Page 29 - Acoustics Today

heard use this great analogy). Animals in the ocean (depend- ing on their anatomy, composition, and physiology) scatter sound with different efficiencies at different frequencies (Chu et al., 1992; Horne and Jech, 1999; Benoit-Bird, 2009a;). Using different acoustic frequencies provides us with the ability to identify and discriminate between different sizes such as year classes of a long-lived species or different types, and in some cases even species, of nekton and zooplankton. In general, with more frequencies one has a greater ability to identify and discriminate different scatterers (Holliday, 1977; Foote and Stanton, 2000).
Discrete broadband systems have been around for decades (Napp et al., 1993; Holliday and Pieper, 1995; Holliday et al., 1989; Au and Benoit-Bird, 2008;) using multi- ple single-frequency or narrowband transducers to span sev- eral decades in frequency. The use of multiple acoustic fre- quencies with ground truthing have allowed scientists to use acoustic backscatter data to identify and discriminate scatter- ing from multiple size classes (Greenlaw, 1979; Kristensen and Dalen, 1986; Warren et al., 2003) or types of scatterers (Pieper et al., 1990; Martin et al., 1996) and even scattering from biological and physical sources (Warren et al., 2003). A thorough review of acoustic species identification is provid- ed by Horne (2000). Recently however, we have begun to see the development of new systems that allow broadband data to be collected and studied providing insights into predator foraging (Stanton et al., 2010; Lavery et al., 2010). As these technologies continue to develop, the data collected during acoustic studies will continue to increase our ability to inter- pret acoustic backscatter data.
Sampling platforms for the assessment of fish and zoo- plankton have traditionally been ship-based allowing scientists to cover large areas over the course of several weeks, but limit- ing the temporal duration of these studies. The development of autonomous underwater vehicles (AUV) and Glider-based samplers allows scientists to cover much larger areas although there are power and engineering constraints which will limit the number and types of frequencies of echosounders used in these systems (Brierley et al., 1998). Observatory and moored systems while limited in spatial coverage can provide unique insights into daily, seasonal, or annual changes in the ecosys- tem that cannot always be observed during shorter-term sur- veys occurring once a year. A very different application of underwater sound (compared to echosounder surveys) for biological assessment has been the use of the ocean water col- umn as a wave guide with lower frequency sources and receiv- er arrays. These large-scale experiments cover 100s of square kms which can ensonify multiple schools of fish nearly instan- taneously (Makris et al., 2006).
At the other spatial scale extreme, some scientists like myself have become interested in ecological interactions between nekton and zooplankton and their predators which occur at very small space (10–100 m2) and time scales (sec- onds to minutes) (Warren et al., 2009; Hazen et al., 2011). If answering the very simple question “How many krill does a whale swallow at a time?” can’t be answered currently with- out very large uncertainties, the calculated estimate may not useful for ecological studies. Collecting co-located (in both
time and space) data on marine predators and their prey is challenging (and in some cases involves being very close to very large animals in boats that are substantially smaller than the predators). Baleen whale and odontocetes can be tracked underwater using instrumented tags. Concurrent measure- ments of these animals’ prey from nearby vessels can provide insights into predator behavior and prey availability (Hazen et al., 2009; Parks et al., 2011).
These engineering advances have greatly increased our ability to collect vast amounts of data covering greater vol- umes of water and acoustic bandwidths. Thus our volume backscatter measurements cover more regions and provide more information about the scatterers in the water column. However, to convert the acoustic data into biological infor- mation, there are still many unknowns on the biological side of things when we try to predict how much sound specific organisms will scatter in the ocean.
One fish, two fish, big fish, tilted fish
If all fish were identical biologically (e.g., their size, shape, internal organs, etc.) and their position and orienta- tion in the water column were constant, acousticians would have a much easier time counting them in the ocean. Fortunately (particularly if you enjoy eating seafood), fish and other animals in the ocean are very diverse spanning many different types and sizes (and tastes).
Animal size has one of the strongest impacts on the amount of acoustic energy they scatter. Specifically the ratio of the acoustic wavelength to the size of the scatterer determines whether the scattering follows Rayleigh or geometric scatter- ing theory (Greenlaw, 1977; Holliday, 1977). In the Rayleigh regime (where scatterers are much smaller than a wavelength), scattering is primarily a function of animal size, with larger animals scattering more energy than smaller ones. However in the geometric regime (where the acoustic wavelength is much smaller than the scatterer), there are strong frequency depend- encies on the scattered energy and animal characteristics like shape and orientation play an important role. In general, lower frequencies allow scientists to look deeper into the water col- umn, but often limit the minimum size of scatterer that can be observed. But there are other factors that need to be consid- ered. Two fish of similar size can have different scattering effi- ciencies depending on their taxonomic identity. Many fish (but not all) have a swim bladder. These are gas-filled organs that help animals regulate their buoyancy and position in the water column. Acoustically, these organs are extremely strong scat- terers (often more important than the rest of the animal) of sound due to the acoustic impedance difference between sea- water and the gas-filled bladder. Some species of fish can con- trol or adjust the volume of gas in their bladder while others cannot. Thus, a fish that changes depth may adjust its swim- bladder volume accordingly which can cause a shift in its scat- tering spectra.
In a very simplistic sense, acoustic scattering from fish and zooplankton could be divided into two approaches. For ani- mals with a swim-bladder or an elastic-shell, the scattering from the animal can be described by modeling the scattering from the structure which typically dominates the backscatter
28 Acoustics Today, July 2012