Physis

Journal of Marine Science

CIEE Research Station Bonaire Tropical Marine Ecology & Conservation Program

Vol. VI Fall 2009

i

“The least movement is of importance to all nature. The entire ocean is affected by a pebble.”

- Blaise Pascal

Physis is the growth, change, adaptation, and recovery of natural populations. Human populations have had a significant effect on the ocean and the organisms it supports. The ocean has provided the origin of life and sustenance for the billions of people that inhabit Earth today; controlling our climate and governing the areas deemed habitable. Conservation, protection and a greater understanding of this life source is vital. One small step towards this accomplishment is our exploration.

Our aim in Bonaire is to witness the complex interactions within the marine and terrestrial habitats in an attempt to better understand. Through our studies and experiences on the island with CIEE, we have been lucky enough to be immersed in a community with a passion for preserving marine environments. We were all drawn to the sea for one reason or another and have left with a stronger appreciation than before. As Jacques-Yves Cousteau once said, “The sea, once it casts its spell, holds one in its net of wonder forever.”

Photo Credits: Front Cover: Grant M. Frank Forward: Grant M. Frank (top 3), Mollie Sinnott Table of Contents: Mollie Sinnott, Noelle Hawthorne, Alison Masyr, Mollie Sinnott, Maggie Thomas, Aurora Schramm, Jill Lau, Pamela Williams, Noelle Hawthorne Student Profiles: Grant M. Frank Staff Profiles: Anouschka van de Ven; Rita Peachey, PhD Intern Profiles: Anouschka van de Ven (3 interns); (bottom left to right) Mollie Sinnott, Noelle Hawthorne, Alyssa Adler Inside Back Cover: Mollie Sinnott

ii

FOREWORD

The Council on International Educational Exchange (CIEE) is an American non-profit organization with 122 study abroad programs in 40 countries around the world. Since 1947, CIEE has been guided by its mission…to help people gain understanding, acquire knowledge, and develop skills for living in a globally interdependent and culturally diverse world.

The Tropical Marine Ecology and Conservation program in Bonaire is a one-of-a-kind program that exemplifies the ability of CIEE to foresee the need for science-based programs abroad. The goal of the CIEE Research Station Bonaire is to provide a world-class learning experience in Marine Ecology and Conservation. The field- based program is designed to prepare students for graduate programs in Marine Science or for jobs in Natural Resource Management. Student participants enroll in six courses: Coral Reef Ecology, Marine Ecology Field Research Methods, Advanced Scuba, Tropical Marine Conservation Biology, Independent Research and the Cultural & Environmental History of Bonaire. In addition to a full program of study, this program provides dive training that prepares students for certification with the American Academy of Underwater Scientists, a leader in the scientific dive industry.

The marine research reported in this journal was conducted within the Bonaire National Marine Park with permission from the park and the Department of Environment and Nature, Bonaire, Netherlands Antilles. There was also a terrestrial project this semester with implications for the marine environment. Many of the students were involved in collaborative studies with CIEE’s long-term research program, Sea Turtle Conservation Bonaire, the Light and Motion Sensor Program and the Bonaire National Marine Park. Students presented their findings in a public forum on 24 November 2009 at the station for more than 50 members of the general public.

The proceedings of this journal are the result of each student’s Independent Research project. The advisors for the projects published in this journal were Rita B.J. Peachey, PhD and Amanda Hollebone, PhD. In addition to faculty advisors, each student had an intern that was directly involved in logistics, weekly meetings and editing student papers.

iii

TABLE OF CONTENTS

Habitat use, nocturnal behavior, and differences between phases of five common parrotfish species in Bonaire, N.A.

Alyssa Adler…………………………………...……..1

Habitat choice, size distribution, color variance, and feeding behavior of spotted moray eels, Gymnothorax moringa, in coastal waters of Bonaire, N.A.

Grant Marshall Frank…………………………………7

Hawksbill turtle (Eretmochelys imbricata) nests: Possible nutrient sources and drivers of community structure in a tropical marine system

Noelle Hawthorne……………………………..……14

Traditional datu cactus (Ritterocereus griseus) fences reduce run-off rates and transport of sediment and nutrients on hillsides in Bonaire, N.A.

Alison Maysr…………………………………..……20

The effects of the lunar cycle on plankton density, diversity, and diel migration in the coastal waters of Bonaire, N.A.

Aurora Schramm…………………………………….28

iv

TABLE OF CONTENTS

How does water quality correlate with coral disease, bleaching, and macroalgal growth on coastal reefs? A comparative study of various anthropogenic threats on Bonaire, N.A.

Mollie Sinnott……………………………………….36

A comparative study of benthic and coral reef fish communities on artificial versus natural reefs of Bonaire, N.A.

Maggie Thomas…………………………………….45

Parrotfish mucus cocoon production at night in Bonaire, N.A.

Chelsey Weathersbee……………………………….54

Christmas tree worms (Spirobranchus giganteus) and their role as bioindicators of environmental stress on coral reefs of Bonaire, N.A.

Pamela Williams…………………………………….59

Determining how coral reef habitat structure correlates with fish species richness at six dive sites in Bonaire, N.A.

Carolina Yanson……………………………………67

v Map created by Linda Kuhnz

vi Physis: Journal of Marine Science

Habitat use, nocturnal behavior, and differences between phases of five common parrotfish species in Bonaire, N.A.

Alyssa Adler Oregon State University

Abstract Parrotfish are a common and important component of the fringing reefs ecosystem surrounding Bonaire, N.A. In the reef environment, large herbivores like parrotfish graze on macroalgae, allowing for higher coral diversity and abundance. This research studies habitat use among five common species of parrotfish, Scarus guacamaia, Sparisoma viride, Scarus taeniopterus, Sparisoma aurofrenatum, and Scarus vetula found on the leeward coast of Bonaire. The study was performed between Playa Lechi and Something Special dive sites, where transect tapes were placed at three depths representing different habitat types (shallow ridge, reef crest, and reef slope). After a brief recovery period (~ 1 hour), abundances of initial phase and terminal phase parrotfish were determined using SCUBA during midday, and again at night.

Parrotfish density was higher during the day than at night and was significantly different among the three depths. Fisher’s PLSD post-hoc test showed that parrotfish density was significantly higher at 12 m than at 1 or 20 m. During the day, density of initial phase parrotfish was significantly higher than terminal phase, but there was no difference among the three depths. At night, there was no difference between the density of initial phase and terminal phase parrotfish, but there were more parrotfish found at 12 m than at 1 or 20 m. Based on the results of this study, more parrotfish are spotted during the day, parrotfish are found most often at 12 m during the night and day, and there are more initial phase parrotfish than terminal phase at all three depths during the day. Overall, significant findings include information about parrotfish habitat and differences between phases, with the additional note that 12 m depth seems to be an important habitat range for parrotfish in Bonaire.

Introduction component of the reef community because they consume macroalgae that poses a threat to Coral reefs are an important component of coral diversity and viability (Steneck 2007). the world oceans (Steneck 2007) and provide a While parrotfish are abundant on Caribbean variety of benefits to island communities reefs, patterns of habitat use are not well (Burke and Maidens 2004). Reefs are the understood. economic driver of tourism and coral reef fish Most parrotfish are hermaphroditic and provide income for local fisherman as well as change sex between the initial (female) and vital protein for surrounding island terminal (male) life phases. While different communities (Burke and Maidens 2004). phases of the same species feed close to each Coral reef habitats including shallow reef other during the day (Bonaldo et al. 2006), the shelves, mangroves, seagrasses, reef flats, reef terminal phase of Scarus iserti (striped crests, and reef slopes, and each may be parrotfish) will not allow initial phase striped important to different life stages of a variety of parrotfish to feed in the same area (Clifton coral reef fish species. Often juvenile fish will 1989), indicating that there may be differences begin life in seagrass beds, moving onto in feeding habitats between phases within a mangroves and then finally to coral reefs as species. While different phases within a they become more mature (Mumby et al. species may have feeding barriers, individuals 2004). within a phase are usually accommodating, The fringing reef surrounding Bonaire, even among different species (Overholtzer N.A. continues to be among the best in the 1999). A variety of parrotfish species will also Caribbean (Steneck 2007). In Bonaire, large often overlap feeding areas during the day predators are not common, but large (Overholtzer 1999). Parrotfish have been herbivores, such as parrotfish, are (Steneck found to often feed along shallow areas such as 2007). Parrotfish are among the most the reef crest during the day (Ogden 1973), an abundant herbivores and are an important

1 [email protected] Physis: Journal of Marine Science area that is consistently around 12 m depth off and the three transects, on that day, were of the leeward coast of Bonaire. started at that distance north of the Yellow Little is known about the nocturnal Submarine pier. Daytime and nighttime behavior of parrotfish. However; it has been surveys were conducted over the same transect noted that parrotfish prefer deeper areas on lines on each sampling date. Daytime surveys coral reefs rather than shallow reef were conducted between the hours of 11:00 environments at night (Winn 1959). Striped and 13:00 and nighttime surveys were parrotfish migrate from shallow feeding conducted between the hours of 20:00 and grounds to deeper areas to rest nocturnally 21:00. (Ogden 1973). Fish of the family Scaridae Underwater volumetric belt transect have been deemed largely inactive at night, surveys 30 m in length, 1 m high, and 2 m in often resting in coral cavities (Winn 1959). width, were conducted at three depths (1, 12, The purpose of this study was to gain a and 20 m). Because parrotfish are seen better understanding of habitat use by initial foraging during the day along the shallow shelf and terminal phase parrotfish during the at about 1 m depth, a transect was placed daytime and nighttime. The following within the shallow shelf habitat, at a depth of 1 hypotheses were tested: m, and data were collected by snorkeling. The H1: The highest density of terminal phase second transect was placed at 12 m, which is parrotfish will be at 12 m during the day around the reef crest, because parrotfish have because of noted feeding habits and been found to be more abundant at shallow indication of preferred daytime habitat reefs than deep reefs during the day (Ogden 1973). (Bruggeman et al. 1996). The third transect H2: The highest density of initial phase was placed at 20 m because parrotfish are parrotfish will occur at 1 m during daytime thought to go deeper on the reef at night observations due to competitive feeding (Ogden 1973). At 12 and 20 m, data were interactions between phases (Clifton 1989). collected with SCUBA. On sampling days, 3 - This study investigated difference in habitat 30 m transects were placed at 1, 12, and 20 m, use between diel periods and tests these and following a brief recovery period additional hypotheses: parrotfish abundance, species, depth, and H3: Parrotfish density will be higher during phase were recorded during the day and again daytime observations than nighttime at night. At night, divers searched the belt observations, because they will be more transect area with flashlights, searching for fish active, thus easier to observe. in the reef crevices. H4: Terminal and initial phase parrotfish will To determine the effect of time of day and be more abundant in deeper waters (20 m) depth on parrotfish densities a 2-factor analysis at night, based on observations of previous of variance (ANOVA) was used, followed by studies (Winn 1959, Ogden 1973). Fisher’s PLSD post-hoc test. A 2-tailed t-test This study was the first to investigate was used to analyze difference in parrotfish parrotfish habitat use by initial and terminal density in the daytime versus during the phases at night. The purpose of this study is to nighttime. gain a better understanding of night habitat use of parrotfish, as well as to improve the Results understanding of habitat use by different phases during the day. In total, 220 parrotfish were observed during this study. Of those, 35 were seen at Materials & Methods night and 186 were recorded during the day. There were 66 Scarus taeniopterus (princess Visual underwater transect surveys were parrotfish), 31 Scarus vetula (queen conducted between Playa Lechi, near the parrotfish), and 20 Sparisoma aurofrenatum Yellow Submarine Dive Shop (12°09’30.87” (redband parrotfish) seen during the daytime, N, 68°16’52.77” W), and 200 m north of the making them the 3 most abundant species for pier, toward the dive site Something Special, daytime observation (Fig. 1). In total, 13 on the leeward side of Bonaire, N.A. To redband parrotfish, 12 Sparisoma viride determine the start location of the surveys, a (stoplight parrotfish), and 1 Scarus guacamaia random number between 1 and 200 was chosen (rainbow parrotfish) were seen during the Physis: Journal of Marine Science

nighttime, making them the 3 most abundant 0.16 Day species for nighttime observation (Fig. 1). 0.14 ³) Night - Data collected during the daytime resulted in 0.12 higher total initial phase parrotfish than total 0.1 terminal phase parrotfish in every species 0.08 except for stoplight (Fig. 2a), a trend that 0.06 reversed during nighttime data collection (Fig. 0.04 2b). During the day, more parrotfish were Parrotfish density (# density m Parrotfish observed at 12 m depth than at 1 m or 20 m 0.02 depth (Fig. 3; Table 1). 0 11220 70 Depth (m) Day 60 Night Figure 3. Mean density of parrotfish observed during 10 daytime dives and 10 nighttime dives at 3 depths south of Something 50 Special dive site in Bonaire, N.A. 40 Table 1. ANOVA comparing the mean (± SD) density of 30 parrotfish over daytime and nighttime periods and across 3 depths (α = 0.05).

Total parrotfish parrotfish Total 20 Source 10 of Variation SS DF MS F P-value F crit 0 Time of Day 350.416 1 350.416 66.511 < 0.001 4.019 Stoplight Redband Rainbow Princess Queen Depth 50.633 2 25.316 4.805 0.012 3.168 Figure 1. Total abundance of parrotfish species during 10 Time of day daytime and 10 nighttime observations south of Something x Depth 1.033 2 0.516 0.098 0.906 3.168 Special dive site in Bonaire, N.A.

14 Initial phase parrotfish were, however, a Initial significantly more abundant during daytime 12 observation than terminal phase parrotfish (p < Terminal 10 0.001; Fig. 4; Table 2). The interaction between depth and phase is not significant (p = 8 0.358; Fig.4; Table 2), though initial phase 6 parrotfish densities were higher than terminal

Total parrotfish Total 4 phase densities at each depth. There were not significantly more terminal phase parrotfish at 2 12 m than at 1 m or 20 m during the day (p = 0 0.270; Fig. 4; Table 2). Stoplight Redband Rainbow Princess Queen 7 Initial Phase Terminal Phase ) 6 80 Initial -3 5 70 b Terminal 60 4 50 3

40 2

30 1 Total parrotfish Total Parrotfish density (# density 30 m Parrotfish 20 0 10 1 m 12 m 20 m 0 Depth (m) Stoplight Redband Rainbow Princess Queen Figure 4. Mean density of parrotfish observed during 10 daytime dives at 3 separate depths and with regard to parrotfish phase Figure 2. a) Abundance of parrotfish observed during 10 day south of Something Special dive site in Bonaire, N.A. dives at 3 depths south of Something Special dive site in Bonaire, N.A. b) Abundance of parrotfish observed during 10 However, in total data collected, more night dives at 3 depths south of Something Special dive site in Bonaire, N.A. parrotfish were observed at 12 m depth during the day than at 1 or 20 m depth (Fig. 5). Physis: Journal of Marine Science

Table 2. a) ANOVA comparing the mean (± SD) density of 0.14 parrotfish when considering terminal and initial phase and across

3 depths in the daytime (α = 0.05). b) Fisher’s PLSD post-hoc ) 0.12 test for density of parrotfish when considering terminal and -3 initial phase and across 3 depths in the daytime. 0.1

a) 0.08 Source SS DF MS F P-Value F crit 0.06 of Variation Phase 48.600 1 48.600 13.39 < 0.001 13.390 0.04 Depth 9.733 2 4.867 1.341 0.270 2.682 Phase x Depth 7.600 2 3.800 1.047 0.358 2.094 (# density m Parrotfish 0.02

b) 0 Day Night Source Mean Critical of Variation Difference Difference P-Value Figure 6. Mean density of parrotfish observed during 10 daytime Phase 1.800 0.986 <0.001 dives and 10 nighttime dives south of Something Special dive Depth site in Bonaire, N.A. (2 - tailed t-test, p < 0.001). 1 m vs 12 m -0.900 1.208 0.141 1 m vs 20 m -0.100 1.208 0.869 12 m vs 20 m 0.800 1.208 0.189 0.06 Day 0.05 Night When comparing density of parrotfish ) -3 during daytime sampling with density during 0.04 nighttime sampling, regardless of phase, there were significantly higher densities during the 0.03 day (0.103 ± 0.031 m-3) than at night (0.019 ± -3 0.013 m ) (p < 0.001; Fig. 6). On average, 0.02

more parrotfish were recorded in daytime (# density m Parrotfish observation than in nighttime observations, 0.01 regardless of phase or depth (p < 0.001; Fig. 7; 0 Table 3). Terminal Initial

45 Figure 7. Mean densities of initial and terminal phase parrotfish 40 Terminal collected during 10 day dives and 10 night dives south of 35 Something Special dive site in Bonaire, N.A. 30 Initial 25 Table 3. ANOVA comparing the mean (± SD) density of 20 terminal phase and initial phase parrotfish over daytime and 15 nighttime periods (α = 0.05). Total Parrotfish Total 10 5 Source 0 of SS DF MS F P-Value F crit Variation 1 12 20 1 12 20 1 12 20 1 12 20 1 12 20 Time of 245.025 1 245.025 17.68 < 0.001 4.113 Day Princess Queen Rainbow Redband Stoplight Phase 9.025 1 9.025 0.651 0.424 4.113 Time x Species and Depth (m) 42.025 1 42.025 3.032 0.09 4.113 Phase

Figure 5. Abundance of initial phase and terminal phase parrotfish with regard to species observed during 10 day dives at 3 depths south of Something Special dive site in Bonaire, N.A. Discussion

The highest density of parrotfish was at 12 m The first hypothesis of this study predicted during daytime and nighttime observation (Fig. that terminal phase parrotfish would be more 8; Table 4). Regarding phase and nighttime abundant at 12 m depth than at 1 or 20 m, densities, there were more terminal and initial because parrotfish have been found to prefer phase parrotfish at 12 m depth than at 1 m or shallow reef during the day (Ogden 1973). 20 m (p < 0.001; Fig. 8; Table 4). There were more parrotfish at 12 m than 1 or Additional findings show nighttime 20 m during the day, but the difference in parrotfish densities with regard to depth; once terminal phase parrotfish density was not again it is found that there is a significant significant between 12 m and 1 or 20 m. In difference between parrotfish density at 12 m other studies the 12 m depth was also when compared to density at 1 or 20 m (p < important for understanding parrotfish habitat 0.001; Fig. 3). use. Physis: Journal of Marine Science

2.5 Initial Phase hypothesis. The difference in parrotfish abundance could be due to difference in

) Terminal Phase -3 2 activity during the day versus night (Winn 1959). The fact that parrotfish were more 1.5 active during the day than at night made sampling problematic; fish hiding in coral 1 cover were much more difficult to observe and record than those freely swimming. Future 0.5 studies might develop different methods of Parrotfish Density (# 30 m Density Parrotfish observation (such as tagging with transmitters) 0 in order to follow individual fish and record 1 m 12 m 20 m habitat use at night at different depths. Depth (m) The final hypothesis predicted that

Figure 8. Mean density of parrotfish observed during 10 parrotfish would be more abundant at 20 m nighttime dives at 3 separate depths and with regard to parrotfish than 1 or 12 m at night. Statistical analysis of phase south of Something Special dive site in Bonaire, N.A. data showed a significantly higher parrotfish

Table 4. a) ANOVA comparing the mean (± SD) density of density at 12 m than 1 m and at 12 m than 20 parrotfish when considering terminal and initial phase and at 3 m during nighttime observations, which did depths in the nighttime (α = 0.05). b) Fisher’s PLSD post-hoc not support the hypothesis. There was no test for density of parrotfish when considering terminal and initial phase and at 3 depths in the nighttime. significant difference between 1 and 20 m parrotfish density at night. These results a) differed from the results found in other studies Source Mean Critical of Variation Difference Difference P-Value of nocturnal behavior, where diurnal migration Phase -0.367 0.483 0.134 from shallow to deeper reefs occurred (Ogden Depth 1973). This difference could have been due to 1 m vs 12 m -1.200 0.592 <0.001 1 m vs 20 m -0.400 0.592 0.181 variance of substrate with depth between 12 m vs 20 m 0.800 0.592 0.009 Bonaire and prior locations studied such as off the Caribbean coast of Panama (Ogden 1973). b) Coral cavities are utilized as nocturnal Source SS DF MS F P-Value F crit of Variation parrotfish shelter (Winn 1959); perhaps the 12 Phase 2.017 1 2.017 2.312 0.134 2.312 m range in the study site provided better Depth 14.933 2 7.467 8.561 <0.001 17.121 Phase x protection than the habitat at 20 m. 2.533 2 1.267 1.452 0.243 2.904 Depth This study is important to the understanding of parrotfish habitat. Based on For example, Bruggeman (1996) found a this study, it can be said that the 12 m depth is higher rate of coral erosion on the shallow reef important to parrotfish in Bonaire, which may (6 - 12 m) when compared to deeper reefs, help efforts of habitat and ecosystem demonstrating a higher density of parrotfish in conservation in the future. these areas as was seen in this study. The second hypothesis predicted the Acknowledgements highest abundance of initial phase parrotfish would be present at 1 m depth during daytime I would like to thank Dr. Rita Peachey, my observations. However, terminal and initial advisor, for the many hours of work she phase were most abundant at 12 m. Also, at dedicated to my independent research project. each depth initial phase was more abundant Thank you to Caren Eckrich for making sure I than terminal phase. This could be due to the dove safely, Kate Jirik and Lauren Saulino for similar foraging habits between terminal and reading drafts of my paper that were more than initial phase (Bonaldo et al. 2006). Perhaps rough, and Maggie Thomas for helping me the reef at 12 m contains nutrition that the numerous times with Excel. Also thank you to limestone shelf at 1 m does not. Lina Yanson, Mollie Sinnott, and Maggie The third hypothesis predicted higher Thomas for diving with me multiple times a parrotfish abundance during daytime day. observations than nighttime observations. Parrotfish density was significantly higher during the day than at night, supporting the Physis: Journal of Marine Science

References

Bonaldo R. M., J. P. Krajewski, C. Sazima, and I. Ogden, J. C. and N. S. Buckman. 1973. Sazima. 2006. Foraging activity and resource Movements, foraging groups, and diurnal use by three parrotfish species at Fernando de migrations of the striped parrotfish Scarus Noronha Archipelago, tropical West Atlantic. croicensis bloch (Scaridae). Ecology 54:589- Marine Biology 149:423-433. 596. Bruggeman, J. H., M. van Kessel, J. M. van Rooij, Overholtzer K. L. and P. J. Motta. 1999. A. M. Breeman. 1996. Bioerosion and Comparative resource use by juvenile sediment ingestion by the Caribbean parrotfish parrotfishes in the Florida Keys. Marine Scarus vetula and Sparisoma viride: Ecology Progress Series 177:177-187. implications of fish size, feeding mode and Steneck, R. S. and D. E. Olson. 2007. Trends in habitat use. Marine Ecology Progress Series macroalgae abundance in Bonaire, 2003-2007. 134:59-71. A Report on the Status of the Coral Reefs of Burke, L. and J. Maidens. 2004. Reefs at Risk in Bonaire in 2007. Chapters 2-3. the Caribbean. World Resources Institute, Winn H. E. and J. E. Bardach. 1959. Differential Washington, D.C. food selection by moray eels and a possible Clifton, K. E. 1989. Territory sharing by the role of the mucous envelope of parrotfishes in Caribbean striped parrotfish, Scarus iserti: reduction of predation. Ecology 40:296-298. patterns of resource abundance, group size and behavior. Animal Behavior 37:90-103. Mumby, P. J., A. J. Edwards, J. E. Arias-Gonzalez, K.C. Lindeman, P.G. Blackwell, A. Gall, M. I. Gorczynska, A. R. Harborne, C. L. Pescod, H. Renken, C. C. C. Wabnitz, and G. Llewellyn. 2004. Mangroves enhance the biomass of coral reef fish communities in the Caribbean. Nature 427:533-543. Physis: Journal of Marine Science

Habitat choice, size distribution, color variance, and feeding behavior of spotted moray eels, Gymnothorax moringa, in coastal waters of Bonaire, N.A.

Grant Marshall Frank Colorado College

Abstract The spotted moray eel, Gymnothorax moringa, is one of the most abundant moray eels found in the coastal waters of Bonaire, N.A. However, little is known regarding the factors that contribute to its choice of habitat, behavior, and times of activity. Contradictory evidence has been reported for many species of Gymnothorax as to whether they are nocturnal or diurnal, yet little is known concerning color and size, which may be correlated to diet and choice of habitat. This study sought to determine how size, behavior, and color correlate with reef flat and reef slope habitats and at what time (morning or evening) G. moringa is most active. Observations of G. moringa were conducted in the westward coastal waters of Bonaire. A “U”-shaped search pattern was utilized in locating spotted moray eels in 5 adjacent study areas extending perpendicular from the shore to a depth of ~ 15 m. Once an individual was located behavior, jaw size, and color, were recorded in order to assess differences among individuals on differing habitats (reef flat or reef slope), and times of day (morning: 6:00 - 7:30 or evening: 18:00 – 19:30). G. moringa was found to be in greater abundance on the reef flat in the evening displaying exposed venting behavior and individuals were predominantly white in coloration. In the morning G. moringa were found to be in greater densities on the reef slope, displaying foraging behavior, and were predominantly black in coloration. Representatives of all size classes were distributed on the reef flat regardless of time, however, small individuals were not observed on the reef slope in the evening.

