8/15/2019 Deteksi Biosonar

1/7

Target Strength of a Nylon Monofilament and an Acousticallyn ance net: re ct ons o osonar etect on anges

errence A. Mooney,1,2 Paul E. Nachtigall,1 and Whitlow W. L. Au1

awa nst tute o ar ne o ogy, n vers ty o awa , . . ox , a ua, , Department of Zoology, University of Hawaii, Edmondson 152, 2538 The Mall, Honolulu, HI 96822, USA

A stract

Thousands of marine mammals die each year ins er es-re ate entang ements. su stant a

number of these animals entangle themselves ing nets. wo popu at ons n mme ate angerare the coastal stock of the mid-Atlantic bottle-nose o p n, urs ops truncatus, an t e u oMaine harbour porpoise, Phocoena phocoena. einvestigated the efficacy of using an alternative netmaterial made with barium sulphate hypothesizedto be acoustically more reflective than traditionalnets. By using simulated dolphin echolocationclicks, the target strength of the experimental netwas compared with the target strength of a simi-lar gauge nylon net. Results demonstrated that atangles greater than normal incidence, but less than40°, the new barium sulphate net was acousticallymore re ect ve t an t e ny on net; owever, t erewas no significant difference in the target strengtho t e two nets at . t ang es greater t an ,both nets were difficult to discern from back-groun no se. arget strengt s o t e nets wereused to calculate detection ranges for . truncatusan . p ocoena. ot spec es s ou e a e todetect the experimental nets at a distance greatert an t e ny on nets. or . truncatus, t s stance

may be enough to reduce entanglement; however,ecause o . p ocoena s ower source eve ec o-

location signals, they may not detect either netwith echolocation in time to avoid contact.

Key Words:  entanglement, gillnet, take reduc-tion, urs ops truncatus, Phocoena phocoena,target strength, acoustic reflectivity

ntro uct on

The bycatch of marine mammals by fisheries isa pro em resu t ng n t reatene or en angerepopulations and species of cetaceans around theg o e. part cu ar nterest are sma ec o ocat-ing odontocetes, which often entangle and drownn nets, part cu ar y g nets ar ng et a ., ).

In the USA, the incidental take and mortalityt e coasta m - t ant c ott enose o p n,

urs ops truncatus, and Gulf of Maine harbourporpo se, ocoena p ocoena, ave een en-

tified by the National Marine Fisheries Service) as excee ng y g . e ycatc es o

these two populations by bottom-set gillnet fish-r es are greater t an w at s eeme susta na e

to the populations (Cain, 2002; Waring et al.,1999). Several measures have been tested, withvariable success, in attempts to reduce the gillnetbycatch of the bottlenose dolphin and the harbourporpoise.

Reducing bycatch of odontocetes in gillnetsis a particularly challenging problem because ofthe properties of the nets, and how they are setincreases target catch along with dolphin entan-

lements. The nets often are set in waters of poorv s ty or at ept , w ere g t eve s are ow(Kastelein et al., 2000). Thus, animals may not bea e to etect nets w t t e r v s on. t ona y,

illnets are constructed of nylon monofilament,w c tra t ona y as a wea target strengt ;therefore, regular gillnets reflect sounds, such asc o ocat on c c s, poor y, an ec oes may eifficult to perceive by odontocetes. It is possible

t at ec o ocat ng cetaceans, n part cu ar, w not

perceive gillnets as an obstacle because the echoesrom t e nets are re at ve y wea u ones,

1991).Previous research on reducing porpoise and

olphin bycatch has had variable success (Cox etal., 2001; Gearin et al., 2000; Kraus et al., 1997;Trippel et al., 1999). Because herring catch andharbour porpoise bycatch peaks coincide tempo-rally (Trippel et al., 1996), fishing ground clo-sures might reduce bycatch; however, these clo-sures w ave e eter ous soc a an econom cimpacts on local fisheries. Acoustic alarms, orp ngers, appear to re uce ar our porpo se mor-tality (Kraus et al., 1997), but many concerns existregar ng t e r ong-term e ect veness. otent a

rawbacks include cost, practicality, and vari-a ty n success o re uc ng ycatc awson

