Fisheries - East Carolina Universitycore.ecu.edu/BIOL/luczkovichj/publications/Fisheries.pdfFisheries VOL 31 NO 9 SEPTEMBER 2006 435 2005), and perhaps squid (Iversen et al. 1963)

Fisheries Fish NewsLegislative UpdateJournal HighlightsCalendarJob Center

American Fisheries Society • www.fisheries.org

SEPTEMBER 2006VOL 31 NO 9

LISTENING TO FISH: APPLICATIONS OF

PASSIVE ACOUSTICS

EVALUATING HUMAN ANDBIOLOGICAL OBJECTIVES OF

COOPERATIVE RESEARCH

Cover: Passive acoustics can be an excellent tool for public education. Herechildren listen to underwater sounds as they learn about the estuarineenvironment.

Credit: North Falmouth Congregational Church

Fisheries • VOL 31 NO 9 • SEPTEMBER 2006 • WWW.FISHERIES.ORG 433

INTRODUCTION

Listen to fish? Although most fish-eries biologists are generally aware thatsome fishes are soniferous (sound pro-ducing), relatively few have stopped toconsider the potential importance of lis-tening to fish sounds to their fields ofstudy. In fact, few realize that manyimportant recreational and commercialfishery species are highly soniferous,including many in the cod and drumfamilies. Fishes produce sounds to com-municate with one another while theyare feeding, mating, or being aggressiveand also make incidental noises associ-ated with feeding, swimming, and otherbehaviors (Fine et al. 1977b). Far fromJacques Cousteau's The Silent World(Cousteau and Dumas 1953), our estuar-ies, seas, and oceans are teaming withthe sounds of marine life.

Fish are difficult to see and study inthe ocean. Scuba techniques can help inshallow waters and a range of activeacoustic and optical techniques canassist in deep water, but we are stilllargely ignorant of the distribution andbehavior of the great majority of marinefish. Possibly one of the greatest chal-lenges to researchers attempting to studythe behavioral ecology of fishes is that of

finding the fish in the first place. Oftena scientist must go to great lengths con-ducting expensive and time-consumingbiological surveys simply to determinethe locations or habitats where a fishcan be found, before any attempt tostudy its biotic and abiotic interactionscan be made. After all, you can't studysomething you can't find. Any tool thatcan help scientists to locate fish is there-fore valuable.

A new field of passive acoustics israpidly emerging in fisheries science andmarine biology, in which scientists useunderwater technology to listen in onthe noisy aquatic realm. Passive acous-tics is distinguished from other types ofbioacoustics, because it uses naturallyoccurring sounds to gather informationon fishes and other marine organisms,rather than using artificially generatedsounds. Passive acoustics provides anumber of important benefits for fish-eries research. First, it provides amethod of non-optically observing fishactivity and distribution. That is, it canbe used to find and monitor fishes (andother animals) that produce sound.Second, passive acoustics is a non-inva-sive and non-destructive observationaltool. Third, it provides the capability of

Rodney A. RountreeR. Grant GilmoreClifford A. GoudeyAnthony D. HawkinsJoseph J. LuczkovichDavid A. MannRountree is a senior scientist at MarineEcology and Technology Applications, Inc.,Waquoit, MA. He can be reached [email protected]. Gilmore is asenior scientist, Estuarine, Coastal, andOcean Science, Inc., Vero Beach, FL.Goudey is director, Center for FisheriesEngineering Research, MIT Sea GrantCollege Program, Cambridge, MA.Hawkins is director, Loughine LtdAberdeen, Scotland. Luczkovich is anassociate professor at the Institute forCoastal and Marine Resources, Departmentof Biology, East Carolina University,Greenville, NC. Mann is an assistantprofessor, University of South Florida,College of Marine Science, St. Petersburg.

Passive acoustics is a rapidly emerging field of marine biology that until recentlyhas received little attention from fisheries scientists and managers. In its sim-plest form, it is the act of listening to the sounds made by fishes and using thatinformation as an aid in locating fish so that their habitat requirements andbehaviors can be studied. We believe that with the advent of new acoustic tech-nologies, passive acoustics will become one of the most important and excitingareas of fisheries research in the next decade. However, a widespread lack offamiliarity with the technology, methodologies, and potential of passive acous-tics has hampered the growth of the field and limited funding opportunities.Herein, we provide an overview of important new developments in passiveacoustics together with a summary of research, hardware, and software needs toadvance the field.

FEATURE: FISHERIES RESEARCH

Listening to Fish: Applications of PassiveAcoustics to Fisheries Science

Joe Luczkovich monitors fish sounds in thelaboratory at East Carolina University.

Scott Holt listens to red drum sounds from a fish pier in Texas.

SCO

TT HO

LT, UN

IVERSITY

OF TEX

AS A

T AU

STIN

JOE LU

CZK

OV

ICH

,, EAST C

ARO

LINA

UN

IVERSITY

434 Fisheries • VOL 31 NO 9 • SEPTEMBER 2006 • WWW.FISHERIES.ORG

continuous or long-term monitoring aswell as remote monitoring. Such long-term monitoring can provide importantinformation on daily and seasonal activ-ity patterns of fishes and other marineorganisms. In addition, passive acousticscan also be used to simultaneously mon-itor sources of noise pollution, and tostudy the impact of man's activities onmarine communities. Anthropogenicsources include noise generated by boat-ing activity, seismic surveys, sonars, fishfinders, depth finders, drilling for oil andgas, and military activities. As recentlyreviewed by Popper (2003) theseanthropogenic noise sources all havepotentially important impacts on marinefauna.

The ability to listen to fish and othermarine life allows scientists to identify,record, and study underwater animalseven in the absence of visual informa-tion. Coupling passive acoustics withconventional fishery sampling tech-niques provides a powerful newapproach to fisheries research. We seekto promote a better understanding ofpassive acoustics applications to fish-eries among fisheries professionals, andfeel that passive acoustics will become amajor area of fisheries research in thenext decade. Towards this purpose, weprovide, herein, an overview of someimportant developments in passiveacoustics applications to fisheries andsummarize areas of research and devel-opment that we believe are needed tocatalyze advancement in the field.

BACKGROUND

Over 800 species of fishes from 109families worldwide are known to besoniferous (Kaatz 2002), though this islikely to be a great underestimate. Ofthese, over 150 species are found in thenorthwest Atlantic (Fish and Mowbray1970). Amongst the soniferous fishes aresome of the most abundant and impor-tant commercial fish species, includingmany codfishes, drum fishes, grunts,groupers, snappers, jacks, and catfishes.Some invertebrates with important fish-eries also produce sounds, includingmussels (Mytilus edulis), sea urchins(Fish 1964), white shrimp (Penaeussetiferus, Berk 1998), spiny lobsters(Moulton 1957; Fish 1964; Patek 2002),American lobster (Homarus americanus,Fish 1966; Henninger and Watson

Figure 1. Representative calls of sciaenid fishes. Energy distribution patterns associated withharmonic frequency bands and fundamental frequencies are species specific and can be used toidentify species presence based on the call characteristics.

Figure 2.Representativesciaenid internalanatomy revealing thegas bladder and sonicmuscles of a maleweakfish, Cynoscionregalis.

Figure 3. Temporal overlap pattern of soniferousactivity among sciaenid fishes in a Florida estuary.

Figure 4. Spatial distribution of spawningaggregations of sciaenid fishes asdetermined by soniferous activity in aFlorida estuary. Spawning locations ofdifferent species are represented by ellipseswith different shading patterns.

GRA

NT G

ILMO

RE, ECO

US


2005), and perhaps squid (Iversen et al.1963).

Passive acoustics has been used forover 50 years in fish biology and fish-eries surveys (see Fish et al. 1952 andFish and Mowbray 1970 for a summaryof early work) and is being used rou-tinely today to determine habitat use,delineate and monitor spawning areas,and study the behavior of fishes(Hawkins 1986; Rountree et al. 2003a,2003b). Using hydrophones, marineecologists and fishery biologists havebeen able to listen to the sounds fishesproduce and identify species specific(Hawkins and Rasmussen 1978; Myrbergand Riggio 1985; Lobel 1998), and evenindividual specific (Wood et al. 2002),signatures using signal processing andspectral analysis computer algorithms.Often these sounds dominate the acous-tic environment where they occur, as inthe drum family (Sciaenidae), so muchso that they interfere with military andpetroleum prospecting operations thatinvolve acoustic monitoring (e.g., Fishand Mowbray 1970). In other situations,such as damselfishes on coral reefs, thesounds are not loud and require special-ized techniques to detect them (Mannand Lobel 1995a,b).

