Research on a Bottom Camera Bait Station (BotCam) for Underwater Fish Observation

The Need for Research on Sampling Methods

Effective management of deepwater bottomfish resources in the Pacific Islands Region requires accurate stock assessments. These depend on accurate information about bottomfish distribution, relative abundance, stock structure, size and age composition, and other biological characteristics. Such information is derived from various sources using a variety of methods. Fishery-dependent data include catch, effort, or biological statistics from commercial or recreational fishing collected at sea or at landing sites. Fishery-independent data include research data derived from fish captured using hook and line gear or observed underwater by Scuba divers, cameras, and other methods. Some data needed for stock assessment may be derived from a combination of sources and collection methods.

Each of the methods for collecting stock assessment data has its advantages and shortcomings. Research is needed to develop and test newer methods to provide alternative ways of collecting essential information or to enable collection of data under specific constraints. For example, new methods may enable more comprehensive or cost-effective sampling or monitoring. Or in cases where it is desirable to limit the sampling mortality in certain populations, new methods may make it feasible to collect some kinds of data without capture or removal of the fish. The essential need is to be able to collect all data required for robust stock assessments. Having a wide range of survey and sampling methods, each with known attributes, will ensure that adequate assessments can be conducted under increasingly complex circumstances.

Development of the BotCam

Scientists are deploying BotCam from the NOAA vessel Oscar Elton Sette.

To address this need, the PIFSC Coral Reef Ecosystem Division is developing a prototype, deep water (350 m) camera station that can be used as a cost-effective and non-extractive method to assess and monitor bottomfish and other commercially important deep water species. This Bottom Camera ("BotCam") system includes programmable control functions which allow for the activation of imaging systems, bait release mechanisms, image scaling indicators, and acoustic release to enable recovery of the camera. The camera bait station can be deployed repetitively during a survey of a site or can sit dormant on the seafloor ready for activation at a preset time. Further, the stereo-video configuration of the camera system allows for the sizing and ranging of both fish and benthic features

Development of a field-tested deep-water camera bait station, coupled with a standard method to analyze the collected image data, will provide a cost-effective and non-extractive alternative method to assess the abundance and size composition of bottomfish populations in deepwater habitats.

Types of Data Collected

The BotCam can be deployed to collect many kinds of information useful for population assessment:

  • Counts of fish to estimate relative abundance and density of fish species --- such data can be used to make spatial and temporal comparisons.
  • Measurements of fish and features of benthic habitat.
  • Measurement of distances, sampling area.
  • Measurement of fish swimming speed, current speed and turbulence.
  • Identification of habitat --- characterization of bottom by hardness, cover, slope, structure and composition.
  • Ground validation for multibeam surveys, passive acoustic surveys, and defined bottomfish regions.
  • Observation of fish behavior, such as reaction to bait/lures, intra- and cross-species interactions, aggressive/passive behavior.
  • Identification and tracking of tagged fish.

Testing of a BotCam Prototype

Location: Arakane, CNMI Depth: 241 meters Species: Pritipomoides auricilla (Yellowtail Kale), Caranx lugubris (Black Jack).

Between April and August 2005, a prototype BotCam was tested at several locations around the island of Oahu, Hawaii, including deployments in 4 restricted fishing areas (RFAs). Depths ranged from 70 to 350 meters and bottom types ranged from soft mud flats to wire coral covered pinnacles. The BotCam was also tested on slopes greater than 45 degrees and in high current regions. Several species of fish were identified, including Kahala ( Seriola dumerili ), Butaguchi (Pseudocaranx dentex ), Hogo (Pontinus macrocephalus ), Opakapaka (Pritsipomoides filamentosus ), Gindai (Pristipomoides zonatus ) , Kalekale (Pristipomoides sieboldii ), Ehu (Elelis carbunculus ), Onaga (Etelis coruscans ) and Hapu'upu'u (Epinephelus quernus ).

Statistical Analysis of BotCam Data

In early work with camera bait stations in productive regions, statistical methods of analysis were established and tested by Ellis and DeMartini (Ellis and DeMartini 1995). They used 4 parameters: time to first arrival of a target fish (TFAP), maximum number counted in any frame (MAXNO), total duration in sequence (TOTTM), and species present and duration of bait attachment (BTM).

In their tests, they found that the maximum number (MAXNO) parameter correlated best to the traditional catch per unit effort" (CPUE) parameter used in fishing surveys. Use of the MAXNO parameter avoids the potential problem of counting the same fish multiple times as it exits and re-enters the camera's field of view and yields conservative relative density estimates.

Location: Oahu, Hawaii Depth: 73 meters Species: Seriola dumerili (Kahala), Epinephelus quernus (Hawaiian Grouper)

The current version of BotCam improves upon earlier camera bait stations because it provides the means to not only estimate relative abundance and species composition of target fish, but also their size. Further, in preliminary studies, the BotCam has been shown to provide order of magnitude estimates of current speeds and the potential to directly measure turbulence, allowing the area of influence of the bait to be quantified (Merritt 2005).

Between September and October 2005, a second prototype BotCam was deployed during a Pacific Island Fisheries Science Center (PIFSC) Reef Assessment and Mapping Program (RAMP) cruise of the NOAA Ship Oscar Elton Sette in waters of the Commonwealth of the Northern Mariana Islands, Guam, and Wake Atoll. The unit was deployed 45 times to depths ranging from 100 to 350 m. Bottom slopes reached well over 50 degrees on several drops. Seventeen fish species, including all of the commercially important species were identified; several other fishes seen remain unidentified. Images of deep water corals and algae were also collected.

Future Directions of BotCam Research

Wake Island Depth: 347 meters Species: Etelis carbunculus (Ehu)

Research on the BotCam system is continuing. Several near-term research objectives have been identified:

  • Simplify deployment and recovery operations.
  • Improve the BotCam's field of view.
  • Reduce the size/weight of the overall BotCam package.
  • Update electronics and software modules to improve battery life, image quality and data uploading.
  • Develop automated video analysis methods.
  • Investigate alternatives for acquiring higher resolution images.
  • Engage the larger fisheries science and management community to develop BotCam sampling protocols (e.g., bait mixture, set times, sample sizes, and fish sizing parameters) applicable to specific survey requirements.


Ellis, D. M., and E. E. DeMartini.
1995. Evaluation of a video camera technique for indexing abundances of juvenile pink snapper, Pristipomoides filamentosus, and other Hawaiian insular shelf fishes. Fish. Bull., U.S. 93:67-77.
Harvey, E., M. Cappo, M. Shortis, S. Robson, J. Buchanan, and P. Speare.
2003. The accuracy and precision of underwater measurements of length and maximum body depth of southern bluefin tuna (Thunnus maccoyii) with a stereo-video camera system. Fisheries Research, 63: 315-326.
Merritt, D.
2005. BotCam: Design, testing and development of a fully automated stereo-video bottom camera bait station for ecosystem monitoring of bottom fish species. Master's Thesis, University of Hawaii, School of Ocean and Earth Science and Technology, Department of Ocean and Resources Engineering.