 Aquatic Mammals 2004, 30(2), 220-226, DOI 10.1578/AM.30.2.2004.220

 

8/15/2019 Deteksi Biosonar

2/7

8/15/2019 Deteksi Biosonar

3/7

t en re ates to source eve s. ssum ng t e sameDI and DTE, and inserting the newly calculatedsource levels, equation (2) can be rewritten asfollows:

 TLDL = (SL L – SLKB) + ({TStt}DL – {TStt}KB)– DL - B) + KB  )

his equation is used to calculate detectionrange for the bottlenose dolphin. In equation (4),SL refers to source level and subscripts DL and KBrefer to a different location and Kaneohe Bay,respectively. Because bottlenose dolphins havet e a ty to vary ntens ty o ec o ocat on c c s(Moore & Pawloski, 1990), predictions of detec-t on ranges or ott enose o p ns were con uctefor a wide range of source levels. The noise leveln a erent ocat on was var e rom re at ve y

calm, quiet seas (27 dB) to rougher seas (33 dB) toaccount or erent sea states. wenty-seven

of noise is roughly the ambient noise in Beaufortsea states - , w c s e ore w te caps appearon the water. Thirty-three dB is the ambient noiseo eau ort sea states > , w en w te caps arevisible.

quation (4) is solved for the transmission lossin a different location using an α of 0.03 dB/m fordeep water temperatures of 5 °C. This value wasinserted into equation (3) to determine detectionranges of both nets for a bottlenose dolphin in

aneohe Bay and in locations where take reduc-tion needs to be implemented.

P. phocoena  predictions employed a slightlydifferent method because the harbour porpoise’sec o ocat on c c pea requenc es are c arac-teristically around 130 kHz, and ambient noisen t s porpo se s env ronment s typ ca y muc

lower (Kastelein et al., 1999). Further, targetetect on exper ments ave not een pu s e

on the harbour porpoise in a noisy environment.us, we on y app e ar our porpo se pre c-

tions to a quiet environment using a simple esti-mate o two-way transm ss on oss ncorporate

into the following equation: = og + α   )

dditionally, we used anα of 0.038 dB/m basedon the peak frequency of 130 kHz and ambientwater temperatures of 15 °C.

revious research (Kastelein et al., 1999)showed that for a 5.08-cm diameter stainless-steelsphere, the 50% correct detection threshold forharbour porpoises was 15.9 m and the 90% cor-rect detection threshold was 12.0 m. The TS ofthis sphere was measured at -36.6 dB re: 1µPa.Inserting these values into equation (3), we deter-m ne t e transm ss on oss t at wou occur asa harbour porpoise is echolocating on the sphere.

en, was eterm ne to e . w en t erange was 16 m and 44.1 dB when the range was

m.

e erence n etween t e nets ansphere can be added to transmission loss to resolvethe transmission loss that would occur when P.

hocoena echolocates on a net. With transmissionloss, equation (3) can then be solved for the range

f detection for a regular monofilament nylon netr barium sulphate net. Range calculations wereetermined for each net at the respective angles

measured. Additionally, because detection rangeistances have been established for the 5.08-cmiameter target at both 50% and 90% correctetect on eve s, we pre cte ar our porpo se0% and 90% detection distances for both nets.

esu ts

Target Strengt Resu tst norma nc ence, t ere was no s gn cant -

ference in target strength between the two nets> . ). ean o t e ny on net was - .