Introduction previous studies suggests that foraging activities among moray eels have a strong Moray eels are one of the most notable fish correlation with habitat choice and that activity predators on coral reefs (Winn and Bardach levels are differential among species and not a 1959). Although moray eels have been shared trait by the entire genus. frequently studied there is little consensus There is also contradictory evidence as to among researchers as to the activity patterns of the behaviors utilized by moray eels to acquire these animals. Hobson (1968) and Bohlke and prey. Several studies suggest that morays sit Randall (2000) suggested that moray eels are and wait for their prey (“ambush” predation) opportunistic predators feeding during both the until it comes within striking distance (Hiatt day and night, especially on injured or and Strasburg 1960; Randall 1967), while disturbed fish. During the day Chave and others have observed morays actively foraging Randall (1971) observed Gymnothorax spp. at differing times of the day (Chave and sitting with their heads exposed but little else Randall 1971; Karplus 1978; Young and Winn was seen suggesting primarily nocturnal 2003). Bardach et al. (1959) found two species activity. Dublin (1982) found contradicting of moray to remain only partially exposed evidence in Gymnothorax moringa stating that unless enticed to forage by a rotten piece of individuals were often seen during the day meat, and Randall (1967) went on to describe displaying forms of foraging behavior. morays eels in general to be ambush predators. Whereas, Winn and Bardach (1959) observed However, Hobson (1968) observed an Indo- morays moving over reefs from dusk to dawn, Pacific species of Gymnothorax actively searching for carrion and live prey. Many foraging on reef flat habitats. Similar behavior have described eels of the genus Gymnothorax, was observed by Young and Winn (2003) in in particular, to have increased nocturnal two Gymnothorax spp. on a shallow back reef activities (Hiatt and Strasburg 1960; Starck and in Belize, while Chave and Randall (1971) Davis 1966; Randall 1967) with specific even observed a species of Gymnothorax leave species such as the spotted moray, G. moringa, the water in an attempt to feed on beach being entirely nocturnal (Bardach et al. 1959). dwelling crabs. In addition to behavior, even The wide array of results stemming from these less is known about the combined effects of [email protected] 7

Physis: Journal of Marine Science behavior with size or color variation in relation to habitat choice of moray eels. In Bonaire, N.A. one of the most common reef-dwelling eels is the G. moringa, the spotted moray eel (Humann and Deloach 2002). With a relative absence of sharks and marine mammals, Bonaire’s marine ecosystem consists of higher trophic levels dominated by groupers, snappers, tarpon and eels (G. Frank, personal observation). Apex predators keep food webs in balance and increase resilience to trophic cascades as well as phase shifts in marine ecosystems (Hughes et al. 2007). Despite an apparent prevalence of moray eels on both reef flat and reef slope type habitats (G. Frank, personal observation), it is currently unknown what factor or combined factors drive habitat selection. Therefore this study Figure 1. Map of Bonaire, N.A and the location of all 5 sample addressed the following questions: areas. 1) is G. moringa nocturnal or diurnal, 2) does G. moringa prefer the reef flat or reef slope habitats, 3) does color variation within G. moringa correlate with habitat selection, 4) does size correlate with habitat selection, and 5) does behavior correlate with habitat selection?

Materials & Methods

This study took place along the leeward coast of Bonaire, N.A. between the Pal’i Lechi and Figure 2. “U”-shaped search pattern followed during morning Something Special dive sites (N 12.16030 W and evening observations of each sample area. 068.28210 and N 12.16117 W 068.28310, respectively; Fig. 1). Five rectangular plots (Fig. 2). (approx. 30 x 76 m) were designated along the A red light was used during evening shoreline and ran perpendicular to the shore to observations of eels to avoid altering their a depth of ~ 15 m, thus encompassing both natural behaviors. For each Gymnothorax shallow sand plain and deeper reef habitats. moringa observed data was collected on depth, Boundaries of these sampling sites were location, behavior, coloration, and jaw size marked with rebar stakes for ease of from a distance  50 cm. Depth was recorded navigation. Underwater SCUBA surveys were using a dive computer placed at the same depth used to observe eels in each defined study area as the individual being observed, while twice within a 24 - hour period (morning: 6:00 location was recorded as either reef flat – 7:30; evening: 18:00 – 19:30). Surveys (dominated by sand within a depth range of 0 – followed a “U”-shaped search pattern starting 5 m) or reef slope (dominated by hard corals at the southwestern corner of the study area at within a depth range of 5 – 15 m). a depth of ~ 15 m. Observers swam north 30 Behavior of each eel was observed for 3 m until the opposite boundary of the study site min from a distance  1 m. An acclimation was reached, at which point they swam 4 m period was determined to be negligible because towards the shoreline (East) followed by a 30 of the distance maintained between the m swim south until the opposite boundary was observer and the specimen resulted in no reached. This was conducted over the entirety observable change in behavior despite human of the study site until the shoreline was reached

8

Physis: Journal of Marine Science presence. Behaviors were recorded as surveys. Jaw size could not be quantified 3 foraging, exposed venting, or hiding. An eel times and color 6 times due to a structural was “foraging” if it actively moved from obstruction or unapproachable foraging structure to structure in an apparent search for behavior. Location and behavior were potential prey. “Exposed venting” occurred recorded for all individuals observed. when an eel’s body was obscured by structure, but its head was exposed and the mouth was Density open. “Hiding” occurred when an individual Overall, both location (reef flat or slope) was observed in or under structure without its and time (morning or evening) did not have head exposed. significant effects on eel density (p = 0.743 At the end of each observation period, and p = 0.781, respectively), however, the measurements of jaw and color variation were interaction of location and time did have a conducted as to avoid any behavioral significant effect (p = 0.030). In the morning alterations as a result of the proximity of the more individuals were present on the reef slope measuring device. Jaw size was measured (0.003  0.005 individuals m-2) than on the reef using a ruler attached perpendicular to the end flat (0.002  0.001 individuals m-2), whereas, of a 50 cm PVC pipe and was defined as the in the evening higher eel densities were distance from the tip of the mandible to the end observed on the reef flat (0.003  0.001 of the jaw (below the eye). Eels were individuals m-2) than on the reef slope (0.001  categorized into one of three sizes: small (0.0 - 0.001 individuals m-2; Fig. 3). 1.5 cm), medium ( 1.5 - 3.0 cm) and large ( 0.007 Morning 3.0 cm). Color variation of each eel ) -2 0.006 encountered was determined by counting the Evening number of black and white spots intercepted by 0.005 a ~ 5 cm segment on the body directly 0.004 posterior to the head. Dominant color was 0.003 calculated based on the abundance of white 0.002 0.001 and black spots and eels were categorized into (individuals m Density one of three color variations: white (white > 0 black), black (white < black), or neutral (white Flat Slope = black) if neither black nor white was Figure 3. Mean densities ( SD m-2) of G. moringa on the reef determined to be dominant. flat and reef slope for morning and evening observation times. P- To assess preferred habitat type, eel values derived from 2 - factor ANOVAs (location p = 0.743; time p = 0.781; location x time p = 0.030). density was calculated based the number of -2 individuals m observed on the reef flat and Behavior reef slope. Frequency of occurrence was Eel behavior appeared to differ between calculated for represented categories within the reef flat and reef slope. In the morning behavior, jaw size, and color variation as to there was a significant difference in overall determine the percentage of individuals frequency of behaviors (p = 0.001) and in the displaying the characteristic on the reef flat interaction of behavior and location (p  and reef slope. Frequencies were arcsine 0.001). On the reef flat ~ 83 % of eels were transformed resulting in data with an exposed venting and ~ 17 % were hiding but underlying distribution that is nearly normal foraging was never observed. On the reef slope (Zar 1999). Two–factor ANOVAs were run to ~ 56 % of eels were foraging, ~ 36 % were assess any difference in density (location), exposed venting and ~ 8 % were hiding. The behavior, jaw size, and color. The above foraging behavior was exclusively observed on perameters were analyzed as either morning or the reef slope where as the exposed venting evening observations and were never was observed more often than the hiding combined. behavior on the reef flat (Fig. 4). In the evening, there was a similar pattern Results as seen in the morning. There was a significant difference in overall frequency of In total 54 Gymnothorax moringa were behaviors and between the interaction of observed over the 3-week study period, 24 location and behavior (both p  0.001). during morning surveys and 30 during evening

9

Physis: Journal of Marine Science

1 1 Small a) a) 0.9 Foraging 0.9 Medium 0.8 Exposed Venting 0.8 Large 0.7 Hiding 0.7 0.6 0.6 0.5 0.5 Frequency Frequency 0.4 Frequency 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0 0 Flat Slope Flat Slope Foraging 1 1 Small 0.9 Exposed Venting 0.9 b) Medium

0.8 Hiding 0.8 Large 0.7 0.7 0.6 0.6 0.5 0.5

Frequency Frequency 0.4 0.4 Frequency 0.3 0.3 0.2 0.2 0.1 0.1 0 0 Flat Slope Flat Slope

Figure 4. Mean frequencies ( SD) of behaviors displayed (foraging, exposed venting, and hiding) by G. moringa on the Figure 5: Mean frequencies ( SD) of jaw sizes (small, medium, reef flat and reef slope during the a) morning (behavior p = and large) represented by G. moringa on the reef flat and reef 0.001; location x behavior p < 0.001) and b) evening (behavior p slope during the a) morning (size p = 0.169 location p = 0.936; < 0.001; location x behavior p < 0.001). P - values derived from size x location p = 0.204) and b) evening (size p = 0.035; 2 - factor ANOVAs. location p = 0.732; size x location p = 0.230). P - values derived from 2 - factor ANOVAs. Foraging was observed exclusively on the reef slope (100% of individuals) while exposed equally (~ 21.5 % to ~ 46 % of individuals) venting was the dominant behavior on the reef while on the slope, no small eels were flat (90% of individuals; Fig. 4) observed and medium individuals were observed twice as frequently as large Jaw Size individuals (Fig. 5). In the morning size, location, and the interaction between the two had no significant Color affect on the frequency of occurrence of small, In the morning, there was no significant medium or large individuals (p = 0.169, p = difference in the frequency of colors 0.936, and p = 0.204, respectively). On the reef represented (p = 0.105), however, the flat, all sizeb) classes were represented nearly interaction of location and color did have a equally (~ 28 % to ~ 37 % of individuals) significant effect on frequency (p = 0.004). while on the slope, medium sized eels White coloration was represented on the reef dominated (~ 73 % of individuals) and both flat whereas black coloration was represented small and large eels represented  30 % of the on the reef slope (both 75 % of individuals). population (Fig. 5). The remaining eels on the reef flat were neutral In the evening, location and the interaction of in color, and they were white on the reef slope size and location did not have significant (Fig. 6). effects on frequency (p = 0.732 and p = 0.230, As seen in the morning, there was no respectively). However, there was a significant difference in the frequency of significant difference in overall jaw size of colors overall (p = 0.126), but the interaction individuals observed (p = 0.035). On the reef of location and color had a significant effect on flat, all size classes were represented nearly frequency (p < 0.001) in the evening. White

10

Physis: Journal of Marine Science

1 Black who observed diurnal forays away from a) White structure. 0.9 Neutral 0.8 While foraging was never seen on the reef flat, exposed venting was the dominant 0.7 0.6 behavior regardless of time. Exposed venting, 0.5 while seemingly a stationary behavior, may be

0.4 linked to ambush predation. Several studies Frequency 0.3 have suggested that a “sit and wait” method of 0.2 predation is frequently utilized by moray eels

0.1 (Hiatt and Strasburg 1960; Randall 1967). It is 0 possible that feeding occurs on both the reef

Flat Slope slope and reef flat habitats, but the strategy (foraging or ambush predation) differs. 1 Black White Predation strategy may be driven by the 0.9 Neutral availability of prey in a particular habitat. 0.8 Foraging by G. moringa on the reef slope 0.7 maybe driven by prey richness and as a result 0.6 may not be energetically feasible. Huey and 0.5 Pianka (1981) described foraging predators to 0.4 feed primarily on sedentary prey items while Frequency 0.3 ambush predators typically feed on active 0.2 (foraging) prey items and further suggest that 0.1 foraging mode within a species varies with 0 prey availability. Flat Slope In addition to foraging, black color Figure 6: Mean frequencies ( SD) of color variants (black, variants were only found on the reef slope, white, and neutral) represented by G. moringa on the reef flat while white individuals dominated on the reef and reef slope during the a) morning (color p = 0.105; location x color p = 0.004) and b) evening (color p = 0.126; location x flat (Fig. 6). These differences in color and color p < 0.001). P - values derived from 2 - factor ANOVAs. location appeared in both the morning and evening and therefore does not suggest a Individuals were observed exclusively on the daylight – darkness shift for camouflage. It is reef flat (~ 98 % of individuals), while black possible that individuals of a defined color are individuals were observed exclusively on the selected by their preferred habitat (or vice reef slope (~ 33 % of individuals). The versa) and/or the diet found in that habitat remaining individuals for both locations were plays a role in skin pigmentation similar to that neutral in coloration (reef flat: ~ 2 %, reef found in flamingos and other shore birds slope: ~ 67 % of individuals; Fig. 6) (Ralph 1969). Further research is necessary in determining the exact factors that contribute to Discussion color variations observed. Analysis of stomach content of individuals representing This study revealed that greater densities color variations may determine if diet is indeed of Gymnothorax moringa occur on the reef a driving factor. slope in the morning (~ 66 % of individuals) G. moringa of all size ranges were present and on the reef flat in the evening (~ 74 % of on both reef flat and reef slope habitats in the individuals; Fig. 3). Additionally, foraging morning, however, small individuals were was the dominant behavior displayed on the never observed on the reef slope in the reef slope and was never observed on the reef evening. Numerous studies have found apex flat regardless of time. In fact, most eels (~ 83 predators such as groupers and snappers to be % to ~ 90 % of individuals) on the reef flat highly nocturnally active (Longley and were found exposed venting. These data Hildebrand 1941; Schroeder 1964; Hobson suggest that G. moringa is diurnal, feeding 1965), which may contribute to the reduction during morning and evening hours but only on of small eels on the reef slope in the evening. the reef slope. This supports the findings of It is possible that due to the difficulty of Dublin (1982) and Bohlke and Randall (2000) observing small eels in the evening, some individuals may have been overlooked.

11

Physis: Journal of Marine Science

However, if this is the case it can still be the nocturnal reef predators Gymnothorax concluded that small individuals are not moringa and G. vicinus. Copeia 1959(2):133- actively foraging on the reef slope and may be 139. concealed by refuge structure. In addition, Bohlke E. B. and J. E. Randall. 2000. A review of intermediate sized individuals were found to the moray eels (Angulliformes: Muraenidae) of the Hawaiian Islands, with descriptions of two dominate the reef slope habitat. Reasoning for new species. Proceedings of the Academy of this is unclear, however, reef complexity may Natural Sciences of Philadelphia 150:203-278. lend itself to support intermediate sized eels by Chave, E. H. N., and H. S. Randall. 1971. Feeding providing appropriately sized refuge structure behavior of the moray eel, and prey items. This may also be a factor of Gymnothorax pictus. Copeia 1971(3):570-574. population demographics and may show that Dubin, R. E. 1982. Behavioral interactions between the representative G. moringa population Caribbean reef fishes and eels (Muraenidae around Bonaire to have predominantly medium and Ophichthidae). Copeia 1982(1):229-232. ( 1.5 - 3.0 cm) jaw size. Hiatt, R. W. and D. W. Strasburg. 1960. Ecological relationships of the fish fauna on coral reefs of Furthermore, Young and Winn (2003) the Marshall Islands. Ecological Monographs found evidence of larger G. moringa eating 30(1):65-127. smaller conspecific individuals in Belize. This Hobson, E. S. 1965. Diurnal-nocturnal activity of suggests that increased activity of relatively some inshore fishes in the Gulf of California. larger G. moringa on the reef slope may Coepia 1965(3):291-302. contribute to the absence of smaller individuals Hobson, E. S. 1968. Predatory behavior of some in the evening. shore fishes in the Gulf of California. U.S. Fish Overall, this study provides evidence that Wildlife Services 73:1-92. G. moringa is diurnal and that both morning Hughes, T. P., M. J. Rodrigues, D. R. Bellwood, D. and evening foraging activities are displayed Ceccarelli, O. Hoegh-Guldberg, L. McCook, N. Moltschaniwskyj, M. S. Pratchett, R. S. primarily on reef slope habitats, while possible Steneck, and B. Willis. 2007. Phase shifts, ambush predation occurs on reef flat habitats. herbivory, and the resilience of coral reefs to Size does not appear to correlate with habitat climate change. Current Biology 17:1-6. choice on the reef flat but intermediate sized Humann, P. and N. Deloach. (2002). Reef fish individuals are dominant on the reef slope. identification: Florida, Caribbean, Bahamas. This study has given new insight into a Jacksonville: New World Publications. previously un-described color variation in G. Karplus, I. 1978. A feeding association between the moringa as well as evidence that G. moringa is grouper Epinephelus fasciatus and the moray indeed diurnal, contradicting, previously eel Gymnothorax griseus. Copeia. reported nocturnal tendencies. Further 1978(1):164. Longley, W. H. and S. F. Hildebrand. 1941. research is necessary to determine the exact Systematic catalogue of the fishes of Tortugas, factor or factors that contribute to color Florida. Carnegie Institution of Washington variants and mode of predation. It is still Publication 535(34):1-331. unclear how habitat plays a role in selection of Ralph, C. L. 1969. The control of color in birds. behavior and color and if diet contributes to American Zoologist 9:521-530. these varying traits and activities. Randall, J. E. 1967. Food habits of reef fishes of the West Indies. Studies in Tropical Oceanography Acknowledgements 5:665-847. Schroeder, R. 1964. Photographing the night Thank you to Lauren Saulino and Mollie creatures of Alligator Reef. National Geographic 125(1):128-154. Sinnott for helping in data collection. A Starck, W. A., II and W. P. Davis. 1966. Night special thanks to Maggie Thomas and Mollie habits of fishes of Alligator reef, Florida. Sinnott for aiding in the structure and Ichthyologica 38(4):313–356. formatting of this paper and to Dr. Amanda Winn, H. E., and J. E. Bardach. 1959. Differential Hollebone for all of her hard work, without her food selection by moray eels and a possible this would not have been possible. role of the mucous envelope of parrot fishes in reduction of predation. Ecology 40(2):296-298. References

Bardach, J. E., H. E. Winn, and D. W. Menzel. 1959. The role of the senses in the feeding of

12

Physis: Journal of Marine Science

Young, R. F. and H. E. Winn. 2003. Activity Zar, J. H. (1999). Biostatistical Analysis 5th edition. patterns, diet, and shelter site use for two Upper Saddle River, New Jersey: Pearson species of moray eels, Gymnothorax moringa Education, Inc. and Gymnothorax vicinus, in Belize. Copeia 2003(1):44-55.

13

Physis: Journal of Marine Science

Hawksbill turtle (Eretmochelys imbricata) nests: Possible nutrient sources and drivers of community structure in a tropical marine system

Noelle Hawthorne Bucknell University

Abstract Beach environments are considered nutrient poor systems that support limited abundances of life due to the lack of attainable nutrients. Since the surrounding environment is nutrient limited, plants and organisms residing in sandy beach communities take advantage of available nutrients whenever possible, for example, nests laid on the beach. This study assessed whether nesting hawksbill sea turtles (Eretmochelys imbricata) are transporters of nutrients from ocean systems to nutrient-poor beaches of Bonaire, N.A. It was hypothesized that nitrogen (N) and phosphorus (P) levels would be elevated, infaunal organisms would be more abundant, and plant cover would be higher in nest plots compared to areas without nests. To determine the input of nutrients from nesting and the potential effects of nutrient enrichment on the plants and infauna, five experimental arrays, including nest, mechanically disturbed (no nutrient addition), and undisturbed treatments were sampled from September to October 2009 on Klein Bonaire. Five days following hatching events, sediment cores were taken to assess concentrations of N and P, as well as to determine the abundance of infauna. Plant percent cover was also determined for each plot. Nutrients did not differ significantly among plot type, with both N and P consistently at low concentrations. For all nest plots, 2.5 X more taxonomic groups, including known predators, were detected than in undisturbed or mechanically disturbed plots. No plants were found in any plot type for the duration of the study. This study suggests that hawksbill sea turtle nests are not strong drivers of coastal community structure in Bonaire. It is believed that the CaCO3 composition of the sand and the limestone base of the island do not allow for nutrient retention and thus excess nutrients are not available for exploitation by beach plants or infauna.

Introduction hatched egg clusters were attributed to the increase in meiofauna density, which altered Nutrient transfer from one ecosystem to the infaunal community structure by causing another has been an area of study frequently increases in some species and decreases in visited by scientists, particularly land-sea others. interactions (Hummon et al. 1976; Anderson Not only can infaunal community structure and Polis 1999; Bouchard and Bjorndal 2000). be altered by nest presence, there are other There are many vectors for nutrient transport, organisms that have been shown to take ranging from those that enter a system directly advantage of excess nutrient availability from (e.g. sea lion defecation; Fariña et al. 2003) to nests on beaches. The nutrients deposited by indirectly (e.g. runoff from human activities; green sea turtle (Chelonia mydas) nests in Smith 1996). Nesting sea birds on the Gulf of Oman have supported the growth of fungal California directly contributed levels of species that are not able to grow in other nitrogen (N) and phosphorus (P) to the soil that locations on the beach (Elshafie et al. 2007). were six times higher on islands where the These saprophytic soil fungi, known to birds nested versus where bird nesting was decompose organic matter in soils, were found absent, simply through defecation (Anderson in eggshells, on egg membranes containing and Polis 1999). A co-occurring doubling of N dead embryos, and in the immediate and P levels in two common dune plants surrounding sediments where eggs were laid appeared to be due to assimilation of the (Elshafie et al. 2007). Thus, under temperate excess nutrients in the soil. In Delaware, nesting conditions, green sea turtle nests U.S.A., infauna (within the sediments) significantly contributed to sedimentary fungal increased in density after horseshoe crab community structure. (Limulus polyphemus) egg clusters were Beach communities are typically nutrient deposited in the sandy beaches (Hummon et al. poor and thus, limited in the type and density 1976). Unidentified substances from of plants they can support (Kachi and Hirose [email protected] 14

Physis: Journal of Marine Science

1983). On the beaches of Japan, a common leeching by fungi, plants, and animals dune plant (Oenothera erythrosepala) is potentially occurs with other species of sea primarily limited by N and P availability turtles as it does in green sea turtle nests. (Kachi and Hirose 1983). It is possible that The southern Caribbean island of Bonaire beach and dune plants with similar limitations supports a relatively small population of could benefit from nutrients leeched by sea nesting sea turtles (~ 80 - 100 nests) each year turtle eggs. Additionally, where loggerhead (G. Egbreght, personal communication). A (Caretta caretta) and green sea turtles nest in majority of those nests (~ 60) are laid by the sand dunes of Florida, U.S.A., soil N hawksbill sea turtles, whereas loggerhead, a concentrations were found to increase as nest few green, and occasional leatherback sea density increased (Hannan et al. 2007). Dune turtles account for the remaining nests. Green plants, such as sea oats (Uniola paniculata), sea turtles nest mostly on the East coast of the near sea turtle nesting areas had higher 15N main island of Bonaire at Playa Chikitu concentrations in leaf tissue than those plants (12˚16’47.74” N, 068˚20’53.74” W), whereas farther away from nest sites. Since tissue 15N loggerhead and hawksbill sea turtles tend to levels tend to be elevated when an organism lay eggs on the small island of Klein Bonaire obtains nutrients from a marine source off the west coast of the main island. As compared to terrestrial sources, it was previously suggested for green and loggerhead concluded that the sea oats were exploiting sea turtles in more temperate habitats residual nutrients left by loggerhead and green (Bouchard and Bjorndal 2000; Hannan et al. sea turtle nests. 2007), Bonaire’s nesting sea turtle population It has been estimated that an average may have a more profound effect than loggerhead sea turtle nest contains ~ 18,724 kJ previously realized by directly providing of energy, 72.0 g of N, and 6.5 g of P essential nutrients to dune and beach (Bouchard and Bjorndal 2000). Only 27 % of stabilizing plants as well as infaunal that energy, 29 % of N, and 39 % of P leaves communities. If the nesting sea turtle the beach in the form of hatchlings, but 34 % population declines or moves to different of energy, 51.0 g of N (40 %), and 4.0 g of P nesting beaches, the once stable bionetwork of (32 %) remain in the sand immediately post- plants that retain beach sediments may be hatching and are available for utilization by thrown off balance, causing drastic changes in detritivores, decomposers, and plants (23 % of the coastal ecosystem (Bouchard and Bjorndal nests sampled contained roots growing in or 2000; Hannan et al. 2007). around the eggs; Bouchard and Bjorndal The goal of this study was to address 2000). Witherington (2006) also suggested whether nutrients (N, P, and potassium [K]) that loggerhead sea turtle nests are major from the laying of sea turtle nests are sources of N and P, contributing “thousands of detectable in beach sediments of Bonaire and pounds of fertilizer each year” to nutrient- whether any evident nutrients support the deprived beach plants. Furthermore, Lutz et al. growth of nearby plants and/or infaunal (2003) reinforced the importance of sea turtles communities. It is hypothesized that relatively transporting nutrients from highly productive high levels of nutrients will remain in the nest areas to “nutrient-starved beach sands.” sediments within several days after a hatching Crooks and Sanjayan (2006) advocate that this event compared to baseline nutrients in contribution of “fertilizer” to nutrient-depleted undisturbed sediments. It is further systems is an incentive to “save” sea turtles. hypothesized that the percent cover of beach Detailed chemical and structural analysis plants and species richness of infauna will be of green sea turtle eggshells show that they higher in nest sites than undisturbed sites due contain small gaps as a result of the formation to these organisms exploiting excess nutrients of aragonite crystals, which allow for gas and left by the sea turtle nests. water exchange (Phillott and Parmenter 2007). If an embryo dies, the eggshell flakes over Materials & Methods time, eventually exposing and releasing the internal nutrients to the external environment. Study Site Since eggshell structure of the world’s seven This study was carried out on Klein 2 sea turtle species is similar (Phillott and Bonaire, a 6 km uninhabited tropical island ~ Parmenter 2007), exploitation of nutrient 0.8 km from the western coast of Bonaire, N.A

15

Physis: Journal of Marine Science

(Fig. 1). It is an island with a limestone base plots included the surface and substrate sands surrounded by fringing reefs and beaches (~ 0.6 m deep) that were not handled prior to composed of calcium carbonate sands (De sampling, whereas mechanically disturbed Freitas et al. 2005). The beaches support plant plots were churned up by gloved hand prior to species such as the fleshy carpetweed, sea sampling to simulate a nesting attempt by a purslane (Sesuvium portulacastrum) and dune- hawksbill sea turtle and to control for physical stabilizing beach morning glory (Ipomoea pes- disturbance apart from nutrient addition. Each caprae). Other vegetation include wild olive array was at least 10 m from adjacent arrays. trees (Bontia daphnoides), calabash Nutrient content and infaunal communities (Crescentia cujete), buttonwood (Conocarpus were assessed by taking two randomly placed erectus), and rubber vine (Cryptostegia cores from each experimental plot five days grandiflora) (De Boer 1996). The island is after a hatching event or mechanical frequently visited by nesting loggerhead and disturbance between the hours of 10:30 and hawksbill sea turtles that tend to lay their eggs 11:30 (Table 1). Disturbance events and within and under the vegetation of Klein’s sampling were separated by five days due to beaches (N. Hawthorne, personal observation). accessibility to the island and scheduling with Nesting loggerhead and hawksbill sea turtles the patrolling of Sea Turtle Conservation visit Klein Bonaire from mid-July to Bonaire (STCB – a local conservation and November (G. Egbreght, personal research group). The corer was made of a ¾ m communication), however site selection for long, 0.08 m diameter PVC pipe, marked at 0.6 this study was limited to hawksbill nests as m. Upon sampling, the pipe was driven into they were the only nests hatching during the the sand with a mallet to the mark and the sampling period from 16 Sept. to 28 Oct. 2009. sediments removed. Each core was stored separately and returned to the laboratory for analysis.

