strength, while at more sheltered beaches McPherson et al. (1999). Assessment of alarm
ambient noise may be dominated by snapping signal propagation and ambient noise levels is
shrimp with spectral levels between 65-80dB re conducted for each beach within QSCP contract
l(iPa"/Hz at 3kHz irrespective of weather areas, or commercial fishery areas. Under most
conditions. Ambient levelsin fishingareas inside alarm,propagation and ambientlevel conditions,
theGreatBarrierReefwherewaterdepthis>20m a+15dB SNR is achieved 15mout from each net
appeartobedominatedby fishchorusesthatmay between adjacent alarms, ifalarms are spaced
reach spectral levels of 65dB re luPa7Hz at 50m along the net. As QSCP nets are 186m in
-3kHz (R. McCauley, pers. comra.). length, contractors are currently required to
There are few biological data to determine the position five alarms on gillnets a minimum of
most appropriatepositioning ofalarmson nets in 45m apart, to achieve this SNR/distance out
relation to auditory capacityofmarine mammals scenario.
andbackgroundnoise. Krausetal. (1995) spaced Whether the +15dB SNR at 15m from the net
d1r0okpHpzedatloaramSsNaRt odfis+t1a5ncdeBsanwdhedreemonSsPtLr*aStehdada sisceanmairinoiimsuampporrokprnioawteniasconuosttkicnoswtnan,dahrodwaegvaeirnsitt
significant reduction in bycatch of harbour whichwhaleentrapments.orlackofthem,canbe
porpoise. Gearin et al. (1999) placed alarms a compared.
distance apart thai permitted harbour porpoise to
hear 3kHz alarms at a SNR of +10dB up to a
FUTURE RESEARCH
Beaufort sea state of4 (i.e. 11-16 knots).
As spacing between alarms increases it
heightens thechance ofan acoustic 'hole' occur- Environment Australia has funded DPI,
ringforananimal approachingapointonthe net. University ofQueensland, Memorial University
or gear, midway between two alarms. The only of Newfoundland, SEANET and Queensland
discernible acoustic cues wouldbeon eitherside Parks and Wildlife Service to examine the
oftheapproachinganimal,butnotdirectlyahead. behavioural responses ofdugongs and dolphins
Acoustic 'holes' would be more significant to acoustic alarms. Funding has also been
where the range from the line ofsources is less provided forthe furtherdevelopmentofthe Lien
than the source spacing, which would normally (Cairns) alarm for deployment throughout
be the case of interest. In this situation, the Queensland's gillnet fisheries, including those
received signal would be dominated by the that may interact with humpback whales. It is
contributions ofthe closest two alarms, and the hopedthroughtheseexperimentswewillcometo
contributions from other alarms could be more fully assess bycatch in gillnet fisheries and
neglected. Thereceivedsignalislowestwhen the develop effective means to minimise it.
rleinceeivoefra(laanirmmasl)atisroinghta lainngelwehsicahndcrmosisde-swtahye coDnPdIucdtoesacnooutsbteilcieavelaitrwmourledsebaeracphprtohpartiatmeatyo
between two adjacent alarms. jeopardise the lives ofmarine mammals simply
The minimum distance from the net that in order to achieve more rigorous experiments
provides humpback whales sufficient time or that would demonstrate >95% probability of
space to avoid a collision was considered to be effectiveness for alarms. Gribble et al. (1998)
15m based on the maximum length for the described the level of bycatch of marine
species. Lien et al. ( 1990) and Lien et al. (1992) mammals in Queensland gillnet fisheries as
indicated that the circumstances in which probably minor and there will be no attempt to
humpback whales were caught in both alarmed raise fishing effort to increase bycatch numbers
and non-alarmed nets suggested that in some simply to achieve a statistical probability.
