26 July 2001
By Malcolm Schongalla, Doug Escribano, Sarah Oberhauser, Dave Smith
Abstract
Fish populations and fish behavior were observed on a small artificial reef in Graham’s Harbor, San Salvador, Bahamas. Fish populations, probably entirely juvenile, were counted daily for 3 consecutive days. Physical features of the reef were categorized and compared to fish movement and preference of microhabitat. Fish populations remained highly stable during the study period. Different fish species preferred different microhabitats and exhibited unique movement patterns and zonation. Our observations generally do not disagree with published literature on artificial reefs or fish behavior.
Introduction
Artificial reefs are excellent subjects for scientific study for many reasons. Reef fish are well known to be rapidly attracted to artificial reefs (Bohnsack et. al., 1994). An artificial reef with high complexity increases abundance of juvenile and adult reef fishes (Gorham and Alevizon, 1989). The fish populations attracted to artificial reefs can even be larger and more diverse than those found on nearby natural reefs (Rilov and Benayahu, 2000). Complexity of the reef is important because it affects opportunities to study fish territoriality, zonation, and usage of microhabitats.
Since “complexity” is a broad term, it is difficult to objectively characterize it for the purpose of scientific comparison. Previous researchers have taken several approaches, including describing crevice space, surface features, and general topography (Zahary and Hartman, 1985; Gorham and Alevizon, 1989). This study attempts to describe complexity in terms of a combination of the above characteristics.
This study had two purposes: to evaluate the consistency of the fish population from day to day and to study territoriality, zonation, and usage of microhabitats by fish living on the reef.
Methods and Materials
Reef Characterization
The artificial reef selected for use was located in southeastern Graham’s Harbor, on the north side of San Salvador, Bahamas. The reef was approximately 30 feet from shore. The water depth at low tide was approximately six feet. The reef was surrounded by a sandy floor. Approximately 20 feet further out from shore, the sea floor changed to a seagrass bed of Thalassia testudinum.
The reef was a preexisting structure constructed with concrete subunits. Each subunit was made of rectangular lengths of concrete that intersected orthogonally in the middle. Each subunit was approximately 3 feet long in each direction. These subunits were arranged side by side in a rectangle approximately 3 feet by 10 feet by 1 foot high. The bottom parts of the subunits were sunk into the sea floor. The long axis of the rectangle ran parallel to the shore. At one end, an additional subunit was placed on top of the others to make a pinnacle. The arrangement of the concrete blocks created numerous crevices within the structure.
Based on reef appearance and general trends in fish behavior, parts of the reef were divided into microhabitats. The “top” category described the water within approximately 3 feet above the top of the reef. “Crevice” referred to the space in between or underneath blocks of concrete, generally protected from open water. “Periphery” referred to the water surrounding the reef on the sides, excluding the water directly above the reef. “Pinnacle” was used to describe the water immediately above or adjacent to (i.e. within approximately 1 foot) of the tallest block of the reef.
Data Collection
Fish observations were made on three consecutive days at different times, in mid-June 2001. Observations on the first day were made at 2:30 PM, on the second day at 4:30 PM, and on the third day at 3:00 PM. On each day the depth of the water was recorded. On the first day, water temperature was also recorded.
On days 1 and 3, four observers simultaneously counted the numbers of each species of fish present. The observers floated on the surface within 5 to 10 feet horizontally from the reef. Each observer counted the number of each species. For some species, the population was estimated to the nearest five individuals. For each species, the population counts obtained by the observers were averaged and rounded off to obtain the final value used as the population count. On day 2, only one observer was present and the species counts obtained by that observer were used as the final population count for that day. In addition, each population was labeled according to the type of microhabitat in which it was most commonly found. Labels were determined by consensus of the observers based upon each observer’s subjective impression of where the fish spent the most time. Some species were felt to be frequent in multiple microhabitats. The category “In Between” refers to these species and does not exclude a species from being included under another label.
Results
Populations
Nine species of fish were identified and included in the data set. All fish were probably juveniles. The grunts (Family Haemulidae) were not identified to the specific species. A tenth species was witnessed with one representative on the last day but was not included in the data set. The two most common species were Psuedupeneus maculatus (Spotted Goatfish) and the grunts. The two least common species were Lutjanus griseus (Gray Snapper) and Mulloidichthys martinicus (Yellow Goatfish). Figure 1 lists the population counts for each species on each day.
A Chi-Square test was performed to compare the population of fish each day with the population of fish on the other two days. The test failed to reject the null hypothesis that species populations remained constant day to day, with a p-value of 0.999.
Species were categorized by their choice of habitat, in order to determine whether different microhabitats were characterized by different fish populations. Figures 2a and 2b detail the species found in each microhabitat. The Index of Similaries between two samples was used to comparitively evaluate the species composition of each microhabitat:
S = 2 C / (A + B)
A = Number of species in the first habitat
B = Number of species in the second habitat
C = Number of species common to both habitats being compared
Crevice and Periphery were found to have no species in common (S=0) while the Top and Periphery had the highest relative similarity (S=0.8). Figure 3 contains the S values for each comparison between microhabitats.
