Volume 53

A Versatile and Inexpensive Method for Training and Testing Observers Conducting Underwater Visual Censuses Requiring Size Estimates


Authors
Kadison, E.; Addison, C.; Dunmire, T.; Colvocoresses, J.
Download PDF Open PDF in Browser

Other Information


Date: 2002


Pages: 581-590


Event: Proceedings of the Fifty Third Annual Gulf and Caribbean Fisheries Institute


City: Fort Pierce, Florida


Country: USA

Abstract

Underwater visual-based population monitoring programs that include the collection of size-structure information of the organisms surveyed require that observers collecting this data are able to accurately estimate the recorded size parameter. Training and testing of observers conducting size estimates during underwater visual surveys have traditionally employed very simple linear artificial targets of known dimensions or more complex and costly life-sized models of the organisms surveyed. As part of its efforts to monitor reef fish populations in the Florida Keys, the Florida Marine Research Institute has developed a new visual target type that is inexpensive to build, easy to maintain and deploy in-situ, and allows for almost infinite variability in target image size as well as accurate representation of species-specific visual aspects. These new targets are created by placing images of organisms photocopied onto waterproofpaper between two sheets of Plexiglas. A large variety of sizes can be generated from the same original image by using the photocopier to reduce or enlarge copy size. Teflon bolts and wingnuts are used to hold the two Plexiglas sheets together around the inserted images. Targets are deployed in the field using an anchor and float arrangement. New observers can be trained and tested with the images before collecting program data, and experienced observers retested on a regular basis. This new target system, which should overcome the estimation biases associated with the use of simple linear target systems without incurring the high expenses of constructing solid, lifelike models of the organisms surveyed, should be tested and evaluated more thoroughly for its applicability as a training tool

PDF Preview