To identify the species of fish produc-ing a sound, one first must do “soundtruthing.” There are two main ways thishas been accomplished: (1) captive fishrecordings and (2) in situ (i.e., in thefield under natural conditions) record-ings. Although acoustic complicationsin a tank or aquarium, combined withunnatural behavior and sound produc-tion, make captive fish recordingsproblematic, especially for larger fishes,such problems can be overcome(Okumura et al. 2002). Field recordings,on the other hand, sometimes sufferfrom the difficulty of matching soundsto species and behaviors. Field methodshave been particularly successful intropical waters where environmentalconditions allow the use of advancedscuba and underwater video technolo-gies (Lobel 2001, 2002). Knowledge ofsound source levels is important for cal-culating the detection limits ofhydrophones (e.g., Sprague andLuczkovich 2004). Precise measurementof sound source requires knowledge ofthe location of the fish and of thehydrophones’ characteristics. Despitethese difficulties, biologists have been

able to link the aggressive and spawningbehaviors of some fisheries species totheir sound production using a combina-tion of in situ and tank studies. Forexample sounds produced by haddock(Melanogrammus aeglefinus) duringcourtship and mating have beenrecorded and analyzed in this manner(e.g., Hawkins 1986). Once the associa-tion of sounds to specific species andbehaviors has been established, passiveacoustics provides a rapid way of estab-lishing the spawning component ofessential fish habitat (EFH).

EXAMPLES OF FISHERIESAPPLICATIONS

Passive acoustics studies using rela-tively simple techniques have beensuccessful in locating concentrations ofimportant fish species, opening the wayfor further, more detailed studies of theirbehavior, distribution, and habitat use.Below we provide a brief review of pastand current research on two groups ofsoniferous fishes that support large fish-eries, the drumfishes (sciaenids) andcodfishes (gadids).

Drum fishes (Sciaenids)

History—Sciaenid fishes have beenknown to produce sound for centuries(Aristotle 1910; Dufossé 1874a, 1874b)and the association of sciaenid soundswith spawning has been known nearly aslong (Darwin 1874; Goode 1887). Forhundreds of years the Chinese havelocated sciaenid spawning sites fromtheir water craft by listening to drum-ming sounds emanating from the waterthrough the hull of their boats (HanLing Wu, Shanghai Fisheries Institute,pers. comm.). The location of sciaenidspawning sites using underwater tech-nology is recent and dependent on theavailability of underwater transducers,hydrophones, and acoustic recordersused to access and study underwatersounds (Fish and Mowbray 1970).Hydrophone tape recordings of soundsproduced by large sciaenid aggregationsduring spawning were pioneered byDobrin (1947), Dijkgraaf (1947, 1949),Knudsen et al. (1948), Protasov andAronov (1960), Tavolga (1960, 1980),and Fish and Mowbray (1970).

The location and description ofsoniferous sciaenid aggregations usingmobile hydrophones moving along asound transect at spawning sites was

conducted by Takemura et al. (1978),Mok and Gilmore (1983) and Qi et al.(1984). A portable hydrophone andrecording system was carried via a boatfrom one site to another along a mea-sured transect with recordings madealong a preset grid or in a linear series(Mok and Gilmore 1983; Gilmore2002). Recordings were made for30–300 seconds at each site dependingon transect length. Recorded soundswere verified by auditioning of capturedspecimens. This technique allowed spa-tial and temporal isolation andidentification of species-specific soundsproduced by sciaenid fishes, particularlyunder conditions of high sound attenua-tion for large group sounds (lowfrequency, high intensity sounds). Usingdetailed sonographic analyses of fieldrecordings made on transects, Mok andGilmore (1983) described the character-istic sounds of black drum (Pogoniascromis), spotted sea trout (Cynoscionnebulosus), and silver perch (Bairdiellachrysoura, Figure 1). Subsequent tothese observations, considerable addi-tional work has been done on soundcharacterization in these species, as wellas the weakfish (C. regalis) and the reddrum (Sciaenops ocellata). Passive acous-tic transect techniques have been usedby several investigators to locate spawn-ing sciaenid groups in the field (Saucierand Baltz 1992, 1993; Connaughton andTaylor 1994, 1995; Luczkovich et al.1999, 2000).

Function of Sound Production inSciaenids—The most predictable androbust sounds produced by many fishesare those associated with reproduction.As in many soniferous animals, it is themale that must attract a mate andinduce her to donate eggs for fertiliza-tion, and, therefore, it is often only themale that produces sound. Large choralaggregations of male sciaenid species areformed by spotted seatrout, weakfish,red drum, and silver perch specifically toattract females with which to spawn.

Sciaenid Sound ProductionMechanisms—The most robust andenergetic sciaenid sounds are producedby sonic muscles indirectly or directlyvibrating the membrane of the gas blad-der. When a freshly captured, recentlycalling, male seatrout is dissected, thebright red sonic muscles surrounding thegas bladder can be easily differentiatedfrom the exterior lateral body muscles


(Figure 2). The muscle vibratory rate isdirectly associated with the fundamentalfrequency of the characteristic seatroutcall produced by the gas bladder.

Most of the 270 species in this familyprobably produce sound using sonicmuscles associated with the gas bladder.Using the species specific muscle con-traction rates and the gas bladder shape,sciaenids produce sounds that can beused to identify species within the fam-ily (Mok and Gilmore 1983), as hasbeen demonstrated in amphibians andbirds. The characteristic shape of thesciaenid gas bladder is so conservativethat it has been used as one of the pri-mary morphological characters toclassify sciaenids and to determine theirphyletic relationships (Chu 1963; Chao1978).

When and Where Do SciaenidsProduce Sound?—Mok and Gilmore(1983) demonstrated that sciaenidsound production was specifically associ-ated with crepuscular and nocturnalcourtship and spawning activities.Pelagic eggs and larvae of the spotted

seatrout were collected with planktonnets at spawning sites during periods ofsound production (Mok and Gilmore1983; Alshuth and Gilmore 1993, 1994,1995). The seasonal mating calls weredirectly associated with primary spawn-ing activity in east-central Floridasciaenids. Soniferous activity of varioussciaenid species exhibited distinct sea-sonal patterns, based on an analysis ofover 300 acoustic transects collectedbetween 1978–2002 (Figure 3). Thesesame studies of soniferous spawningaggregations have demonstrated longterm stability of spawning site locations(Figure 4), with the principal spawningsites identified by Mok and Gilmore(1983) being used for over 20 years(Gilmore 2002).

Codfishes (gadids)

A number of gadid species are sonif-erous and the trait is likely widespreadwithin the family (Figure 5; Hawkinsand Rasmussen 1978; Almada et al.1996). A strong correlation betweensoniferous activity and the spawning

cycle has been observed in most speciesstudied to date (Hawkins andRasmussen 1978). Although Atlanticcod (Gadus morhua) have a limitedsound production repertoire, haddockhave been shown to exhibit differentcalls at different stages of courtship(Figure 6; Hawkins et al. 1967; Hawkinsand Rasmussen 1978; Hawkins 1986).Both species have specialized sonic mus-cles that vibrate the swim bladder toproduce sounds for communication(Hawkins 1986). The sonic muscle issexually dimorphic in haddock andAtlantic cod, with males having signifi-cantly larger muscles than females(Templeman and Hodder 1958;Templeman et al. 1978; Rowe andHutchings 2004). In addition, the sonicmuscle undergoes seasonal maturationin concert with the gonad maturationcycle. Evidence for sexual selection ofsonic muscle size has been reported(Engen and Folstad 1999; Nordeide andFolstad 2000; Rowe and Hutchings2004).

