NOTE Sluka and Sullivan. Spear fishing on species composition and size of groupers on reefs 
389 
expected to differ between sites under differing fish- 
ing intensities owing to the biology of the species, 
their site-attached nature, and the susceptibility of 
these species to fishing pressure. 
Materials and methods 
Sampling on four patch (shallow-water, small) reefs 
occurred in February 1992, April 1993, September 
1993, January 1994, April 1994, and September 1994. 
Nine additional patch reefs were sampled in Septem- 
ber 1994. A total of 13 patch reef sites were sampled 
(Fig. 1). There were no significant seasonal differences 
in the size distribution of all grouper species combined 
at these sites (Sluka and Sullivan, 1996). Thus, data 
from all seasons were combined for analyses. 
Researchers were trained to estimate lengths of 
fish consistently and accurately using methods out- 
lined in Bell et al. (1985). Observers had approxi- 
mately five minutes to estimate the length of a series 
of fish models of varying lengths. The length of each 
model was recorded in one of five categories: <5 cm, 
5-15 cm, 16-25 cm, 26-35 cm, and >35 cm. The fre- 
quency distribution of estimated model lengths was 
compared to the known distribution by using a chi- 
square test. The bias of each observer was determined 
as either consistently underestimating or overestimat- 
ing the size of the fish models. The information on bi- 
ases was given to each observer. The observer then re- 
peated the length estimation procedure until there was 
no significant difference between the observed and ex- 
pected distributions (P>0.05). Observers were found to 
be competent for length estimation after 2-3 trials. 
At four of the sites (MPR, TS1, TS2, and HOP), 
transect lines of 20 m or 25 m in length were used to 
sample the number, species, and length of groupers. 
The transect lines were laid in representative por- 
tions of each patch reef. The transect line was 
searched 6 m out from each side for a total width of 
12 m (width by visual estimation). Within each 
transect all groupers were enumerated and their 
length category and species recorded on underwater 
paper. Observers using SCUBA searched through- 
out the width of the transect, examining all crevices, 
caves, and holes. At the other nine sites, observers 
sampled the entire patch reef. Patch reef size was 
not quantified at these sites; thus density could not 
