Robert Gisiner

Acoustic identification of female Steller sea lions (Eumetopias jubatus)

Steller sea lion (Eumetopias jubatus) mothers and pups establish and maintain contact with individually distinctive vocalizations. Our objective was to develop a robust neural network to classify females based on their mother-pup contact calls. We catalogued 573 contact calls from 25 females in 1998 and 1323 calls from 46 females in 1999. From this database, a subset of 26 females with sufficient samples of calls was selected for further study. Each female was identified visually by marking patterns, which provided the verification for acoustic identification. Average logarithmic spectra were extracted for each call, and standardized training and generalization datasets created for the neural network classifier. A family of backpropagation networks was generated to assess relative contribution of spectral input bandwidth, frequency resolution, and network architectural variables to classification accuracy. The network with best overall generalization accuracy (71%) used an input representation of 0-3 kHz of bandwidth at 10.77 Hz/bin frequency resolution, and a 2:1 hidden:output layer neural ratio. The network was analyzed to reveal which portions of the call spectra were most influential for identification of each female. Acoustical identification of distinctive female acoustic signatures has several potentially important conservation applications for this endangered species, such as rapid survey of females present on a rookery.

The Office of Naval Research Marine Mammal Program, 1990–2006

ONR first stood up a dedicated marine mammal research program in 1990. The initial effort was led by Dan Costa, with the support of experienced ONR program leaders like Mel Briscoe, Bernie Zahuranec, and Steve Zornetzer. The author was privileged to manage the program from 1993 to 2006; a period of rapid expansion of concern about the marine environment accompanied by the desire to put tools for improved understanding in the hands of scientists and decision makers. The emphasis from the start was on multi-disciplinary collaborations. Major research themes included hearing and exploration of Temporary Threshold Shift as a metric of auditory risk; advancement of animal-borne scientific instruments (“tags“); advancement of ocean acoustic monitoring technologies, refinement of field sound exposure experimental methods, and improvement of sound exposure models. Frequent external reviews, including four National Research Council (NRC) reviews, were vital to the quality and credibility of the program. The 2005 NRC review on the Population Consequences of Acoustic Disturbance (PCAD) is a textbook example of how ONRs basic research contributions have had far-reaching consequences that greatly exceeded initial expectations. The convergence of growing concern about manmade sound in the ocean and the ONR model for quality in basic research has had a profound impact on our ability to understand ocean bioacoustics and to make better informed decisions about the potential risks posed by human activities in the marine environment.ONR first stood up a dedicated marine mammal research program in 1990. The initial effort was led by Dan Costa, with the support of experienced ONR program leaders like Mel Briscoe, Bernie Zahuranec, and Steve Zornetzer. The author was privileged to manage the program from 1993 to 2006; a period of rapid expansion of concern about the marine environment accompanied by the desire to put tools for improved understanding in the hands of scientists and decision makers. The emphasis from the start was on multi-disciplinary collaborations. Major research themes included hearing and exploration of Temporary Threshold Shift as a metric of auditory risk; advancement of animal-borne scientific instruments (“tags“); advancement of ocean acoustic monitoring technologies, refinement of field sound exposure experimental methods, and improvement of sound exposure models. Frequent external reviews, including four National Research Council (NRC) reviews, were vital to the quality and credibility of the program. The 2005 N...

Popper, Hawkins, Fechner, Weber: Recalling the importance of experimental psychophysics in the study of environmental noise

To bring the study of internal experiences into the reach of science, pioneers like Fechner and Weber framed a new branch of science called psychophysics. Current leaders such as Art Popper and Tony Hawkins have illustrated through their research the power of psychophysical methods in parsing a complex external and internal world into lawlike patterns of sensory performance. Recent studies of human influences at ecosystem and global scales have focused increasingly on our ability to use sophisticated statistical methods to analyze large complex data sets collected opportunistically, often without an a priori hypothesis. This celebration of the careers of two leading scientists in this field is an excellent opportunity to remind ourselves of the value of parsing complex phenomena into their constituent parts; framing simple hypotheses for relatively simple statistical analysis; and perhaps most important of all, using experimental methods to minimize sources of bias inherent in the study of internal states...

Recent advances in scientific understanding of the effects of sound from seismic surveys

The E&P Sound & Marine Life Joint Industry Programme (SML JIP), a partnership of 13 oil and gas companies and associations, funds independent scientific research to increase understanding of the potential effects of E&P sound on marine life. To advance understanding of the interaction between sound from oil and gas operations and marine life, the JIP identifies and commissions research to: (1) support planning of E&P projects and risk assessments, (2) provide the basis for appropriate operational measures that are protective of marine life, and (3) inform policy and regulatory development. SML JIP research categories include sound source characterization and propagation, physical and physiological effects and hearing, behavioral impacts and biological significant effects, and technologies for monitoring and mitigation. Highlights of projects to better characterize the source properties of seismic air sources, understand the hearing of Arctic pinnipeds, assess hearing recovery in marine mammals exposed to ...