Brawn (1961a, b, c) provided the bestdescription to date of the role of soundcommunication in courtship and spawn-ing behavior in Atlantic cod based onlaboratory studies. She found that sonif-erous activity is most common duringthe spawning season, being rare at othertimes, except for a fall “aggressionperiod.” She hypothesized that the fallsoniferous period was related toincreased aggressive interactions prior todispersal to the foraging grounds. In thespawning season soniferous behavior isstrongly associated with spawning andboth spawning activity and call fre-quency peaked during the early eveninghours. Interestingly, soniferous activityis most frequent at night during thespawning season, but most frequent dur-ing the day during the fall “aggressionperiod.” Brawn (1961a, b, c) attributedthis to nocturnal spawning in the win-ter, and diurnal feeding interactionsduring the fall. Unfortunately, Brawndid not provide us with a detailed statis-tical description of Atlantic cod soundcharacteristics. However, later studiesindicate that Atlantic cod producesounds predominantly in the frequencyrange of 80–500 Hz (Fish and Mowbray1970; Hawkins and Rasmussen 1978;Finstad and Nordeide 2004). Nordeideand Kjellsby (1999) recently recordedsounds of Atlantic cod from the spawn-

Figure 5. Representative calls of gadid fishes. Each species can be identified by its unique pulsepattern. For scale, the horizontal bar under each graph represents 100 ms. (Redrawn fromHawkins 1986).


ing grounds off of Norway, and suggestedthat passive acoustics could be used tostudy spawning in the field.

The soniferous behavior of haddockis somewhat better described than thatof Atlantic cod (Figure 6; Hawkins et al.1967; Hawkins and Rasmussen 1978;Hawkins 1986). They have a similar fre-quency range, but can be distinguishedby differences in pulse characteristics(Figure 5; Hawkins and Rasmussen1978). Hawkins and his colleagues arecurrently conducting studies aimed atusing passive acoustics as a tool to iden-tify spawning habitat of haddock inEuropean waters (Hawkins et al. 2002;

Hawkins 2003). In an Arctic fjord innorthern Norway, workers from the FRSMarine Laboratory, Aberdeen and theUniversity of Tromsø have located aspawning ground of haddock. Passive lis-tening has revealed that this species,previously thought to spawn offshore indeep water, can also form large spawningconcentrations close to shore (Hawkinset al. 2002; Hawkins 2003).

Norwegian researchers at theInstitute of Marine Research have pio-neered the use of remote controlledplatforms to obtain video and audio dataon the spawning behavior of Atlanticcod and other gadids important in

European fisheries (Svellingen et al.2002).

A PRIMER ON TECHNIQUES

The success and development of pas-sive acoustics applications to fisheriesdepends on high quality recording sys-tems and analysis software. Thetechnology should be matched to thequestions being asked. For most ques-tions, the needed technology exists foradvancing the field and the main imped-iment is insufficient knowledge on theuse of technology.

Hydrophones

Hydrophones are the most basic ele-ment of any recording system. They areunderwater microphones that typicallyconvert sound pressure into an electricalsignal that can be recorded by a dataacquisition system. Many commerciallyavailable hydrophones can be used forfish bioacoustics studies. When choos-ing a hydrophone, researchers mustconsider its sensitivity and frequencyresponse characteristics to ensure it iswell suited to their needs. Since mostfish sounds range in frequency from 20Hz for the largest fishes to 4 kHz for thesmallest fishes, the hydrophone shouldcover this frequency range. Thehydrophone sensitivity should be suchthat a loud fish sound produces a signalof about 1 V, and thus, a sensitivityrange of around -160 to -170 dBV/�Paat 1 m (decibel volts/micro pascal) isgood for most fisheries applications.

Data storage

Data storage devices include analogand digital tape recorders, audio videorecorders (Lobel 2003), and computerswith sound cards. Digital systems pro-vide obvious advantages over analogsystems in terms of greater frequencybandwidth and dynamic range, and willbe the most commonly used systems inthe future. Which system is chosen willdepend on the recording situation.Computer systems that may be practicalfor recordings made in the laboratorymay not be practical in a field situation,because of power, portability, and envi-ronmental issues. One importantdevelopment is the use of remotely oper-ated vehicles (ROVs), telemetry, andunderwater listening stations for moni-toring sound producing fishes (Mannand Lobel 1995b; Sprague and

Figure 6. Different stages of the courtship behavior of haddock, Melanogrammus aeglefinus, canbe distinguished by differences in the pulse repetition rate of the call. The time-base, scale bar,for the calls is 50 ms. (Redrawn from Hawkins 1986).


Luczkovich 2004; Locascio and Mann2005). These systems will be importantin characterizing which species producewhich sounds, especially for species thatare difficult to maintain in a laboratorytank. They will also prove useful for doc-umenting behavior of fish aggregationswhen multiple individuals call simulta-neously.

One needs to be aware of severalcaveats of recording systems.

1. Data compression: Some recorders(such as mini disc and MP3) use datacompression techniques that alter therecorded sound frequency and level.These would not be appropriatewhen detailed descriptions of soundcharacteristics are required, but couldbe useful for ecological monitoring oftemporal and spatial patterns of wellknown sounds.

2. Automatic Gain Control (AGC):Many systems (especially many ana-log and digital tape recorders andvideo cameras) use AGC to keep therecorded volume within the samerange. If a system uses AGC, it willnot be possible to determine thereceived sound level.

3. Bit resolution: Systems that recordwith a higher bit resolution will havea larger dynamic range (the range ofthe quietest and loudest sounds thatcan be recorded).

One problem that many researchershave encountered is boat induced noiseon recording systems, either throughelectrical noise on the boat, or the phys-ical movement of the boat causing thehydrophone to move. Bungee cords havebeen successfully used to decouple boatmovement and hydrophone movement,and acceleration canceling hydrophonesare commercially available.

Data loggers

Acoustic data loggers are useful forrecording over long periods of time inmany locations simultaneously. Dataloggers provide a way to gather informa-tion on the temporal distributions ofsound producing fish that would not bepossible otherwise without considerableinvestment of human resources. A goodexample of this are the pop-up recorders(Calupca et al. 2000). Computers arethe best option for recording where con-tinuous power is available, such as fromshore or on a large boat. Commercial

software (see below) exists for recordingon a given duty cycle. However, contin-uous power is rarely available in fieldsituations where one would like to makerecordings. In these situations, lowpower battery operated acoustic dataloggers are required.

To date all acoustic data loggers thathave been used for passive acousticshave been engineered by individual lab-oratories to perform this task. Theseinclude analog tape recorders that havebeen modified to record on a particularduty cycle (Luczkovich et al. 2000), anddigital dataloggers that have been pro-

grammed to record as desired (Mannand Lobel 1995b; Calupca et al. 2000;Sakas et al. 2005). No data loggers canbe purchased and used directly withouteither engineering or software program-ming (usually both). This lack of an off-the-shelf product has greatly limitedtheir use in passive acoustic applicationsto fisheries.

Telemetry

Telemetry systems broadcast ahydrophone signal to a boat or shorebased receiver. They perform the samefunction as data loggers in allowingrecordings over a large area for longperiods of time. Several types of teleme-try systems are available includingsonobuoys (VHF), cell phone systems,and short range microwave systems. Allof these systems require line of sightbetween the transmitter and receiver,and a relatively high level of engineer-ing to set up and maintain. Telemetry isalso capable of delivering video to docu-ment behavior during sound production(Svellingen et al. 2002). Satellite sys-tems generally do not support thebandwidth needed for transmittingacoustic data. At this point, someamount of preprocessing would berequired, so that limited data on soundcharacteristics (e.g., root-mean-squareamplitude, frequency spectra) could betransmitted.

Hydrophone Arrays

Fixed and towed hydrophone arrayshave been used for determining thelocations of vocalizing whales in manydifferent situations (e.g., Watkins andSchevill 1972; Clark 1980), but haveonly recently been applied to fishes(Barimo and Fine 1998; Mann andJarvis 2004). Hydrophone arrays holdpromise in answering questions aboutfish distributions that could not be oth-erwise obtained with single hydrophonerecordings (D'Spain et al. 1996), butrequire a high level of sophistication forsetting up, operating, and analyzing thedata. The minimum spacing betweenhydrophones in an array needed forlocalization is calculated as the speed ofsound in water (approximately 1,500m/s) divided by the sound frequency. Forexample, to locate a fish calling at 1,000Hz (e.g., silver perch), a minimum spac-ing of 1.5 m is required, while a spacing

Examples of data logger use in passiveacoustics.