Table 1. Hatching, mechanical disturbance, and sampling dates No Name Beach of the five experimental array sites. The array sites are denoted using their location on the beach, from the east end of the beach (2000 m), to the west end (0 m). Sampling took place between 9:00 and 12:00 on days specified in 2009.

marker marker Hatched Sampled Sampled Disturbed Nest beach Mechanical Mechanically Mechanically Nest Sampled

1465 25 Sept. 30 Sept. 9 Oct. 14 Oct.

1865 4 Oct. 9 Oct. 9 Oct. 14 Oct. Figure 1. Map of Klein Bonaire (inset of Bonaire, N.A.). Stars designate experimental site locations. 1854 14 Oct. 19 Oct. 19 Oct. 23 Oct. 1130 18 Oct. 23 Oct. 23 Oct. 28 Oct. Sampling 260 18 Oct. 23 Oct. 23 Oct. 28 Oct. In order to assess the impacts of nutrient loading from hawksbill nesting on Klein Bonaire, replicated nest sites (n = 5) were In the lab, individual samples were first 2 identified along a 2 km stretch of beach. Each sieved using window screen (0.03 x 0.03 cm ). site was based upon the location of a hawksbill Most organic plant matter was removed and sea turtle nest that hatched during the timing of live organisms were separated from the the study. At each true nest plot (nest cavity sediments, preserved (10 % formalin and and the disturbed area on the surface), transferred to 70 % EtOH after > 24 h), undisturbed plots and mechanically disturbed counted, and identified to Order for plots (¼ m2) were randomly assigned to a comparison of taxonomic richness among position along the same tidal height as the nest treatments. The sediments were dried at room and neighboring treatments. Each treatment temperature ( 28 C) for 24 - 48 h and then within an array was separated by ≤ 1 m and analyzed for nitrogen (N), phosphorus (P), and -2 any variation in this distance was due to potassium (K) content (in kg m ) using a impediments such as vegetation. Undisturbed LaMotte Soil Test Kit (Model 3-5880). A

16

Physis: Journal of Marine Science single-factor ANOVA was run for each drainage. The bucket was sampled in the same nutrient to determine if concentrations differed manner as the field experiments, but at 18 h among nests, undisturbed, and mechanically and 5 d after the nutrients were added. It was disturbed experimental plots. not possible to sample the bucket immediately Percent cover and identification of plants after the pour because the water did not filter was determined for all plots of each out quickly and some drying time appeared experimental array at using a ¼ m2 quadrat. necessary. The same solution was also taken Percent cover of plants was also assessed over to Klein Bonaire and poured over a ¼ m2 area time from the hatching event to the end of the 2 m from a previously sampled site. This plot study. was sampled as above at 5 and 60 min after the Upon preliminary assessment that nutrient nitrate was added. concentrations in true nest plots were low, both a laboratory and field-based nutrient addition Results study were set up to determine whether natural concentrations of nitrates (proxy for N) from No difference in nutrient content was an average-sized hawksbill sea turtle nest found among nest, undisturbed, and could be detected by the LaMotte Soil Test Kit mechanically disturbed plots for N, P, or K (p used. Since there is limited data for exact > 0.995, p = 0.397, p = 0.920, respectively; nutrient contents of hawksbill sea turtle eggs Fig. 2). N concentration readings for all available in the literature, it is assumed for this experimental plots consistently read as “low” study that nutrients deposited by each readings on the test kit scale. P concentrations hawksbill sea turtle egg are directly varied little, with only the mechanically proportional to that deposited by a loggerhead disturbed plots showing a slight elevation in sea turtle egg (Bouchard and Bjorndal 2000). the mean (0.96 ± 0.00 103 kg m-2) above nest Results from the Bouchard and Bjorndal and undisturbed plots (0.89 ± 0.00 103 kg m-2). (2000) study were fitted to the average egg K varied the most of the measured nutrients. volume of a hawksbill turtle in order to The highest concentrations of K were found in estimate the amount of nutrients that may enter the undisturbed plots (16.14 ± 4.01 103 kg m-2), a beach habitat following hatching. A typical while they were equal in the nest and hawksbill clutch includes 130 eggs that contain mechanically disturbed plots (15.24 ± 4.01 103 28.70 mL of liquid each and a typical kg m-2). For both the lab and field nutrient loggerhead clutch includes 109 eggs with addition studies, nitrate levels were as low as 36.20 mL each (van Buskirk and Crowder those from the experimental nest study after 1994). With this information, the amount of only 18 h and 5 min, respectively. nitrate to be used in the experiments was calculated: 22 Mean P 20 Mean N 1) 1 loggerhead sea turtle nest = 72.00 g N 18 Mean K nest-1 16 2) 1 loggerhead sea turtle egg = 72.00 g N / -2 14 -1 -1 12 109 eggs clutch = 0.66 g N egg Kg m 10 3

3) 1 hawksbill sea turtle egg = (28.70 mL / 10 8 36.20 mL) * 0.66 g N = 0.52 g N egg-1 6 4) 1 hawksbill sea turtle nest = 0.52 g N * 4 -1 -1 2 130 eggs clutch = 67.60 g N nest 0 5) 1 hawksbill sea turtle nest = 28.70 mL * Nests Undisturbed Mechanical 130 eggs clutch-1 = 3731 mL nest-1

For convenience, the volume of liquid added to Figure 2. Mean ( SD) nitrogen (N), phosphorus (P), and potassium (K) levels in kg m-2 detected in the sands from the the beach sands in this study was rounded up three plot types. to 4 L. Thus, 72.5 g of nitrate were diluted in 4 L of distilled water. In the lab, the solution Of the infauna found within the cores, was poured evenly over an 18.9 L (0.4 m high) there were a total of five taxonomic groups in bucket of sand from Klein Bonaire with ten the nests, including organisms from Araneae, holes drilled in the bottom to allow for slow Hymenoptera, and Amphipoda as well as two

17

Physis: Journal of Marine Science types of pupae, two in the undisturbed plots A higher taxonomic richness over all five (Annelida and one type of pupae), and two in nest plots suggests that the variety of infaunal the mechanically disturbed plots (Annelida and organisms attracted to these areas is higher Amphipoda). Thus, there was a 2.5 X than those mechanically disturbed or difference between nest plots and each control undisturbed plots. The presence of spiders and plot type (Fig. 3). ants solely in the nests may be indicative of generalist and opportunistic predators quickly 6 exploiting what is available (Sanders 2007). 5 Since nutrients were not detectable, it may be 4 the organics, such as proteins, of the remaining 3 turtle material that is attracting these predators, rather than N or P (Mo et al. 1990). Chemical 2 analyses of the organics present in surrounding 1 sediments after a sea turtle hatching event Total taxonomic richness richness taxonomic Total 0 could provide answers as to why these Nest Undisturbed Mechanical consumers were found, but not excess plant Figure 3. Total taxonomic richness of all replicates for each plot cover or roots. type. Additionally, this study did not support the hypothesis, as well as what has been found Plants were never found growing over the previously (Bouchard and Bjorndal 2000; Lutz experimental plots for the entire duration of the et al. 2003; Hannan et al. 2007), that percent study, nor have roots been detected during plant cover would be higher in nest plots due other on-going monitoring of Klein Bonaire to nutrient enrichment. At least in the short (G. Egbreght,personal communication). term, there does not appear to be a lasting supply of nutrients to support the growth of Discussion beach flora in Bonaire. Furthermore, a replicate study of Bouchard The results of this study do not support the and Bjorndal (2000) with hawksbill sea turtle hypothesis that relatively high levels of nests (and other species) that quantifies the nutrients would be found in nests. Although amount of nutrients would provide baseline sea turtle nests provide nutrients to the data on what additions of energy and nutrients surrounding environment, those nutrients are would be expected following hatching events. not necessarily assimilated into the flora and If sediment samples, for both nutrient and fauna within the beach ecosystem (Hannan et organic analyses, could be taken during and al. 2007). The lack of nutrients found in the immediately after a hatching event as well as at nest plots of this study may be attributable to later time intervals, the questions of whether or the calcium carbonate sand found on Klein not they are inputting nutrients, or if it is Bonaire (De Freitas et al. 2005). This sand is simply organics, and how long they remain in highly porous compared to silicate sand the sand could be answered. This would also (Rasheed et al. 2003), the sand found on more lend clarification as to whether the type of sand temperate beaches (Shawl et al. 1995). in which the nests are laid determines the Although nutrients have been shown to remain contribution of nest nutrient loading to plant from sea turtle nests in temperate beaches and infaunal communities. Finally, it would be (Bouchard and Bjorndal 2000; Hannan et al. ideal to conduct studies comparative of sea 2007), the nature of calcium carbonate sand turtle nesting in both calcium carbonate and does not seem to allow for nutrient retention. silica-based sands in order to clarify the role This, in combination with the porous limestone sand composition plays in land-sea nutrient of islands like Klein Bonaire, suggests that any transfer via sea turtles. liquids and dissolved compounds percolate For Bonaire, it appears unlikely that through the sediment faster than silica-based hawksbill sea turtles have a profound effect on sands; additional nutrients from hawksbill sea the coastal beach communities. Any turtle nesting (or other sea turtle species) may reductions in the nesting of hawksbill sea never remain in the system long enough for turtles as a result of increasing human assimilation by plants and/or animals. encroachments and threats may not necessarily affect stabilizing plant growth and thus the

18

Physis: Journal of Marine Science erosion of sand from the island already Hannan, L. B., J. D. Roth, L. M. Ehrhart, and J. F. occurring, but could have an effect on infaunal Weishampel. 2007. Dune vegetation sand community structure and composition. fertilization by nesting sea turtles. Ecology This could alter the function of the system as a 88:1053-1058. whole due to different taxonomic groups filling Hummon, W. D., J. W. Fleeger, and M. R. Hummon. 1976. Meiofauna macrofauna different ecological niches and roles. interactions: I. Sand beach meiofauna affected by maturing Limulus eggs. Chesapeake Science Acknowledgements 17:297-299. Kachi, N. and T. Hirose. 1983. Limiting nutrients I would like to thank Dr. Amanda for plant growth in coastal sand dune soils. Hollebone, my advisor and guide throughout Journal of Ecology 71:937-944. this research experience. Thank you to Jill Lutz, P. L., J. A. Musick, and J. Wyneken. 2003. Lau, Lauren Saulino, Gielmon Egbreght, and The Biology of Sea Turtles: Volume II. CRC Mabel Nava for their support and wealth of Press LLC, Boca Raton, Florida. knowledge. Mo, C. L., I. Salas, and M. Caballero. 1990. Are fungi and bacteria responsible for olive ridley’s egg loss? Proceedings of the 10th Annual References Workshop on Sea Turtle Biology and Conservation. NOAA Technical Memorandum Anderson, W. B. and G. A. Polis. 1999. Nutrient 278:249-252. fluxes from water to land: seabirds affect plant Phillott, A. D. and C. J. Parmenter. 2007. nutrient status on Gulf of California islands. Deterioration of green sea turtle (Chelonia Oecologia 118:324-332. mydas) eggs after known embryo mortality. Bouchard, S. S. and K. A. Bjorndal. 2000. Sea Chelonian Conservation and Biology 6:262- turtles as biological transporters of nutrients 266. and energy from marine to terrestrial Rasheed, M., M. I. Badran, and M. Huettel. 2003. ecosystems. Ecology 81:2305-2313. Particulate matter filtration and seasonal Crooks, K. R. and M. Sanjayan. 2006. Conservation nutrient dynamics in permeable carbonate and Connectivity. Cambridge University Press, silicate sands of the Gulf of Aqaba, Red Sea. Cambridge, UK. Coral Reefs 22:167-177. De Boer, B. A. 1996. Our Plants and Trees: Sanders, D. 2007. Ants and spiders in grassland Curacao, Bonaire, Aruba. Stichting food webs: top-down control and intraguild Dierenbescherming, Curacao, Netherlands interactions. Thesis. Georg-August University, Antilles. Göttingen, Germany. De Freitas, J. A., B. S. J. Nijhof, A. C. Rojer and A. Shawl, S. L., A. Schulman, and P. L. Lutz. 1995. A O. Debrot. 2005. Landscape Ecological comparison of Florida silicate and bahamian Vegetation Map of the Island of Bonaire aragonite sand as substrates for sea turtle (Southern Caribbean). Royal Netherlands nesting. NOAA Technological Memorandum Academy of Arts and Sciences, Amsterdam, NMFS 361:128. the Netherlands. Smith, S. V., R. M. Chambers, and J. T. Egbreght, G. Personal communication. 23 Hollibaugh. 1996. Dissolved and particulate September 2009. Klein Bonaire, Bonaire, nutrient transport through a coastal watershed- Netherlands Antilles. estuary system. Journal of Hydrology 176:181- Elshafie, A., S. N. Al-Bahry, A. Y. AlKindi, T. Ba- 203. Omar, and I. Mahmoud. 2007. Mycoflora and van Buskirk, J. and L. B. Crowder. 1994. Life- aflatoxins in soil, eggshells, and failed eggs of history variation in marine turtles. Copeia Chelonia mydas at Ras Al-Jinz, Oman. 1994:66-81. Chelonian Conservation and Biology 6:267- Witherington, B. E. 2006. Sea Turtles: An 270. Extraordinary Natural History of Some Fariña, J. M., S. Salazar, K. P. Wallem, J. D. Uncommon Turtles. Voyageur Press, Witman, and J. C. Ellis. 2003. Nutrient Stillwater, MN. exchanges between marine and terrestrial ecosystems: the case of the Galapagos sea lion Zalophus wollebaecki. Journal of Animal Ecology 72:873-887.

19

Physis: Journal of Marine Science

Traditional datu cactus (Ritterocereus griseus) fences reduce runoff rates and transport of sediment and nutrients on hillsides in Bonaire, N.A.

Alison Masyr Oberlin College

Abstract Most hard corals require seawater with low nutrients and sediment loads to thrive. Unfortunately, on a global scale, increases in both are currently occurring due to poor coastal zone management practices. This causes damage that is often fatal to reef building corals. Plants living in near shore areas provide natural filters for sediments and nutrients, and recently managers have been harnessing the filtering capabilities of plants to protect aquatic ecosystems. In the marine environment, mangroves provide protection by filtering sediments and absorbing nutrients from runoff before it reaches coral reefs. Ritterocereus griseus, a common cactus species on Bonaire, N.A., has similar capabilities in the terrestrial realm. The following hypotheses were tested regarding cactus fences in Bonaire: smaller amounts of phosphate and sediments would be transported, and lower amounts of runoff would be collected down-slope of plots with cactus fences than plots without cactus fences. Experimental plots with cactus fences were compared to control plots without cacti. To construct plots, steel guides were used to direct simulated rainfall across plots with and without cactus fences into a collection cup at the base of the set-up. This study determined that R. griseus reduces the volumes of runoff and the amount of sediment and nutrients transported down-slope. The use of cactus fences could increase the resilience on Bonaire’s reefs by decreasing sediment and nutrient inputs to near shore waters and are a sustainable resource on the small island.

Introduction nutrient levels. Excess nutrients in near shore waters is another major anthropogenic factor Sediment runoff (Rogers 1990) and excess that can kill corals (Rabalais 2002). Nutrients nutrients (Hallock and Schlager 1986) are reduce coral growth by encouraging algal damaging coral reefs in the Caribbean (Rivera- abundance, which may cover corals and block Monroy et al. 2004). The world-renowned sunlight needed by symbiotic zooxanthellae coral reefs in Bonaire are declining and are at (Rogers 1990). Nutrients can also inhibit the risk of being destroyed (BNMP 2006). In the formation of calcium carbonate, CaCO3, the past few years, many construction sites have molecule corals use to build their skeletons appeared on the hillsides and along the coast (Hallock and Schlager 1986; Rogers 1990). of Bonaire (BNMP 2006). Bonaire has not Studies have shown that high concentrations implemented sediment retention practices and of phosphate result in decreased rates of heavy rainfall events cause large amounts of calcification (Kinsey and Davies 1979; sediment to be washed into the ocean (BNMP Snidvongs and Kinzie 1994). High phosphate 2006). In addition, feral ungulates, such as levels can also have a direct relationship with donkeys and goats, graze on the ground cover coral mortality rates because they increase reducing its presence and increasing erosion water turbidity, complicating light absorption (BNMP 2006). The fine particles in runoff for zooxanthellae (Walker and Ormond 1982). cloud the water and damage corals in two In addition to direct damage to corals by ways. First, by attenuating light, particles sediments due to smothering and the effects of prevent photosynthetic coral symbionts, nutrients on algal growth that can allow algae zooxanthellae, from absorbing light needed for to outcompete corals, increases in sediments photosynthesis (Rogers 1990). Secondly, too and nutrients via runoff can stress reef- much sediment can push corals beyond their building corals, making them susceptible to usual cleansing threshold whereby they can no disease (Rabalais 2002). longer rid themselves of sediments, which can Vegetation filters are economical, self- result in mortality (Tomascik and Sander sustaining and have proven to be effective in 1987). retaining sediments and in absorbing nutrients While all living things require nutrients to (Abu-Zreig et al. 2003). Several species of thrive, some organisms are adapted to low scrubs are successful in preventing runoff by

[email protected] 20

Physis: Journal of Marine Science physically blocking sediments and impeding fences could increase coral resilience by the water’s path (Casermeiro et al. 2004). reducing the harmful effects of runoff. Casermeiro et al. (2004) concluded that areas densely packed with Rosemary shrubs are Materials & Methods successful at preventing erosion and runoff. On Bonaire, a practice of building living Study Site fences using cactus to indicate borders and This study took place on Bonaire, an contain livestock is utilized. The cactus fences island that is part of the Netherlands Antilles may provide similar benefits in terms of (Fig. 1). Bonaire is approximately 280 km2 reducing runoff and erosion of sediments as (BNMP 2006) and is located 87 km north of other vegetation like scrubland (Casermeiro et Venezuela (De Freitas et al. 2005). The island al. 2004) or mangroves (Roberston and has both semi-arid and arid regions, with Phillips 1995). under 80 and 50 cm of annual rainfall, This study focused on Ritterocereus respectively (De Freitas et al. 2005). Despite griseus, an abundant cactus species on Bonaire arid conditions, there is a rainy season during (De Boer 1996). R. griseus, commonly known which major rainfall events occur (De Freitas as the datu, begins dividing into many et al. 2005). The majority of the island ranges branches at the ground level (De Boer 1996) in altitude from 4 to 15 m above sea level (De and the island’s residents often make living Freitas et al. 2005). Although there is not fences out of it because it is plentiful, great topographical relief, many houses in affordable and unattractive to goats and Nord Saliña, Republic, and other donkeys (F. Simal, personal communication). neighborhoods are built on hillsides where R. griseus branches are cut and lined up erosion is occurring due to the removal of perpendicular to the thin layer of topsoil (De ground cover by feral ungulates or Freitas el al. 2005), where they regenerate construction of new homes (BNMP 2006). shallow roots. Shallow root systems have This study took place in Republic, a been known to be effective in quickly storing residential neighborhood located north of the large amounts of water (Cody 2002) and capital, Kralendijk (Fig. 1). A 9 m long nutrients in dry climates (Nobel 1989). section of 13-month old cactus fence was Presently, cactus fences, which have identified along a dirt road in Republic (12° proven effective at keeping livestock inside or 11' 2.70" N, 68° 16' 40.39" W; Fig. 2). outside of yards and fields, may have other Approximately 0.08 km away from the cactus valuable functions such as sediment retention, fence along the same dirt road, three nutrient removal or reduction of runoff, nonconsecutive sections of 9 m total without especially on sloping landscapes. The cactus or other vegetation were chosen for no following hypotheses were tested regarding cactus treatment plots (12° 11' 2.83" N, 68° 16' cactus fences in Bonaire: 43.00" W; Fig. 3). Experimental plots at each H1: Smaller amounts of phosphate would be location, cactus and no cactus, were chosen at transported by runoff on plots with R. random and each plot was only used once. griseus fences than plots without fences; Treatments were carried out by sequentially H2: Smaller amounts of sediment would be alternating between the cactus and no cactus collected down-slope of plots with cactus locations to reduce the possible effects of time fences than plots without cactus fences; or day. and H3: Runoff rates would be lower on plots with Experimental Design R. griseus fences than plots without cactus Prior to experimentation, the slope of each fence. plot was determined using a level. The level The results of this study could be of great was placed in the middle of the plot with the interest and consequence for Bonaire. Cactus up-slope end located 15 cm before the cactus fences are an widespread natural resource that fence or 30 cm up-slope of the middle of the could be used to retain sediments on hillsides plot in control treatments. Mean slope for the and potentially aid the island in managing cactus and no cactus areas were 12.3  1.7  runoff during rain events. Runoff carries with and 12.0  2.3 , respectively. The shape of it nutrients, herbicides, and pesticides that can the experimental set-up was similar to Abu- harm coral reefs (BNMP 2006). R. griseus Zreig et al. (2003). Trials were run in an

21

Physis: Journal of Marine Science

Figure 3. Photograph of no cactus area located about 0.08 km from the cactus fence along the same road in Republic, Bonaire, NA (12°11'2.83"N, 68°16'43.00"W).

a collection cup (567 mL). The cup was

Figure 1. Map indicating the no cactus control site indicated by buried in the ground, up to its rim, at the base the pink tab (12°11'2.83"N, 68°16'43.00"W ) and cactus fence of the ‘V’ to collect runoff water (Fig. 4). A site indicated by the blue tab (12°11'2.70"N, 68°16'40.39"W ). rectangular piece of latex, approximately 14 x Inlay: Map of Bonaire with study site indicated by the black box. 22 cm, was placed under the guides at the base of the ‘V’ to further aid in funneling water into the collection cup. This prevented the soil that had been disturbed while digging a hole for the collection cup from being carried in the runoff. The simulated rainfall lasted 45 s. Runoff was collected in the cup during the same period of time or until the collection cup was full, whichever came first. If the cup filled before the end of the simulated rainfall, the time to fill the cup was recorded. The same method was used for the “no fence” control trials, except in order to compensate for the distance that would have been taken up

by cactus in the experimental plot, water was Figure 2. Photograph of cactus fence located along a road on a poured 30 cm before intersection of the hillside in Republic, Bonaire, NA (12°11'2.70"N, parallel and ‘V’ shaped guides in the control 68°16'40.39"W). plot (Fig. 4). Ten replicates were completed identical manner for both cactus treatment for the cactus fence treatment and nine and no cactus treatment plots. Each plot was replicates were completed for the control or no approximately 0.5 m wide (Fig. 4). A 0.5 m fence treatment because the steel guides failed length of PVC pipe with twenty-four holes in the tenth replicate for the control plots. (0.3 cm diameter each) spaced approximately Phosphate Transport Rate 1.3 cm apart, was used to simulate a rainfall Prior to each trial, 12 g of Scott’s Miracle- event on the plots (Fig. 5). For each trial, 3.79 L of water was poured Gro was added to the water, as instructed by into the rain simulation apparatus, which was the manufacturer. Miracle-Gro Water held approximately 10 cm above the ground Soluble All Purpose Plant Food contains 8 % and 15 cm uphill of the cactus fence. Two phosphate (Scotts.com). The amount of pieces of stainless steel, 60.7 x 5 cm, were phosphate dissolved in the water was tested placed along the ground on each side of the using the Red Sea Phosphate Marine and plot to guide the water down-slope (Fig. 4). Freshwater Test Lab kit to ensure that each Runoff was directed by two additional sample contained the same amount of phosphate. Mean level of phosphate before stainless steel pieces of the same dimensions -1 that came together in the shape of a ‘V’ at each trial was 0.5 mg . After each trial, the

22

Physis: Journal of Marine Science

calculated for cactus treatment and no cactus plots.

Statistical Analysis Two sample t-tests assuming equal variances were performed to compare the amounts of phosphate, sediments, and runoff rates between plots with cactus fences and those without. P-values from a one-tailed t- test were used. Figure 4. Diagram of cactus treatment plots (left) and no cactus control plots (right) showing the upslope simulated rainfall, direction of runoff, steel guides, scale of plots, cactus fence and Results collection cups.