B re: 1µPa (n = 20; SD = 1.1) and mean TS oft e ar um su p ate net was - . n = ;

2.1) (Figure 1). When the angle of incidenceincreased to 10°, 20°, and 30°, there was a signifi-ant difference (Table 1) between the mean TS of

the two nets, regardless of angle (  < 0.001). At anangle of 40°, there was no significant difference inTS between the two nets ( p  > 0.05). The mean TS

f the nylon net was -61.9 dB (n = 20; SD = 5.7)and the TS of the barium sulphate net was -61.1

B (n = 20; SD = 6.1). Beyond 40°, both nets hadlittle to no discernable echo relative to the back-

roun no se measure n aneo e ay.

222  Mooney et al.

Figure 1. Target strength in dB re: 1µPa of barium sulphate

nd nylon nets; solid line: barium sulphate net; dashed line:

nylon net. Error bars represent standard deviations of the

mean for TS measurements. Statistically significant results

re marked with a star.

8/15/2019 Deteksi Biosonar

4/7

P ocoena p ocoenaWe predicted very little difference in detectionrange o t e two nets at rom norma a e ).At 50% detection rates, maximum detection rangeo t e exper menta ar um su p ate net was .

m whereas it was 6.8 m for the nylon net detec-t on range. e max mum range o etect on ssubstantially less at 90% detection; for the bariumsu p ate net, t e est mate etect on range was4.9 m and the regular nylon was 5.0 m.

s t e ang e o nc ence ncrease to ,

20°, and 30°, the predicted detection ranges oft e ar um su p ate net were cons era y greaterthan the nylon net.

e pre cte etect on range o ot netswas similar at 40° from normal incidence. At

pro a e etect on, ranges were as ow as3.1 m for the barium sulphate net and 2.9 m fort e ny on net. re cte etect on ranges wereslightly higher at the 50% detection probability:. m or t e ar um su p ate net an . m or

the nylon net.

urs ops truncatusTo predict detection ranges for the bottlenosedolphin, both source and noise level were varied.Source level was varied because bottlenose dol-phins have the ability to produce clicks over a

range o ntens t es oore aw os , ),and the source levels used predominantly in thewild have not been established. We used two dif-ferent noise levels to compensate for varying envi-ronmental noise conditions.

At normal incidence, the maximum estimatedetection range for a bottlenose dolphin was 77.8

m for the nylon net and 76.4 m for the barium sul-phate net, calculated with a source level of 210

B and noise level of 27 dB (Table 3). Althoughthe difference in predicted detection range is rela-t ve y sma , t e ar um su p ate net ma nta nea greater predicted detection range compared tot e ny on net w en t e ang e o nc ence wasincreased to 10°, 20°, and 30°.

n mum est mate etect on range or otnets was at the lowest source level (170 dB),

reatest no se eve ), an o norma .

At this distance, we predicted the barium sulphatenet wou e etecte at . m an t e ny on netat a similar 4.2 m (Figure 2).

Predicted Detection Ranges of Acoustically Enhanced Gillnet 23

a e .  ean target strengt s o exper menta an contro

nets; data in dB re: 1µPa. Y: p

8/15/2019 Deteksi Biosonar

5/7

scuss on

t ang es greater t an t e norma nc ence, t ebarium sulphate net had an increased target strengthcompare to t e ny on net. us, t appears t at t ebarium sulphate nets do reflect dolphin echoloca-t on s gna s etter t an regu ar mono ament nets.Both nets had relatively weak target strengths fromo p n- e c c s, owever, even at , w en

reflected acoustic energy was the greatest.t ona y, at t ere was no s gn cant er-

ence in TS between the two nets and the predicteddetection distance. At angles 40°, the TS of bothnets was essentially the same as the backgroundnoise and, thus, probably difficult for a dolphin toperceive. We predicted that the barium sulphatenet should be detected at a greater range by bot-tlenose dolphins and harbour porpoises when thenet is approached from angles greater than 0°, butess t an . t , ot nets re ecte t e same