TOP: Autonomous underwater listeningsystem (AULS) designed by CliffGoudey.

MIDDLE: Commercial fishermen John andMoe Montgomery of the F/V Chandelledeploying an AULS on Jefferies Ledgein the Gulf of Maine as part of acooperative study of haddockspawning.

BOTTOM: AULS deployed in theMacGregor Point coral reef in Maui aspart of the Sanctuary Sounds project.

CEN

TER FOR FISH

ERIESEN

GIN

EERING

RESEARC

H,

MIT SEA

GRA

NT

KATIE A

ND

ERSON

, UN

IVERSITY

OF M

ASSA

CH

USETTS AT

AM

HERST

NO

RM G

ARO

N, H

AW

AII ISLA

ND

HU

MPBA

CK

WH

ALE N

ATIO

NA

L MA

RINE SA

NC

TUA

RY


of 12 m is required to locate a fish call-ing at 125 Hz (e.g., red drum).

Signal Processing Software

Many packages are available for dataacquisition and signal processing. Somesuch as MATLAB (Mathworks, Inc.)have a great deal of flexibility andpower, but require a high level of pro-graming knowledge. Others are targetedspecifically at bioacousticians includingSignal (Engineering Design), Raven(and its precursor Canary; CornellUniversity), SASLab Pro (AvisoftBioacoustics), and Adobe Audition (for-merly Cool Edit). The manuals to thesesoftware programs are often the clearestsource of information for learning signalprocessing techniques and their applica-tion.

FUNDING PRIORITIES ANDRESEARCH NEEDS

Research presented at an interna-tional workshop on the “Applications ofPassive Acoustics to Fisheries” in April2002 underscored the great strides thathave been made in the application ofpassive acoustics to fisheries and relatedissues over the last two decades(Rountree et al. 2003a, b). The work-shop was followed by a specialsymposium on passive acoustics applica-tions to fisheries held at the AmericanFisheries Society Annual Meeting inQuebec in August 2003 and organizedby Luczkovich, Mann, and Rountreethat again highlighted the rapidadvances in the field (Luczkovich et al.unpublished). And most significantly,important new initiatives in passiveacoustics have begun in many areas ofthe United States. Although passiveacoustics is currently underutilized as aresearch tool, the success of the work-shop and symposium demonstrates howrapidly the field is emerging and suggestsgreat promise for future research.

As a result of discussions and corre-spondences initiated at the workshopand symposium, we have prepared acomprehensive description of areas thatparticipating scientists felt needed to beaddressed to advance the field of passiveacoustics in fisheries. These areas can becategorized as research, software, hard-ware, and education/outreach needs.

Research Needs

Research needs can be simplisticallysummarized as the questions: what,when, where, and how many?

What?—What fishes and inverte-brates are making sounds? To date weknow of approximately 800 species offishes worldwide that produce sounds.More soniferous species are reported ona frequent basis, but there are no currentsystematic efforts to catalogue soniferousfishes. The landmark studies by MarieFish and William Mowbray (1970)ended over 30 years ago. No one hastaken up the gauntlet since then.Knowledge of what produces underwatersounds is critical to the success of higherlevel studies. Currently the number ofunidentified underwater soundsattributed to fishes is far greater thanthose that can be positively identified.Rountree and Goudey (unpublished)recently recorded unknown sounds onthe commercial fishing grounds off thecoast of Massachusetts. Even moreremarkably, Rountree and Juanes(unpublished) and their studentsrecorded many unknown sounds in areasthat have been extensively studied by

conventional means (e.g., Woods Hole,Massachusetts), or that are in the veryheart of the industrial world (the dockson Manhattan Island, New York;Anderson et al. unpublished). How canit be that we know so little about fishesin these areas that we can not evenidentify some of the most frequent andwidespread fish sounds? As biologists, wefind the lure of the unknown so close tothe doorstep to be a powerful motivator.Although some progress is being madein this area, the most pressing needs are:

1. an effort to catalogue historic recordsof known and unknown sounds;

2. systematic efforts to identify and val-idate sound sources;

3. studies to determine the correlationbetween sound production and spe-cific behaviors; and

4. studies to determine under whatbehavioral conditions fish/inverte-brates make sounds.

Efforts are now underway to digitizeand catalogue sound archives from labo-ratories across North America and inEurope (Rountree et al. 2002; Bradburyand Bloomgarden 2003). The MaCaulayLibrary of the Cornell Lab ofOrnithology has recently made a largecollection of fish sounds from thesearchives available to the public online(www.birds.cornell.edu/MacaulayLibrary/).Unfortunately, many of these historicrecordings are poorly documented andwere made with long outdated technolo-gies. More importantly, they areinsufficient for the needs of fisheriesresearchers, as they are only partiallycomplete. Many of our most importantcommercial fishes that are soniferoushave yet to be recorded. Information onsoniferous freshwater fisheries species isalmost entirely lacking in NorthAmerica.

To expand the catalogue of fish andinvertebrate sounds, field and laboratorystudies are needed to identify unknownsounds, and audition fishes and inverte-brates for sound production. Systematicefforts to identify sounds in each geo-graphic region, including estuaries andtidal fresh waters, are critical to theadvancement of passive acoustics.Evidence that a particular sound is madeby, and unique to, a given species is ofcritical importance. Lack of, or per-ceived lack of, such evidence is the most

Data loggers have been used to studydeep sea fishes in cooperation withcommercial red crab fishermen.

TOP: Cliff Goudey demonstrates datalogger to Daher Jorge of the F/VKrystal James.

BOTTOM: Juan Espana aboard the F/VHanna Boden prepares to deploy adata logger attached to the inside of ared crab pot.

ROD

NEY

ROU

NTREE


common criticism encountered in fund-ing proposals.

There are two major ways to go aboutcataloguing fish sounds in a particularregion. The first, is through the system-atic auditioning of animals collected inthe field, in aquaculture facilities, and inpublic aquaria. Fish and Mowbray(1970) used this method to study thesounds of over 200 species of fishes. Onemajor problem with such efforts is thatfailure to record fish sounds during audi-tions does not indicate the species is notsoniferous, because soniferous behavioris often controlled by sex, age, matura-tion stage, and behavior. Sometimes fishare not physiologically ready for soundproduction outside of the spawning sea-son. Additionally, auditioning programslike that carried out by Fish andMowbray (1970) are sometime criticizedas producing artificial sounds throughelectrical stimulation; however, we notethat at least 47 of 150 species regardedas soniferous by Fish and Mowbray(1970) produced sounds spontaneously.

The second major method of cata-loguing fish sounds is to conduct fieldsurveys to identify temporal and spatialpatterns of sound production. Onceunknown sounds are identified and theirtemporal and spatial patterns described,research can be conducted to determinethe identity of the fish, including theauditioning of fish captured in theappropriate place and time. The advan-tage of this approach is that the field ofpotential sound producers is narroweddown, and the researcher knows that thetarget species is at least physically capa-ble of sound production at the time ofauditioning. However, efforts byresearchers to catalogue underwatersounds in freshwater and marine habi-tats are hampered by the failure offunding agencies to recognize the impor-tance of such studies. Data onunidentified fish sounds are especiallydifficult to publish, even when extensiveobservations are available. We suggestthat such studies are valuable in theirown right as they provide the basicinformation necessary to systematicallysurvey soniferous fishes by providingdata on when and where soniferousactivity occurs. Armed with such data,scientists can devise studies to deter-mine the identity of the soundproducers. The only alternative is tosimply blindly capture fishes and audi-

tion them for sound production. Weliken the denial of the usefulness of cat-alogues of unidentified fish sounds tothe absurd notion that collections ofichthyoplankton for which identifica-tions are not currently available are notuseful.

Incidental sounds made by fishes canbe just as important as soniferous behav-ior. For example, feeding sounds can beused to determine foraging times, loca-tions, and consumption rates (Sartoriand Bright 1973; Mallekh et al. 2003;Anderson et al. unpublished).

Often investigators have used termssuch as “grunts,” “knocks,” “snaps,”“pops,” “staccato,” “drumming,” “hum-ming,” “rumbles,” “percolating,”“purring,” etc., to describe the soundsheard; the names often being ono-matopoeic. Standardizing these sounddescriptions would allow rapid commu-nication between biologists and otherobservers (Anderson et al. unpublished).