Phosphate Transport Rate The amount of phosphate collected down- slope of the no cactus treatment was 3 X that of the cactus treatment plots (Fig. 6). The mean phosphate collected from runoff on each cactus treatment plot was 0.13  0.13 mg L-1 s-1 whereas, the mean phosphate collected from runoff on no cactus plots was 0.40  0.14 mg L-1 s-1 (Fig. 6). Mean phosphate transport rate was found to be significantly lower from cactus treatment plots than from no cactus treatment plots (p < 0.001). Of the 2 mg of

phosphate that was contained in the simulated Figure 5. Rainfall simulation apparatus; 0.5 m length of PVC rainfall, 80.6 % was delivered to the collection pipe with an opening on the top for the addition of water and 24 cup in the no cactus treatments versus 26.7 % holes on the bottom spaced approximately 1.3 cm apart to dispense water on to experimental plots (0.3 cm diameter each). of the phosphate for the cactus treatment. amount of phosphate in the collection cup was Sediment Transport Rate -1 -1 calculated in mg L s . Mean phosphate The mean sediment transport rate was transport rate and standard deviation were significantly lower over cactus treatment plots calculated for cactus treatment and no cactus than no cactus treatment plots (p  0.012). plots. The rate of sediment transport in no cactus treatment plots was 8 X higher (52.1  57.4 g Sediment Transport Rate L-1 s-1) than cactus treatment plots (6.3  8.8 g Water samples were held for 1 h at 4 C to L-1 s-1; Fig. 7). allow the sediment particles to settle. Water was decanted into a beaker and the volume Runoff Rate was recorded. To determine sediment weight, The mean runoff rate across no cactus sediments were placed in pre-weighed fence plots was 3 X that of the cactus aluminum foil boats and dried in an oven at 40 treatment plots (Fig. 8). The mean runoff oC for 24 h. After drying, the amount of -1 -1 across no cactus fence plots was 3.9  19.9 mL sediment collected was calculated in g L s , s-1 whereas, the mean runoff on cactus by subtracting the initial weight from the final treatment plots was 12.4  18.6 mL s-1 (Fig. weight then accounting for the volume (L) and 8). In 45 s, 2.1 % of the water added to cactus time (s). Mean sediment transport rate and treatments was collected, while 6.6 % of the standard deviation were calculated for cactus water was collected from no cactus treatments. treatment and no cactus plots. The t-test indicated a highly significant difference between runoff rates from cactus Runoff Rate treatments versus no cactus treatments (p < The runoff rate was calculated by dividing 0.001). the runoff volume by the trial time (mL s-1). Mean rate and standard deviation were

23

Physis: Journal of Marine Science

reduce the effects of runoff on coral reefs in 0.6 Bonaire. ) -1 s -1 0.5 Phosphate Transport Rate 0.4 Phosphate can negatively influence coral reefs because it can inhibit the calcification 0.3 processes of corals and promote algal growth (Kinsey and Davies 1979; Hallock and 0.2 Schlager 1986; Rogers 1990). Algae can

0.1 overgrow corals, blocking sunlight needed by zooxanthellae for photosynthesis (Rabalais

Phosphate in collected run-off (mg L 0 2002). The results of this study support the Cactus Fence Control hypothesis that significantly lower levels of phosphate would be transported across Figure 6. Comparison of mean phosphate transport rate (± SD mg L-1 s-1) collected from runoff on 0.5 m cactus fence (n = 10) experimental plots with cactus fence than and no fence control (n = 9) (t-test, p < 0.001). control plots (p < 0.001; Fig. 5). This experiment did not measure phosphate 120

) absorption but rather the mechanical -1 s withholding of phosphate by the presence or -1 100 absence of cactus fence. By allowing less water to flow through the fence, Ritterocereus 80 griseus prevented most of the phosphate from

60 passing through as well. Cactus fence plots retained 73 % of the phosphate applied as a 40 component of simulated rainfall, whereas plots without cactus retained only 19 % of the 20 phosphate. Therefore, R. griseus fences are

Sediment in collected run-off (g L (g run-off in collected Sediment capable of reducing the amount of phosphate 0 transported to the ocean. Cactus Fence Control The use of vegetation as a method for Figure 7. Comparison of mean sediment transport rate (± SD g L-1 s-1) collected from runoff on 0.5 m cactus fence (n = 10) and decreasing water flow and thereby increasing no fence control (n = 9) (t-test, p = 0.012). nutrient uptake has been documented in other studies (Gopal 1999; Abu-Zreig et al. 2003). 18 Abu-Zreig et al. (2003) concluded that filter 16 length and plant density are important factors 14

) in increasing filter efficacy for phosphorus. -1 12 The effects of cactus density were not 10 investigated in this study but could affect the 8 overall absorption rates. R. griseus has 6 shallow roots, a root system type known to 4

Runoff rate rate Runoff (mL s quickly absorb nutrients (Nobel 1989). This is 2 an area that warrants more study because 0 cactus fences may also reduce nutrients Cactus Fence Control through uptake by the roots, further reducing Figure 8. Comparison of mean runoff rate (± SD mL s-1) from the amount of phosphate reaching the ocean. 0.5 m cactus fence (n = 10) and no fence control (n = 9) (t-test, p < 0.001). Sediment Transport Rate

Sediments harm corals because they Discussion increase the turbidity of the water (Walker and

Ormond 1982; Rogers 1990). Corals are not It is vital to reduce the amounts of only suffocated, but cannot attain sunlight for phosphate and sediments reaching the ocean their algal symbionts (Rogers 1990). The via runoff in order to sustain coral reef results of this study support the hypothesis that ecosystems. The results of this study show less sediment would be transported down- that traditional cactus fences can be utilized to slope in plots with cactus fence when

24

Physis: Journal of Marine Science compared with areas without cactus (p = around the island and requires little in terms of 0.012). The weight of sediments transported upkeep when used as a fence. Use of the across no fence treatments was 8 X higher cactus as fencing appears to be a sustainable than the cactus fence plots (Fig. 6). The cactus practice because sections are removed from fence demonstrated effectiveness at retaining existing plants that continue to grow and the sediments in comparison to non-vegetation section in the fence will root and also thrive. areas. Decreasing sedimentation is an Erosion is evident on the hillsides of Nord essential to maintaining normal recovery Saliña and Republic and runoff is entering the intervals for coral reefs (Paine et al. 1998). nearshore environment from construction sites Studies have already shown the plant located on the waterfront, which are typically density is a useful tool in decreasing cleared of all vegetation prior to construction sedimentation (Abu-Zreig et al. 2003; (A. Masyr, personal observation). The use of Casermeiro et al. 2004). Casermeiro et al. cactus fences in the neighborhoods located on (2004) concluded that increased plant cover hillsides would likely be effective in reducing decreases erosion in studies with different runoff to the ocean, especially if the fences types of scrubs, rather than cacti that were were constructed on lots prior to building new used in the current study. However, the same houses. It may be less plausible to use cactus effect should apply because while the plants fences on the waterfront for large construction differ in morphology and height, they both sites because fences would need to be have shallow root systems. Pearce et al. constructed in advance of the project to give (1998) showed that plant height has no effect the cacti time to root. on sediment retention. The use of cactus fences is an example of how local practices, derived from tradition, as Runoff Rate opposed to science, provide insight to As predicted, runoff rates were lower for ecological processes and sustainable solutions areas with cactus fence compared to the no for local problems. For decades, Robert cactus controls. The mean runoff rate from Johannes compared the practical knowledge of cactus plots was significantly smaller than that artisanal fishermen to scientific knowledge of control plots (p < 0.001), reducing runoff accepted by research communities. For from the plot by 3 X (Fig. 7). The results of example, it was determined that the fishermen this study suggest that R. griseus may be a of Marovo Lagoon, Solomon Islands, useful resource in protecting Bonaire’s coral possessed information pertaining to local fish reefs. Decreasing the amount of runoff behavior that was untested by the scientific entering the ocean is crucial as runoff often world until 2000 (Johannes and Hviding carries nutrients, herbicides, pesticides and 2000). By observing sea birds gathering sediments (BNMP 2006), which are all above the water, fishermen could tell which harmful to coral reefs (Walker and Ormond predatory fish were below the surface 1982; Rogers 1990). An indirect relationship (Johannes and Hviding 2000). Traditional, between plant density and runoff rate has non-science based knowledge has been passed already been shown for woodland areas down from generation to generation in the (Greene et al. 1994). Greene et al. (1994) Netherlands Antilles. In Curaçao, people who concluded that this is because stems can create could not afford wells formerly used the areas for water to quickly seep into the soil. Kadushi cactus, Subpilocereus repandus, as a Estimated over a 10 m area, a cactus fence water purifier (Morton 1967). They placed would allow 78.2 mL s-1 runoff whereas, a 10 pieces of despined cactus in muddy water and m no cactus area would allow 248.6 mL s-1. when they returned days later the water was This evidence suggests that Ritterocerues reported to be potable (Morton 1967). These griseus can slow runoff rate and could be traditional practices capitalize upon applied on a larger scale on hillsides in ecologically important occurrences that have Bonaire. R. griseus fences are capable of only recently been recognized by scientific holding back large amounts of runoff and communities. Ritterocereus griseus fences are therefore phosphate and sediments. Cactus among these beneficial customs as they reduce fences may be a useful tool in reducing the nutrient transport rates, sediment transport anthropogenic influences occurring on coral rates and runoff rates. reefs of Bonaire. The cactus is abundant

25

Physis: Journal of Marine Science

Acknowledgements aggregating behavior. SPC Traditional Marine Resource Management and Knowledge I would like to thank Dr. Rita Peachey for Information Bulletin 12:22-29. her invaluable advising and countless hours of Kinsey, D. W. and P. J. Davies. 1979. Effects of field work, Kate Jirik, Lauren Saulino, and Jill elevated nitrogen and phosphorous on coral reef growth. American Society of Limnology Lau for their revisions and insight, and Evelyn and Oceanography 24:935-940. Anthony for use of her cactus fence. "Miracle-Gro water soluble all purpose plant food - Scotts Miracle-Gro." 25 Sept. 2009. Gardening References & Lawn Care Advice for Enthusiasts - Scotts Miracle-Gro. http://www.scotts.com/smg. Abu-Zreig, M., R. P. Rudra, H. R. Whiteley, M. N. Morton, J. F. 1967. Cadushi (Ceres Repandus Lalonde, and N. K. Kaushik. 2003. Mill.), a useful cactus of Curacao. Springer Phosphorous removal in vegetative filter strips. 21:185-189. Journal of Environmental Quality 32:613-619. Nobel, P. S. 1989. Temperature, water availability, Bonaire National Marine Park. 2006. BNMP and nutrient levels at various soil depths- Management Plan. consequences for shallow-rooted desert Casermeiro, M. A., J. A. Molina, M. T. de la Cruz succulents, including nurse plant effects. Caravaca, J. H. Costa, M. I. H. Massanet, and American Journal of Botany 76:1486-1492. P. S. Moreno. 2004. Influence of scrubs on Paine, R. T., M. J. Tegner, and E. A. Johnson. runoff and sediment loss in soils of 1998. Compounded perturbations yield Mediterranean climate. Catena 57:91-107. ecological surprises. Ecosystems 1:535-545. Cody, M. L. 2002. Growth form variations in Pearce, R. A., G. W. Frasier, M. J. Trlica, W. C. columnar cacti (Cactaceae: Pachycereeae) Leininger, J. D. Stednick, and J. L. within and between North American habitats, Smith.1998. Sediment filtration in a montane p. 164-188. In T. H. Fleming and A. Valiente- riparian zone under simulated rainfall. Journal Banuet (eds.), Columnar Cacti and Their of Range Management 51:309-314. Mutualists: Evolution, Ecology, and Rabalais, N. 2002. Nitrogen in aquatic systems. Conservation. University of Arizona Press, Ambio 31:102-112. Arizona. Rivera-Monroy, V. H., R. R. Twilley, D. Bone, D. De Boer, B. A. 1996. Candle Cactus, p. 54-55. _In L. Childers, C. Coronado-Molina, I. C. Feller, B.A. De Boer, B.A. Our plants and trees: J. Herrera-Silveira, R. Jaffe, E. Mancera, E. Curacao, Bonaire, Aruba. Stichting Rejmankova, J. E. Salisbury, and E. Weil. Dierenbescherming Curacao, Curacao, 2004. A conceptual framework to develop Netherland Antilles. long-term ecological research and management De Freitas, J. A., B. S. J. Nijhof, A. C. Rojer, and objectives in the wider Caribbean region. A. O. DeBrot. 2005. Landscape ecological BioScience 54:843-856. vegetation map of the island of Bonaire Roberston, A. I. and M. J. Phillips. 1995. (Southern Caribbean). Caribbean Research Mangroves as filters of shrimp ponds effluent: and Management of Biodiversity Foundation Predictions and biogeochemical research 9-64. needs. Hydrobiologia 295:311-321. Gardner, T. A., I. M. Côté, J. A. Gill, A. Grant, and Rogers, C. S. 1990. Responses of coral reefs and A. R. Watkinson. 2003. Long-term region- reef organisms to sedimentation. Marine wide declines in Caribbean corals. Science Ecology Progress Series 62:185-202. 301:958-960. Simal, F. 22 Sept 2009. Manager. Washington Gopal, B. 1999. Natural and constructed wetlands Slagbaai National Park, Bonaire, N.A. for wastewater treatment: Potentials and Communicated by e-mail: washingtonpark problems. Water Science and Technology @stinapa.org 40:27-35. Snidvongs, A. and R. A. Kinzie. 1994. Effects of Greene, R. S. B., P. I. A. Kinnell, and J. T. Wood. nitrogen and phosphorus enrichment on in vivo 1994. Role of plant cover and stock trampling symbiotic zooxanthellae of Pocillopora on runoff and soil-erosion from semi-arid damicornis. Marine Biology 118:705–711. wooded rangelands. Australian Journal of Soil Tomascik, T. and F. Sander. 1987. Effects of Research 32:953-973. eutrophication on reef-building corals. Marine Hallock, P., W. Schlager. 1986. Nutrient excess Biology 94:53-75. and the demise of coral reefs and carbonate Walker, D. I. and R. F. G. Ormond. 1982. Coral platforms. Palaios 1:389-398. death from sewage and phosphate pollution at Johannes, R. E. and E. Hviding. 2000. Traditional Aqaba, Red Sea. Marine Pollution Bulletin knowledge possessed by the fishers of Marovo 13:21–25. Lagoon, Solomon Islands, concerning fish

26

Physis: Journal of Marine Science

The effects of the lunar cycle on plankton density, diversity, and diel migration in the coastal waters of Bonaire, N.A.

Aurora Schramm Eckerd College

Abstract The lunar cycle is a key environmental factor influencing the feeding, reproduction, and migration of many marine organisms, including fish, invertebrates, and zooplankton. To investigate the influence of lunar stage on zooplankton density and diversity in Bonaire, Netherlands Antilles, samples of zooplankton were collected from surface waters at midday and at night at each stage of a complete lunar cycle. The purpose of this research was to determine 1) whether zooplankton in surface waters are more abundant during the night or day, 2) during which stage in the lunar cycle zooplankton densities are the highest, 3) whether diel period has an effect on the biodiversity of plankton, and 4) whether lunar stage has a effect on biodiversity of zooplankton. It was found that microzooplankton and macrozooplankton occurred in higher densities at night. The highest microzooplankton density occurred during the waning gibbous phase and the highest macrozooplankton density occurred during the first quarter (289.7 individuals m-3, 28.4 individuals m-3, respectively). Organisms from the classes maxillopoda, malacostraca, and chaetognatha were most prevalent in all samples. The full moon showed both the greatest and least taxonomic diversity among samples with 19 different classes (avg. 13.4) found during the nighttime sample and 7 appearing during the midday sample. Due to an abundance of eggs during the waning gibbous lunar stage it is suggested that lunar spawning has an impact on plankton density and composition.

Introduction has the greatest gravitational pull. This happens during either the new or full moon, The lunar cycle, which completes a full and switches every seven months (Skov et al. rotation every 29.5 d is made up of 8 distinct 2005). In the northeastern coast of the United phases: new moon, waxing crescent, first States, the beginning of the spawning cycle of quarter, waxing gibbous, full, waning gibbous, the American oyster (Crassostrea virginica) 3rd (or last) quarter, and waning crescent correlates with four of the lunar stages, with (NOAA 2007). These phases refer to the the main spawning events occurring illuminated portions of the moon as viewed approximately 8 d after the full moon and from Earth while it completes its immediately after the last quarter (Loosanoff approximately month long orbit (Fig.1). and Nomejko 1951). In a study by Robertson Because of the gravitational pull of the moon et al. (1990), 14 out of 17 species of reef fish, on Earth, the lunar cycle has both direct and were found to have spawning cycles, which indirect effects on many marine organisms. correlated to the lunar phases. Direct effects such as feeding pattern modification (Heidelberg et al. 2004), migration pattern modification (Naylor 2001), and reproductive cycle regulation (Takemura et al. 2004) have influences on abundance, and presence of zooplankton in the water column. Indirect effects such as nocturnal light levels (Ohlhorst 1982), and tides (Naylor 2001) have an influence on predation and transport of zooplankton. Many marine organisms such as reef fish and invertebrates spawn during specific lunar phases (Robertson et al. 1990; Seitz and Schaffner 1995; Lamare and Stewart 1998; Figure 1. Diagram of sunlight (from right) and sequence of Takemura et al. 2004), the highest tides of the moon phases as seen from Earth’s northern hemisphere, year occur when the moon is at perigee and www.moonconnection.com.

27 [email protected] Physis: Journal of Marine Science

Zooplankton are heterotrophic organisms abundances of fish eggs. Zoea, fish eggs, that are associated with the first few meters of chaetognaths, polychaete, and molluscan the ocean where they feed on phytoplankton, larvae were found to increase fivefold in the marine detritus, or other zooplankton night samples with the larger zooplankton (> (Heidelberg et al. 2004). Although some 710 µm) increasing by more than one order of zooplankton contain photosynthetic magnitude (Yahael 2005). In the Southern symbionts, they are not confined to the Caribbean, specifically Bonaire, N.A., there is euphotic areas of the ocean during the day minimal information on the abundance and when predation pressures by visual predators diversity of zooplankton in surface waters are highest (Loose and Dawidowicz 1994). during different phases of the moon (Foster Instead, zooplankton utilize diel vertical 1987; Robertson et al. 1990). migration (DVM), the nocturnal vertical Because Bonaire is located 12 north of ascent of plankton through the water column the equator, it experiences an average sea (Dodson 1990). temperature of 27.8 °C, minor seasonal DVM is primarily influenced by light changes and monthly high and low tide levels, turbidity and food availability; it is also fluctuations of less than a meter (Calvert employed as an anti-predation mechanism 2004). Because of reduced seasonal against vision-dependent hunters (Dodson variability and tidal ranges, if lunar phases do 1990). A study by Gliwicz (1986) found that in fact affect patterns of migration then it may in the lower Zambezi River, in South Africa, be more readily identified in the tropical the lowest densities of plankton per lunar cycle environment of Bonaire. occur during the last quarter moon due to The purpose of this research is to test the predation in the first light of the full and last following hypothesis concerning moon phase quarter moons by the sardine (Limnothrissa and the presence of microzooplankton and miodon). This behavior indicates the driving macrozooplankton in surface waters. force behind DVM is significantly greater than H1: Total zooplankton density and diversity that for predator avoidance. Zooplankton will be highest at night. migrate to the surface in greater abundance H2: Macrozooplankton organisms will show during a total solar eclipse than during a the greatest diel movement by density. standard night, demonstrating that the change H3: During the new moon the density of in light intensity is a primary migration cue macrozooplankton (organisms > 500 µm) (Bright 1972). Larger or egg-carrying such as larval organisms from the classes organisms have a stronger tendency to migrate malacostraca (megalope), and polychaetae and stay deeper during the day (Wright and will be greatest. Vinyard 1980). Although ascent is typically at H4: During the new moon the density of night, some varieties of plankton ascend microzooplankton (organisms < 500 µm) during the day; this movement is termed such as organisms from the class “reverse migration” (Lampert 1989). Despite maxillopoda (copepods), chaetognatha the additional energy costs to undergo vertical will be greatest. migrations, zooplankton cannot remain at H5: The highest biological diversity among the depth for extended periods due to decreased zooplankton classes is predicted to occur growth and fecundity when exposed to colder at night, with peaks after significant temperatures and reduced food resources spawning events, occurring around the (Lampert 1989; Dodson 1990; Loose and main phases of the moon: full, new, first Dawidowicz 1994). quarter, and last quarter. Zooplankton have been found to increase in abundance during the evenings following a Materials & Methods lunar spawning event due to the presence of newly hatched larvae and predation on the Zooplankton Collection newly released gametes (Lamare and Stewart Zooplankton collections were made at 1998). Yahael (2005) found a significant Something Special (12° 9' 43.69" N, 68° 17' difference between day and night zooplankton 7. 79" W), a dive site located on the leeward abundance and diversity over a reef coast of Bonaire, Netherlands Antilles (Fig. 2). environment. High abundances of copepods The sampling site consists of a sandy reef flat were present during the day along with low with a fringing reef that drops off 40 m from

28

Physis: Journal of Marine Science

Figure 4. Dates and moon phases of sample collection.

Sample Processing and Zooplankton Identification Plankton samples were fixed in a 10 % formalin solution within 2 h of collection for a maximum of 3 d before being transferred to a 28 % ethanol solution. To split the macrozooplankton (plankton > 500 µm) Figure 2. Google Earth image of sample site Something Special. Yellow lines represent path followed, white lines represent fraction from the microzooplankton direction, yellow circle represents site marker buoy and the red (organisms < 500 µm) fraction of the sample, star is the entrance point. the following procedure was used. The the coast. There is a deep channel between the sample was washed over a 500 µm mesh sieve reef slope and the island of Klein Bonaire to remove the macrozooplankton which were (Hall 1999). Current flow is predominantly then held in 60 % ethanol. The remaining from south to north with offshore winds fraction was then washed over a 30 µm sieve typically blowing from the east. At night there to remove microzooplankton. is some light pollution from the residences Microzooplankton were transferred via pipette bordering the shore at this site. to a 100 mL sample bottle containing 35 mL A 250 µm mesh plankton net (30.5 cm of a 28 % ethanol solution. diameter, 122.0 cm long) and a flowmeter All macrozooplankton were enumerated (Ocean Test Equipment Inc. model MF315) and identified to class using a Fisher dissecting were attached to opposite ends of a 2 m pole microscope at 0.9 X magnification. using nylon rope (Fig. 3). The apparatus was Microzooplankton were suspended by swirling pushed through the water while snorkeling. and 3 - 1 mL subsamples were transferred via The pole was held in front of the snorkeler and pipette onto a counting slide for enumeration the top of the net was held at a depth of and identified to class by using a dissecting approximately 16 cm. The snorkeler swam microscope at 1.3 X magnification. The over the sand flat heading north and above the number and taxonomic diversity at the class reef crest heading south, against the current. level of each organism in both micro- and The route took 10 min to complete (Fig. 2). macrozooplankton samples were determined. During September and October 2009, samples were collected during the day and at night Data Analysis (12:00 h and 21:00 h, respectively) during Daytime and nighttime zooplankton each of the 8 lunar phases (+ 1 d; Fig. 4). densities were compared using a paired t-test. Plankton densities were standardized per unit volume by multiplying the distance swam by the size of the net opening. Additional inferences about the possible influence of lunar phase on zooplankton density were made through graphical examinations of the abundance and diversity of zooplankton.

Results

Zooplankton

Figure 3. Zooplankton sampling apparatus consisting of a 2 m Plankton densities were significantly higher pole, flow meter and plankton tow net (mesh size 500 µm) used during the night then the day (130.3 ± 73.6 to collect samples by snorkeling. The snorkeler held the pole in individuals m-3, 43.8 ± 32.8 individuals m-3, front and swam a predetermined path during each moon phase.

29

Physis: Journal of Marine Science

150 30

Am -3 25 125 -3 Pm 20 100 15

75 10 # of organisms m organisms # of

5 50 # of individuals m # of individuals 0 25

0 Macro Micro

Figure 5. Mean (± SD) zooplankton abundance during the day Figure 7. Macrozooplankton density for nighttime samples (light bars) and night (dark bars) for macrozooplankton (lunar collected the the dive site Something Special over one lunar phases pooled n = 8) and microzooplankton (t-test p =0.010; n = cycle from 30 Sept. to 24 Oct. 2009 in Bonaire. 8).

(118.9 ± 83.4 individuals m-3, 43.3 ± 32.6 respectively; t-test, p = 0.003; Fig. 5). Overall -3 the mean zooplankton density was highest individuals m , respectively; p = 0.080). Microzooplankton density was significantly during the waning gibbous lunar phase (291.1 -3 individuals m-3). The full moon showed the greater at night (289.7 individuals m ± 53.2; p highest taxonomic diversity with eggs and = 0.010). organisms from the classes maxillopoda (copepods), and malacostraca (amphipods and Macrozooplankton euphausiids) appearing most frequently Macrozooplankton density was highest macrozooplankton (Fig. 6). Nighttime during the first quarter moon and lowest during the waxing gibbous (28.4 ± 155.4 densities of macrozooplankton were not -3 -3 significantly higher than daytime densities individuals m , 0.4 ± 1.6 individuals m , respectively; Fig. 7). Organisms from the

20 Micro PM Macro PM 18 Micro AM 16 Macro AM 14

12

# of classes 10

8

6

4

2

0 Waxing Full Moon Waning 3rd Waning New Moon Waxing First Gibbous Gibbous Quarter Crescent Crescent Quarter

Figure 6. Mean number of taxonomic classes observed in microzooplankton samples and number of macrozooplankton taxonomic classes present in samples collected during the night and day of each lunar stage from September- October 2009 at the dive site Something Special in Bonaire.

30

Physis: Journal of Marine Science

classes maxillopoda (copepods) and 100 Copepods malacostraca (euphausiids) had the highest 90 Amphipods densities. The highest macrozooplankton -3 80 Chaetognaths taxonomic diversity (14 classes) occurred 70 during the first quarter moon (Fig. 6) and 60 lowest diversity (7 classes) occurred during 50 40 the full moon. m individuals # of 30 Microzooplankton 20 Microzooplankton density was highest at night 10 during the waning gibbous and lowest during 0 the waxing crescent moon phase (289.3 ± 53.2 individuals m-3, 54 0 ± 3.6 individuals m-3, respectively; Fig 8). Microzooplankton showed the highest taxonomic diversity (19 classes) during the full moon with organisms Figure 9. Mean number of copepods, amphipods, and from the classes maxillopoda (copepods) and chaetognaths present in nighttime samples of microzooplankton malacostraca (amphipods) appearing most at each moon phase from 30 Sept. to 24 Oct. 2009 in Bonaire. frequently. Of the three most commonly 30 a) found organisms, copepods and amphipods -3 25 peaked at both the full moon and new moon 20 rd while chaetognaths peaked during the 3 15 Quarter (Fig. 9). 10 Individuals m Individuals 350 5 300 0 -3 250 200 150 100 # of organisms m organisms # of 50 300 b) Micro 0 250 -3 200 Eggs 150 100 Individuals m Individuals 50 0 Figure 8. Microzooplankton density (± SD) for nighttime samples collected at the dive site Something Special over one lunar cycle from 30 Sept. to 24 Oct. 2009 in Bonaire.