amount of energy and would be detected at the

same range. eve s o ec o ocat on- e s gna sreflected from either net at angles ≥ 40° were notgreater t an t e ac groun no se, at east n ourexperimental situation, and potentially in other

g no se env ronments.n this study, source levels were varied for

ott enose o p ns ecause t e amp tu e o s g-nals produced by wild animals in the area of thenets s un nown. or t e ar our porpo se, w cproduces lower source levels, we predicted theyw on y etect nets at a re at ve y c ose stance.Maximum detection distance is predicted to beabout 10 m. In noisy seas, detection distance maybe considerably less. Harbour porpoises swim atspeeds up to 4.3 m/sec. At this speed, and in areasof high noise and using low echolocation sourcelevels, a harbour porpoise may not be able to

etect e t er a mono ament or ar um su p atenet before making contact with the net. Even so,

ur predictions should be taken as only an indi-ator of detection range because Phocoena-likelicks were not used to estimate harbour porpoiseetection distances. TS values from both dolphin

and harbor porpoise signals are predicted to bevery similar, however (Au, 1994). Thus, it is validto apply TS results obtained with a dolphin signalto predict harbour porpoise detection ranges aslong as the basis of the prediction is understood.

ott enose o p ns genera y em t g ercholocation source levels than the harbour por-

po se. ea amp tu es o ott enose o p nshave been measured from 170 to 210 dB re: 1µPa

oore aw os , ). en source eve sare high, detection distances may be as far as 80 mor ot types o nets. en ott enose o p ns

mit echolocation clicks of higher SLs, a bariumsu p ate net may re uce ycatc entang ementrates over monofilament nets. It is important tonote t at t e etect on range epen s on a ogquation, so detection ranges vary quite a bit with

source level and, thus, the detection range pre-icted for bottlenose dolphins can be as low as 6

m in noisy seas.Bottlenose dolphins travel at speeds of up to

4 km/hr for short bursts of speed (Lockyer &Morris, 1987); however, for sustained effort andminimum energy expenditure, a dolphin wouldhave to travel at an average 2.1 m/sec (Williamst al., 1992). Observations of a wild lone dolphinst mate average spee to e - m r or

about 2.7-5.5 m/sec (Lockyer & Morris, 1987).en trave ng at ower ve oc t es ust over

m/sec), bottlenose dolphins should be able toetect a g net an c ange course e ore t con-

tacts the net, even when its peak SL is only 170. en ee ng an presuma y trave ng at

reater speeds (5 m/sec), this may still be possi-e. t e an ma appens to ma e a urst o spee

at the wrong time, however, the animal might beuna e to etect a net n t me to avo t.

At lower peak source levels, the net materialmay become more important. For instance, at lowervelocities, when the source level is 180 dB, noiselevel is 27 dB, and angle of incidence is 20°, thebarium sulphate net was predicted to be detected 3m further or almost 1 s sooner than a nylon net. Ata 170 dB source level, 20° from normal incidence,and 33 dB noise level, the difference in predicted

etect on range s m, or a most s more t me toetect and avoid a barium sulphate net as opposed

to a ny on net. ese exper menta nets may pro-vide the extra distance needed for an animal toavo entang ement. t g er source eve s, tappears that echolocating animals may detect bothnets at a reasona e stance.

224  Mooney et al.

Figure 2. Predicted maximum detection range of a barium

su p ate net at norma nc ence at ncreas ng source eve s

(dB re: 1µPa) by urs ops truncatus; diamonds: 27dB of

no se; tr ang es: o no se

8/15/2019 Deteksi Biosonar

6/7

t rema ns to e eterm ne w et er t ereduced harbor porpoise bycatch observed whenbarium sulphate nets were used (Cox & Read,2001; Trippel et al., 2003) was due to detection viaecholocation and consequent avoidance or otherfactors. Because detection of a barium sulphatenet by an echolocating animal might not alwaysbe possible, at least in time to avoid net contact,other nonmutually exclusive factors might havecaused the observed reduction in harbour porpoisebycatch. Cox & Read (2001) suggested that stiff-ness or co orat on may e t e om nant actor nreduced bycatch. For example, nets constructed ofst er ne, suc as ar um su p ate n use ny on,may have a reduced tendency to collapse aroundan entang e a o p n. tu es are necessary toexplore these explanations. Further, the hypothesist at ar um su p ate nets re uce ycatc y e ng