When?—Temporal patterns in soundoccurrence are needed for two main rea-sons. First, knowledge of when fishesand invertebrates make sounds leads toknowledge about their biology and ecol-ogy. Second, we need this informationto effectively use passive acoustics as atool to locate fish, to identify their habi-tat requirements, and to conductpresence/absence and abundance sur-veys (see below). Some of the mostimportant needs here are:

1. Studies to determine the relationship ofsound production to fish size. Earlystudies have shown that sound char-acteristics sometimes change withfish size (e.g., dominant frequency)and hence the relationship betweenfish size and sound characteristicsmay be used to determine length fre-quency data for soniferous fishes(Fish and Mowbray 1970; Lobel andMann 1995; Connaughton et al.2000).

2. Studies to determine the relationship ofsound production to sex and maturitystage (e.g., Connaughton and Taylor1995). Do both male and females in apopulation produce sounds? Are thesounds the same or different? Doimmature fish make sounds? Are thesounds made by immature and matureindividuals different? The answers tothese questions can provide scientistswith valuable information on the

temporal and spatial distribution pat-terns of fishes by sex and maturitystages. They can also provide datauseful in studies of reproductive ecol-ogy and studies specifically importantfor fisheries assessment, such as thequantification of fish fecundity.

3. Quantification of daily and seasonalpatterns in sound production (Breder1968; Fine et al. 1977a;Connaughton and Taylor 1994). Ifwe know when a fish produces soundsthen studies of the daily and seasonalpatterns of that soniferous activitycan be correlated with daily and sea-sonal patterns of that specificbehavior. For example, if it is knownthat a particular sound is only pro-duced when a fish is spawning, thenthe determination of the daily pat-tern in soniferous activity can beused to infer the daily pattern ofspawning (i.e., to determine whattime of day the fish spawns).Similarly, seasonal patterns in sonif-erous activity can provide an index ofthe spawning season period (Figure3).

Where?—Identifying where a fishoccurs is one of the most fundamentallyimportant topics in fish ecology andfisheries management. It is the first steptowards identifying essential fish habitat(EFH) as mandated of fisheries managersby the reauthorization of the MagnusonStevens Fishery Conservation andManagement Act (Oct. 1996). EssentialFish Habitats are defined as “thosewaters and substrate necessary for fishfor spawning, feeding or growth to matu-rity.” The lowest level criterion foridentification of EFH is simply presence/absence. At the minimum, passiveacoustics surveys can be used to identifythe presence of soniferous fishes in habi-tats. If the behavior associated with aparticular fish sound is known, thenhigher levels of identification, such asthe identification of spawning habitat,can be achieved.

To accurately associate underwatersounds with habitats, it is necessary toknow the range at which the sound canbe detected by the survey instruments.A lack of quantification of sound sourcedetection ranges is the second mostcommonly cited criticism encounteredin funding proposals. To determinesound source detection ranges several


types of studies are needed. (a) Howloud is it?—Quantification of soundsource levels for a givenspecies/stage/environment is the firststep. (b) Sound Propagation—how fardoes sound travel under a given set ofenvironmental conditions. This is thestudy of tomography. Studies in shallowwater are especially needed. (c) FishShoals-studies are needed to examinehow the multiple sound production ofmany fish in a shoal combine. How dosound pressure levels vary with shoal sizeand distance from the source?

Passive acoustic mapping of the spa-tial distribution of sounds is one of themost important potential products ofpassive acoustics technology. Studies tomap the distribution of fish sounds, andhence the distribution of soniferousfishes and their essential fish habitat arestrongly needed. Passive acoustics offersmany advantages over traditional meth-ods of mapping fish distribution. It isnon-destructive, non invasive, and rela-tively inexpensive over the long runcompared to traditional methods such astrawl surveys. If the “what” and “when”are known, then passive acoustics can beused to map EFH based onpresence/absence for soniferous species.The “when” is important because spatialsurveys with passive acoustics are onlyvalid if conducted during the window oftime when fish are soniferous. In manycases, fishes restrict their soniferousactivity predominately to narrow win-dows of time. For example, spottedseatrout (Cynoscion nebulosus) calls pre-dominantly from just before sunset intothe early evening (Mok and Gilmore1983), so attempts to use passive acous-tics to map their distribution wouldhave to be conducted within that timeframe.

How many?—Quantification of fishabundance at a location and time is animportant objective of many fisheriesstudies and monitoring programs. Oncethe what, when, and where questionshave been addressed, additional studiescan quantify abundance through passiveacoustics techniques. The two mostimportant study areas are: (1) correla-tion between the “amount” of soundproduction and the abundance of fish,and (2) determination of the proportionof fish that are soniferous at a givenplace and time.

In the simplest situation, the numberof fish calls can be counted and corre-lated to the true amount of fish present.This requires the ability to distinguishindividual calls. More rigorously itwould also require the ability to distin-guish individuals that call repeatedlyfrom those that don't. Often, however,fish and calls are so numerous that indi-vidual calls cannot be distinguished. Inthese cases, studies are needed to deter-mine the relationship between soundpressure level and fish abundance. Forexample, Luczkovich et al. (1999)showed a relationship between soundpressure level and egg abundance, andindicated that a relationship shouldexist between sound production andspawning stock size.

As mentioned above, soniferousbehavior in many fishes is sex depen-dent. Often, only the males call. Inaddition, in some cases soniferous activ-ity may be related to size and maturationstage, so that at a given place and time,the number of fish calling is a functionof the size and maturation stage distribu-tion of the population. However,individual fishes fail to produce soundsin a given situation for many reasons.Studies are needed to allow researchersto estimate the proportion of soniferousand non soniferous fishes over a giventime period.

SOFTWARE NEEDS

To carry out the research prioritieslisted above, software must be developedin several areas. Programs should bepackaged to allow biologists to processsound data to the fullest extent possible.

Automatic processing of soundrecordings—The passive acoustics stud-ies described above can generate largeamounts of acoustic data (often thou-sands of hours of recordings) that arelaborious to process fully to obtain thebasic biological data sought (e.g., tem-poral activity patterns, spatialdistribution, etc.). We suggest that soft-ware designed to address the followingprocessing needs would greatly enhancethe utility of passive acoustics to fish-eries.

1. Automatic signal detection softwareused to detect the call of a given fishspecies amid long time series of soundrecordings (thousands of hours) with

multiple sound sources and high lev-els of “noise” is essential.

2. Temporal pattern analysis software totrack the temporal occurrence pat-terns of hundreds to thousands ofdetected calls within long time seriesis critical to the efficient determina-tion of daily and seasonal patterns insound production.

3. Call characterization software tospeed up the process of measuring callcharacteristics such as duration, pulserate, pulse repetition, fundamentalfrequency, pulse width, sound level,etc., in large numbers of calls wouldincrease statistical power in studies oflocation, behavior, and environmen-tal effects on call characteristics.

4. Call classification software to aid inthe identification of a new unknownsound through comparison of its callcharacteristics with the characteris-tics of calls contained in standardreference libraries of known andunknown sounds.

Localization of sound sources—Software and hardware developmentsare needed to simplify the localization ofsound sources received by multiplehydrophone arrays. Automatic signaldetection and classification software areimportant prerequisites. Currently local-ization is a laborious process ofdetermining the starting points of thesame sound received among multiplehydrophones and then using various sta-tistical techniques such as time delaydifferences to triangulate on the calls.To develop the ability to produce realtime GIS plots of soniferous fishes over-lain on environmental parameters, thesesteps must be automated. One importantproduct of localization, other than habi-tat association mapping, is thedetermination of true sound source lev-els (see discussion “Where?”) undervarying environmental conditions (e.g.,Sprague and Luczkovich 2004).

HARDWARE NEEDS

Acoustic hardware technology israpidly evolving; however, there are sev-eral specific needs that would enhancethe development of passive acoustics infisheries.