Eggs Eggs were present in the greatest densities during the night of the waning gibbous and first quarter moon phases (159.7 ± 98.6 Figure 10. a) Density of macrozooplankton for each nighttime -3 -3 sample obtained over a full lunar cycle. b) Density (± SD) for individuals m , 144.4 ± 278.8 individuals m nighttime samples of microzooplankton (excluding eggs) and respectively), and were least dense during the eggs collected at the dive site Something Special over one lunar 3rd quarter lunar phase (8.7 ± 3.1 individuals cycle from 30 Sept. to 24 2009 in Bonaire.. m-3). In the absence of eggs macrozooplankton (141.3 ± 58.4 individuals m-3, 49.6 ± 5.0 densities peaked during the waxing crescent individuals m-3; Fig. 10 B). and were lowest during the waxing gibbous

(16.9 ± 5.0 individuals m-3, 0.4 ± 4.6 Discussion individuals m-3; Fig. 10 A). Microzooplankton densities peaked during the full moon and Zooplankton were lowest during the waxing crescent When comparing zooplankton densities in

31

Physis: Journal of Marine Science the surface waters of Bonaire between diel an unusually high number of eggs that may periods, higher densities were found during the have been present due to the approaching lunar night, as hypothesized (H1). This supports the apogee, which has been correlated with theory that, similar to many other oceanic increased spawning activity of reef organisms. plankton populations, plankton inhabiting The lunar apogee occurs once every month, Bonairean waters exhibit the typical DVM when the moon is farthest away from the Earth pattern (up at night and down during the day). causing a lower than average tidal range The highest zooplankton density occurred (NOAA 2007). The apogee during this study during the waning gibbous, not on the new fell on the 25th of October; the evening after moon, as was hypothesized (H2). A high the first quarter moon sample was collected so abundance of eggs greatly contributed to the both moon phase and apogee may have peak density of zooplankton that was observed influenced macrozooplankton densities during during the waning gibbous moon phase. This this study. After separating eggs from the total finding was likely influenced by spawning evening macrozooplankton densities peaked events that occurred during the full moon, four during the waxing crescent, which falls days prior to the waning gibbous. immediately after the full moon, and were The number of taxonomic classes of total lowest during the waxing gibbous moon. This zooplankton was also higher at night, as post-new moon abundance could be due to a hypothesized (H5). The most taxonomically response by macrozooplankton to the first diverse evening was during the full moon, the light following the new moon and release from stage of the lunar cycle which produces the predation pressure by larval fish following the most light and which has been found in new moon. freshwater habitats to strongly influence some zooplankton feeding and reproductive Microzooplankton behaviors (Canepa 1996). The full moon may Microzooplankton density peaked during be a major influence on organisms such as the waning gibbous moon phase and not copepods that utilize DVM (Dodson 1990). during the new moon, as hypothesized (H4; The density of copepods was highest on the Fig. 8). This difference was most likely due to full moon during this study. When eggs were a spawning event because eggs made up a separated from the total evening zooplankton large portion (72 %) of the microzooplankton sample it was found that microzooplankton collected. Because this sample was obtained 4 densities peaked during the full moon lunar days after the full moon, it also falls within the phase and were at their lowest during the previously noted lunar spawning time period. waxing crescent, this contradicts my After separating eggs from the total evening hypothesis that the new moon would have the microzooplankton densities peaked during the greatest densities and indicates that full moon and new moon and were lowest microzooplankton may be responding to the during the waxing crescent (Fig. 10). This light, which enhances DVM. data confirms my hypothesis and suggests that true microzooplankton peaks were masked by Macrozooplankton a lunar spawning event. The highest overall macrozooplankton The most taxonomically diverse lunar density occurred during the first quarter moon phase was the full moon, which contained a (4 Oct. 2009), instead of the new moon as total of 19 microzooplankton classes. hypothesized (H3). Heidelberg et al. (2004) Copepods and amphipods were present during found that the diversity and abundance of all lunar phases, but interestingly, abundances demersal zooplankton that exhibit DVM, in peaked during the full and new moon lunar Discovery Bay, Jamaica was highest during phases, perhaps indicating that celestial light, the first quarter moon. The first quarter moon moon light for full moon and starlight for new phase occurs following the new moon and moon, is a major driving force behind diel proceeding the full moon, when organisms that vertical migration in Bonaire. On the other exhibit lunar spawning (such as corals) are hand, abundances of chaetognaths peaked most affected by environmental factors, such during the 3rd quarter moon phase, following as tides, and light level (Foster 1987; Naylor the abundance of eggs present during the 2001). The nighttime sample collected during waning gibbous. This may indicate that the first quarter moon of this study contained chaetognaths prey upon eggs in surface waters

32

Physis: Journal of Marine Science following spawning events, leading to References predation as a possible driving force for that class. The detection of copepod and amphipod Bright, T. 1972. Effects of a total solar eclipse abundance during specific lunar cycles on the vertical distribution of certain contradicts Ohlhorst’s (1982) finding of no oceanic zooplankters. Limnology and correlation between behavior or taxonomic Oceanography 17:296-301. diversity with specific lunar cycles. However, Calvert, J. B. 2004. Tidal Observations : Tides further research involving multiple lunar in the Gulf of Mexico and the Caribbean cycles may yield different findings. Sea.http://mysite.du.edu/~jcalvert/geol/tid In the future, research that examines the es.htm#Gulf. relationship between lunar phase and Canepa, J. L. 1996. Neuston Composition and spawning activity of Caribbean organisms the effects on fresh water sediment plumes would help to clarify its effect on zooplankton in paopao and Opunohu Bays, (Moorea, communities. Eggs comprised a large portion French Polynesia) Biology and of overall zooplankton density, therefore it is geomorphology of tropical islands, important to determine what organisms are Student Research Papers University of spawning and during which lunar cycles and California:38-40. compare to plankton diversity and density. Dodson, S. 1990. Predicting diel vertical This study shows that zooplankton are migration of zooplankton. American present in the surface waters of Bonaire in Society of Limnology and Oceanography higher densities at night than during the day. 35(5): 1195-1200. The largest variety of taxa were found during Foster, S. A. 1987. Diel and lunar patterns of the full moon (Fig. 6). The greatest reproduction in the Caribbean and Pacific zooplankton density was present during the sergeant major damselfishes Abudefduf first quarter moon with the highest egg saxatilis and A. troschelii. Marine Biology densities occurring during the waning gibbous 95:333-343. moon. This peak in zooplankton density Gliwicz, M. Z. 1986. A lunar cycle in during the first quarter also coincided with a zooplankton. Ecology 67:883-897. lunar apogee. This is shown in the sudden Hall, D. B. 1999. The geomorphic evolution of jump in egg density (shown in Fig. 10 b) that slopes and sediment chutes on forereefs results in a decrease in microzooplankton Geomorphology 27(3):257-278. (Fig. 7) and an increase in macrozooplankton Heidelberg, K. B., K. P. Sebens, and J. E. (Fig. 8) densities during the last two lunar Purcell. 2004. Composition and sources of phases. near reef zooplankton on a Jamaican After egg data was separated from the forereef along with implications for coral night macrozooplankton and feeding. Coral Reef 23: 263-276. microzooplankton samples, a peak in InterKnowledge Corp. 2005. Bonaire Dutch microzooplankton was observed during both Carribean. Last accessed November 13, new moon and full moon phases (Fig. 10 a), 2009 http://www.geographia.com/bonaire/ and macrozooplankton density peaked during Lamare, M. D. and B. G. Stewart. 1998. Mass the waxing crescent (Fig. 10 b), which directly spawning by the sea urchin Evechinus follows the new moon. This suggests that the chloroticus (Echinodermata: Echinoidea) lunar cycle may in fact have an effect on in a New Zealand fiord. Marine Biology density of zooplankton in the Caribbean 132(1):132-140. waters of Bonaire. Lampert, W. 1989. The adaptive significance of diel vertical migration of zooplankton. Acknowledgements Functional Ecology 3:21-27. Loosanoff, V. and C. Nomejko. 1951. Thanks to Dr. Rita Peachey for editing, spawning and settling of the American assisting with Excel files, and advising. The oyster, C. Virginica, in relation to lunar interns, Kate Jirik, Lauren Saulino, and Jill phases. Ecology 32:113-134. Lau for sitting on the sea wall and making sure Loose, C. and P. Dawidowicz. 1994. Trade- I didn’t get hit by a boat during my night offs in diel vertical migration by sampling. I also want to thank Prof. Cohen for zooplankton: The costs of predator assistance with plankton identification. avoidance. Ecology 75:2255-2263.

33

Physis: Journal of Marine Science

Moon Connection. 2009. Understanding The Skov, M. W., R. G. Hartnoll, R. K. Ruwa, J. P. Moon Phases. http://www.moon Shunula, M. Vannini, and S. Cannicc. connection.com/moon_phases.phtml 2005. Marching to a different drummer: Naylor, E. 2001. Marine animal behavior in crabs synchronize reproduction to a 14- relation to lunar phase. Earth Moon month lunar-tidal cycle. Ecology 86:1164- Planets 85:291-302. 1171. NOAA. 2007. Our restless tides: A brief Takemura A., S. Rahman, S. Nakamura, Y. Ju explanation of the basic astronomical Park and K. Takano. 2004. Lunar cycles factors which produce tides and tidal and reproductive activity in reef fishes currents. http://tidesandcurrents.noaa.gov with particular attention to rabbit fishes. /restles4.html Fish and Fisheries 5(4):317 – 328. Ohlhorst, S. L. 1982. Diel migration patterns Wright D., W. J. O’Brian and G. L. Vinyard. of demersal reef zooplankton. Journal of 1980. Adaptive value of vertical Experimental Marine Biology and Ecology migration: A simulation model argument 60:1-15. for the predation hypothesis. Evolution Robertson D. R., C. W. Petersen, and J. D. and Ecology of Zooplankton Communities Brawn. 1990. Lunar reproductive cycles of 29:138-137. benthic-brooding reef fishes: reflections of Yahel R., G. Yahel, T. Berman, J. S. Jaffe, and larval biology or adult biology? A. Genin. 2005. Diel Pattern with Abrupt Ecological Monographs by Ecological Crepuscular Changes of Zooplankton over a Society of America 60(3):311-329. Coral Reef. Limnology and Oceanography Seitz, R. C. Schaffner. 1995. Population 50(3):930-944. ecology and secondary production of the polychaete Loimia medusa (Terebellidae). Marine Biology 121(4):701-711.

34

Physis: Journal of Marine Science

How does water quality correlate with coral disease, bleaching, and macroalgal growth on coastal reefs? A comparative study of various anthropogenic threats on Bonaire, N.A.

Mollie Sinnott Wake Forest University

Abstract Coral reefs worldwide are currently jeopardized by anthropogenic factors such as land-based pollution, coastal development, and sediment erosion. In the Caribbean alone, nearly two-thirds of coral reefs have been deemed as threatened. This study investigated the potential negative effects of water quality and eutrophication, Enterococci bacteria (found in human gut), and sedimentation on coral disease, bleaching, and macroalgal growth on the near shore reefs of Bonaire, N.A. Monitoring sites were defined according to their proximity to anthropogenic activity: “more impacted” or “less impacted” (< 200 m and > 200 m from coastal development, respectively). Water samples at 5 m were collected weekly and at 12 m biweekly from each site and tested for nutrient concentrations - (NO3, NO2 , NH4-N, PO4), Most Probable Number of Enterococci bacteria, sedimentation rates, and particle size distributions. Video transects (100 m) were also taken at defined depths and analyzed for live coral cover and diversity, percent disease and bleaching, and macroalgal cover. Data showed elevated NH4-N levels at all sites, Enterococci bacteria present at 3 of the 4 sites (mainly at 5 m), and sediment particle counts showed significant differences among sizes at both depths and between the interaction of size and impact at 12 m. There was also a strong trend of finer grained sediments at high impact sites and coarser grained sediments at low impact sites. Very little overall coral disease (1.105 ± 1.563 % at more impacted sites and 0.400 ± 0.566 % at less impacted sites in 12 m) and bleaching (3.245 ± 0.615 % at more impacted sites and 1.390 ± 1.966 % at less impacted sites in 12 m) was found on the reefs however, neither were present at 5 m. There was significantly more macroalgae at 12 m and a strong trend of more macroalgae at the deeper, more impacted sites. This study suggests that increased anthropogenic activity on Bonaire is contributing to the increased NH4- N levels, Enterococci bacteria presence, and finer particle sediments, which future studies may correlate significant interactions between these parameters and coral disease, bleaching, and macroalgal growth.

Introduction algal growth, which inversely affect light Coral reefs worldwide have faced availability and photosynthetic activity of increasing anthropogenic stressors over the zooxanthellae in corals, thereby slowing coral past century, including habitat destruction, growth rates (Larsen and Webb 2009). eutrophication, erosion and sediment runoff, Pollution in the form of near shore sewage and global climate change (Burke and Maidens runoff could also be a cause of coral decline. 2004). These threats can cause coral reef In Puerto Rico, Bonkosky et al. (2009) found degradation through increased disease that pooled bacteria counts at sites > 0.5 km susceptibility and mortality and through from shore showed significantly lower decreased coral recruitment and resilience to microbial concentrations than at sites closer to temporary disturbances (Smith et al. 2008). shore. This fecal pollution caused increased Water quality is an important factor in coral stress within the benthic community and determining survivorship because coral growth decreased water transparency and sunlight and reproduction are limited to a narrow range penetration (Bonkosky et al. 2009). Similarly of temperature, salinity, and nutrient levels in the Florida Keys, U.S.A., where sewage has (Burke and Maidens 2004). Changes in historically been disposed of in septic tanks, nutrient concentrations and the overall water the thin mucosal layer of corals in coastal chemistry around coral reefs are commonly waters tended to accumulate enteric bacteria (9 influenced by runoff and discharge from rivers of 15 coral heads sampled were positive for and areas with increased anthropogenic one or more fecal indicators) with 97% of the activity (Larsen and Webb 2009). Increased coral disease in the area likely caused by nutrient concentrations can result in increased human activity (Lipp et al. 2002). [email protected] 35

Physis: Journal of Marine Science

In addition to sewage runoff, sediment A rising problem in Bonaire is that the deposition on reefs has become a problem due island has no centralized waste water treatment to excess terrestrial runoff from land-clearing, system, only septic tanks (both lined and removal of coastal habitats, construction, and unlined). When lined septic tanks are drained, limestone mining (Burke and Maidens 2004). human waste is deposited into unlined trenches Normally, corals can remove small amounts of at the center of the island (Reynolds 2008). sediment but cannot survive high rates of This waste then leaks through the thin layer of sedimentation (Nugues and Roberts 2002). soil and porous limestone and is believed to However, it is not only the amount of make its way into near shore waters, similar to sediment, but the sediment grain size that can the previously mentioned situation in the harm corals. It has been found that coral tissue Florida Keys. More commercial construction damage under a layer of sediment increases for the expanding tourism industry is also with small grain size compared to large grain noticeable on the island (M. Sinnott, personal size (Nugues and Roberts 2002; Fabricius observation), leading to a potential increase of 2005). Sedimentation decreases a coral’s sediment runoff. Previous work in Bonaire has overall fitness because it requires more energy already shown evidence of Enterococci expenditure than normal to remove sediments, bacteria and finer grained sediments at 12 m making it more susceptible to disease (Nugues depth near areas of higher coastal and Roberts 2002). development, but few significant conclusions Coral cover in the Caribbean has were made (Rini 2008). drastically decreased over the past three The objective of this study was to use decades and many reefs have begun water chemistry monitoring, sediment undergoing a phase shift from coral-dominated collection, and video surveys to determine to macroalgal-dominated systems (Kline 2006; whether coral disease, bleaching, and Steneck et al. 2007). This shift can reduce a macroalgal growth correlated with human reef’s resilience to both abiotic and biotic activities that could increase nutrient levels, threats, making it more susceptible to disease fecal bacteria, and sedimentation in the coastal and bleaching. The majority of reported coral waters of Bonaire. The questions addressed diseases have come from the Caribbean with include: 1) how does water chemistry, fecal 23 known types, however, the sudden spread bacteria counts, sedimentation rates, and of disease and impact of added stresses are still particle size distributions compare between relatively unknown (Burke and Maidens sites with greater and lesser human activity, 2) 2004). Bleaching is a coral’s reaction to stress, how do these same parameters compare such as microbial infections, increased between deeper and shallower depths, and 3) temperatures, or changing water quality. does the percent cover and diversity of live Corals can typically recover from mild corals, frequency of coral disease and bleaching events, but in the Caribbean, bleaching, and abundance of macroalgae differ reoccurring bleaching over the past few between depth, impact, and their interaction? decades has caused widespread damage to Reef environments with elevated nutrient coral structures and overall reef condition concentrations, noticeable Enterococci (Burke and Maidens 2004). This decline in bacteria, and finer sediment particles as a coral cover gives way to possible macroalgae result of coastal development are predicted to dominance. A mean cover of 23 % have a higher percentage of coral disease, macroalgae was recorded for the Caribbean in bleaching, and macroalgal growth than reefs 2003, which suggests increased macroalgae is with less human impact. an indicator of reef degradation; increasing anthropogenic impacts could be one possible Materials & Methods reason (Steneck and Olson 2007). In Bonaire, N.A., Steneck and Olson (2007) found an 8 % Study Sites mean cover of macroalgae, which although Four sites along the western coast of low for the Caribbean, was a significant Bonaire were determined as either “more increase from previous years (4 % in 2003 and impacted” or “less impacted” sites. More 2 % in 2005) — suggesting the beginning of a impacted sites (Bari’s Reef and 18th Palm) possible phase shift on the island’s were characterized as being < 200 m from a surrounding reefs. commercial establishment (i.e. a resort or

36

Physis: Journal of Marine Science construction site) and “less impacted” sites returned to the laboratory and analyzed for - (Witches Hut and White Slave) were > 200 m nitrate (NO3), nitrite (NO2 ), ammonia (NH4- from development, typically near rural areas N), and phosphate (PO4) within 2 h of - (Rini 2008; Fig. 1). Sites were chosen in collection. NO3, NO2 , and NH4-N were tested accordance with CIEE’s ongoing water quality following the methodology of the LaMotte monitoring project at 12 m depths. Water SaltWater AquaCulture Test Kit (Model AQ - chemistry analysis, fecal contamination, and 4) and PO4 was assessed using a Red Sea sedimentation were assessed using SCUBA at Marine and Freshwater Test Lab. 5 and 12 m depths for each site. The 5 m depth was chosen because it was the average Enterococci Bacterial Analysis depth just above the reef crest for all sites and Enterococci bacteria counts were the 12 m depth was chosen for its certainty of determined using the site-specific water high coral cover and diversity and easy acquired at all locations and depths, at 5 m accessibility. weekly and 12 m biweekly. Following the Enterolert® detection methodology, 10 mL of site-specific water was acquired prior to chemical analyses to avoid cross- contamination and diluted with 90 mL of distilled water in a sterile container. One unit of Enterolert® fluorescing substrate was mixed with this solution and poured into an IDEXX Quanti-tray®. This tray was vacuum sealed and incubated at 41 ± 0.5 °C for 24 h. Following incubation, bacterial fluorescence was visually assessed under a black light and the number and size of the tray wells that fluoresced were recorded and converted into a Most Probable Number (MPN) using Enterolert’s MPN generator table. MPN is defined as the number of bacteria populations per 100 mL of sample water.

Sediment Analysis Sediment collection and analysis, including particle size count, were modified from Gleason’s (1998) methodology. Sediments were collected at each site and Figure 1. Map of Bonaire, N.A. Closed circles represent each study site. Witches Hut and White Slave are less impacted, depth using PVC traps (7.5 cm diameter, 15 Bari’s Reef and 18th Palm are more impacted. cm long). These traps were capped at the base and open at the top, thus allowing for Water Chemistry sediments to settle inside. The PVC traps were In order to determine the composition of attached ~ 10 m above the substrate in an the water at the chosen sites and depths, water upright position to a rebar stake (1 m). On a samples were collected weekly at 5 m and weekly basis for 5 m and biweekly for 12 m, biweekly for 12 m (in accordance with CIEE the traps were capped, collected, and replaced monitoring) for several weeks (Oct. – Nov. using SCUBA. These traps were returned to 2009). Sample bottles (250 mL Nalgene) were the laboratory where sediments were left to acid washed with a 10 % HCl solution and settle for ~ 1 h before the overlying saltwater filled with distilled water before entry into the was decanted off. The sediments were gently water. At depth, site-specific water was rinsed with tap water three times (water obtained by inverting the bottle, filling it with decanted between rinses) to dissolve and air from a secondary SCUBA air source, remove any salts, poured into a pre-weighed turning it upright again to be filled with water, aluminum foil container, and placed in a and repeating this process for a total of three drying oven (~ 40 °C) for 48 h. Once dry, times before capping. All water samples were visible organic material was removed with

37

Physis: Journal of Marine Science forceps and the dry sample was re-weighed for Statistical Analyses the calculation of sedimentation rate (grams A series of 2-factor ANOVAs were used to day-1) before being transferred into a small assess any differences among depth, impact, vial. To determine particle size composition, and their interaction for water chemistry, the dried sample was re-suspended in 25 mL of sedimentation rates, and CPC results. Two- tap water. Following suspension, a 1 mL sub- factor ANOVAs were used to assess particle sample of this solution was removed using a size to determine size, impact, and interaction 10 mL syringe and placed onto a counting cell significance. Two paired t-tests were used to (with visible gridline pattern) and viewed show possible significance of phosphate and under a light microscope (40 X). One square ammonia concentrations between impact sites of the counting cell grid was randomly chosen at each depth. and the first 200 particles encountered (from left to right) were measured using a rotating Results ruler (µm). Particles were assigned to one of the following size categories: < 10, 11 - 50, 51 Water Chemistry - 100, 101 - 250, 251 - 500, and > 500µm. Water analyses showed minimal NO3 and - - This was repeated for two more 1 mL sub- NO2 at all sites and depths. NO2 was never samples from each original sediment sample. detected (Fig. 2a and 2b) and NO3 was only The mean size class distribution for each depth present on one day at 5 m of a more impacted and site was then determined. site (Fig. 2a). Mean PO4 levels were also relatively consistent throughout the study; Video Transect Analysis levels never exceeded 0.1 ppm for any site Video transects were utilized in order to type or depth (Fig. 2a and 2b). There was no assess the status of coral disease and difference in PO4 between site types at either 5 bleaching, as well as macroalgal growth at m or 12 m (p = 0.423 and p = 0.423, each study site. Permanent transects were 1.4 marked at 5 and 12 m depths of all sites, with a) More the sediment traps acting as center markers. 1.2 Following the specified depth contour, Less 1 transects were run North and South of the traps and marked with rebar stakes (1 m) at 30 and 0.8 50 m. Video documentation of these transects 0.6 was performed once during the study (early November 2009) using SCUBA and a 0.4 Concentration (ppm) Concentration waterproof camcorder housing with metal 0.2 wand (2 m). Videos were shot perpendicular to the reef at a slow swimming speed with the 0 tip of the wand staying just above the contour Nitrite Nitrate Ammonia Phosphate of the substrate. Once acquired, videos were trimmed and cut into 50 frames using Picture 1.4 Motion Browser then analyzed using Coral b) More 1.2 Point Count (CPC) software. In order to allow Less for the analysis of one frame per 2 m of 1 transect, the number of total frames was 0.8 chosen based on limited time constraints. For each frame, 15 points were chosen randomly 0.6 under which the substrate was identified as 0.4 species of coral, type of disease, presence of (ppm) Concentration 0.2 bleaching, or presence of macroalgae (methods modified from Rini 2008). CPC evaluated 0 each transect in terms of the frequency of Nitrite Nitrate Ammonia Phosphate identifiable substrate as a percentage of the - Figure 2. a) Mean NO3, NO2 , NH4-N, and PO4 concentrations (± overall points and the Shannon-Weaver Index SD ppm) at 5 m between more and less impacted sites. b) Mean - was used to quantify coral diversity. NO3, NO2 , NH4-N, and PO4 concentrations (± SD ppm) at 12 m between more and less impacted sites.

38

Physis: Journal of Marine Science

0.6 respectively). At 5 m, NH4-N levels were More higher at the more impacted sites than at low 0.5 ) impact sites (Fig. 2a). Whereas at 12 m, levels -1 Less were lower overall but a higher concentration 0.4 of NH4-N was observed at less impacted sites 0.3 (Fig. 2b). There was no difference between site type at either 5 or 12 m (p = 0.551 and p = 0.2 0.508, respectively). 0.1

Sedimentation rate rate (g day Sedimentation 0 Bacterial Analysis 5 m 12 m Enterococci bacteria were present at 3 of the 4 sites at both depths. It appears that there Figure 4. Mean sedimentation rate (± SD g day-1) for 5 and 12 m are greater mean MPN concentrations of and between more and less impacted sites. (ANOVA, depth p = bacteria at shallower depths in more impacted 0.316, impact p = 0.540, depth x impact p = 0.542). sites (6.588 ± 0.088; Fig. 3). However, there is size and impact (p = 0.985 and p = 0.185, no statistical difference in MPN for depth, respectively). At 12 m, there was a significant impact, or the interaction of the two (p = difference for both particle size and the 0.173, p = 0.234, and p = 0.942, respectively). interaction between particle size and site impact (p < 0.001 and p = 0.005, respectively), 10 More but not for site impact alone (p = 0.986). 9 Less There were a greater overall number of 8 particles at more impacted sites in all size 7 categories except for the > 10-50 µm size class 6 (Fig. 5b). 5 160 a) More 4 140 Less 3 120 MPN concentrations 2 100 1 80 0 60 5 m 12 m 40

Figure 3. Mean MPN concentrations (± SD) for Entercocci of particles number Mean 20 bacteria at 5 and 12 m between more and less impacted sites. (ANOVA, depth p = 0.173, impact p = 0.234, depth x impact p = 0 0.942).

Sedimentation Particle size (µm) Mean sedimentation rates appeared 160 More slightly higher at 5 m of less impacted sites 140 b) -1 Less (0.249 ± 0.269 g day ) versus more impacted 120 -1 sites (0.121 ± 0.013 g day ) and 12 m more 100 and less impacted sites (0.074 ± 0.010 g day-1 80 and 0.075 ± 0.043 g day-1, respectively; Fig. 60 4). However, there was no statistical difference among depth, impact, and their 40 interaction (p = 0.316, p = 0.540, and p = of particles number Mean 20 0.542, respectively). 0 Assessment of particle size distributions showed that higher impact sites at 5 m have more fine particles (< 10 and > 10-50 µm), Particle size (µm) whereas lower impact sites have more coarse Figure 5. a) Mean particle sizes (± SD µm) represented at 5 m particles (> 50-100, > 100-250, and > 250-500 between more and less impacted sites (ANOVA, particle size p < µm; Fig. 5a). There was a significant 0.001, impact p = 0.985, particle size x impact p = 0.185). b) Mean particle sizes (± SD µm) represented at 12 m between difference among particle sizes (p < 0.001) but more and less impacted sites (ANOVA, particle size p < 0.001, not between site impact or the interaction of impact p = 0.986, particle size x impact p = 0.005).

