Introduction

[Phil Pepe (28)5-7106]

The following abstract is of a paper presented by the authors at the 26th 
Annual Meeting of the Association of Marine Laboratories of the Caribbean in 
1994. We have since used visual methods for censusing parrotfish and 
surgeonfish populations on the outer reef flats of Roatan, Honduras. We will be 
returning to Roatan this summer to continue our work.

Comments would be appreciated!

	A NEW VISUAL CENSUSING TECHNIQUE FOR ENUMERATING
		LARGE, DAY ACTIVE REEF FISHES

	Philip Pepe (Phoenix College, Phoenix, AZ 85013 
	and Oceanic Society Expeditions, San Francisco, CA 94123),  

	Nicole Crane (Hopkins Marine Station, 
	Stanford University, Pacific Grove, CA 93950 
	and Oceanic Society Expeditions, San Francisco, CA 94123), 

	Geoff Stein, Monica Cervinek, Kevin Kittredge, 
	Tasha Bair and Nancy Osborne (ReefWatch Volunteers) 

				ABSTRACT

	The Kick Cycling Visual Technique (KCVT) for censusing fishes  was 
developed and tested. The KCVT does not require the advance deployment of 
markers as does the Strip Transect Visual Technique STVT.  Divers need only a 
compass and a slate to carry out a census and use a standard timed kick cycle 
method for estimating the distance covered while swimming along a compass 
heading. This is of great advantage when time is limited and the goal of a 
census is to visit multiple sites over large reef areas.
	The KCVT was used to census parrotfish and surgeonfish at four 
locations at the northwest end of Lighthouse Reef Atoll, Belize. All censusing 
was done in outer reef flat environments on the seaward side of the reef crest 
at depths of 30 to 45 feet.  
	The census data was used to study the precision of the method. The 
Precision Statistic (P=SE/M) ranged from 10-26%. 
 	There is a strong similarity in the distribution of parrotfish and 
surgeonfish between sites. There were no significant differences between the 
KCVT surgeonfish censuses at all four sites (Kruskall-Wallis p=0.05). The 
sample distributions from three sites were not significantly different from one 
another in both parrotfish and surgeonfish (Kruskall-Wallis p=0.05). However 
the  mean number of parrotfish was greater and significantly different at 
Lighthouse Wall (Kruskall-Wallis p=0.05).
	Paired STVT and KCVT censuses produced similar sample distributions at 
Manta Point. There was no significant difference between the STVT and KCVT data 
pairs (Mann-Whitney U Test, p=0.05). The results support the hypothesis that 
the KCVT provides a similar degree of accuracy and precision as the STVT for 
censusing large, day active reef fish.