PAPER Acoustic Signatures of Three Marine Arthropods, Squilla mantis (Linnaeus, 1758), Homarus americanus (H. Milne Edwards, 1837), and Nephrops norvegicus (Linnaeus, 1758) (Arthropoda, Malacostraca) AUTHORS ABSTRACT John A. Fornshell The acoustic signatures of three marine crustaceans, Squilla mantis (Linnaeus, NATO Undersea Research Centre 1758), Homarus americanus (H. Milne Edwards, 1837), and Nephrops norvegicus (NURC), Italy (Linnaeus 1758) (Arthropoda, Malacostraca), were experimentally determined in National Museum of Natural History measurements using the calibration tank at the NATO Undersea Research Center, Department of Invertebrate Zoology, La Spezia, Italy. The specimens were insonified at 45° rotational intervals with a Smithsonian Institution, sound source emitting pings from 30 to 120 kHz. For all three species, the value of Washington, DC the nondimensional parameter ka (where k is the acoustic wave number and a is the Alessandra Tesei characteristic dimension of the object) was >5. The absorption spectra, defined as the Paul D. Fox frequencies at which the intensity of the reflected sound was less than 5% of the in- NATO Undersea Research Centre cident intensity, were determined. These spectra changed with the changing aspects (NURC), Italy and were unique for each animal in this study. Two of the species were in the same Institute of Sound and Vibration infraorder, Astacidea. Our results contribute to the development of an acoustic iden- Research (ISVR), University tification system for surveys of marine animals. of Southampton, Highfield, Southampton, United Kingdom know which organisms are living in a local availability and ability to tolerate Introduction given region of the sea, how many are lowered salinities, this species was he detection and identification present, and how they are distributed selected for use in these experimental of marine animals by means of active relative to each other and to various studies. In addition, the American lob- T fi Homarus americanus SONAR methods offers several im- benthic habitats. As a rst step in ster (H. Milne portant possibilities for providing the development of such sampling Edwards, 1837) and Nephrops population data of improved quality techniques, controlled experiments norvegicus (Linnaeus 1758) were also as compared with more traditional in the calibration tank at the NATO selected because of their availability. sampling methods. Acoustic sampling Undersea Research Center (NURC) There have been several studies of methods allow the sampling of a may yield important useful information the sounds produced by crustaceans much larger proportion of the water on the acoustic signatures of three such as those in this study. Stomatopods column without the disadvantages of species, which have been selected for (Patek and Caldwell, 2006) and avoidance and clogging of nets and this study. nephropid lobsters (Mendelson, 1969 bottom trawls. The results of acoustic Squilla mantis (Linnaeus 1758) is and Henninger and Watson, 2006) are sampling are achieved without the found throughout the Mediterranean known to produce sounds by rapid destruction of habitat or removal of Sea (Ungaro et al., 2005; Ambella muscular contractions causing the individuals from the environment. et al., 2006; Atkinson et al., 1997, carapace to vibrate resulting in low Ideally, the researcher would like to Maynou et al., 2004). Because of its frequency sounds <300 Hz. These September/October 2010 Volume 44 Number 5 67 sounds are believed to be defensive methods to sample the plankton. The H. americanus are all members of the in nature (Staaterman et al., 2010; ability to differentiate between taxa class Malacostraca. The last two are Bouwma and Herrnkind, 2009; Patek has been achieved (Wiebe et al., 1996; in the same order, Decapoda, and fam- and Caldwell, 2006). In this study, ac- Stanton et al., 1998a,b; Lavery et al., ily, Nephropidae (Ruppert et al., tive SONAR is being used to identify 2007; Roberts and Jaffe, 2008). Identi- 1996). The morphology of the three the animals. fication to taxonomic categories below species is similar in that they display The density, compressibility, and the phylum or the class level requires a the basic crustacean body plan but speed of sound in these animals are significant amount of data processing sufficiently different that it may be not know with certainty, but Greenlaw and sampling at multiple frequencies expected to produce significantly dif- and Johnson (1982) and Foote (1990) (see Stanton et al. in Medwin, 2005). ferent acoustic signatures. provide measurements for several dif- In our current study, we have Experimental studies on the acoustic ferent marine arthropods showing used the absorption between 30 and signature of various manmade targets, that