Something just clicked: Whitlow Au and the sonar of dolphins

I suspect that no one, especially not Whit, imagined in 1970 that he would become the worlds authority on dolphin biosonar, also known as active echolocation. Since that time Whit has amassed an astonishing list of hundreds of publications on the subject, including the seminal text on marine mammal echolocation, The Sonar of Dolphins, published in 1993. It may surprise many of the younger members of ASA to know that there was a Cold War “arms race” for knowledge about dolphin sonar and that Whit and his U.S. colleagues carried on a lively and mutually enriching correspondence with Soviet colleagues against a background of government intrigues and limitations on travel and conference attendance out of concerns about a “biosonar gap” every bit as worrisome to military leaders as the nuclear arms race (well, maybe not quite as worrisome). Since then Whit has branched out to many other fields of bioacoustics and mentored a diverse array of students at the University of Hawaii, but I will focus on the years a...

Putting tags in the researcher's toolkit: An examination of the strengths, limitations, and added-value from animal tagging

The US Navy, through the Office of Naval Research and other offices, has focused on improving animal tags, reducing cost, and increasing availability. From data-rich packages like video and acoustic dataloggers to simple location-only tags, tags provide a variety of new data to studies of marine animals and their ecosystems. Tags realize their full potential when calibrated or validated against other existing alternative sensor systems like visual surveys, photo-identification, genetics, and acoustic monitoring. When tag cost, cost of delivery and recovery or monitoring are weighed against the data uniquely available from tags, an integrated data collection strategy involving animal tagging can be developed to generate the best data at the optimal total cost for a given research or resource management scenario.

Terminology and conceptual frameworks for cumulative effects analysis.

In the short history of underwater noise regulation most efforts have focused on the risk posed by a single source over a short period. However, noise in the workplace, cityscape, and terrestrial environment are typically expressed in terms of the cumulative consequences of repeated exposures from multiple sources over long periods. Recent publications and workshop reports on cumulative effects’ analysis are reviewed, and a terminology and conceptual approach are proposed to facilitate analysis of the cumulative effects of underwater noise. The term cumulative effect is reserved for the accretion of effects over time while the term aggregate effect is proposed for the effect of multiple concurrent stressors, both acoustic and nonacoustic. Possible metrics for comparing effects across individual stressors are discussed, including measures of caloric cost and physiological stress. A modeling framework is proposed, by which the synergistic effect of combined stressors might be quantitatively assessed.

The nature and nurture of seeing with sound: The role of learning in biosonar.

Echolocating dolphins achieve performances in detecting and classifying components of their environment that rival the performances we typically associate with vision. Many researchers have referred to dolphins “seeing with sound” or forming internal “images” of the world via acoustics, implying a presumed isomorphy of sensory inputs and a common representation at higher levels of processing in the brain. But some aspects of acoustics do not translate into visual equivalents (hollow objects and objects of different materials) and some visual aspects of an object do not translate into acoustic equivalents (color and brightness). So, how is this cross‐modal sensory translation achieved? Is it hard‐wired into the anatomy of the brain, learned through association, or is it some combination of the two? Experimental tests of nature versus nurture in cross‐modal sensory performance are reviewed, as first explored by Schusterman and Kastak. Further studies are suggested to reveal the respective roles of neuroanat...

Marine bioacoustics in a changing Arctic.

The retreat of summer ice in the Arctic during the past 5 years has been accompanied by a surge of alarm about iconic ice‐associated marine mammal species: polar bears, walrus, ringed seals, and others. Changes for subarctic species of mammals, fishes, and planktonic organisms have not received as much attention, but will also need to be monitored. The paucity of data about Arctic marine mammals is now more than ever a serious impediment to good decision making. One of the greatest sources of hope is the tremendous advance in marine bioacoustic sensing technology that has occurred within the past decade. Passive and active acoustic technologies offer new means of obtaining biological data in a summer Arctic Ocean where ice is less available as a platform for animals and scientists, and in a winter Arctic Ocean that will continue to be one of the most difficult research environments in the world.

Biosonar performance of a false killer whale (Pseudorca crassidens) improved with practice

Cylinder wall thickness discrimination tasks have been used to assess the limits of dolphin biosonar. An attempt to replicate the benchmark data from a bottlenose dolphin (Tursiops truncatus) resulted in a surprising order of magnitude better performance by a false killer whale. The improvement came over multiple testing episodes, which suggests that initial limits to discrimination were overcome by learning; either through better understanding of the reinforcement contingencies, attention to novel sources of information within the returning echoes, or both. These results offer important insights into methodological considerations for testing animal psychophysical performance, especially for sensory performances like echolocation in which the subject exercises active control over the sensory input. The ability to improve sensory performance with practice also offers insights into the way the echo information is processed into an internal representation of external physical reality by the central nervous s...