39

Physis: Journal of Marine Science

Video Analyses 0.4 More Percent live coral cover and coral diversity 0.35 Less both displayed significant difference between 0.3 depths (p = 0.003 and p = 0.010, respectively), but between neither site impact nor the 0.25 interaction of the two factors. Live coral cover 0.2 at 5 m was minimal for both more and less 0.15 impacted sites (2.065 ± 0.007 % and 3.65 ± Coral diversity 0.1

5.162 %, respectively), but was much higher at Index) (Shannon-Weaver the 12 m depth with more and less impacted 0.05 sites showing similar mean percentages 0 5 m 12 m (22.075 ± 4.137 % and 16.295 ± 2.001 %, respectively; Fig. 6). Coral diversity showed Figure 7. Mean coral diversity (± SD Shannon-Weaver Index) within the 100 m video transects at 5 and 12 m and between to be consistently low at 5 m between more more and less impact sites. (ANOVA, depth p = 0.010, impact p and less impacted sites (0.080 ± 0.000 and = 0.810, depth x impact p = 0.602). 0.095 ± 0.134, respectively), and significantly higher within the 12 m depth between more 4.5 More and less impacted sites (0.335 ± 0.021 and 4 a) Less 0.295 ± 0.021, respectively; Fig. 7). 3.5 For percent of coral with disease and 3 bleaching, there was no difference among 2.5 depth, impact, or their interaction with the exception of bleaching with significantly 2 greater occurrence at 12 m than 5 m (p = 1.5

0.034). Occurrence of disease and bleaching % of livecoral diseased 1 were both relatively low, with neither found at 0.5 5 m; 1.110 ± 1.563 % diseased at 12 m at more 0 impacted sites and 0.400 ± 0.566 % at less 5 m 12 m impacted sites (Fig. 8a), 3.245 ± 0.615 % 4.5 bleaching at 12 m at more impacted sites and b) More 4 1.390 ± 1.966 % at less impacted sites (Fig Less 8b). Despite these low occurrences of disease 3.5 and bleaching, there was a significant 3 difference in percent macroalgal cover 2.5 between depths (p = 0.010) with ~ 1.1 % at 5 2 m and ~ 3.3 - 7.0 % at 12 m (Fig. 9). The data also suggest strongly higher macroalgal cover 1.5 at deeper, more impacted sites (p = 0.100). % of livecoral bleached 1 0.5 30 More 0 5 m 12 m 25 Less Figure 8. a) Mean percent of live coral with disease (± SD) in the 100 m video transects at 5 and 12 m and between more and 20 less impacted sites (ANOVA, depth p = 0.270, impact p = 0.581, depth x impact p = 0.581). b) Mean percent of live coral affected 15 by bleaching (± SD) within the 100 m video transects at 5 and 12 m and between more and less impacted sites (ANOVA, depth p = 0.034, impact p = 0.272, depth x impact p = 0.272). 10

% live coral Discussion 5 Minimal levels of NO , NO -, and PO at 0 3 2 4 all depths and sites suggest that these 5 m 12 m compounds may have had little impact on the Figure 6. Mean percent live coral cover (± SD) within the 100 m status of corals on the reefs of Bonaire for the video transects at 5 and 12 m and between more and less duration of this study. These results did not impacted sites. (ANOVA, depth p = 0.003, impact p = 0.439, depth x impact p = 0.206). support the prediction that increased nutrients

40

Physis: Journal of Marine Science

9 More attractiveness (Bonkosky 2009). Human waste 8 Less may also be associated with coral morality. 7 Kline et al. (2006) found that excess dissolved organic carbons, such as glucose, galactose, 6 lactose, and starch, deposited on reefs 5 significantly increased coral death. Results 4 showed that when compared to the control 3 group, coral treated with increased levels of % macroalgae % 2 dissolved organic carbons had 5 X higher 1 mortality rates. These same organic carbons 0 were found to be present in human fecal 5 m 12 m samples used in a separate study to isolate Figure 9. Mean percent macroalgal cover (± SD) within the 100 metabolites in human intestinal bacteria (Wang m video transects at 5 and 12 m and between more and less impacted sites. (ANOVA, depth p = 0.010, impact p = 0.099, et al. 2000). Dissolved organic carbons may depth x impact p = 0.100). disrupt the chemical balance within corals and increase microbial growth rates in the outer would be present; however, NH4-N levels were mucous layer, causing death from oxygen relatively high (Fig. 2a and 2b) even though depletion or accumulating poisons (Kline et al. they showed no significant difference between 2006). Testing human fecal waste specifically depths of impact sites. NH4-N is a common on corals could provide more information chemical found in sea water because more than about how increased amounts of these organics half of the nitrogen excreted as waste by affect reefs. marine organisms is in the form of NH4-N and Sedimentation analysis revealed that more because it is a by-product of the breakdown of and less impacted sites have similar rates of organic matter and decomposing organisms sedimentation, which was not as predicted, but (Degobbis 1973). Levels varied throughout this variation could be due to site-specific the study but reached a peak of 2.0 ppm at 5 m wave action, currents, or storm disturbance on several days. This raises concern because a dispersing particles along the reef. Particle safe level of ammonia in marine systems size distributions however, showed that more ranges from 0.2 - 0.4 ppm (Alken Murray impacted sites have more fine sediments and Corp. 2006). Levels only spiked occasionally less impacted sites have more coarse grains within the study duration, but prolonged particularly at 5 m (Fig. 5a). This could be due exposure can block oxygen transfer in fish gills to the increased amount of construction in and cause permanent gill damage (Alken more impacted areas versus in less impacted Murray Corp. 2006). Many factors can affect areas. Development sites require land nutrient concentrations in marine systems and excavation for building foundations, concrete a possible contributing factor to elevated NH4- mixing, and removing foliage; this breaking up N concentrations in Bonaire could be from the of soil and fine concrete mix creates large breakdown of human waste water from septic amounts of dust for rain to wash into the tanks and unlined trenches (as previously nearby ocean, which could explain the finer described). particles found at more impacted sites (BNMP Although Enterococci bacteria levels were 2006; Rini 2008). The reefs could be at even not significant as originally predicted, the fact more of a disadvantage because Bonaire does that bacteria were present on the reefs not implement sediment retention practices at surrounding Bonaire is cause for alarm. construction sites (BNMP 2006). Grain sizes Enterococci appeared more often at sites with at 5 m depths were more consistent than at 12 more anthropogenic impact (Fig. 3), which m, but this variability could be attributed to the could be a result of increased faulty sewage low number of study sites or to larger particles disposal or septic tank leakage into the water sinking and settling at the bottom before column as previously suggested. Enterococci drifting to deeper depths. on Bonaire’s reefs affect humans because it Total coral cover and diversity were could lead to potential public health risks from significantly higher at 12 m depths but no ingesting fecal toxins, decreased fish difference was found between impact sites. populations due to habitat loss, and decreased Low diversity for all sites at 5 m correlates tourism because of poor visibility and reef with the minimal amount of live coral cover

41

Physis: Journal of Marine Science found there. Unfortunately, coral disease and Acknowledgements bleaching were found to be present on Bonaire’s reefs, but the percent cover was I would like to thank Pamela Williams for relatively low. It should also be noted that, accompanying me on countless dives, even in although other diseases are likely present on strong currents in the dark, Lauren Saulino, Jill the reefs, the only disease found with CPC was Lau, and Kate Jirik for driving me around and Dark Spot Disease. This disease primarily for their help with data collection in the lab, affects scleractinian corals, but the pathogen Albert Bianculli for helping me find the that causes this disease remains unknown infamous sediment traps, and most of all I (NOAA 2006). This relates to Kaczmarsky et would like to thank Dr. Amanda Hollebone for al.’s (2005) study linking increased human her guidance throughout this entire process. sewage to both Black Band Disease and White Plague Type II; significantly more diseased References stony corals were found along the coastal reefs of St. Croix in areas with more constant Alken Murray Corp. 2006. Interpreting Water exposure to sewage runoff. Future studies Analysis Test Results. http://www.alken- examining the microbial composition within murray.com/TESTS01.htm Dark Spot Disease and the distribution around Bonkosky, M., E. A. Hernandez-Delgado, B. areas with sewage runoff could create another Sandoz, I. E. Robledo, J. Norat-Ramirez, and H. Mattei. 2009. Detection of spatial link between Enterococci and coral disease. fluctuations of non-point source fecal pollution Despite the low occurrence of bleaching in coral reef surrounding waters in and disease, macroalgal growth appeared to southwestern Puerto Rico using PCR-based differ between depths and was found at greater assays. Marine Pollution Bulletin 58:45-54. abundance in deeper, high impact sites (Fig. Bonaire National Marine Park. 2006. BNMP 9). Since this strong trend was noticed within Management Plan. a study of low site replication, further Burke, L. and J. Maidens. 2004. Reefs at Risk. monitoring could lend more support for this World Resources Institution: Washington, D.C. result and may also reveal a stronger link Degobbis, D. 1973. On the storage of seawater between increased macroalgal cover and samples for ammonia determination. Limnology and Oceanography 18:146-150. higher NH4-N concentrations. Since ammonia Fabricius, K. E. 2005. Effects of terrestrial runoff is a nutrient for algae and plant growth (Alken on the ecology of corals and coral reefs: Murray Corp. 2006), there is potential for the Review and synthesis. Marine Pollution high levels of both to be related. Excess Bulletin 50:125-146. nutrients in coastal areas can potentially cause Gleason, D. F. 1998. Sedimentation and harmful algal blooms, changes in community distributions of green and brown morphs of the structure, and decreased biodiversity, all of Caribbean coral Porites asteroides Lamarck. which inhibit larval colonization and decrease Journal of Experimental Marine Biology and overall coral cover (Burke and Maidens 2004). Ecology 230:73-89. This study suggests that high Kaczmarsky, L. T., M. Draud, and E. H. Williams. 2005. Is there a relationship between proximity concentrations of NH4-N, the presence of to sewage effluent and the prevalence of coral Enterococci bacteria, and finer sediment disease? Caribbean Journal of Science 41:124- particles are all possible contributors to 137. increased coral disease, bleaching, and Kline, D. I., N. M. Kuntz, M. Breitbart, N. macroalgal growth. Improper sewage disposal, Knowlton, and F. Rohwer. 2006. Role of increased runoff from development elevated organic carbon levels and microbial construction, and expanding tourism activity in coral mortality. Marine Ecology negatively impact near shore reefs (Burke and Progress Series 314:119-125. Maidens 2004). Coastal reefs in Bonaire still Larsen, M. C. and R. M. T. Webb. 2009. Potential have more percent live coral cover and effects of runoff, fluvial sediment, and nutrient discharges on the coral reefs of Puerto Rico. diversity than the average amount in the Journal of Coastal Research 25:189-208. Caribbean, but the potential threat of a phase Lipp, E. K., J. L. Jarrell, D. W. Griffin, J. Lukasik, shift to a macroalgae–dominated system J. Jacukiewicz, and J. B. Rose. 2002. increases the urgency of controlling harmful Preliminary evidence for human fecal anthropogenic practices (Steneck and Olson contamination in corals of the Florida Keys, 2007). USA. Marine Pollution Bulletin 44:666-670.

42

Physis: Journal of Marine Science

NOAA. 2006. Coral disease identification and Smith, T. B., R. S. Nemeth, J. Blondeau, J. M. information: Dark spots. http://www.coral. Calnan, E. Kadison, and S. Herzlieb. 2008. noaa.gov/coral_disease/dark_spots.shtml Assessing coral reef health across onshore to Nugues, M. M. and C. M. Roberts. 2002. Partial offshore stress gradients in the US Virgin mortality in massive reef corals as an indicator Islands. Marine Pollution Bulletin 56:1983- of sediment stress on coral reefs. Marine 1991. Pollution Bulletin 46:314-323. Steneck, R. S. and D. E. Olson. 2007. Trends in Reynolds, T. 2008. Letters to the editor. The macroalgae abundance in Bonaire, 2003-2007. Bonaire Reporter 15:7-9 Sea Monitor A Report on the Status of the Coral Reefs of Foundation, comp. 2008. LMSP Bonaire Bonaire in 2007. Chapter 3. Phase #1 and #2. Sea Monitor Foundation. Wang, L., M. R. Meselhy, Y. Li, G. Qin, and M. Rini, A. 2008. Is #2 the number one problem in Hattori. 2000. Human intestinal bacteria Bonaire? An examination of fecal capable of transforming secoisolariciresinol contamination and sedimentation from runoff. diglucoside to mammalian lignans, enterodiol Physis 4:25-29. and enterolactone. Chemical Pharmaceutical Bulletin 48:1606-1610.

43

Physis: Journal of Marine Science

A comparative study of benthic and coral reef fish communities on artificial versus natural reefs of Bonaire, N.A.

Maggie Thomas Trinity College

Abstract Caribbean coral reef ecosystems are threatened by anthropogenic impacts such as pollution, overfishing, and habitat destruction. In an effort to alleviate these pressures and restore habitat, artificial reefs such as marina breakers, Reef Balls, and mooring blocks have been deployed and consequently colonized by marine species. Many studies have investigated the benthic and fish communities developing on these artificial structures as compared to adjacent natural structures. Results have shown that artificial reefs can successfully be colonized by benthic and pelagic communities but are not always comparable to the associated communities. The purpose of this study was to compare the composition of benthic habitat and the use of this habitat by fish between man- made mooring blocks functioning as artificial reefs and natural coral reefs of Bonaire. Quadrats were used to estimate and compare percent cover of benthic organisms on the top and west faces of mooring blocks versus the top and west faces of physically paired natural reef sites (n = 8). An 8 min visual census was conducted on each face of each site pairing to estimate fish abundance and diversity for those species interacting with the habitat. Results showed greater percent live benthic cover on the natural versus artificial reef. Benthic diversity was highest on the west face of the artificial reef when comparing the interaction of face and reef type, but did not differ significantly between reef types. Fish community diversity also did not differ between reef types. However, the composition of both benthic and reef fish community diversity differed greatly between the natural and artificial reefs. It was found that Montastrea annularis and sponges dominate the natural while the brain corals (Diploria labrinthiformis and Diploria strigosa) dominated the artificial reef. Bicolored damselfish (Stegastes partitus) and brown chromis (Chromis multilineata) were found in the highest densities on the natural reef, while sergeant major (Abudefduf saxatilis) and bluehead wrasse (Thalassoma bifasciatum) were found in the highest densities on the artificial reef. This study provides evidence that placement of artificial reefs does not cause a shift in overall benthic and reef fish community diversity on the natural reef, but may change the composition of this diversity.

Introduction from nearby reefs (Perkol-Finkel and Benayahu 2004; Langhamer et al. 2009), thus Caribbean coral reefs are hotspots of becoming artificial reefs beyond their intended biodiversity that are increasingly threatened by use. the presence of humans and their interactions The introduction of man-made habitats with the sea (Burke and Maidens 2004). into marine systems has led to altered Overfishing, pollution, increased abundances and distributions of marine sedimentation, terrestrial runoff, and organisms that inhabit those structures versus recreational activities are negatively impacting that found in natural habitats (Perkol-Finkel Caribbean coral reef ecosystems (Burke and and Benayahu 2004; Clynick 2008; Burt et al. Maidens 2004; Perkol-Finkel and Benayahu 2009b). For example, two unintentional 2004). Thus, deployments of artificial artificial reefs located in the Red Sea were structures have been proposed to help alleviate dominated by soft corals (approximately 90 % anthropogenic stressors (e.g. Reef Balls to cover) after 14 and 34 years, whereas the restore habitat, mooring blocks to avoid anchor adjacent natural reef was characterized by damage; Perkol-Finkel and Benayahu 2004; stony corals (approximately 40 % cover), due Burt et al. 2009b). Many times breakers, to the vertical orientation present in the mooring blocks, shipwrecks, wave power artificial reef, (absent in the natural reef) and foundations, and dock pillars are colonized by the life history traits of the soft corals (Perkol- current-driven planktonic larvae encountering Finkel and Benayahu 2004). Clynick (2008) and settling on the hard substrate or by the found that fish migrating between near-shore migration of juvenile and adult individuals reefs and marina-based artificial reefs in [email protected] 44

Physis: Journal of Marine Science

Sydney, Australia were characterized as large, benthic and associated pelagic community. It while the resident fish on the artificial reef is further hypothesized that as a result of were characterized as small. These results competition for space to reproduce, common were best explained by the absence of large reef fish such as the sergeant major (Abudefduf predators within the marina. saxatilis) will be more abundant on the It has also been suggested that substrate artificial reefs (mooring blocks) versus the composition, orientation, size, and/or associated natural reef. positioning of the artificial reef play a critical role in the successful colonization of benthic Materials & Methods organisms (Glasby and Connell 2001; Burt et al. 2009a; Langhamer et al. 2009). For Study Site example, Burt et al. (2009a) tested the Comparative observations of mooring difference in coral recruitment on four block (artificial reef = AR) and natural reef common materials (gabbro, concrete, granite, (NR) communities took place at eight sandstone) used in construction of artificial moorings located between the Playa Lechi and reefs, as well as terra cotta tiles that are often Something Special dive sites on the leeward used in scientific experiments to assess benthic side of Bonaire (block 1: 12° 09’ 30.87” N 68° recruitment. Coral recruitment appeared 16’ 52.77” W, block 8: 12° 09’ 36.55” N 68° highest on the gabbro tiles, but site-specific 16’ 58.16” W). The moorings (n = 8) are differences such as larval supply, water comprised of three 1 m3 concrete blocks within temperature, and light availability (driven by ½ m of each other, 27 - 35 m from the adjacent reef positioning), played a more important role mooring line, and all blocks approximately 10 in recruitment success than tile composition. m due east of the reef crest. This study area Glasby and Connell (2001) tested epibiotic was chosen because of the homogeneity of the growth on vertical pontoons, acting as artificial eight moorings, providing uniform artificial reefs, located within the harbor and associated reef structure for comparison with the NR. It rocky reefs of Sydney, Australia. They found is recognized that the AR and NR are at that orientation had little effect on the different depths, but any effects of depth were recruitment, but rather it was reef type (vertical minimized by staying within a depth range of pontoon versus natural reef) that was most 3 - 11 m (3 m at the top of the blocks and 8 m influential in the growth and success of at the top of the NR). Of the three blocks epibionts. associated with one mooring, the block with On the island of Bonaire, concrete the most westerly lying face was used in this mooring blocks have been deployed at dive study to compare with a physically paired area site locations along the west coast to of NR lying directly perpendicular to each discourage the use of anchors (BNMP 2009). mooring (270°). Both the top and west faces These blocks have become colonized by of the mooring blocks were used for benthic organisms, such as hard corals and comparison with the NR because the reef slope sponges, and are readily used by many reef also faces west with a horizontal face, thus fish species, thus acting as artificial reefs. The providing the most similar orientation for purpose of the present study is to assess the comparison. community compositions of structure-forming To assess the community composition of benthic organisms and reef fish species structure forming benthic organisms on the inhabiting and/or frequently utilizing mooring mooring blocks, a 1 m2 quadrat was separately blocks versus the adjacent natural coral reef. It placed on both the top and west-facing is hypothesized that there will be greater surfaces of a randomly chosen mooring block percent live benthic cover found on the natural and its paired NR site using SCUBA during reef than artificial reef. Additionally, it is the hours of 11:00 - 15:00. All corals and predicted that there will be greater benthic and sponges found within the quadrat were reef fish community diversity found on or identified and live percent cover was utilizing the natural reef than artificial reefs, determined. With these data total live percent and the composition of this diversity will vary cover was calculated and the mean percent live because the artificial reefs are much younger cover of each face per reef type (AR or NR) than the natural reefs, thus allowing for a was found. Diversity of benthic organisms different and less temporally developed found on each reef type was calculated using

45

Physis: Journal of Marine Science

Simpson’s Diversity Index. Diversity was then 80 top face broken down by mean density of each species 70 west face observed on the top and west faces of the two 60 reef types. Two-factor ANOVAs were used to 50 statistically analyze all comparisons between 40 documented characteristics of benthic 30

communities present on the top and west faces % cover Mean of the mooring blocks and those on the natural 20 reef. 10 Immediately following the assessment of 0 benthic community composition, fish Natural Artificial assemblages actively interacting with or found within ½ m of the face (top and west) and reef Figure 1. Mean percent live benthic cover of NR and AR (n = 8) type (AR or NR) being observed were between the Playa Lechi and Something Special dive sites, Bonaire, N.A. Reef: p = 0.009, Face: p = 0.185, Reef x Face: p = recorded. Observers located approximately 2 0.179. Error bars indicate standard deviation. P-values from a m away from the AR or NR took a two-factor ANOVA. comprehensive visual census of fish for 8 min, following a 2 min acclimation period to limit top face west face the effects of disturbance from the presence of 4 the divers (methods modified from Clynick 2008; Langhamer et al. 2009). Fish identity 3 and abundance were recorded. Diversity of fish found on each face of the two reef types 2 was calculated using Simpson’s Diversity Index. Diversity was then broken down by 1 mean density of each species observed on the Index Diversity Simpon's top and west faces of the two reef types. Two- 0 Natural Artificial factor ANOVAs were used to statistically analyze all comparisons between documented characteristics of fish communities present on Figure 2. Diversity of benthic organisms on NR and AR (n = 8) between the Playa Lechi and Something Special dive sites, the top and west faces of the AR and those Bonaire, N.A. Reef: p = 0.248, Face: p = 0.361, Reef x Face: p = found on the NR. 0.047. Diversity was calculated using Simpson’s Diversity Index. Error bars indicate standard deviation. P-values from a two-factor ANOVA. Results Despite this apparent similarity in benthic Benthic Community diversity, the composition of the reefs differed. The NR showed significantly more live Montastrea annularis and sponges comprised benthic cover than the AR (p = 0.009; Fig. 1). ~ 30 - 35 % of live benthic cover on the top Live benthic cover on the top faces ranged and west faces of the NR while these two from 14 - 80 % on the NR and 8 - 22 % on the organisms comprised only ~ 0 - 5 % on the top AR and live benthic cover on the west faces and west faces of the AR (Figs. 3a and 3b). ranged from 13 - 99 % on the NR and 17 - 50 On the other hand, the brain corals (Diploria % on the AR (Fig. 1). labrinthiformis and Diploria strigosa) Additionally, there was no difference in comprised 0 - 1 % of live benthic cover on the the overall diversity of live benthic corals and top face of the NR, while these two species sponges between the NR and AR (p = 0.248; comprised ~ 10 - 15 % on the top face of the Fig. 2) or the specific faces of both reef types AR (Fig. 3a). D. labrinthiformis and D. (p = 0.361; Fig. 2). The only difference was strigosa comprised ~ 0 - 1 % of live benthic significantly higher diversity on the west face cover on the west face of the NR while these of the AR than the top or west face of the NR two species comprised 15 - 20 % on the west (p = 0.047; Fig. 2). face of the AR (Fig. 3b). Fire coral (Millepora complanata) was the second most abundant

46

Physis: Journal of Marine Science

70 a) Natural 60 Artificial 50 40 30 20 10 0 Mean % Mean % live benthic cover

70 b) 60 50 40 30 20 10 0 Mean % Mean % live benthic cover

Figure 3. Mean percent live benthic cover on NR and AR (n = 8) between the Playa Lechi and Something Special dive sites, Bonaire, N.A. Error bars indicate standard deviation. a) top face b) west face.

14 benthic species observed on the west face of top face 12 the AR with 11.25 % mean percent live west face benthic cover, while it was only observed at 10 1.25 % on the top face of the AR. 8 Fish Community 6 There was no effect of reef face on the 4 diversity of fish utilizing the AR and the NR (p Simpson's Diversity Index Diversity Simpson's 2 = 0.127 and p = 0.910, respectively; Fig. 4). Twenty-five species of fish were observed on 0 the top face of the NR, while 19 were observed Natural Artificial on the top face of the AR (Fig. 5a). Bicolored Figure 4. Diversity of fish observed on NR and AR (n = 8) damselfish (Stegastes partitus) and brown between the Playa Lechi and Something Special dive sites, Bonaire, N.A. Reef: p = 0.127, Face: p = 0. 9102, Reef x Face = chromis (Chromis multilineata) were observed 0.365. Diversity was calculated using Simpson’s Diversity in the highest densities on the top face of the Index. Error bars indicate standard deviation. P-values from a NR with mean densities of 5.250 ± 4.359 fish two-factor ANOVA. m-2 and 1.250 ± 1.292 fish m-2, respectively. Twenty-nine fish species were observed on Sergeant majors (Abudefduf saxatilis) and the west face on the NR, while 16 were bluehead wrasses (Thalassoma bifasciatum) observed on the west face of the NR (Fig. 5b). were observed in the highest densities on the Again, S. partitus and C. multilineata were top face of the AR with mean densities of observed in the highest densities on the west 8.625 ± 8.595 fish m-2 and 4.375 ± 5.560 fish face of the NR with mean densities m-2, respectively.