values for the ratio of the density 120 kHz, notably a significantly nar- some with internal inconsistencies in of the organism to the density of sea row range of frequencies, compared their physical and acoustic properties, water vary between 1.010 and 1.088, withearlierworkwheretwoorthree have displayed acoustic signatures, that values for the ratio of celerity in orders of magnitude frequency ranges which could be related to nonuniformi- the organism to celerity in seawater were employed and generally speaking ties in internal structure of the SONAR vary between 0.997 and 1.075, and implying reduced hardware, signal target. In earlier research projects, that values for the ratio of the com- processing, and data storage demands Whispering-gallery Waves or Rayleigh pressibility in the organism(s) to the than higher frequency methods. The Waves with wave spectra character- compressibility of seawater vary be- species in our study are much more istics of physical nonuniformities in tween 0.850 and 1.075. Maaβ and closely related to each other than the targets were used to identify internal Kuhnapfel (2009, personal communi- those in earlier works where taxa were flaws in the structure of the SONAR tar- cation) give sound speed values for differentiated. S. mantis, N. norvegicus, gets. This resulted from the condition various organs in the human body ranging from 1400 to 1600 m/s. FIGURE 1 These values correspond to 0.93 to 1.07 for a ratio of sound speed in soft The experimental set up in the calibration tank at the NURC. The arrival times at the receiving transducer for the initial signal, target, surface, and bottom reverberation are shown. Data col- tissues to sound speed in seawater. It lection was always terminated before sidewall reverberations arrived. seems reasonable then to assume a value of 1.04 to 1.12 for the ratio of sound speed in the animal to that of fresh water in the tank. A considerable amount of research into the acoustic signature(s) of marine organisms has been accumulated, demonstrating the potential to identify marine animals by their acoustic signa- ture. This work has included experimen- tal studies in tanks, field observations, computer modeling, and theoretical work (Greenlaw and Johnson, 1982; Foote, 1990; Stanton et al., 1998a,b; Stanton and Chu, 2000; Roberts and Jaffe, 2008; Fornshell, 2008; Jones et al., 2009). The work done so far has clearly established the efficacy and the potential of using acoustic detection 68 Marine Technology Society Journal that Rayleigh Waves display dispersion FIGURE 2 in the presence of density/sound velocity (A) Normalized acoustic intensity as a function of wavelength from 20 to 120 kHz for Squilla variations in the medium. Such waves mantis. (B) Normalized intensity as a function of wavelength from 20 to 120 kHz for Homarus are relatively easy to detect in the rever- americanus. (C) Normalized intensity as a function of wavelength from 20 to 120 kHz for Nephrops beration from an object in water (Tesei norvegicus. et al., 2007, 2008). The back scattering may also include Lamb waves, a special form of Rayleigh Waves resulting from the displacement of a boundary between layers of significantly differing densities and sound velocities (Oraevsky, 2002). These displacements propagate both in the plane of the boundary layer and per- pendicular to it. These waves are pro- duced in experimental studies and in field studies in pelagic and benthic en- vironments (Zampolli et al., 2008). The acoustic backscattering from the periwinkle, Littorina littorea, has been shown to be characterized by Lamb waves (Warren et al., 2002). In the pres- ent work, we take advantage of this phe- nomenon to observe the backscattering spectrum of living organisms. A clear view of the knowledge of the spectrum of the backscatter from marine crusta- ceans can best be obtained in carefully controlled tank experiments. With such knowledge, it may eventually be possible to identify objects on or im- bedded in bottom sediments. Experimental Methods The experimental setup and proce- dures follow closely those described in experiments on solid spheres (Tesei et al., 2008). The acoustic signature of each of the three species was mea- sured in the calibration tank at the NATO Undersea Research Center, whichis4.5×3×2.3m,hassteel walls and bottom, and is filled with fresh water. The animals were sus- water at one end of the tank. Although The receiver was an omni-directional pended with a thin nylon wire with its sensitivity is roughly flat between hydrophone with a roughly flat response their dorsal side up and long axis of 40 and 100 kHz, the high signal-to- between 1 and 300 kHz. As shown in the body in the horizontal. The source noise ratio made it possible to obtain Figure 1, it was located