more detectable relies on the assumption that thean ma s ec o ocat ng or searc ng or an acous-tic target. It is important to note that previous workas s own t at t ese ar um su p ate nets a so

significantly reduce seabird bycatch (Trippel etal., 2003) and also alleviate marine turtle bycatch.Thus, even if acoustic reflectivity does not appearto be the actual contributing factor to reducingbycatch in the gillnet fishery, barium sulphate netsshow promise as a socially and economically fea-sible method of reducing incidental take of marineanimals.

Ac now e gments

This work was supported by the National Marines er es erv ce grant num er

to Paul E. Nachtigall from the Southeast Regionrotecte esources v s on s at y ang,

with assistance from Aleta Hohn. Additionalun ng came rom , nc., an t e

Project AWARE Foundation to T. Aran Mooney.e aut ors wou e to t an orm o y or

information on his barium sulphate net and fors comments on t e manuscr pt; on aste e n,

Arturo Serrano, and Jeanette Thomas in reviewerroles; and Finn Larsen and Ed Trippel for helpfulinformation.

Literature Cited

Au, W. W. L. (1988). Detection and recognition models

f dolphin sonar systems. In P. E. Nachtigall &

. . . oore s. , n ma sonar: rocesses an

 performance (pp. 753-768). New York: Plenum Press.

u, . . . . e sonar o o p ns. ew or :

pringer.

u, . . . . onar etect on o g nets y o p ns:

heoretical predictions.  Report of the International

a ng omm ss on   pec a ssue , - .

u, . . ., ones, . . coust c re ect v ty o

nets: Implications concerning incidental take of dol-

phins. Marine Mammal Science, 7 (3), 258-273.

ain, W. L. (2002). Marine mammals: Notice of intent to

prepare an environmental impact statement for a take

reduction plan for the western north Atlantic coastal

stock of bottlenose dolphins. Federal Register , 67 (140;

fr22jy02-34), 47772-47773.

ox, T. M., & Read, A. J. (2001). Echolocation of porpoises

around acoustically enhanced gillnets.  14th Biennial

Conference on the Biology of Marine Mammals, Society

or ar ne amma ogy, ancouver, ana a.

ox, T. M., Read, A. J., Solow, A. J., & Tregenza, N. (2001).

ar our porpo ses ocoena p ocoena   a tuate

to pingers? Journal of Cetacean Research Management ,

, - .

Dawson, S., Read, A. J., & Slooten, E. (1998). Pingers,

porpo ses an power: ncerta nt es w t us ng p ng-

ers to reduce bycatch of small cetaceans.  Biologicalonservat on , - .

earin, P. J., Gosho, M. E., Laake, J. L., Cooke, L., Delong,

. ., ug es, . . . xper menta test ng o

acoustic alarms (pingers) to reduce bycatch of harbour

porpoise,Phocoena phocoena, in the state of Washington.

 Journal of Cetacean Research Management , 2, 1-9.

astelein, R. A., Au, W. W. L., & de Haan, D. (2000).

Detection distances of bottom-set gillnets by harbour

porpoises (Phocoena phocoena) and bottlenose dolphins

urs ops truncatus .  Marine Environmental Research

49, 359-375.

astelein, R. A., Au, W. W. L., Rippe, H. T., & Schooneman,

N. M. (1999). Target detection by an echolocating

ar or porpo se ocoena p ocoena . ourna o t e

 Acoustical Society of America, 105(4), 2493-2498.

raus, . ., ea , . ., o ow, . ., a w n, .,

Spradlin, T., Anderson, E., & Williamson, J. (1997).

coust c a arms re uce porpo se morta ty. ature

525.

oc yer, ., orr s, . . servat ons on v ng

behaviour and swimming speeds in a wild juvenile

urs ops truncatus   quat c amma s , - .