,
M mi STIC Al ARMS
ACKNOWLEDGEMEN'IS n \\z i,v tfl educe incidental moilalily of
The at sea assistance of QSCP contractors harbouu.rmoplwrpFoisitsai mIDduillwnefisldlIifiin.il RcmpOfdlatti<3<th
ton (Gold Coast)i Noel Walker : ... D.C.April 1995. I i
(mSuonbsthaiinneinCgoaRsetl)dadnadtaJeafnfdBafetwknoibasnenrv(aMtiaocnksayo)f ii sjUI , BARNEJY. IWW.JjTjOlcDcDt,soSlMad5dEinTg0soundsto
humpback whale behaviouraroundnets has been cA traps on the probability of collisions by
essential foi the developtneni of an effective humpbackwhales.Pp. U1-7M.In rbomasj '•
asccioeunsttiifcicalaadrvmiCsftlraOteHgy.acoWuestiarces garnatdefwulhafloei man tl »cRn-oA\ A. S>umpiiu (P'leinum Presa;i Ni.ie.iw
behaviour provided by Doug Cato (Defence Yoik).
Sriencc and Technology Organisation). Mike LIEN.BOJ.R.GGHAOAORDD,,DC..XHPlIST!T!MOAIVMiCv, WRlUlESLN.TRP.,,
NRooabdMacnCdauKlecnySc(hCumltizm(CUnniivveerrssiitt)y)ofa'nSduinIev), t1estsMoAfHaOcNou,stiTc. &devMicIeTsCoHnELgLro,unDd.lis1h995g.illFnieetlxd;
Hague (University of Queensland). Olivia assessment ofeffectiveness is rcdiw inj tiartx u
oWrhiygibnialrdpiaes/soisateldarmwsitmhadcoensitnruCcatiironns. oDfenliitse pIohropnoiiasse. I Aalc&fe NaPcph.i1g-.2il2l., IPn.P.Kas(attesl)emS.enKs.oAr.y.
i
Ballnm ofSkANl-T and the scnmi authors &ot)S cmsofa<malii noamxc Is 1 1 fcSpil Publi
Craigand Rohan assi&ted withtheconstructionof Woetden).
the Lien (t kirns) alunns i ii N C LAM B, GR1BB1 I N & McPHERSON.
We thank Barry Anderson (BASA Technical GR. 199S. Use ol acoustic alartiis u> reduce
Services. Hnsivtnc) tor assisiauce wnh develop- ohnnniQpbuaecelns^lhaalnid'bsy GttoclndinsChonatsktc.ontPrpo.lg4il5bi-e4ti6
ment ol'that company's alarm and Siegfried Eigl (Ahsiiaeu In pribblc N. v.. WcWttetscm,CiR. (&
Iv.'rPtohs-dieov-eNleogpmEleencttroofnitchse,LCiaeinrn(sC)airfnusrjaaslsairstma.nce cioansnecrvatiion.(edS'esc)onSdhaWiorkldi fiishieriesiCcontn^aen^ds
Workshop Proceedings, Brisbane August 1996
LITERATURE CITED Queensland Dept of Pritnarv Indusitiea
CORWIN. J.T. 198!. Audition in clasmobranchs. Pp. [ onk-rence and Workshop Series No QC9 I
81-105. In TivoUn, W.N., Popper, A.N. & Fay, LIEN. J., TODD, S. & CFUIGNE, .1. 199ft fofaencew
K(islKic.--o.-d(.S,p)riInkgeiri-nVnfgeralnadg:sNouenwdYcoormkm)u.nicaii. nhbuomupibapcerkeepwthiadlnes,in flroorrrgte ienectisdceenatnasl. ceastpcthes Wm
DUDLEY. S.F.J. 1998. Shark netting in Natal fixedi :" al mi 1 1 ieitfofnjflsb
Kwa/ulu: an update, and a personal perspective devtcex, and the acpi^liCS of fishing eKcai lvp
\.A.. MicphaeirLrsooI.n,Pp.G3R7-4&4(LAabnster,acBt.),(etdosu)riSbIble, S3e4n7s-o3r6;2. tIbni tFtkhsomofseeIlac&eanKss.st(elPelienn,um Piieds)
FBirsihslebaragineeemscnACiuorgiaugnsrdiexscLio9n9s(WejorivaLQt-uideoineo.npsSlPearcnodocnedDeedWniotnrglsO,df Mcph1e.9r9m9s.oYnoAr,eko)guMri;e, cp,imopde.rini.etsi.ri&f l(oiwmBcBontIe,alanr.ma-..