Data storage—Low cost, high capac-ity digital recorders are needed to allowfield sampling at the rates and intensi-ties necessary to study temporal and


spatial patterns of sound production.Devices that can be programmed torecord at varying sampling rates andtime settings, including continuous andtime lapse would be particularly valu-able. Existing low-cost recordersdeveloped by the music industry arerapidly becoming scarce due to theindustry shift from wave file storage for-mats to MP3 file format storage (MP3 isless desirable because it compresses anddistorts the sounds).

Coupled “visualization” technolo-gies—To validate the identification ofsound sources in the field, devices thatcouple passive acoustics with othertechnologies that can be used to identifya fish are needed. Coupled acousticoptic systems include devices that arecapable of recording both underwatervideo and acoustic data are among thesimplest to use and are highly reliablefor identification (Lobel 2001).Amazingly, most underwater video tech-nology marketed to researchers lacksacoustic recording capabilities. In ouropinion this is entirely due to a failure ofthe manufacturers to recognize the mar-ket potential of such devices, ratherthan to technical issues. Another diffi-culty of acoustic-optic systems is theneed to use artificial light sources for thevideo recording, and resulting dramaticchange in fish behavior and probabilityof detection. In addition, artificial lightprovides a limited range of visibility.Infrared lights are only effective for afew feet, and even state-of-the-art lowlight level cameras are ineffective atnight even at moderate depths. Finallystudies are needed to understand theeffect of light of various wavelengthsand intensities on fish behavior todevelop the optimum optic system.

Coupling passive acoustics with othervisualization technologies such as activeacoustics and laser line-scanning is alsopromising. Active acoustic technologyhas the advantage of providing rela-tively long range observationcapabilities compared with video, butrequires the input of acoustic energyinto the environment. Laser line-scan-ning can provide high resolutionimaging at short ranges (<10 m; Careyet al. 2003), but is currently relativelyexpensive.

Portable passive acoustic surveydevices—Small, self-contained packagesof acoustic gear would allow mobile

recording of acoustic and video duringshore or small boat based studies.Currently investigators have to purchaseseparate components and assemble theminto a package themselves. Becausecomponents were not manufactured towork together, numerous in-lineadapters often must be used to transferthe acoustic signal among components.Researchers need to be able to drop ahydrophone over the side of a small boator dock, together with a small underwa-ter video camera, and be able tosimultaneously record and monitorvideo and acoustic data. Ideally, theaudio and video data would be automat-ically converted to digital form forstorage. Availability of low cost devicesof this type would facilitate the feasibil-ity of small-scale research projects for alarger scientific community.

EDUCATION ANDOUTREACH

Most fish bioacousticians are biolo-gists first and engineers second. Theyhave arrived at fish bioacoustics becauseit is a powerful tool for studying fishes.This means that engineering and signalprocessing principles must be learned onthe job. Unfortunately, there is no onegood source of information aboutrecording and signal processing that isaccessible and practical for the fish bioa-coustician. This gap can be bridged both

by producing these targeted materials,conducting training workshops, and byattracting engineers with a biologicalinterest to the field. The workshop par-ticipants identified a strong need toeducate scientists, managers, and thepublic on the uses of passive acoustics.Some specific needs include:

1. Manuals and other literature intendedto introduce biologists to the field;

2. Workshops that bring biologists, acous-ticians, and engineers together are vitalas they stimulate the transfer of knowl-edge and ideas among the disciplines;

3. Development of passive acoustic train-ing centers for researchers andstudents; and

4. Incorporation of passive acoustics intofisheries curricula.

Passive acoustics technologies pro-vide a unique public outreach potential.Scientists and laymen alike are oftenfascinated by the phenomenon of under-water sounds. Passive acousticstechnologies are amenable to multime-dia display via the Internet and havegreat potential as public education andoutreach tools.

ACKNOWLEDGMENTS

We wish to acknowledge the contri-bution of numerous participants in thePassive Acoustics Workshop and AFSSymposium. Numerous discussions atthese meetings, and in subsequent corre-spondences since, were invaluable to theauthors in the development of thismanuscript. The workshop and publica-tion of the workshop proceedingsreceived major funding from MIT SeaGrant, the Office of Naval Research,and from the Northeast Great LakesCenter of the National UnderseaResearch Program. Travel for someworkshop participants was funded inwhole, or in part by: Connecticut SeaGrant, Florida Sea Grant, Hawaii SeaGrant, Louisiana Sea Grant, NorthCarolina Sea Grant, the South CarolinaSea Grant Consortium, Texas SeaGrant, and the Woods Hole Sea GrantPrograms. This is SMAST ContributionNo. 06-0601.

Selected examples of soniferous fishes thatsupport important recreational or commercialfisheries (fish sketches courtesy ofNOAA/NMFS Northeast Fisheries ScienceCenter).



REFERENCES

Almada, V. C., M. C. P. Amorin, E. Pereira,F. Almada, R. Matos and R. Godinho.1996. Agonistic behaviour and soundproduction in Gaidropsarus mediterraneus(Gadidae). Journal of Fish Biology49:363-366.

Alshuth, S., and R. G. Gilmore. 1993. Eggidentification, early larval developmentand ecology of the spotted seatrout,Cynoscion nebulosus C. (Pisces:Sciaenidae). International Council for theExploration of the Sea, ConferenceManuscript 1993/G 28.

_____. 1994. Salinity and temperaturetolerance limits for larval spotted seatrout,Cynoscion nebulosus C. (Pisces:Sciaenidae). International Council for theExploration of the Sea, ConferenceManuscript 1994/L:17.

_____. 1995. Egg and early larvalcharacteristics of Pogonias cromis, Bairdiellachrysoura, Cynoscion nebulosus (Pisces:Sciaenidae), from the Indian RiverLagoon, Florida. International Council forthe Exploration of the Sea, ConferenceManuscript 1995/L:17.

Aristotle. Historia Animalism, IV, 9. Trans.by D. D'A.Thompson. 1910. ClarendonPress, Oxford.

Barimo, J. F., and M. L. Fine. 1998.Relationship of swim-bladder shape to thedirectionality pattern of underwater soundin the oyster toadfish. Canadian Journal ofZoology 76:134-143.

Berk, I. M. 1998. Sound production by whiteshrimp (Penaeus setiferus), analysis ofanother crustacean-like sound from theGulf of Mexico, and applications forpassive sonar in the shrimp industry.Journal of Shellfish Research17:1497-1500.

Bradbury, J. W., and C. A. Bloomgarden.2003. Creating a web-based library ofunderwater biological sounds. Pages103-106 in R. A. Rountree, C. Goudey,and T. Hawkins, eds. Listening to fish:proceedings of the international workshopon the applications of passive acoustics tofisheries. 8-10 April 2002. Dedham, MA.MIT Sea Grant Technical Report.

Brawn, V. M. 1961a. Aggressive behaviour inthe cod (Gadus callarias L.). Behaviour18:107-147.

_____. 1961b. Reproductive behaviour of thecod (Gadus callarias L). Behaviour 18:177-198.

_____. 1961c. Sound production by the cod(Gadus callarias L.). Behaviour 18:239-245.

Breder, C. M., Jr. 1968. Seasonal and diurnaloccurrences of fish sounds in a smallFlorida Bay. Bulletin of the AmericanMuseum of Natural History 138:278-329.

Calupca, T. A., K. M. Fristrup and C. W.Clark. 2000. A compact digital recordingsystem for autonomous bioacousticmonitoring. Journal of the AcousticSociety 108:2582.

Carey, D. A., D. C. Rhoads and B. Hecker.2003. Use of laser line scan for assessmentof response of benthic habitats anddemersal fish to seafloor disturbance.Journal of Experimental Marine Biologyand Ecology 285-286:435-452.

Chao, L. N. 1978. A basis for classifyingwestern Atlantic sciaenidae (Teleostei:Perciformes). NOAA Technical ReportNMFS Circular 415.

Chu, Y. T., Y. L. Lo, and H. L. Wu. 1963. Astudy on the classification of the sciaenoidfishes of China, with description of newgenera and species. 1972 Reprint,Antiquariaat Junk, Netherlands.

Clark, C. W. 1980. A real-time directionfinding device for determining the bearingto the underwater sounds of southern rightwhales, Eubalaena australis. Journal of theAcoustical Society of America 68(2):508-511.

Connaughton, M. A., and M. H. Taylor.1994. Seasonal cycles in the sonic musclesof the weakfish, Cynoscion regalis. FisheryBulletin 92:697-703.

