Results

A description of the five habitat types surveyed during the study is described in Table 1 and illustrated in figures 4,5,6,7. In general the area was composed of five different habitat types: a mixed rubble and sand bottom (type I), which graded into a coral and rubble area (type IV) on the north side of the bay; a sand and silt bottom (type II), which extended from the shore out to offshore patch reefs (type V) up the middle of the bay; and a basalt pavement and rubble area (type III) which also graded into a coral and rubble area (type IV) on the south side of the bay. An estimate of the area of each habitat type is listed in Table 2. The habitat types surveyed were very similar in description and area to those sampled by Chaney et al. (1977) and did not appear to have varied since the previous survey. However, since the previous study made no quantitative estimates of the area of each habitat type, no statistical comparisons were possible.

The relative abundance of all species in each habitat type is presented in Table 3. In general, the number of species of plants and invertebrates observed in each habitat type were greatly reduced compared to the earlier studies. Only one of the 13 subtidal plant species listed in Ball (1977), Porolithon onkodes, was seen. Moreover, almost all of the species of sponges, flatworms, sipunculans, echiurians, ectoprocts, annelids, arthropods, molluscs and echinoderms listed by Chaney et al. (1977) were not seen during this survey. Thus, the number of species of invertebrates declined from 106 species in 1976 to 21 species in 1996, a decline of over 80%. In contrast, the number of species of fishes experienced a small decline over the 20-year period: 64 species were seen in 1976 and 57 in 1996, a decline of 11%

Results of quantitative surveys of substratum types on patch reefs is presented in Table 4 and Figure 8. Although there were marked changes in substratum composition between the two surveys, most of these changes were not statistically significant due to low sample size (2) in the 1976 survey. Only two species of corals, Cyphastrea ocellina and Pavona varians, were significantly less abundant in 1996 relative to 1976 (Figure 9). Overall, however, there was a marked decrease in the diversity and percent cover of all species of coral.

Overall, the number of species of coral seen declined from 10 in 1976 to five in 1996, a 50% decline (Figure 9). Five coral species seen in 1976, Cyphastrea ocellina, Leptastrea bottae, Montipora patula and Pavona varians, were not observed on transects in 1996. Total percent coral cover declined from 44% in 1976 to 6.7% in 1996 and there was a concurrent increase in the percent cover of dead coral from 15.6% in 1976 to 54.2% in 1996, suggested high mortality of live corals between the two surveys. The percent cover of mud and sand declined from 41% cover in 1976 to 30.5% in 1996.

The dominant species of coral in 1976, finger coral (Porites compressa), declined from 16% to 1.3% cover. Similarly, the cover of lobe coral (Porites lobata) declined from 11% to 4%, rice coral (Montipora verrucosa) declined from 7% to 0.4%, and cauliflower cover (Pocillopora meadrina) declined from 3.5% to 0.4%. Other substrate types, such as encrusting coralline algae, turf algae, and the red coralline algae Porolithon onkodes, were not recorded on transects in 1976 but present in 1996 (Figure 9).

The abundance of fishes along transects on patch reefs is presented in Table 5 and Figure 10. Overall, the number of species of fish changed from 35 in 1976 to 39 in 1996, an 11% increase. The species composition of the fish community in 1976 was markedly different from that in 1976: the overall percent similarity between these two communities was 30.2% (Figure 10). There were a total of 31 species not common to either survey (Table 6). Fourteen of the 35 species counted on transects in 1976 were not recorded on transects in 1996. In particular, these included the abundant species Chromis ovalis and Scarus sordidus. Similarly, 17 species counted on transects in 1996 were not recorded on transects in 1976. These included the abundant species, Lutjanus fulvus, Acanthurus leucopareius and Chaetodon auriga. These differences are also reflected in changes in species diversity indices: the Shannon-Weiner diversity index changed from 1.07 in 1976 to 1.17 in 1996 and the evenness index changed from 0.69 in 1976 to 0.73 in 1996. Thus, the fish community in 1996 was both slightly more diverse and slightly more evenly distributed than the community in 1976. In contrast however, the mean abundance of all fish declined from 27.9 fish/100m² in 1976 to 18.1 fish/100m² in 1996, a 35% decline.

The five most abundant fishes seen on reefs in 1976 were, in rank order of abundance: Mulloidichthys flavolineatus, Chromis ovalis, Scarus sordidus, Thallosoma duperrey, and Abudefduf abdominalis. These species accounted for 72% of the fishes seen in 1976. In contrast, in 1996, the five most abundant fishes seen on reefs were, in rank order of abundance: juvenile Scarus, Ctenochaetus strigosis, Gomphosus varius, Thallosoma duperrey and Acanthurus triostegus. These species accounted for 61% of the fishes seen in 1996.

Overall, however, the only statistically significant change in abundance occurred in five species (Table 5). One species, Thallosoma duperrey, exhibited a significant decline in abundance from 2.9 fish/100m² in 1976 to 1.4 fish/100m² in 1996, a 51% decline in abundance. In contrast, five species, Acanthurus nigroris, Acanthurus triostegus, Chaetodon lunula, Parupeneus cyclostomus, and juvenile Scarus, exhibited significant increases in abundance. Lack of significance was mostly due to the small sample size (two transects in 1976 and three in 1996) associated with sampling fishes. Based on alpha = 5% and a power of 0.90 (beta =10%) the average minimum detectable difference for all species was ± 1.8 fish/100m². Thus, although changes in many species were obvious, due to the small minimum detectable difference, these changes could not be detected statistically.

In order to examine ecological changes in the fish communities analyses were also conducted among fish families and feeding guilds (Figure 11). In general, the abundance of Damselfishes (Pomacentridae), Goatfishes (Mullidae), Wrasses (Labridae) and Parrotfish (Scaridae) declined from 1976 to 1996 while the abundance of Surgeonfishes (Acanthuridae) and Butterflyfishes (Chaetodontidae), increased over the same 20-year period. Overall, however, there were no significant changes in either fish families or feeding groups between the two surveys. These results were probably due to low statistical power as stated earlier. Similar changes were also noted in the abundance of fish feeding guilds. In general the abundance of carnivores and herbivores decreased between 1976 and 1996 while the abundance of corallivores, omnivores, and planktivores increased over the same 20-year period. However, none of these differences were statistically significant.

Last update: 1/25/2005