Discussion
The results of the Chi-square analysis provide evidence that the fish population was extremely stable over the time period of the study. Small variations in population counts from day to day may be a result of census errors rather than fish movement. This is not surprising given the short time frame of the study. Interestingly, Spanier (2000) reports that over a 10 year period, an established artificial reef in the southeastern Mediterranean gained only one new species of fish. Our study would need to be extended for many additional years to support or dispute Spanier’s findings.
The index of similarities tests provide evidence that the microhabitats we picked are each generally characterized by different species. The two most statistically similar habitats, Top and Pinnacle, are visually similar and adjacently located. Using their relative similarity of 0.8 as a reference point, comparisons of other microhabitats have much lower similarity values. Four of the six comparisons have an S value below 0.3. The comparison between the Top and the Periphery is the next highest, at 0.5, which does not is not surprising considering the two microhabitats are also collocated and spatially similar. Overall, the S values usually agreed with observers’ subjective impressions of microhabitat boundaries. However it is important to note that 3 species were categorized as “in between,” because it was difficult to conclusively declare them to spend the most time in a certain microhabitat.
The ordered, regular design of this reef may have played a role in its attraction of fish. A reef with regular, modular construction and internal voids can attract significantly more fish than irregular dispersed reef modules (Brock and Norris, 1989). Our reef utilized regular concrete modules. While internal voids were not large, there were many internal spaces large enough to accommodate almost any of the fish present on the reef. One characteristic of the reef was that it was relatively low to the harbor floor and lacked strong vertical zonation. A stronger vertical component of the reef design would probably make the reef more attractive to fish (Rilov and Benayahu, 2000).
Many qualitative observations were made regarding fish behavior. All fish appeared juvenile, which certainly has an important impact on observed behavior. While little interspecies aggression was observed, certain larger, dominant grunts seemed to be aggressive towards smaller grunts in a territorial fashion. Goatfish also seemed to be consistent schoolers and the Slippery Dick would often accompany the Goatfish schools on forays. When a piece of shell or other debris was dropped in the water close to the reef, a goatfish school would quickly descend upon it, accompanied by a few Slippery Dick. Also, predation, which has been an important factor in other reef fish studies (Kock, 1982), was not apparent to us.
On the last night of the survey, observers watched the reef under night conditions at approximately 10:00 PM. While no data from this observation was incorporated in the study, personal observations indicate that there was virtually no fish activity present on the reef. Some of the species which were observed during the day are nocturnal hunters and presumably away from the reef to feed. The few fish that were remaining (Squirrelfish, Blue Tang, Blackbar Soldierfish) were deep inside the Crevice region and did not appear to move.
These observations from the night snorkel give the impression that the artificial reef serves as a daytime “safe haven” for juvenile fish, which leave during the night to feed. Rooker et. al. (1997) report similar observations on artificial reefs in Flower Garden National Marine Sanctuary, Florida. They specifically note that, “at night, all daytime herbivores and omnivores were either completely absent or greatly reduced in number.” Our limited observations support theirs.
Figure 1. Population counts of each species on each day.
Figure 2a. Number of different species found in each microhabitat
Figure 2b. Microhabitats in which each species was commonly observed.
Figure 3. Index of similarity comparisons between each microhabitat
Acknowledgements
Dr. Hays Cummins and the TME 2001 T.A.s provided guidance on formulating our question, applying statistics, and provided equipment for research use.
References
Bohnsack, J.A., D.E. Harper, D.B. McClellan, and M. Hulsbeck. 1994. Effects of reef size on colonization and assemblage structure of fishes at artificial reefs off southeastern Florida, USA. Bulletin of Marine Science 55(2-3): 796-823.
Brock, R.E. and J.E. Norris. 1989. An analysis of the efficacy of four artificial reef designs in tropical waters. Bulletin of Marine Science 44(2): 934-941.
Gorham, J.C., and W.S. Alevizon. 1989. Habitat complexity and the abundance of juvenile fishes residing on small scale artificial reefs. Bulletin of Marine Science 44(2): 662-665.
Kock, R.L. 1982. Patterns of abundance variation in reef fishes near an artificial reef at Guam. Env. Biol. Fish. 7(2): 121-136.
Rilov G., and Y. Benayahu. 2000. Fish assemblage on natural versus vertical artificial reefs: the rehabilitation perspective. Marine Biology 136: 931-942.
Rooker, J.R., Q.R. Dokken, C.V. Pattengill, and G.J. Holt. 1997. Fish assemblages on artificial and natural reefs in the Flower Garden Banks National Marine Sanctuary, USA. Coral Reefs 16: 83-92.
Spanier, E. 2000. Changes in the ichthyofauna of an artificial reef in the southeastern Mediterranean in one decade. Scientia Marina 64(3): 279-284.
Zahary, R.G., and M.J. Hartman. 1985. Artificial marine reefs off Cataline Island: recruitment, habitat specificity and population dynamics. Bulletin of Marine Science 37(1): 387-395.
Next Article
Previous Article
Return to Topic Menu
It is 7:21:38 PM on Friday, May 9, 2008. Last Update: Saturday, May 4, 2002