_____. 1995. Seasonal and daily cycles insound production associated withspawning in the weakfish, Cynoscionregalis. Environmental Biology of Fishes42:233-240.

Connaughton, M. A., M. H. Taylor, and M.L. Fine. 2000. Effects of fish size andtemperature on weakfish disturbance calls:implications for the mechanism of soundgeneration. Journal of ExperimentalBiology 203:1503-1512.

Cousteau, J. Y., and F. Dumas. 1953. Thesilent world. Perennial Library, Harperand Row, Publishers, New York.

Darwin, C. 1874. The descent of man. 2ndedition. H. H. Caldwell, New York.

Dijkgraaf, S. 1947. Ein tone erzeugenderfisch in Neapler Aquarium. Experimenta3:494.

_____. 1949. Untersuchungen uber diefunktionen des ohrlabyrinths beimeeresfischen. Physiologia ComparataOecoloia 2:81-106.

Dobrin, M. B. 1947. Measurements ofunderwater noise produced by marine life.Science 105:19-23.

D’Spain, G. L., W. A. Kuperman, L. P.Berger, and W. S. Hodgkiss. 1996.Geoacoustic inversion using fish sounds.The Journal of the Acoustical Society ofAmerica 100:2665.

Dufossé, A. 1874a. Rescherches sur les brutset les sons expressifs que font entendre lespoissons d'Europe et sur les organesproducteurs de ces phenomenesacoustiques ainsi que sur les appareils del'audition de plusieurs de ces animaux.Annales des Sciences Naturelles. Zoologieet Biologie Animale Series 5, Vol. 19.

_____. 1874b. Rescherches sur les bruts et lessons expressifs que font entendre lespoissons d'Europe et sur les organesproducteurs de ces phenomenesacoustiques ainsi que sur les appareils del'audition de plusieurs de ces animaux.Annales des Sciences Naturelles. Zoologieet Biologie Animale Series 5, Vol. 20.

Engen, F., and I. Folstad. 1999. Codcourtship song: a song at the expense ofdance? Canadian Journal of Zoology77(4):542 550.

Fine, M. L., H. E. Winn, L. Joest and P. J.Perkins. 1977a. Temporal aspects ofcalling behavior in the oyster toadfish,Opsanus tau. Fishery Bulletin 75:871-874.

Fine, M. L., H. E. Winn and B. Olla. 1977b.Communication in fishes. Pages 472-518in T. Sebok, ed. How animalscommunicate. Indiana University Press,Bloomington.

Finstad, J. L., and J. T. Nordeide. 2004.Acoustic repertoire of spawning cod,Gadus morhua. Environmental Biology ofFishes 70:427-433.

Fish, J. F. 1966. Sound production in theAmerican lobster Homarus americanus.Crustaceana 11:105.

Fish, M. P. 1964. Biological sources ofsustained ambient sea noise. Pages175-194 in W. N. Tavolga, ed. Marine bio-acoustics. Pergamon Press, NY.

Fish, M. P., A. S. Kelsey, Jr., and W. H.Mowbray. 1952. Studies on theproduction of underwater sound by NorthAtlantic coastal fishes. Journal of MarineResearch 11:180 193.

Fish, M. P., and W. H. Mowbray. 1970.Sounds of western North Atlantic fishes.Johns Hopkins Press, Baltimore,Maryland.

Gilmore, R. G, Jr. 2002. Sound productionand communication in the spottedseatrout. Pages 177-195 in S. A. Bortone,


ed. Biology of the spotted seatrout, CRCPress, Boca Raton, Florida.

Goode, G. B. 1887. American fishes: apopular treatise upon the game and foodfishes of North America with specialreference to habits and methods ofcapture. L. C. Page Co., Boston.

Hawkins, A. D. 1986. Underwater soundand fish behaviour. Pages 114-151 in T. J.Pitcher, ed. The behaviour of teleostfishes. Groom Hellm, London.

______. 2003. The use of passive acoustics toidentify a haddock spawning area. Pages49-53 in R. A. Rountree, C. Goudey, andT. Hawkins, eds. Listening to fish:proceedings of the international workshopon the applications of passive acoustics tofisheries. 8-10 April 2002. Dedham, MA.MIT Sea Grant Technical Report MITSG03-2.

Hawkins, A. D., L. Casaretto, M. Picciulin,and K. Olsen. 2002. Locating spawninghaddock by means of sound. Bioacoustics12:284-286.

Hawkins, A. D., C. J. Chapman, and D. J.Symonds. 1967. Spawning of haddock incaptivity. Nature 215:923 925.

Hawkins, A. D., and K. J. Rasmussen. 1978.The calls of gadoid fish. Journal of theMarine Biological Association of theUnited Kingdom 58:891 911.

Henninger, H. P., and W. H. III., Watson.2005. Mechanisms underlying theproduction of carapace vibrations andassociated waterborne sounds in theAmerican lobster, Homarus americanus.The Journal of Experimental Biology208:3421-3429.

Iversen, R. T. B., P. J. Perkins, and R. D.Dionne. 1963. An indication ofunderwater sound production by squid.Nature 199(4890):250-251.

Kaatz, I. M. 2002. Multiple sound producingmechanisms in teleost fishes andhypotheses regarding their behaviouralsignificance. Bioacoustics 12:230-233.

Knudsen, V. O., R. S. Alfred, and J. W.Emling. 1948. Underwater ambient noise.Journal of Marine Research 7:410-429.

Lobel, P. S. 1998. Possible species specificcourtship sounds by two sympatric cichlidfishes in Lake Malawi, Africa.Environmental Biology of Fishes52:443-452.

______. 2001. Fish bioacoustics and behavior:passive acoustic detection and theapplication of a closed-circuit rebreatherfor field study. Marine Technology Journal35(2):2-19.

______. 2002. Diversity of fish spawningsounds and the application of passiveacoustic monitoring. Bioacoustics 12:286-289.

______. 2003. Synchronized underwateraudio-video recording. Pages 136-139 inR. A. Rountree, C. Goudey, and T.Hawkins, eds. Listening to fish:proceedings of the international workshopon the applications of passive acoustics tofisheries. 8-10 April 2002. Dedham, MA.MIT Sea Grant Technical Report MITSG03-2.

Lobel, P. S., and D. A. Mann. 1995.Spawning sounds of the damselfish,Dascyllus albisella (Pomacentridae), andrelationship to male size. Bioacoustics6(3):187-198.

Locascio, J. V., and D. A. Mann. 2005.Effects of Hurricane Charley on fishchorusing. Proceedings of the RoyalSociety of London Biological Letters1:362 365.

Description• fellowships for highly qualified Ph.D.-level graduate students

interested in careers in: (1) population dynamics of living marineresources and development and implementation of quantitativemethods for assessing their status, and (2) economics ofconservation and management of living marine resources

• support for up to three years for Population Dynamics fellowships,and up to two years for Marine Resource Economics fellowships

• approximately two fellowships awarded each year in eachdiscipline, with overall maximum of 12 Fellows at any time

• fellows work closely with mentors from NMFS Science Centers orLaboratories and may intern at NMFS facility on thesis research orrelated problem

Program goals• encourage qualified applicants to pursue careers in and increase

available expertise related to: (a) population dynamics andassessment of status of stocks of living marine resources, or (b) economic analysis of living marine resource conservation and management decisions

• foster closer relationships between academic scientistsand NMFS

• provide real-world experience to graduate students and accelerate their career development

Eligibility• must be United States citizen• prospective Population Dynamics Fellows must be

admitted to Ph.D. program in population dynamics orrelated field (applied mathematics, statistics, orquantitative ecology) at academic institution in

United States or its territories• prospective Marine Resource Economics Fellows must be in

process of completing at least two years of course work in Ph.D.program in natural resource, marine resource, or environmentaleconomics or related field

Award• grant or cooperative agreement of $38,000 per year awarded to

local Sea Grant program/host university• 50% of funds provided by NMFS, 33 1/3% provided by National

Sea Grant Office (NSGO), and 16 2/3% provided by university asrequired match of NSGO funds

• disbursement of award for salary, living expenses, tuition, healthinsurance, other fees, and travel determined by university

Relevant dates• application deadline—early February 2007 (see Sea Grant website

for details—www.seagrant.noaa.gov/funding/rfp2006.html)• fellowship start date: 1 June 2007

Contact• Dr. Terry Smith

National Sea Grant College Program1315 East-West HighwaySilver Spring, MD 20910301/[email protected]

• any state Sea Grant program—www.nsgo.seagrant.org/SGDirectors.html

• any participating NMFS facility—www.nmfs.noaa.gov/science.htm

National Marine Fisheries Service (NMFS) / Sea Grant Joint Graduate Fellowship Program in Population Dynamics and Marine Resource Economics


Luczkovich, J. J., M. W. Sprague, S. E.Johnson, and R. C. Pullinger. 1999.Delimiting spawning areas of weakfishCynoscion regalis (Family Sciaenidae) inPamlico Sound, North Carolina usingpassive hydroacoustic surveys.Bioacoustics 10:143-160.