SPretriiieiasi'NvoInQdiCr9.nK0[e0sL (STBHOC ttld ft I I thop i . ol\ovpeedoastlo.irledwuatcfeiTmNaorlinQeuemeansmlmaanld,hAvue-a.ttrcahliai.n
GEAReIRxNp..eLriPA.mJAe..nKtGhaGl.SmJHO.iV,hcHMnUboG.rtH.h,EeCrS'<n.'<W>KakshMh,iLn.gP,tIoWXn-LqqMO;N,rGm McPFDCRoEnRpuSoimOaNs>s,<nt'Gn5R1Sc/iSeCMntMRitIfoiBctBhCL<oPm.m[]itIfcjtreaa\et,iiAM•a-Ov1 N^1E9h9,9il.iiB
ini I1"/SM12mteothothdes tIontreendiuicieiiobuy.eialteh\.\ bRaelpionritg Qiu(oe)esn.slasnhda;rka bcaolnatnrcoelbetriwsekenmaacnceapgteabmleenlteveilsn
Commission Scientific Committee. May IW99. ofbather safety onsibilitj and
GRBBLE. K.A.. McPHERSON, GR. & LAN] eateh. I'p. 21-35. In GribblL-, N.A.. KlePherson.
I&98. htleet of" the Queensland Slunk Control QF St i^nc, fv (edsj Shwl rnani x'meni and
ji.uvi on nun-target species• whale, dUgong con tion. Second SVortd Fisheries ( ou
and dolphin: a review Journal ol Marine and Workshop Proeeedmgs. Brisbane Aueusi 1996
t Research 49; 645 652, Queensland Dcpi ot"Primary Industrie*
KOSCBINSKi s. & CUL1K, B 1997, Hovi to defer (\-uietencc and Workshop SeriesNo.QC98I'01
harhour porpoise [Pho<<><thi phocowa)\ passive PATERSON. R.. PAT^RSON. P. & CATO. D.H. 1994,
tots vs. (ringers, Report SC/48/SM14 lathe The sUtus oJ humpback wluli
Intern.ttfonnl Wholi tj m mi 9< rntific tHnvuuvJiitc 01 east Australia ihim yeans al'lei
. [996 i svhalme. Biological Consenation 70: 135-142.
KRAIBASL,DWSI1N.0.. KK.i.IM5)0,L0WA.A, .ANSDPRKAROSLOINN..T.h.&, REI WILKINSONI. 11. iI\99Nf.i.R.AJe.o.usShIcLBdEeRl,SmG.:.K &
a illjamson.j. 1995 AfieldtestoFtbeuseof harnital marine n..mnn.il Rshi rV intCtai
506 MEMOIRS OF THE QUEENSLAND MUSEUM
Proceedings of the Seattle Workshop. March minke {Balaenoptera acutorostrata) whales to
1966. (National Marine Fisheries Service and acoustic alarm devices designed to reduce
Marine Mammal Commission USA: Seattle). entrapmentinfishinggear.Pp.727-739. InThomas,
RICHaCA.nIRd.Dn&SoiOTseNH.,O(MAWS.cJaO.d,Ne,mGiRDcE.PEHr.NesE1s9,:95NC.e.RMw.arYJioRrn,ke)Mm.AamLmMaEl,s Jm.aA.m,mKaalstesleenisno,rRy.As.ys&temSsu.pi(nP,lAe.nuYm(ePdrse)ssM:aNreiwn-e
STONE, G, KRAUS, S., HUTT, A., MARTIN, S., York).
YOSHINAGA, A. & JOY, L. 1997. Reducing TYACK, P. 1997. Acoustic communication under the
bycatch: can acoustic pingers keep Hector's sea. Pp. 163-220. In Hopp, S.L., Owren, M.J. &
dolphins out offishing nets. Marine Technology Evans,C.S.(eds)Animalacousticcommunication.
SocietyJournal 31(2): 3-7.
TODD,S.,LIEN,J.&VERHULST,A. 1992.Orientation (Springer-Verlag: Berlin).
of humpback {Megaptera novaeangliae) and