Moore, P. W. B., & Pawloski, D. A. (1990). Investigationson t e contro o ec o ocat on pu ses n t e o p n

( urs ops truncatus). In J. Thomas & R. A. Kastelein

(Eds.),  Sensory abilities of cetaceans (pp. 305-316)  

New York: Plenum Press.

Trippel, E. A., Strong, M. B., Terhune, J. M., & Conway,

J. D. (1999). Mitigation of harbour porpoise (Phocoena

 phocoena   bycatch in the gillnet fishery in the lower

Bay of Fundy. Canadian Journal of Fisheries Aquatic

Science, 59, 113-123.

r ppe , . ., ang, . ., trong, . ., arter, . .,

Conway, J. D. (1996). Incidental mortality of harbour

porpo se ocoena p ocoena   y t e g net s ery n

the lower Bay of Fundy. Canadian Journal of Fisheries

quat c c ence , - .

Trippel, E. A., Holy, N. L., Palka, D. L., Shepard, T. D.,

e v n, . ., er une, . . . y on ar um

Predicted Detection Ranges of Acoustically Enhanced Gillnet 25

http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0824-0469()7:3L.258[aid=6155858]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0824-0469()7:3L.258[aid=6155858]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0824-0469()7:3L.258[aid=6155858]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0824-0469()7:3L.258[aid=6155858]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0006-3207()84L.141[aid=5099427]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0006-3207()84L.141[aid=5099427]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0006-3207()84L.141[aid=5099427]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0141-1136()49L.359[aid=6155853]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0141-1136()49L.359[aid=6155853]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0141-1136()49L.359[aid=6155853]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-4966()105:4L.2493[aid=6155856]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-4966()105:4L.2493[aid=6155856]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-4966()105:4L.2493[aid=6155856]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-4966()105:4L.2493[aid=6155856]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-4966()105:4L.2493[aid=6155856]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-0836()388L.525[aid=5099430]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-0836()388L.525[aid=5099430]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0167-5427()13:1L.31[aid=6155855]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0167-5427()13:1L.31[aid=6155855]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0824-0469()7:3L.258[aid=6155858]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-4966()105:4L.2493[aid=6155856]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0001-4966()105:4L.2493[aid=6155856]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0167-5427()13:1L.31[aid=6155855]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0006-3207()84L.141[aid=5099427]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0006-3207()84L.141[aid=5099427]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0141-1136()49L.359[aid=6155853]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0141-1136()49L.359[aid=6155853]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-0836()388L.525[aid=5099430]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-0836()388L.525[aid=5099430]

8/15/2019 Deteksi Biosonar

7/7

su p ate g net re uces porpo se an sea r morta ty.

 Marine Mammal Science, 19(1), 240-243.

Waring, G. T., Palka, D. L., Clapham, P. J., Swartz, S.,

ossman, M. C., Cole, T. V. N., Bisack, K. D., & Hansen,

. J. (1999). U.S. Atlantic marine mammal stock assess-

ment. NOAA Technical Memorandum (NMFS-NE 116).

Williams, T. M., Friedl, W. A., Fong, M. L., Yamada, R. M.,

edivy, P., & Haun, J. E. (1992). Travel at low energetic

ost by swimming and wave-riding bottlenose dolphins.

ature , - .

226  Mooney et al.

http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-0836()355L.821[aid=4982080]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-0836()355L.821[aid=4982080]http://www.ingentaselect.com/rpsv/cgi-bin/linker?ext=a&reqidx=0028-0836()355L.821[aid=4982080]