Project Number: NAGL-01-02A

Page 1 of 4


Inclusive Dates of Operations: 10/17 Ė 10/23/01

Dive System(s): Phantom III S2

Total Dive Time: 35 h 25 m

Support Vessel: R/V Connecticut

Number of Dives: 14

Maximum Working Depth (m): 70 m

Project Title: Identification of Soniferous Fishes on Stellwagen Bank: Validation of Their Sound Production Characteristics and Association of Sounds with Specific Habitats and Behaviors

Principal Investigator: Dr. Francis Juanes

Co-Principal Investigator: Dr. Rodney Rountree

Preliminary Scientific Results:

Data collected during this cruise demonstrates that an ROV can serve as an adequate vehicle for the collection of underwater acoustic data. To our knowledge ROVís have never before been used for this purpose. The Phantom III S2 was equipped with three hydrophones that successfully operated to depths of up to 70 m under sometimes harsh environmental conditions. Extensive testing conducted during this cruise will provide us with data required to develop the Soniferous Fish Locator device. However, it is also clear that even without the SFL capability, the ROV makes a good platform for underwater sound collection and has great potential for providing the means to validate sound source identification for vocals fishes (i.e., make visual determination of vocal species). Because of weather delays, we conducted only 14 of the 21 planned ROV dives, but still collected over 35 hours of underwater sound data. Because of these frequent interruptions, it was not possible to collect sufficient data for detailed habitat or day-night comparisons. However, we collected sufficient data to suggest that both strong habitat and day-night differences occur. For example, pollock was much more abundant during the night dives than during day dives. Adult cunner and pollock obviously avoided the ROV during the day, but pollock were strongly attracted to the ROV at night due to our use of chum and bright lights. The chum attracted swarms of amphipods that in turn attracted a large aggregation of pollock as well as haddock, cod and skates. Spiny dogfish and Longhorn sculpin were more abundant in sand and gravel, and were absent in boulder habitat. Cusk were only observed in boulder habitat and avoided the ROV both during the day and night when the lights were on. When only infrared lights were used, the cusk was clearly attracted to the chum bag on the ROV and showed no avoidance of a stationary ROV. Although fish sounds were infrequent, with only 12 potential events recorded (single or series of calls), we feel that this was a result of the natural seasonal pattern of fish vocalization in the study area, and is not a reflection of operational problems. Indeed, we anticipated that sounds would likely be infrequent during the fall cruise period because of the lack of spawning activity. However, since this type of work had not been conducted in the area previously and consequently the seasonal pattern in vocal behavior is not known, we felt that sampling within the fall period was necessary and provides a useful comparison with other seasons. We fully expect that the planned spring and summer surveys will result in far more prevalent vocalizations because several of the target study species will be spawning.

Cruise Products (including number of videos, still photos or rolls of film, and summary of other samples collected; e.g. number of box cores):

Seventeen (17) copies of VHS, HI-8 black and white, and Hi-8 color tapes were recorded during 14 ROV dives. Extensive audio data from the ROV dives were captured to computer and stored in 225 files amounting to over 10 G of data. This included 48 calibration signal files. Calibration signals were recorded at the beginning of the tapes for each dive. Input voltage was also recorded so that calibrated audio data can be obtained directly from the tapes to obtain signal source levels. Thus additional sound source data can be obtained from the recordings in the future even without our calibration test data. In addition to the ROV dives, we conducted 17 drop samples in which a hydrophone and small infra-red lighted camera were deployed from the ship to record addition audio and video data when the ROV was not in operation. This data was stored to 17 VHS tapes totaling over 21 h of recordings. Although noise from the ship was significant, some useful audio of fish sounds may have been collected pending further analysis.

Safety Problems and Concerns:

I observed no safety problems/concerns during this cruise. All operations were conducted in a highly safety conscious manner.

Dive System Management (e.g. adequacy of dive system to accomplish tasks):