Luczkovich, J. J., J. D. Hall, III., M.Hutchinson, T. Jenkins, S. E. Johnson,R. C. Pullinger, and M. W. Sprague.2000. Sounds of sex and death in the sea:bottlenose dolphin whistles suppressmating choruses of silver perch.Bioacoustics 10:323-334.

Mallekh, R., J. P. Lagardere, J. P. Eneau,and C. Cloutour. 2003. An acousticdetector of turbot feeding activity.Aquaculture 221:481-489.

Mann, D. A., and S. M. Jarvis. 2004.Potential sound production by a deep-seafish. Journal of the Acoustical Society ofAmerica 115(5):2331-2333.

Mann, D. A., and P. S. Lobel. 1995a. Passiveacoustic detection of fish soundproduction associated with courtship andspawning. Bulletin of Marine Science57(3):705-706.

_____. 1995b. Passive acoustic detection ofsounds produced by the damselfish,Dascyllus albisella (Pomacentridae).Bioacoustics 6:199-213.

Mok, H. K., and R. G. Gilmore. 1983.Analysis of sound production in estuarineaggregations of Pogonias cromis, Bairdiellachrysoura, and Cynoscion nebulosus(Sciaenidae). Bulletin of the Institute ofZoology, Academia Sinica 22:157-186.

Moulton, J. M. 1957. Sound production inthe spiny lobster Panulirus argus(Latreille). Biological Bulletin 113:286-295.

Myrberg, A. A., and R. J. Riggio. 1985.Acoustically mediated individualrecognition by a coral reef fish(Pomacentrus partitus). Animal Behavior33:411-416.

Nordeide, J. T., and I. Folstad. 2000. Is codlekking or a promiscuous group spawner?Fish and Fisheries 1:90-93.

Nordeide, J. T., and E. Kjellsby. 1999. Soundfrom spawning cod (Gadus morhua L.) atthe spawning grounds. ICES Journal ofMarine Science 56:326-332.

Okumura, T., T. Akamatsu, and H. Y. Yan.2002. Analyses of small tank acoustics:empirical and theoretical approaches.Bioacoustics 12:330-332.

Patek, S. N. 2002. Squeaking with a slidingjoint: mechanisms and motor control ofsound production in palinurid lobsters.

Journal of Experimental Biology205:2375-2385.

Popper, A. N. 2003. Effects of anthropogenicsounds on fishes. Fisheries 28(10):24-31.

Protasov, V. R., and M. I. Aronov. 1960. Onthe biological significance of sounds ofcertain Black Sea fish. (In Russian).Biofizika 5:750 752.

Qi, M., S. Zhang, and Z. Song. 1984. Studieson the aggregate sound production of mostspecies of croaker (sciaenoid fishes) inBohai Sea, Yellow Sea and East ChinaSea. Studia Marina Sinica, Peking21:253-264.

Rountree, R. A., P. J. Perkins, R. D.Kenney, and K. R. Hinga. 2002. Soundsof western North Atlantic fishes: datarescue. Bioacoustics 12(2/3):242 244.

Rountree, R. A., C. Goudey, T. Hawkins, J.Luczkovich and D. Mann. 2003a.Listening to fish: passive acousticapplications in marine fisheries. Sea GrantDigital Oceans. Massachusetts Institute ofTechnology Sea Grant College Program.MITSG 0301. Available at:http://web.mit.edu/seagrant/aqua/cfer/acoustics/PD9x9FINAL.pdf

Rountree, R. A., C. Goudey, and T.Hawkins. Editors. 2003b. Listening tofish: proceedings of the internationalworkshop on the applications of passiveacoustics to fisheries. 8-10 April 2002.Dedham, MA. MIT Sea Grant TechnicalReport MITSG 03-2. Available at:http://web.mit.edu/seagrant/aqua/cfer/acoustics/PAprocBrFINAL.pdf

Rowe, S., and J. A. Hutchings. 2004. Thefunction of sound production by Atlanticcod as inferred from patterns of variationin drumming muscle mass. CanadianJournal of Zoology 82:1391-1398.

Sakas, C. J., C. Goudey and R. A.Rountree. 2005. Sanctuary soundsmonitoring underwater sounds in theNational Marine Sanctuaries. Oceans2005 Marine Technology Society/Instituteof Electrical and Electronics Engineers,Inc. Conference Proceedings. MTS.

Sartori, J. D., and T. J. Bright. 1973.Hydrophonic study of the feedingactivities of certain Bahamian parrotfishes, family Scaridae. Hydro-LaboratoryJournal 2:25-56.

Saucier, M. H., and D. M. Baltz. 1992.Hydrophone identification of spawningsites of spotted seatrout Cynoscionnebulosus (Ostichthys: Sciaenidae) nearCharleston, South Carolina. NortheastGulf Science 12(2):141-145.

_____. 1993. Spawning site selection byspotted seatrout, Cynoscion nebulosus, andblack drum, Pogonias cromis, inLouisiana. Environmental Biology ofFishes 36:257-272.

Sprague, M. W., and J. J. Luczkovich. 2004.Measurement of an individual silver perchBairdiella chrysoura sound pressure level ina field recording. Journal of the AcousticSociety of America 116(15):3186-3191.

Svellingen, I., B. Totland and J. T. Oevredal.2002. A remote-controlled instrumentplatform for fish behaviour studies andsound monitoring. Bioacoustics 12:335-336.

Takemura, A., T. Takita, and K. Mizue.1978. Studies on the underwater soundVII: Underwater calls of the Japanesemarine drum fishes (Sciaenidae). Bulletinof the Japanese Society of ScientificFisheries 44:121-125.

Tavolga, W. N. 1960. Sound production andunderwater communication in fishes.Pages 93-136 in W. E. Lanyon and W. N.Tavolga, eds. Animal sounds andcommunication. American Institute ofBiological Sciences, Washington, D.C.

_______. 1980. Hearing and soundproduction in fishes in relation to fisheriesmanagement. Pages 102-123 in J. E.Bardach, J. J. Magnuson, R. C. May, and J.M. Reinhart, eds. Fish behavior and its usein the capture and culture of fishes.International Center for Living AquaticResources Management ConferenceProceedings 5, Manila, Philippines.

Templeman,W., and V. M. Hodder. 1958.Variation with fish length, sex, stage ofsexual maturity, and season in theappearance and volume of the drummingmuscles of the swimbladder in thehaddock, Melanogrammus aeglefinus L.Journal of the Fisheries Research Board ofCanada 15:355-90.

Templeman, W., V. M. Hodder, and R.Wells. 1978. Sexual maturity andspawning in haddock, Melanogrammusaeglefinus, of the southern Grand Bank.The International Commission for theNorthwest Atlantic Fisheries ResearchBulletin 13:53-65.

Watkins, W. A., and W. E. Schevill. 1972.Sound source location by arrival-times ona non-ridid three-dimensionalhydrophone array. Deep-Sea Research19:691-706.

Wood, M., L. Casaretto, G. Horgan, and A.D. Hawkins. 2002. Discriminatingbetween fish sounds a wavlet approach.Bioacoustics 12:337-339.

Documents

Fisheries - East Carolina Universitycore.ecu.edu/BIOL/luczkovichj/publications/Fisheries.pdfFisheries VOL 31 NO 9 SEPTEMBER 2006 435 2005), and perhaps squid (Iversen et al. 1963)