Potential for coupling of underwater TV monitoring with passive acoustics
Charles A. Barans1, David Schmidt2, Myra C. Brouwer*3
1, 3 Marine Resources Research Institute, South Carolina Department of Natural Resources, 217 Ft. Johnson Rd., Charleston, SC 29412
2 Jackson and Tull - SMARTech Division, 1211 Ben Barron Lane, Moncks Corner, South Carolina 29462
The temporally and spatially fluid nature of fish associations often confuses interpretations of their underwater sounds. Ideally, visual confirmation of the sound producers and any behaviors associated with sound making should accompany acoustic data. Investigators have long been expanding their inventory of specific sounds that represent, conclusively or potentially, specific underwater sources. These are the acoustic "standards" by which to compare unknown sounds for recognition/identification purposes. Unfortunately, these standards are often not available for acoustic research on fishes. Although some sonic signatures have been corroborated with visual observations, many more are needed to assist in interpretations of sonic data from hydrophones placed in complex habitats with many interacting species of vertebrates and invertebrates. The underwater television (UWTV) is ideal for direct correlations between specific sounds and their causes, if the water visibility is acceptable.
Underwater video devices can provide a wealth of information to scientists and fishery managers including seasonal movements of fishes, the potential for development of indices of abundance for some migrating and resident populations, and any seasonal behaviors associated with the formation of pre-spawning aggregations along a migration route. An UWTV system offshore allows the study of fishes on the bottom throughout the year without the costly trips to a research site in inclement weather. The present visual capability of UWTV should be integrated with acoustic information to enhance fisheries biologists understanding of fish behavior and movements within the region. This paper describes recent results from a permanent installation of an UWTV at an artificial reef offshore.
The Underwater Television System
The research site was established in 25-28 m of water about 72 Km off central Georgia on May 11, 1999 with the deployment of several large fish attraction units (Artificial Reefs, Inc.). On August 24, 1999, the underwater TV cameras, cable and computer were installed and images were transmitted via microwaves to shore. Artificial reef structures are arranged in a circle around the camera system to maintain a resident aggregation of reef fishes, to attract transient species and to focus local fish activities within the view of the cameras.
The video system consists of two main parts. A pressure housing, located on the sea floor and a video capture engine, located remotely. Six monochrome video cameras are housed with a micro-controller and a few basic sensors. The micro-controller provides the means to multiplex the 6 analog video signals to one coaxial cable running between the pressure housing and the video capture engine. The video capture engine is a computer running under the Windows NT operating system. A console application controls a video frame grabber, which takes the analog video signal and converts it to digital images. The system has the ability to capture multi-frame or single frame images based on parameters set by the user. The micro-controller in the pressure housing receives commands from the video capture engine for camera selection, tilting data, and system status updating. The embedded computer also acts as a web server through which the video system is controlled. File transfer and system parameter updates are made possible by an interface between the web server and the console application controlling the video system using pcANYWHERE.
Small black and white security cameras (Supercircuits PC-23C) with low light capabilities (to 0.04 lux) and relatively low resolution (460 lines) were used. Camera lenses of 8 mm allowed a 12 degree angle of view, and the seabed was in view at about 13.7 m from the camera. Daily observations (~65) were conducted between 1200 and 2130 GMT (Greenwich Mean Time). Still images ~15 Kb (jpg) were recorded and logged at 10 minute intervals for 10 sec. Video clips ~400 Kb (avi) were recorded on the hour from camera No. 5, since only camera No. 5 was directed at reef structure with any reef fish activity. Images were downloaded from the remote computer to the laboratory computer for fish counts and long-term data storage.
Observations using UWTV
We have learned how to deploy and maintain the UWTV system and remote operations systems. We have temporally documented species presence and activity. The seasonal dates of the first appearance of various fish species are especially important for identification of any prespawning migration to the south by adult grouper, one of our main target species. Seasonal changes in the makeup of the fish assemblage at the UWTV site appear much greater than previously believed. We are documenting the annual cycle of juvenile recruitment in spring and summer followed by intense predation by transient species later in the year.
Large schools of bait fish have been present in most seasons accompanied by schools of predatory greater amberjacks (Seriola dumerili). The subjects of interest to us, snapper and grouper species, however, have yet to establish resident populations at the site. Year-round resident species included Atlantic spadefish (Chaetodipetus faber), gray triggerfish (Balistes capriscus), and the predators black seabass (Centropristis striata), and great barracuda (Shyraena barracuda). Other resident species may not have been observed due to decreased visibility and/or increased cryptic behavior during winter.
The visibility near the bottom was relatively poor throughout May, at some times in late summer and often after winter Northeaster storms and hurricanes. Swarms of juvenile round scad (Decapterus punctatus) appeared in mid-May and were occasionally accompanied by snapper and grouper species. Transient species in May included loggerhead seaturtle (Caretta caretta), sand tiger shark (Odontaspis taurus), rock hind (Epinephelus adscensionis), nurse shark (Ginglymostoma cirratum), and cobia (Rachycentrum canadum). The settlement and recruitment to an area by juvenile baitfish may attract many temporary predatory species. In 2000, we first sighted small baitfish in April, although periodic clouds of very small juveniles could not be identified to species.