47

Physis: Journal of Marine Science

18 Natural a) Artificial 16

14 ) -2 12

10

8

6 Mean density (# m 4

2

0 graysby bar jack blue tang doctorfish glass goby trumpetfish rock beauty rock french grunt bridled goby blue chromis creole wrasse butter hamlet saddled blenny sergeant major brown chromis spanish hogfish french angelfish scrawled filefish bluehead wrasse queen parrotfish sharpnose puffer redlipped blenny yellowtail hamlet yellowtail smooth trunkfish smooth dusky damselfish dusky bluestriped grunt striped parrotfish ocean surgeonfish smallmouth grunt smallmouth yellowtail snapper yellowtail longfin damselfish princess parrotfish redband parrotfish stoplight parrotfish stoplight blackbar soliderfish blackbar whitespotted filefish banded butterflyfish threespot damselfish bicolored damselfish foureye butterflyfish yellowheaded wrasse yellowheaded yellowtail damselfish yellowtail schoolmaster snapper schoolmaster

18 b) 16

14 ) -2 12

10

8

6 Mean density (# m 4

2

0 graysby bar jack blue tang doctorfish glass goby trumpetfish rock beauty rock french grunt bridled goby blue chromis creole wrasse butter hamlet saddled blenny sergeant major brown chromis spanish hogfish french angelfish scrawled filefish bluehead wrasse queen parrotfish sharpnose puffer redlipped blenny yellowtail hamlet yellowtail smooth trunkfish smooth yellowtail damsel yellowtail dusky damselfish dusky bluestriped grunt striped parrotfish ocean surgeonfish smallmouth grunt smallmouth longfin damselfish snapper yellowtail princess parrotfish redband parrotfish stoplight parrotfish stoplight blackbar soliderfish blackbar whitespotted filefish banded butterflyfish bicolored damselfish foureye butterflyfish threespot damselfish yellowheaded wrasse yellowheaded schoolmaster snapper schoolmaster Fish Species Observed

Figure 5. Mean density of fish species observed on the NR and AR (n = 8) between Playa Lechi and Something Special dive sites, Bonaire, N.A. Error bars indicate standard deviation. a) top face b) west face. of 4.750 ± 4.928 fish m-2 and 2.875 ± 4.615 3.202 fish m-2, respectively. Upon closer fish m-2, respectively. A. saxatilis and T. analysis of the most abundant fish on the AR bifasciatum were observed in the highest and that of the NR, it was found that there densities on the west face of the AR with mean were significantly more A. saxatilis than S. densities of 7.125 ± 6.151 fish m-2 and 1.875 ± partitus observed on the top and west face of

48

Physis: Journal of Marine Science

18 a) Natural and damage from anchors on the natural reef 16 Artificial slope. These concrete blocks have become

) 14 -2 colonized by benthic organisms and are 12 utilized by a diverse array of fish species 10 (Figs. 3a, 3b, 5a, and 5b). 8 This study showed that the NR had 6 significantly more live benthic coral cover Fish density (# m density Fish 4 2 than the AR, data which supports the initial 0 hypothesis (Fig. 1). Differences in benthic sergeant major bicolored damselfish community composition may be explained by the difference in substrate age. The mooring 18 blocks were installed at dive site locations 16 b) along the coastline approximately 30 years ago

) 14

-2 (BNMP - Moorings 2009), while the NR has 12 been developing in accordance with the 10 geologic history of Bonaire (BNMP 2006). 8 Additionally, corals, sponges, and 6 encrusting species must recruit to concrete on Fish density (# m density Fish 4 the AR, a substrate neither natural to nor 2 present on the NR. A study of Japanese 0 scleractinian corals found that physical sergeant major bicolored damselfish complexity of substrata had a significant effect Figure 6. Sergeant major and bicolored damselfish densities on on coral spat survivorship (Nozawa 2008). the NR and AR (n = 8) between the Playa Lechi and Something Special dive sites, Bonaire, N.A. Error bars indicate standard The porous nature of the limestone base of deviation. P-values from a two-factor ANOVA. a) top face. coral reefs in Bonaire provides micro-crevices Reef: p = 0.364, Fish: p = 0.438, Reef x Fish: p = 0.002. b) west that may serve as refuge structures to protect face. Reef: p = 0.620, Fish: p = 0.286, Reef x Fish: p = 0.003. coral spats from predation as compared to the

concrete surfaces of the AR which may lack the AR than the top and west face of the NR (p this small scale rugosity and provide increased = 0.002 and p = 0.003, respectively; Figs. 6a opportunities for predation. The NR on and 6b). The density of A. saxatilis on the top -2 Bonaire is primarily comprised of Montastrea face of the NR ranged from 0 - 3 fish m -2 annularis, a large boulder coral (BNMP 2006). versus 1 - 25 fish m on the top face of the Vertical and horizontal surfaces on the AR. Density of A. saxatilis on the west face of -2 NRwere oftentimes cross-sections of the large the NR ranged from 0 - 6 fish m , while M. annularis coral head, rather than an density of A. saxatilis ranged from 0 - 15 fish -2 uncolonized vertical surface like that of the m on the west face of the AR (Fig. 6a). AR. These findings support the results of this Conversely, the density of S. partitus on the -2 study that show greater percent live benthic top face of the NR ranged from 0 - 12 fish m -2 -2 cover on the NR than AR (Fig. 1). versus 0 - 3 fish m on AR, and 0 - 16 fish m -2 Although there was significantly more live on the west face of NR versus 0 - 4 fish m on cover on the NR, there was no difference in AR (Fig. 6b). overall benthic diversity between the AR and

NR. Interestingly though, there was Discussion significantly higher diversity on the west face

of the AR, than on either face of the NR or the Benthic Community Composition top face of the AR when comparing the Caribbean coral reefs are threatened by interaction between reef face and reef type disturbances caused by human activities such (Fig. 2). This result supports previous findings as overfishing, marine and terrestrial pollution, that that show that benthic species recruit more and increased sedimentation (Burke and readily to the vertical faces of AR than the Maidens 2004). In an effort to relieve some of horizontal faces (Rogers et al. 1984; these anthropogenic stresses, AR have been Langhamer et al. 2009). One possible introduced to natural marine environments explanation for this recruitment pattern can be (Burt et al. 2009). In Bonaire, mooring blocks hydrodynamics and water flow specific to the have been deployed to discourage the use of study site (Maldonado and Young 1996).

49

Physis: Journal of Marine Science

Planktonic larvae drifting from the parent Glasby and Connell 2001; Vermeij 2006; colony followed by settlement and Nozawa 2008). survivorship on the host substrate drives coral recruitment (Vermeij 2005; Wilson and Fish Community Composition Harrison 2005). Vertical surfaces, such as No difference in fish diversity was those found on the mooring blocks, provide a observed between the two reef types, faces, or physical blockade for lateral water current the interaction between reef and face type, movement, increasing the probability of data which does not support the original recruit settlement and in turn increasing the hypothesis (Fig. 4). These results refute the chances of recruit survivorship. initial hypothesis that reef fish diversity would A previous study found that benthic be greater on the NR than AR. Conflicting community composition on an AR versus an evidence exists on the effect of NR versus AR associated NR differed greatly when on fish diversity (Ambrose and Swarbrick controlling for depth, substrate composition, 1989; Clynick 2008; Burt et al. 2009b). and orientation, suggesting it is reef type that Ambrose and Swarbrick (1989) found no is most important in determining success of difference in overall diversity of fish benthic recruitment and colonization (Glasby assemblages on NR and AR off the coast of and Connell 2001). In teasing apart the live Southern California, while Clynick (2008) and benthic diversity observed on the NR and AR, Burt et al. (2009b) found marked differences there was a distinct difference in the in fish diversity between reef types. The close composition of species dominating the two proximity of the NR and AR in this study reef types (Figs. 3a and 3b). Two brain corals, (within 10 m) could be a possible explanation Diploria labrinthiformis and Diploria strigosa, for no difference in diversity between the reef comprised a majority of live percent cover on types, allowing species to easily migrate both the top and west faces of the AR, while between the two reef types. M. annularis and sponges dominated the NR. Although there was no difference in In addition to reef type affecting benthic overall fish diversity, the breakdown of the recruitment, spawning technique and diversity observed at the two reef types was successional order may affect survivorship of distinctly different (Figs. 5a and 5b). These recruits. All four species found to dominate data suggest that the differences in benthic the two reef types in this study are known composition between the two reef types does broadcast spawners (Vermeij 2006). Smith not necessarily have an effect on overall fish (1992) found that while brain corals in diversity, but may lead to the establishment of Bermuda were poor recruiters, they had lower unique fish communities. The physical juvenile mortality than other coral species differences between the two reef types, the NR such as Porites asteroides. Additionally, being a complex, rugose system in contrast to Vermeij (2006) found that both D. the AR that is primarily a flat surface with labrinthiformis and D. strigosa recruited to an significantly less percent live cover (Fig. 1), artificial substrate two years before M. may be a driving force behind this difference annularis. These two studies corroborate in the fish represented. Gratwicke and Speight results found in this study, suggesting that (2005) found that as habitat complexity brain coral larvae may be more generalist than increased so did fish species richness. It is M. annularis or sponge larvae when settling possible that the utilization of NR versus AR on available substrata. by specific coral reef fish may be based on the It is unlikely that the differences between physical complexity of the substrata. live benthic diversity and cover on the NR and Differences in fish behavior were AR can be explained by any one cause, but observed between the two reef sites, rather it may be the interaction of many biotic suggesting substrate-induced behavior (M. and abiotic factors. Water currents, spatial and Thomas, personal observations). A. saxatilis, temporal coral spawning patterns, recruitment as predicted by the original hypothesis, and T. success, coral growth rates, orientation of the bifasciatum dominated the AR. A. saxatilis substrate, and composition of the substrates was observed to be utilizing the AR for may all affect coral recruitment and succession breeding, whereas breeding behavior was (Smith 1992; Maldonado and Young 1996; never observed on the NR. A majority of the T. bifasciatum observed on the AR were

50

Physis: Journal of Marine Science juveniles, suggesting the use of the mooring References blocks as suitable habitat for young fish, perhaps due to protection from predation Ambrose, R. F. and S. L. Swarbrick. 1989. present in the open water (Clynick 2008). S. Comparison of fish assemblages on artificial partitus and C. multilineata were the most and natural reef off the coast of Southern abundant species observed on the NR. High California. Bulletin of Marine Science densities of C. multilineata are expected as 44(2):718-733. Bonaire National Marine Park. 2006. BNMP previous comprehensive REEF fish surveys Management Plan. cite C. multilineata as the most common fish Burke, L. and J. Maidens. 2004. Reefs at Risk in found on the natural reef surrounding Bonaire the Caribbean. World Resources Institute, (Burke and Maidens 2004). Washington, D.C. Densities of A. saxatilis were relatively Bonaire National Marine Park – Moorings. high on the AR while densities of A. saxatilis http://www.bmp.org/moorings.html Accessed were low on the NR (Fig. 6). The opposite 17 September 2009. trend appeared for S. partitus, with high Burt, J., A. Bartholomew, A. Bauman, A. Saif, and densities on the NR and low densities on the P. F. Sale. 2009a. Coral recruitment and early AR (Figs. 6a and 6b). This negative benthic community development on several materials used in the construction of artificial relationship suggests an ecological interaction reefs and breakwaters. Journal of between these two species. It is suspected that Experimental Marine Biology and Ecology S. partitus is outcompeting A. saxatilis for 373:72–78. space on the natural reef. S. partitus is a Burt, J., A. Bartholomew, P. Usseglio, A. Bauman, known territorial damselfish that recruits and P. F. Sale. 2009b. Are artificial reefs quickly (within 1 year) and has been known to surrogates of natural habitats for corals and have high population densities on coral reefs fish in Dubai, United Arab Emirates? Coral (0.23 adults m-2; Robertson 1996). The AR Reefs 28:663–675. may be providing a new habitat for A. Clynick, B. G. 2008. Characteristics of an urban saxatilis, one that is better suited for its fish assemblage: Distribution of fish associated with coastal marinas. Marine Environmental reproductive needs. It is possible the Research 65:18-33. combination of availability of open space Glasby, T. M. and S. D. Connell. 2001. Orientation found on the AR and lack of apparent and position of substrata have large effects on predators of A. saxatilis surrounding the AR epibiotic assemblages. Marine Ecology (M. Thomas, personal observation) is more Progress Series 214:127–135. conducive to the reproductive success of these Gratwicke, B. and M. R. Speight. 2005. Effects of fish. habitat complexity on Caribbean marine fish Differences in coral and fish community assemblages. Marine Ecology Progress Series diversity are evident between the two reef 292:301-310. types. Results of this study suggest that the Langhamer, O., D. Wilhelmsson, and J. Engstrom. 2009. Artificial reef effect and fouling impacts placement of these AR does not change the on offshore wave power foundations and overall diversity of natural reef, but rather buoys – a pilot study. Estuarine, Coastal and causes a shift in diversity composition due to Shelf Science 82:426–432. the availability of open space for benthic Maldonad, M. and C. M. Young. 1996. Effects of species to colonize and reef fish to use physical factors on larval behavior, settlement accordingly. and recruitment of four tropical demosponges. Marine Ecology Progress Series 138:169-180. Acknowledgements Nozawa, Y. 2008. Micro-crevice structure enhances coral spat survivorship. Journal of I would like to thank Alyssa Adler, Noelle Experimental Marine Biology and Ecology 367:127-130. Hawthorne, and Alison Maysr for helping me Perkol-Finkel, S. and Y. Benayahu. 2004. with countless hours of data collection. Community structure of stony and soft corals Additional thanks to Grant Frank and Mollie on vertical unplanned artificial reefs in Eilat Sinnott for the many late nights and written (Red Sea): Comparison to natural reefs. Coral revisions. Finally, this project would not have Reefs 23:195–205 been possible without the guidance and support of Dr. Amanda Hollebone.

51

Physis: Journal of Marine Science

Robertson, D. R. 1996. Interspecific competition Vermeij, M. J. A. 2005. Early life-history dynamics controls abundance and habitat use of of Caribbean coral species on artificial territorial Caribbean damselfishes. Ecology substratum: the importance of competition, 77(3):885-899. growth, and variation in life-history strategy. Rogers, C. S., H. C. Fitz III, M. Gilnack, J. Beets, Coral Reefs 25:59-71. and J. Hardin. 1984. Scleractinian coral Wilson, J. and P. Harrison. 2005. Post-settlement recruitment patters at Salt River submarine mortality and growth of newly settled reef canyon, St. Croix, U.S. Virgin Islands. Coral corals in a subtropical environment. Coral Reefs 3(2):69-76. Reefs 24:418-421.

52

Physis: Journal of Marine Science

Parrotfish mucus cocoon production at night in Bonaire, N.A. Chelsey Weathersbee Wofford College

Abstract Many marine organisms use mucus to catch food, clean themselves, or for protection from predators. Some species of parrotfish (Family Scaridae) use mucus to build cocoons around themselves at night, which is thought to be a form of protection. Although there are reports of parrotfish constructing mucus cocoons, little is known about which specific species produces cocoons, where on the reef cocoons are used, or how prevalent the behavior is in Bonaire, N.A. The purpose of this study was to determine which species and phases of parrotfish construct cocoons, the distribution of cocoons from the reef slope to the shallow subtidal, and whether cocoons are being used for protection from predators. Observations took place between Oct. – Nov. 2009, after 23:00 h. Six depths were surveyed for parrotfish (1, 3, 6.5, 10, 15, and 20 m) and surveys were standardized by time. A guide diver assisted in keeping time, recording predators, and maintaining depth. Seven to 10 min was spent at each depth starting at 20 m and working up to 1 m. This study provides information on the nighttime ecology of parrotfish, which may be important for conservation of the species. During this study, two species of parrotfish, Scarus taeniopterus (princess parrotfish) and Scarus vetula (queen parrotfish), were found in cocoons; cocoons were only built along the reef slope, and none were found on the reef flat. Only terminal phase S. taeniopterus were found in cocoons, whereas terminal and initial phase S. vetula were found in cocoons.

Introduction

In marine environments, organisms such as corals and fish use mucus. One way in which corals use mucus is to trap and remove sediments (Weber et al. 2006). Mucus has many functions for fish as well; it is used in respiration, reproduction, disease resistance, feeding, nest building, and protection from predators (Shephard 1994). Of the 14 species of parrotfish known to the Caribbean, two have been reported inside of mucus cocoons, Scarus taeniopterus (princess parrotfish) and Scarus vetula (queen parrotfish) (DeLoach 1999; Fig. 1). Parrotfish exist in three stages: juveniles, Figure 1. Scarus taeniopterus found in mucus cocoon on 1 Nov. initial phase, and terminal phase (DeLoach 2009 at the Kas di Arte dive site in Bonaire, N.A. Photograph taken by Mollie Sinnott. 1999). Sazima and Ferreira (2006) stated that both initial and terminal phase of Scarus that mucus cocoons made by S. vetula contain zelindae, a species found in the Southwest antibiotics that kill common tropical bacteria Atlantic, produce mucus cocoons. The cocoon that affect fish. Winn and Bardach (1959) originates at the mouth and is produced around hypothesized that predation would be lower on the fish like an envelope (Winn and Bardach parrotfish that make mucus cocoons (Scarus) 1959). The cocoons contain holes, which are when compared with parrotfish that do not located at the mouth and at the back of the (Sparisoma). The results indicated that cocoon caudal fin (Winn and Bardach 1959). builders were preyed upon less often than non- Mucus cocoons may provide protection for cocoon builders (Winn and Bardach 1959). parrotfish against bacteria, infections, parasites Parrotfish are abundant throughout tropical (Videler et al. 1999), or for protection against waters (DeLoach 1999) and are important predators (Winn and Bardach 1959). Videler herbivores in coral reef ecosystems (Lewis et al. (1999) showed that mucu cocoons made 1986). Due to their role as large herbivores in by S. vetula contain antibiotics that kill Bonaire, N.A. (Alvarado et al. 2007) and the

[email protected] 53

Physis: Journal of Marine Science lack of knowledge about nighttime cocoon- building behavior; this study is important because it will provide information on a little studied behavior that could be essential to the conservation of the species. The purpose of this study is to test the following hypothesis related to the nighttime distribution patterns and behaviors of parrotfish. H1: S. taeniopterus and S. vetula will be found using mucus cocoons, and parrotfish from the genus Sparisoma will not because they have not been observed using mucus cocoons (Winn and Bardach 1959). H2: More parrotfish will be found in mucus cocoons on the reef slope compared to the reef flat because there is more coral reef Figure 2. Map of Bonaire, N.A., indicating the 10 study sites and therefore more potential for where nighttime observations were made. concealment. H3: Terminal phase parrotfish will make sites using SCUBA. The observer spent ~ 7 - mucus cocoons more often that initial 10 min searching and recording parrotfish phase parrotfish because they are larger found at each depth, starting at 20 m and and it is harder for them to find places to heading into the current, or south, if there was hide. no significant current, then to 15 m depth H4: When predators are present there will be heading north, and working up the reef to 1 m more parrotfish found using mucus (Fig. 3). Parrotfish species, phase, depth, cocoons compared to not using cocoons presence or absence of a mucus cocoon, and because mucus cocoons are a form of presence or absence of injury was recorded for protection (Winn and Bardach 1959). each specimen observed during the timed H5: Parrotfish found to have an injury will swim. A guide diver assisted the observer in construct a mucus cocoon because the keeping time, recording predators, and cocoons have antibiotics in them that fight maintaining constant depth. against bacteria (Videler et al. 1999). Start Finish Materials & Methods 20 m 15 m 10 m 6.5 m 3 m 1 m

Study Sites N Observations for this study took place at 10 different sites along the west coast (leeward side) of Bonaire, N.A. The sites chosen were similar in rugosity and habitat type. The northern most site was Witches Hut (1212’22.63” N, 6819’00.43” W), and the southern-most site was Bachelor’s Beach (1207’32.79” N, 6817’17.22” W), with the other eight sites located in-between (Fig. 2). S Four of the 10 sites (Oil Slick, Cliff, Bari, Donkey Beach) were chosen through Figure 3. Diagram of the search pattern used to search for parrotfish. Observers spent ~ 7 - 10 min at each depth, communication with a local diver (A. depending on current speed. Bianculli, personal communication) and the other six were chosen due to their proximity to Data Analysis those four sites. Data were standardized by search time, parrotfish min-1. To determine which species Data Collection of parrotfish built mucus cocoons most often a Observations were taken at six depths (1, 3, two-sample t-test ( = 0.05) was used. It was 6.5, 10, 15, and 20 m) after 23:00 h at 10 dive

54

Physis: Journal of Marine Science not possible to run a test on whether parrotfish 0.06 built cocoons on the reef flat (1, 3, 6.5 m) or -1 reef slope (10, 15, 20 m) most often because 0.045 parrotfish in mucus cocoons were never observed on the reef flat. A single factor 0.03 analysis of variance (ANOVA) was used to determine which depth of the reef slope (10, 15 0.015 and 20 m) cocoons were found most often. A

two-sample t-test ( = 0.05) was also used to min cocoons # of fish in mucus 0 determine which phase (terminal or initial) of Scarus taeniopterus Scarus vetula parrotfish made cocoons most often. A trendline (y = mx + b) including the R2 value Parrotfish Figure 5. Mean number of parrotfish min-1 found in mucus was used to show the relationship between the cocoons ( SD) of the 2 species of parrotfish found in mucus mean number of predators and the mean cocoons at 10 dive sites in Bonaire, N.A. number of parrotfish found in mucus cocoons. -1 A test was not run on parrotfish injury because than S. taeniopterus (0.010  0.014 fish min ; there was no way to accurately detect injuries. Fig. 5). Mucus cocoons were only constructed Results along the reef slope; none were found on the reef flat. Although no significant difference was found (p = 0.470), parrotfish build more During the data collection period from -1 Oct. – Nov. 2009, a total of 151 parrotfish cocoons at 20 m (0.084  0.124 fish min ) -1 from four species were recorded, 16 of which than at 15 m (0.040  0.052 fish min ) or 10 m -1 were found in mucus cocoons. The species (0.040  0.084 fish min ; Fig. 6). observed most often was Sparisoma viride 0.25

(stoplight parrotfish), followed by Sparisoma -1 aurofrenatum (redband parrotfish), Scarus vetula (queen parrotfish), and Scarus 0.2 taeniopterus (princess parrotfish; Fig. 4). 0.15

0.3 0.1

0.25

-1 0.05 0.2 # of fish in mucus min cocoons # of fish in mucus 0 0.15 10 15 20 0.1 Depth (m)

# of total fish min -1 0.05 Figure 6. Mean number of parrotfish min found in mucus cocoons ( SD) located on the reef slope (10 m, 15 m, 20 m). 0 Data were collected at night at 10 different dive sites in Bonaire, N.A.

Both initial and terminal phase parrotfish were observed in mucus cocoons. Although no significant difference was found (p = 0.190)

Parrotfish terminal phase parrotfish (0.019  0.028 fish min-1) were found to make cocoons more often Figure 4. Mean number of total parrotfish min-1 ( SD) found at 10 dive sites in Bonaire, N.A. One night dive took place at each than initial phase parrotfish (0.010  0.013 fish site. min-1; Fig. 7). Predators appear to have little effect on Two species were observed using mucus parrotfish and their construction of mucus cocoons, S. vetula and S. taeniopterus. cocoons. There is a slightly negative trend (y Although there was no significant difference in 2 = -0.182x + 0.042) however with an R value which of these species made cocoons most of 0.044 there is no positive relationship (Fig. often (p = 0.193) S. vetula (0.018  0.025 fish -1 8). min ) was found to make cocoons more often

55

Physis: Journal of Marine Science

-1 0.06 reef slope and therefore more opportunity to conceal themselves. 0.045 Terminal phase parrotfish were found in cocoons more often than initial phase 0.03 parrotfish, which supports the hypothesis that terminal phase parrotfish would make cocoons 0.015 more often than the initial phase due to their larger size making it harder to hide in the reef

# of fish in mucus # cocoons of fish in min mucus 0 framework. Another possibility that terminal Terminal Initial phase parrotfish were found in cocoons more Phase often that initial phase parrotfish is that only one species of parrotfish, S. vetula, was found Figure 7. Mean number of terminal and initial phase parrotfish to make mucus cocoons in its initial phase. min-1 found in mucus cocoons ( SD) during nighttime sampling at 10 dive sites in Bonaire, N.A. Because S. vetula and S. taeniopterus are both capable of making cocoons, but only S. vetula 0.14 was found to make cocoons in its initial phase -1 it is possible that the initial phase S. vetula 0.12 found in mucus cocoons could have already 0.1 had the features of a terminal phase S. vetula with exception to the color change. Since 0.08 color is the only feature able to be used to 0.06 distinguish terminal (male) versus initial 0.04 (female) phase without examination in a lab, the initial phase parrotfish found in mucus # of fish in mucus min cocoons # of fish in mucus 0.02 cocoons may possibly be terminal phase and 0 therefore have the terminal phase capability to 0 0.05 0.1 0.15 make cocoons.

# of predators min-1 There was a slightly negative relationship between parrotfish in mucus cocoons and Figure 8. Mean number of parrotfish min-1 found in mucus cocoons compared to mean number of predators min-1 found predator. The hypothesis that when a greater throughout the study (R2 = 0.044; y = -0.183 + 0.042). Data number of predators are present there would be were collected at night at 10 different dive sites in Bonaire, N.A. more parrotfish found using mucus cocoons was not supported. Although predators do not Discussion appear to be the present cause of parrotfish constructing mucus cocoons, there could have Of the four species of parrotfish observed, been higher predation in the past or a specific two were found to make mucus cocoons at predator that lead parrotfish to adapt this night, Scarus vetula and Scarus taeniopterus. behavior as a means of protection. This supports the hypothesis that only S. vetula Originally, a component of the study was and S. taeniopterus would be found in mucus to determine whether injury influenced the use cocoons since parrotfish in the genus of mucus cocoons and it was hypothesized that Sparisoma are not known to make cocoons in Scarus parrotfish found to have an injury the wild (DeLoach 1999). Of the 2 species would also be found in a mucus cocoon; found in mucus cocoons, S. vetula was found however, there was no way to accurately detect to make cocoons more often than S. injuries. Parrotfish, whether in cocoons or not, taeniopterus, which is most likely because sleep within the framework of the reef. If they terminal and initial phase S. vetula were found are not hidden in a crevice or under a coral to using cocoons whereas only terminal phase head, they have one side of their body backed S. taeniopterus were found using cocoons. up against the reef making it impossible to Parrotfish in mucus cocoons were only detect injury unless fish were captured and observed on the reef slope which supports the examined. hypothesis that more mucus cocoons would be The reason for such high standard found on the reef slope compared to the reef deviation (error bars) is due to the small flat because there is more coral cover on the sample size and lack of consistency in which mucus cocoons were observed. It cannot be

56

Physis: Journal of Marine Science determined through this study the cause of Bianculli, A. Personal communication. October parrotfish building mucus cocoons. It may be 2009. Bonaire, Netherland Antilles. possible that some parrotfish just prefer to DeLoach, N. and P. Humann. 1999. Reef Fish sleep in a cocoon while others do not, but since Behavior. New World Publications, Inc. each of the 10 sites were only visited once Jacksonville, Fl. Lewis, S. M. 1986.The role of herbivorous fishes in there is no evidence of whether individual the organization of a Caribbean reef parrotfish practice the same behavior (building community. Ecological Monographs 56:184- cocoons or not building cocoons) nightly. 200. Sazima, I. and C. E. L. Ferreira. 2006. A cocoon- Acknowledgements producing parrotfish in the southwestern Atlantic. Coral Reefs 25:212. I would like to thank Dr. Rita Peachey, Dr. Shephard, K. L. 1994. Functions for fish Amanda Hollebone, Caren Eckrich, and Kate mucus.Reviews in Fish Biology and Fisheries Jirik for all of their support. Thank you to Dr. 4:401-429. Rita Peachey and Kate Jirik for their aid in my Videler, H., G. J. Geertjes, and J. J. Videler. 1999. Biochemical characteristics and antibiotic research dives. Thank you to Mollie Sinnott properties of the mucus envelope of the queen and Jill Lau for photographs, Albert Bianculli parrotfish. Journal of Fish Biology 54:1124- for information on dive sites and timing, and 1127. Jerry Ligon for his help identifying a Weber, M., C. Lott, and K. E. Fabricius. 2006. parrotfish. Sedimentation stress in a scleractinian coral exposed to terrestrial and marine sediments References with contrasting physical, organic and geochemical properties. Journal of Alvarado, N. A., P. Mumby, and R. Steneck. 2007. Experimental Marine Biology and Ecology Chapter 3: Trends in distribution and 336:18-32. abundance of carnivorous and herbivourous Winn, H. E. and J. E. Bardach. 1959. Differential reef fish populations on Bonaire. A Report on food selection by moray eels and a possible the Status of the Coral Reefs of Bonaire in role of the mucous envelope of parrotfishes in 2007 With Results from Monitoring 2003-2007 reduction of predation. Ecology 40:296-298. 14-22.

