of their relatively small contribution in biomass. Experimental 

 removal of these shrimp from isolated coral heads resulted in a 

 "reduced number of reef fishes and a high incidence of fishes with 

 frayed fins and ulcerated sores" (Limbaugh et _al. , 1961), The 

 grazing activities of these reef fishes, in turn, have a; major 

 influence on the nature of the reef community. 



In an attempt to quantify the significance of the activity of these 

 shrimp for resident fish populations, we selected 4 coral heads 

 which were isolated by a distance of 5 to 9 m from similar coral 

 heads. The resident fish populations were determined by recording 

 all fish observed in each of 7 days, with observations made in the 

 morning, afternoon and/or evening. We quickly began to recognize 

 individual fish, and the population estimates were very consistent. 

 One set of observations was made at night, but the majority of the 

 fish were hidden asleep in crevices or away, probably foraging on 

 the algal plain. Thirty to thirty-four species, represented by 

 totals of about 160 to 200 individuals were associated with each of 

 the coral heads which measured 1.6 m to 1.8 m long, 0,7 m to 1,2 m 

 wide and 0.4 to 0.8 m tall. 



After the population estimates had been made, all Periclimenes spp. 

 and Stenopus spp, were removed from two of the coral heads and the 

 changes in population numbers and apparent health of the fish were 

 followed for the remaining 7 days of the mission. From the first 

 coral head 22 Stenopus hispidus , 2 S. scutellatus and 3 Periclimenes 

 pedersoni were removed. From the second coral head 28 Stenopus 

 hispidus , 16 Periclimenes pedersoni and 7 Periclimenes sp. were 

 removed. The other two coral heads were left as controls. 



In all cases, no deterioration in health nor increase in parasite 

 infection was noted and some populations actually increased in 

 abundance. This increase was due to the recruitment of juvenile 

 Chaetodon capistratus (1 to each of 3 coral heads), Pomacanthus 

 paru (1), Eupomacentrus partitus (11, 7, 5 and 2, respectively) 

 and Ocyurus chrysurus (1 to each of 2 heads). The lack of de- 

 terioration in health may indicate that effects of the removal of 

 shrimp do not become apparent for at least a week. This seems quite 

 likely. However, when removing the shrimp, we left the cleaning 

 gobies, Gobiosoma ( Elacatinus ) spp., in order to be certain we were 

 testing the importance of theshrimp themselves. Even before removal 

 of the shrimp, all the cleaning activity we observed was performed 

 by these fish rather than the shrimp. These fish, 2 or 4 individuals 

 per coral head, may be able to buffer what effect the lack of 

 cleaning shrimp may have on the fish populations. 



VI-68 