During 1999, tomtate (Haemulon aruolineatum), especially juveniles, were one of the most abundant and conspicuous members of the fish assemblage in close proximity with the structural reef units. Large schools of round scad often swam in and out of view near the tops of reef units. Immediately above any fish structure and well up into the water column were loose aggregations of adult greater amberjack. Other jacks occasionally passed through the site, including black bar jack (Caranx ruber).
Although resident predators must significantly reduce recruitment of many species, large stochastic predation events appear to have a formidable influence on mortality and survival of small and juvenile reef fishes. Two important large-scale predation events observed in 1999-2000 were the arrival of migrating loons (Gavia imer) and the mid-winter appearance of large populations of ctenophores (believed to be Leucothea milticornus) near the bottom. We observed the loons to visually select fish prey in and near the structural reef units within a meter of the bottom. Also, the large numbers of ctenophores and/or jellyfish in winter corresponded to the temporary residence of an ocean sunfish (Mola mola) and a relatively large population of adult Atlantic spadefish. Both species are known to feed on jellyfish. The importance of predator-prey relationships was confirmed by inference from general observations of the relatively simultaneous arrival of baitfish and some of the higher level predators in the early spring. The infrequent appearance of high level piscivores at the small artificial reef site suggests large feeding ranges of many species, which appear to be "passing through" looking for feeding opportunities. Species that may move back and forth between other habitats within their hunting range and were observed at the site include large adults of loggerhead turtles, sand bar sharks, red snapper (Lutjanus campechanus), gag (Mycteroperca microlepis) and scamp (M. phenax).
Observations from the artificial reef research site can contribute significantly to the understanding of the short-term and long-term temporal changes in an offshore reef fish assemblage. Permanent installations of UWTV systems have the advantage of non-obtrusive observations of fish interactions. The sorts of behaviors that are of most interest to biologists, such as feeding or spawning, are rarely observed in the wild. Documenting such rare events often requires constant, long-term observation that is difficult under natural conditions because the ocean is always changing. Observations by divers are severely limited during seasons of high seas, often including both winter and spring, when migration and/or spawning activities among reef fishes most often occur.
Although the lack of mobility of our UWTV is a spatial limitation, a semi-permanent setting allows temporal investigations. Similar visual studies could develop sound catalogs of transient species and more detailed behavior-related sound patterns of resident species for potential management applications. For instance, estimates of population size of coral-nibbling parrot fish might be made after correlations between the mean rate of munching done per fish, with visual verification of the feeding behavior/sound relationship.
Synergistic information would come from the simultaneous combination of visual and sonic information. The addition of less expensive, passive acoustic data gathering devices could compensate for the lack of spatial coverage by more expensive UWTV systems. It could also provide more complete coverage of the events taking place in the vicinity of the cameras, but beyond their field of view.
Imagine, if you will, the interesting array of sounds that might have accompanied the feeding of loons on cigar minnows near the bottom, Atlantic spadefish biting a chunk off a passing jellyfish or a school of tuna passing through the artificial reef structures. Each of these complex sound series may be far more interpretable with TV documentation during the first several encounters. Behavioral observations correlated with acoustics and environmental conditions and validated over time would contribute to the interpretations of results from other sampling areas for which only acoustic data are available.
Integration of a relatively permanent TV and passive hydrophone offshore to refine the acoustic "dictionary" seems to be nearing reality. The UWTV systems exist associated with Fish Watch, Africam, Aquarius and specific public aquariums throughout the US. The bandwidth necessary to add simultaneous passive acoustic data should be minimal and be waiting only for the enthusiasm to make it happen.
The primary scientific objective of the UWTV system established off Georgia was to document and quantify prespawning aggregations of gag grouper as they move south along the continental shelf. If an associated sound recognition pattern were associated with such fish aggregations and movements, multiple listening stations could be established at key locations across the shelf and along the potential migration path at a cost for monitoring much less than that using other methods. The visual findings of the present UWTV study expand our understanding of the importance of large scale stochastic predation events on relatively localized reef fish aggregations, especially of juveniles and bait species. The sounds generated by the assemblage interactions would have filled chapters of a catalog on reef fish sounds. The scientific community anxiously awaits the development and application of tools that will allow simultaneous visual and sonic investigations of fish associations and behaviors.
Any marine research requires a team effort with many experts and professionals involved to accomplish even a small task offshore. In the case of remote transmissions from an underwater TV system, the numbers of support electronics/computer, divers, vessel operators and agencies willing to loan facilities and/or manpower to get the job done can be staggering. We thank all those multi-agency staff and, in particular, the staff that always can be depended on to accomplish multiple tasks to keep the system functioning: Trent Moore, Cheryl Burden Ross and Travis McKissick. We thank many staff of the following institutions/ agencies/ groups for their undying support, including: the Skidaway Institute of Oceanography, the Grays Reef National Marine Sanctuary, the SC Marine Resources Div, Artificial Reef Section and MARMAP program, the Georgia Coastal Resources Division and the USN Explosive Ordnance Disposal Mobile Unit #12. We are grateful for funding from the US Navy through the Office of Naval Research, National Oceanographic Partnership Program, Grant No. N00014-98-1-0808.