Although we believe the ROV can serve as a useful platform for collection of underwater sounds and vocal behavior data, our experiences on this cruise have identified several areas that require some thought and effort to improve our ability to collect useful acoustic data, including: 1) maintaining a fixed station with minimal noise production (i.e., without thrusters), 2) improve ability to see in low light and near total darkness conditions without affecting fish behavior, and 3) ability to monitor underwater sounds from fixed positions for extended periods of time (1-24 hours). Maintaining a fixed station: The most significant problem is in the ROVís inability to maintain a stationary position on the bottom without the use of thrusters. The ROV offers large advantages over fixed camera systems in that it provides the ability to change positions change direction, and to approach fishes for identification. It would be helpful if ways can be found to allow the ROV to remain stationary on the bottom for periods of ten to twenty minutes at a time with its thrusters and track point system turned off, yet still retain the ability to mobilize to investigate sounds. That would provide us (and others) with the ability to conduct acoustic surveys with various statistical designs in order to investigate spatial and temporal variations in vocal behavior. In addition, the use of an ROV with the SFL for homing on sound sources will ultimately require the ability to frequently turn off noise generating systems for at least 20 sec intervals in order to obtain a bearing to a sound source. Further consideration of methods to provide buoyancy control of the ROV would be helpful. However, it is likely that some fish will be able to avoid an approaching ROV because of the large amount of noise it generates. During our cruise, fish seemed most affected by sounds of the ROV bumping on the bottom, with clear startle behaviors observed on numerous occasions. Further studies to examine the effect of ROV noise on fish behavior would be beneficial in of themselves as ROVs are an important underwater visual census tool. It is clear from our experiences during this cruise that the alternative of using a drop camera system to monitor fish behavior should also be further investigated. The drop camera would allow us to obtain longer time sequences of underwater sounds and fish behavior with less noise production, especially if used in conjunction with chum and various combinations of white light, red light and IR lighting. However, the drop systems sacrifice the significant advantage of mobility provided by the ROV. Often moving just a few feet can mean the difference between identifying a sound source or not. Night observations: Another significant problem with the use of the ROV to study vocal behavior in fishes is the difficulty of operating without the white lights. Typically we could only achieve maximum sampling periods of ten minutes without lights, because of the pilotís need to stabilize a slowly or rapidly drifting ROV and to avoid obstacles. On one occasion we were able to operate without lights for over 20 minutes. That is the only time we were able to observe the behavior of the cusk. We suspect that no-light conditions of at least 10-20 minutes are necessary to observe the more cryptic nocturnal species. We would be interested in investigating the potential of ultra-low light cameras and red filters for the lighting system for their potential to enhance the ROVís ability to obtain data on the undisturbed behavior of fish. Extended observations periods: The final major area we identified for investigation is the ability to collect extended audio and video data on fish vocal behavior in order to examine diel behavior cycles. Many vocal fishes restrict their vocal behavior to specific times of the day, especially when vocal behavior is tied to spawning activity. This makes it difficult to use passive acoustics as a survey tool if daily cycles in vocal behavior are unknown. The possibility of enhancing the current project with the use of a drop camera in alternation with the ROV should be investigated. The principle investigators plan to pursue funding to obtain acoustic equipment to allow us to set up extended observation sites in alternation or conjunction with the ROV cruise. Other considerations: Our experience on this cruise indicated that our current data acquisition equipment is only minimally adequate to the goals of the project. We will investigate obtaining better equipment, including a programmable multichannel I/O board, anti-alias filters and better hydrophone cables. In addition, the use of underwater plug connectors between the hydrophones and the cables will enhance the system.

Logistics and Support Activities (including pre-cruise planning):

We found that the logistical support provided by NURC was outstanding. Our planning meeting in June helped use plan the cruise more effectively, especially in light of changes in the goals of the first cruise that resulted from the collection of underwater fish sounds from the inshore survey component of the project, and from fish sounds data we obtained from other projects. These data made it clear that the development of the SFL would be more difficult than originally anticipated. Of particular concern was the lack of high quality acoustic data for any of the target study species. With that in mind we had to modify the objectives to include the collection of sound source level data through the use of calibrated systems. This greatly complicated the project and made it necessary to obtain additional equipment and software. In the two weeks immediately prior to the cruise we also were able to take advantage of NURCís facilities and staff to conduct tank tests of the ROV, hydrophone system, and quantitative signal acquisitions system on three separate occasions. This greatly improved our preparedness for the cruise. We would be highly interested in continuing frequent consultation with NURC staff in our preparation for the May and July cruises scheduled for 2002. The NURC staff has demonstrated that they can be of invaluable assistance in working out solutions to the problems discussed above. We are also interested in exploring any further avenues of cooperative and collaborative efforts to develop underwater acoustic data collection systems with NURC.

Participants, Affiliations, and Roles (note: participants fully documented in the original cruise plan need only be listed by name and role. All other entries must include full participant name, affiliation, address, phone, fax, email and role in operations. Participant roles include Mission Coordinator, Chief Scientist, Other Scientist, Research Technician, Operations Technician, Post-Doctoral Student, Graduate Student, Undergraduate Student, Teacher/School Representative, or Other.):



Chief Scientist: Rodney Rountree, Phd.

Undergraduate Student/assistant: Meghan Ann Hendry-Brogan

Post-Doctoral Student: Rebecca C. Jordan

Dept. Ecology and Evolutionary Biology

Princeton University

Graduate Student/Research Technician:

David Howe

Dept. Natural Resources Conservation

UMASS- Amherst