57

Physis: Journal of Marine Science

Christmas tree worms (Spirobranchus giganteus) and their role as bioindicators of environmental stress on coral reefs of Bonaire, N.A.

Pamela Williams University of Colorado at Boulder

Abstract The use of biological entities as indicators of environmental stress can provide links between changes in ecological conditions and ecosystem productivity. Historically, bioindicators have been used as a rapid-assessment tool of areas declining in sustainability for the inhabiting organisms. This study investigated the utility of sessile, filter feeding Christmas tree worms (Spirobranchus giganteus) as bioindicators of the presence of potential coral reef stressors. Christmas tree worm density was compared at low impact sites (> 200 m from a commercial establishment) and high impact sites (< 200 m from a commercial establishment). For each site, four quadrats were randomly placed along a 10 m transect at 6, 12 and 18 m depths to assess percent live coral cover and Christmas tree worm density. These data were compared with potential environmental stressors such as excess nutrients (nitrite, nitrate, ammonia and phosphate), human gut (Enterococcus) bacteria, sedimentation rates, and sediment particle size distributions between high and low impacted sites and among depths. Approximately 97% of the worms inhabited live coral. Live coral cover was similar for 12 and 18 m at both high and low impacted sites (~ 17 % - 20 %) but significantly lower at 6 m depth (~ 2 % - 8 %). Despite the similarity in live coral cover at depth, there were significantly more Christmas tree worms at 12 m of high impact sites. At all other sites and depths, the worms never exceeded ~ 1.5 worms m-2. At 12 m, water chemistry analyses did not show any differences between site impact except for phosphate, with significantly greater concentrations at high impact sites. Bacterial loads, sedimentation rate and particle size distributions did not show any differences between site impact although there were finer sediments at high impact sites and coarser sediments at low impact sites. S. giganteus may be found at high densities at high impact sites due to a greater availability of food acquired through filter-feeding biota. Therefore, they may be used as novel indicators of the presence of environmental stressors, such as excess nutrients and finer sediments, in Caribbean coral reef systems.

Introduction and function at historical levels (Cooper et al. 2009). This may be achieved through For both terrestrial and aquatic bioindicators. ecosystems, natural and anthropogenic Reptiles and amphibians have historically stressors can occur through pollutants, been used as rapid-assessment tools to evaluate uncharacteristic weather and an unbalance of the presence of contaminants in terrestrial and predator prey interactions which may freshwater systems. Australian termite- determine the evolution, tolerance, or fatality specialist geckos (Diplodactylus of individuals and ecosystems alike (Smith and conspicillatus) were used as a model organism Budderneier 1992). As organisms in an to assess the hazards of excess sulfur dioxide ecosystem respond to the interactive effects of and salt spray produced by a sulfur mine (Read human population growth and developmental 1998). In areas with high pollutant loads, changes, alterations both harmful and gecko numbers and fecundity were beneficial for the organisms are likely to occur significantly lower than at control sites. This (Smith and Budderneier 1992). The use of could have resulted from the direct effects of biomonitoring for environmental health airborne emissions on gecko survivorship or assessment may provide the data required for the indirect effects of a decline in termites (the legislation directed at reducing these human- geckos’ preferred food) due to industrial sulfur caused environmental stressors (Lange et al. pollution. Read (1998) concluded that geckos 1996). Effective environmental management could be used as sensitive bioindicators of requires monitoring that provides specific links various atmospheric pollutants. between changes in environmental condition For bullfrogs, Rana beiana and Rana and an ecosystem’s ability to provide habitat clamitans, surveyed in New Hampshire, [email protected] 58

Physis: Journal of Marine Science

U.S.A., malformations occurred during diadema, have been used to examine metal and development as a result of several potential pesticide contamination (Dodge and Gilbert environmental causes which included disease, 1984). The worms could not survive the levels parasites, contaminants, climatic changes, acid of mercury, copper and zinc in the synthesized precipitation, and ultraviolet B radiation environment and therefore were useful as (Sower et al. 2000). The discovery of these bioindicators of heavy metal contamination amphibian malformations and mutations in (Dodge and Gilbert 1984). bodies of water that had previously been Christmas tree worms (Spirobranchus considered clean led to increases in public giganteus) may play a similar role in coral reef interest and concern for its affect on the ecosystems of the Caribbean currently under surrounding habitat and human population stress from eutrophication and excess (Burkhart et al. 2000). sedimentation as a result of human population In a review conducted by Lange et al. growth and coastal development (Burke and (1996) citing multiple studies several common Maidens 2009). S. giganteus inhabits varying environmental contaminants and depths of the coral reef in calcareous tubes, biomonitoring techniques were described. filter feeding from a fixed location for their They found that by adding sublethal amounts entire adult life (Humann and Deloach 2003). of arsenic to experimental habitats for a variety Larval settlement assays on the Great Barrier of small bodied crustaceans and fish, including Reef, have shown a clear preference of S. the water flea (Daphnia magna), a copepod giganteus for live Acropora prolifera or (Nitocra spinipes) and the zebrafish Palauastraea ramosa habitat, but never dead (Brachydanio rerio), arsenic could be traced coral rubble or experimental glass controls following the exposure (Lange et. al 1996). (Marsden 1987). The selection of live coral The same review also showed that when species habitat by S. giganteus may even be concentrations of cadmium were added to species-specific. Hunte et al. (1990) found that experimental ponds, carp eggs decreased in the extent to which a coral head was populated size and had reduced success in hatching rates, by S. giganteus in Barbados was not due to making the eggs extremely sensitive indicators surface area, colony size or the coral’s ability to the pollutant (Lange et. al 1996). to overgrow another species. Instead, the most The use of bioindicators may provide a “heavily” and “moderately” colonized corals variety of advantages when measuring water were Diploria strigosa (7.61 m-2), Porites quality in marine systems (Cooper et al. 2008). astreoides (5.33 m-2), Millepora complanata Marine bioindicators may present a time- (6.38 m-2), Montastrea annularis, Madracis integrated measure (from hours to years) that spp. and Agaricia spp. (Hunte et al. 1990). can aid in the evaluation of complex However, other common Caribbean stony ecosystems such as coral reefs (Cooper et al. corals such as Colpophyllia natans, 2008). Cooper et al. (2009) used a variety of Dendrogyra cylindricus, Dichocoenia stokesii, bioindicators in assessing stressors on the Eusmilia fastigata, Meandrina meandrites and Great Barrier Reef. By comparing the percent Mycetophyillia spp. showed no evidence of cover of hard corals, octocorals, and Christmas tree worm recruitment (Hunte et al. macroalgae to changes in water quality and 1990). sedimentation, the elevated sensitivity of these While providing habitat to S. giganteus, organisms generally reflected the ambient corals may actually benefit from the presence conditions. For example, the taxonomic of the worm. For several Red Sea coral richness of zooxanthellate octocorals declined species, Cyphastrea chalcidicum, Favia favus, with increasing turbidity and sedimentation, and Favia laxa, colonization of S. giganteus followed by rapid mortality as the stresses showed coral tissue immediately surrounding continued (Cooper et al. 2009). the polychaete to have little or no damage from Dodge and Gilbert (1984) used corals as disease, bleaching, or predation, while also bioindicators, analyzing the annual growth appearing to recover from stress more quickly rings in Montastrea annularis to determine the (Ben-Tzvi et al. 2005). The presence of filter history of lead pollution and phosphorous feeding worms may provide improved water contamination in the Virgin Islands. circulation, therefore decreasing the Additionally, other benthic marine organisms probability of bleaching, improving the such as the polychaete worm, Ophryotrocha diffusion of waste, and amplifying the

59

Physis: Journal of Marine Science accessibility of nutrients to coral polyps (Ben- there were more nutrients, Enterococci, higher Tzvi et al. 2005). sedimentation rates, and finer particles. If The island of Bonaire, N.A., in the positive correlations between the density of Southern Caribbean, currently faces the Christmas tree worms and the above negative effects of eutrophication and parameters could be established then S. sedimentation as a result of human population giganteus may act as a bioindicator of growth and coastal development (Burke and environmental stressors in Caribbean coral reef Maidens 2004; BNMP 2006). The island has a systems. porous limestone base and is surrounded by a diverse fringing coral reef within 50 m from Materials & Methods shore. The source of excess nutrient loading on Bonaire’s coral reefs may be leaking septic This research was executed at six dive tanks, waste water trenches, loss of mangrove sites along Bonaire’s leeward coast using habitat and sedimentation runoff from areas SCUBA. Three “high impact” sites (Bari under construction (S. Patton, personal Reef, Kas di Arte, and Eighteenth Palm) and communication). Untreated or partially treated three “low impact sites” (Witch’s Hut, Andrea sewage is both directly (pipes leading from II, and White Slave) were assessed (Fig. 1). commercial establishments onto the reef) and “High impact” sites were defined as those < indirectly (unlined septic tanks) emptied into 200 m from a commercial establishment such coastal waters, potentially providing additional as a dive resort, office building, or restaurant nutrients (some of which may appear in the and “low impact” sites were defined as those form of Enterococci) that may alter the water surrounded by > 200 m of rural or residential chemistry of the reefs and support macroalgal land (Rini 2008). In the ongoing study blooms (Hinrichsen 2005; S. Patton, personal conducted by CIEE Research Station Bonaire, communication). Due to recent increases in these sites are continuously monitored at 12 m coastal development and road construction on a biweekly basis for water chemistry with little erosion control on Bonaire (BNMP (including nitrite, nitrate, ammonia, and 2006), an increase of sewage as well as heavy phosphate), Most Probable Number of enteric rain may result in nutrient and sediment runoff human gut bacteria, rates of sedimentation, and onto the island’s coral reefs (S. Patton, particle size distribution. personal communication). This threatens the unique reef habitat of many animals and algal species, the economic benefits of tourism that Bonaire provides, the island’s fisheries, the shoreline protection the reefs provide, and the production of coral sand (BNMP 2006). Currently, the Council on International Education Exchange’s (CIEE) Research Station Bonaire is conducting a long term monitoring study of various components of the coral reefs at ten dive sites along the west coast of Bonaire. As a part of this monitoring, sediment traps are established and water samples taken at 12 m and analyzed for various parameters of the water on a biweekly basis. A quick assessment tool to determine the status of variable stressors on the biota of Bonaire’s reefs is yet to be defined. For the current study, it was hypothesized that there would be more nutrients, higher presence of Enterococci, sedimentation rates, Figure 1. Map of Bonaire, N.A. (BNMP 2006). Location of dive sites for this study indicated with arrows. The sites > 200 m and finer particles at the sites closer to from a commercial establishment are low impact. The sites < commercial establishments. It was further 200 m from a commercial establishment are high impact hypothesized that there would be higher (indicated with an asterisk). densities of S. giganteus in the areas where

60

Physis: Journal of Marine Science

Site Assessments assessed (# and size of well trays counted) In order to evaluate habitat composition of using a black light and converted into a Most each site, a 10 m transect was centered (5 m to Probable Number (MPN) using Enterolert®’s the North and South) on a permanent sediment MPN generator table. The MPN provides the trap (a component of the ongoing CIEE parts per million (ppm) concentration of monitoring project) located at 12 m. In Enterococcus at 12 m at each site and the addition, transects at 6 and 18 m were run number of bacteria populations per 100 mL of parallel and in alignment with the one laid at sample water. The data for the bacterial 12 m. Four ¼ m² quadrats were randomly analysis was analyzed using a one-tailed t-test placed along each transect (two to the North comparing level of impact. and two to the South) to determine live coral cover. Upon encountering a substrate Sediment Analyses inhabited by Christmas tree worms, their Sediments from each site were collected abundance and status of substrate (live or dead using PVC pipes (7.5 cm diameter, 15 cm coral) was recorded. long) with a closed bottom and open top; this allowed sediments from the water column to Water Chemistry continuously collect during the two week To determine the composition of water at period. The PVC trap was attached to a 1 m the chosen sites and depths, samples were long rebar stake with zip-ties, hovering ~ 10 acquired on a biweekly basis from the ongoing cm from the substrate. This trap was used as CIEE monitoring program from 31 Aug. – 12 the central marker for all transects (as Oct. 2009. This encompassed data from one described previously). The trap was collected month prior and during the assessment of the on a biweekly basis by being capped at the site worms and was repeated for the bacterial and and transported to the lab. The sediments sediment analyses. These samples were were left to settle for 1 h before the seawater collected with 250 mL Nalgene bottles which was decanted off and the sediments rinsed with were acid washed with a 10 % HCl solution tap water 3 X to remove excess salts. Once and filled with distilled water for transport. At rinsed, the sediments were placed in an oven at 12 m, a site specific water sample was (40 – 60 °C) and dried for 48 h. Once dried, collected by inverting the open bottle and any organic material was removed with filling it with air from a secondary SCUBA forceps and the remaining particles were source, and repeating this three times before weighed. Sediment rate was determined by capping. These samples were returned to the dividing the final sediment weight by the lab within 2 hours (h) of collection and number of days the trap was in the water - analyzed for nitrite (NO2 ), nitrate (NO3), and (Gleason 1998). These data were analyzed ammonia (NH3) using the LaMotte SaltWater using a one-tailed t-test. AquaCulture Test Kit (Model AQ-4) and Particle composition was determined by phosphate (PO4) using the Red Sea Marine and re-suspending the dry samples into ~ 25 mL of Freshwater Test Lab kit. The data for each tap water. A 1 mL sample of the mixed compound was analyzed using a one-tailed t- solution was extracted from the container test. using a 10 mL syringe, placed on a counting cell slide, and viewed under a compound Bacterial Analysis microscope. One of the 200 cells of the Enteric bacteria counts were determined counting cell slide was randomly chosen and using the same site-specific water collected at the first two hundred particles encountered 12 m on a biweekly basis that was set aside (from left to right) were counted and placed prior to chemical testing to avoid cross- into one of the following size categories: < 10 contamination. One unit of Enterolert® µm, 11-50 µm, 51-100 µm, 101-250 µm, 251- fluorescing substrate was mixed with a pre- 500 µm and > 500 µm. Three sub samples prepared 100 mL solution (10 mL of site were analyzed for each site. The data were specific water and 90 mL of distilled water) evaluated using a two-factor ANOVA. and poured into an IDEXX Quanti-tray®. The tray was vacuum sealed and incubated at 41 °C ± 0.5 °C for 24 h. Following the incubation period, bacterial fluorescence was visually

61

Physis: Journal of Marine Science

Results was significantly higher at high impact sites (0.06 ± 0.04 ppm) than at low impact sites Site and Worm Assessments (0.01 ± 0.02 ppm) (p = 0.048, Fig. 4).

The assessment of habitat composition High Impact showed that live coral cover did not differ 0.7 between type of site Impact or the site Impact .. 0.6 Low Impact x Depth interaction (p = 0.385 and p = 0.793, 0.5 respectively), but there was a significant difference among depths (p = 0.011). At 6 m 0.4 coral cover was between ~ 2 – 8 % whereas at 0.3

12 and 18 m coral cover was ~ 17 – 20 % (Fig. (ppm) Concentration 0.2 2). 0.1 Spirobranchus giganteus was found nearly 100 % of the time embedded in live coral 0 tissue (with the exception of 1 individual out Nitrite Nitrate Ammonia Phosphate of a total of 30). The mean density of S. Figure 4. Mean concentration in parts per million (ppm) (± SD) giganteus among depths and between site of nitrite, nitrate, ammonia and phosphate and between impact impact showed significantly more worms at 12 level (Ammonia: p = 0.195 and Phosphate: p = 0.048). P - values were determined by one-tailed t-tests. m (Impact x Depth interaction: p = 0.003). At these sites worm density was 6.33 ± 1.53 m-2 -2 Bacterial Analysis versus 0 to 1.33 ± 1.53 m at all other site and Mean MPN at high impact sites showed depth combinations (Fig. 3). larger amounts (2.30 ± 3.98), as compared to 0.44 ± 0.51 at low impact sites (Fig. 5), though 35 High Impact between sites, MPN did not show any 30 difference (p = 0.234). 25 Low Impact 7 20 15 6

10 5 % live coral cover 5 4 0 MPN 6 m 12 m 18 m 3

Figure 2. Mean percent live coral cover (± SD) among depths (6, 2 12 and 18 m) and between impact level (Impact: p = 0.385, 1 Depth: p = 0.011, and Impact x Depth: p = 0.793). P - values were determined using a two-factor ANOVA. 0 High Impact Low Impact 8 High Impact

) 7 Figure 5. Mean Most Probable Number (MPN) of Enterococcus -2 6 Low Impact spp. (± SD) between impact level (p = 0.234). P - value was 5 determined by a one-tailed t-test. 4 3 Density (# m Sediment Analyses 2 Between site impact sedimentation rate did 1 0 not show any difference (p = 0.172) with mean -1 6 m 12 m 18 m sedimentation rate of 0.09 ± 0.09 g day at high impact sites and 0.07 ± 0.10 g day-1 at low Figure 3. Mean density of S. giganteus (± SD) among depths (6, impact sites (Fig. 6). However, particle size 12 and 18 m) and between impact level (Impact: p = 0.020, Depth: p = 0.001, and Impact x Depth: p = 0.003). P - values distributions showed differences as a result of were determined using a two-factor ANOVA. site impact (p <0.001). More fine grained particles (< 50 µm) were found at high impact Water Chemistry sites than coarser grained particles (> 50 µm), Water chemistry analyses at all sites never which were primarily found at low impact sites showed nitrites or nitrates. Ammonia was (Fig. 7). found at 0.01 ± 0.01 ppm at high impact sites and 0.22 ± 0.39 at low impact sites. Phosphate

62

Physis: Journal of Marine Science )

-1 0.2 the worms in the current study were found in live coral and because live coral cover was 0.16 similar between the two types of sites among 0.12 depths, availability of live coral habitat did not 0.08 appear to be a factor in explaining the higher 0.04 densities at 12 m at the high impact sites. Water chemistry analysis at 12 m showed Sedimentation rate (g day Sedimentation 0 High Impact Low Impact elevated phosphate levels at high impact sites could contribute to increased worm densities. Figure 6. Mean sedimentation rate in g day-1 (± SD) (p = 0.172). Bizsel et al. (1999) related levels of P - value was determined by a one-tailed t-test. phytoplankton with phosphate in Izmir Bay, Aegean Sea. In spring, as phytoplankton 140 High Impact levels spiked, phosphate concentrations 120 decreased as phytoplankton consumed the 100 Low Impact available phosphorous, removing it from the system. McGlathery et al. (1994) found that 80 there is a positive feedback in phosphorous 60 availability in Bermuda; the percent of 40 .... # of particles phosphorous biologically available for uptake 20 by organisms such as seagrass (Thalassia 0 testudinum) increased as phosphorous loading < 10 > 10-50 > 50-100 >100-250 > 250-500 > 500 itself increased. Furthermore, sites with higher rates of eutrophication had higher levels of Particle size (µm) phosphorous (McGlathery et al. 1994). Such Figure 7. Mean number of particles (± SD) distributed among increased levels of phosphorous most particle class sizes (µm) and between impact level. Particle sizes < 50 µm occurred more frequently at high impact sites (Impact: commonly stem from groundwater seepage p < 0.001, Depth: p = 0.802 and Impact x Depth: p = 0.713). (Szmant 2002), a possible threat to Bonaire’s Particle sizes > 50 µm occurred more frequently at low impact coral reefs because of its influence on presence sites. P - values were determined using a two-factor ANOVA. of alga (S. Patton, personal communication).

In Brazilian reefs, those surrounded by intense Discussion development and influx of nutrients than reefs

with less anthropogenic influence showed 2 X The significantly greater density of worms the macroalgal cover (Szmant 2002). at 12 m at high impact sites than any other site It is also possible that Christmas tree and depth combination supports the original worms in Bonaire have developed biological hypothesis. It suggests that there are adaptations to contaminants such as excess environmental factors at the high impact sites nutrients. S. giganteus are considered to be a contributing to the survivorship of these slow growing worm, with a 15 – 20 year life worms that are not present at the low impacted span in natural populations (Toonen 2002). sites and depths. This study is novel for Wu et al. (2005) showed that many stress Bonaire because it could lead to the use of responses in organisms may decline as they biota as a new assessment tool for determining adapt to a threat. Christmas tree worms at stress on coral reefs. these high impact sites may have adapted to Previous research has shown that the increased level of nutrients and are now Spirobranchus giganteus prefer to settle on thriving, resulting in an increase in their live hard corals because this provides them density. with a better chance of survival; dead coral In addition to runoff as a source of colonies support ~ 18 species of boring phosphorous in the marine environment, it is invertebrates whereas live coral colonies also excreted by bacteria in a dissolved, support ~ 3 (Toonen 2002). When S. organic form (Barsdate et al. 1974). The giganteus inhabit live coral the likelihood of increased levels of enteric bacteria at the high penetration of their calcareous tubes and impact sites may be considered a contaminant therefore attacks from predators, pathogens of the water and could be linked to the increase and other additional stressors is reduced of phosphorus at those sites. Although (Toonen 2002). Additionally, nearly 100 % of systems with filter-feeding organisms such as

63

Physis: Journal of Marine Science

Christmas tree worms show considerably faster Acknowledgements phosphorus cycling when compared to non- ciliate grazed systems, very little is absorbed Thanks to Dr. Amanda Hollebone for by the Christmas tree worms (Barsdate et al. advising throughout the research process. 1974). However, under conditions with these Thanks to Mollie Sinnott, Lauren Saulino and levels of phosphorous, an increase in the Jill Lau for help with collecting data. metabolism of bacteria in the water column is observed (Barsdate et al. 1974). Therefore, References because the bacteria excrete phosphorous, this could explain the increased levels at high Barsdate, R. J., R. T. Prentki, and T. Fenchel. 1974. impact sites. Phosphorus cycle of model ecosystems: Sedimentation rates did not show any Significance for decomposer food chains and significant difference between sites but there effect of bacterial grazers. Nordic Society were finer grained particles at the high impact Oikos 25:239-251. Ben-Tzvi, O., S. Einbinder, and E. Brokovich. versus low impact sites. The ciliary-mucus 2005. A beneficial association between a feeding process used by S. giganteus enables polychaete worm and a scleractinian coral? this species to elaborately sort particles based Coral Reefs 25:1. on size and density, favoring items such as Bizsel, N., H. A. Benli, and K. C. Bizsel. 1999. A phytoplankton (Pagliosa 2005). The highly synoptic study on phosphate and branched radioles lining the stalk pull water phytoplankton in the hypereutrophicated Izmir through the tentacle crown, catching food on bay (Aegean Sea). Tubitak 25:89-99. the tentacles which are moved to a food groove Bonaire National Marine Park Management Plan. in the center where mucus transports the food 2006. STINAPA Bonaire. to the mouth (Pagliosa 2005). Christmas tree Burke, L. and J. Maidens. 2004. Reefs at Risk in the Caribbean. World Resources Institute, worms generally feed on aggregates of Washington, D.C. phytoplankton such as unicellular algae and Burkhart, J. G., G. Ankley, H. Bell, H. Carpenter, detrital floc and larger food items like D. Fort, D. Gardiner, H. Gardner, R. Hale, J. C. zooplankters as they mature (Pagliosa 2005; Helgen, P. Jepson, D. Johnson, M. Lannoo, D. Toonen 2005). The ciliary reversal capture Lee, J. Lary, R. Levey, J. Magner, C. Meteyer, mechanism is thought to be used by Christmas M. D. Shelby, and G. Lucier. 2000. Strategies tree worms to capture food particles by the for assessing the implications of malformed reversed beat of cilia in the ciliated band where frogs for environmental health. Environmental food pieces are confined in the water currents Health Perspectives 108:83-90. passing over the band (Hart 1996). As a result Cooper, T. F., J. P. Gilmour, and K. E. Fabricius. 2009. Bioindicators of changes in water quality of this feeding method, small particles move on coral reefs: Review and recommendations faster in the water layer along the cilium tips of for monitoring programs. Coral Reefs 28:589- the radioles (Hart 1996). This suggests that 606. smaller particles may be easier to collect from Cooper, T. F., P. V. Ridd, K. E. Ulstrup, C. the water column, therefore making sites with Humphrey, M. Slivkoff, and K. E. Fabricius. greater abundances of small particles a more 2008. Temporal dynamics in coral preferable habitat for S. giganteus. bioindicators for water quality on coastal coral There are a number of discrepancies reefs of the Great Barrier Reef. Marine concerning the use of bioindicators in marine Freshwater Research 59:703-716. systems including the limitations of the Dodge, R. E. and T. R. Gilbert. 1984. Chronology of lead pollution contained in banded coral organism and the importance of understanding skeletons. Marine Biology 82:9-13. the details of temporal changes of biological Gleason, D. F. 1998. Sedimentation and responses before their use as bioindicators (Wu distributions of green and brown morphs of the et al. 2005). With this understanding and Caribbean coral Porites asteroides Lamarck. further research identifying the coral species Journal of Experimental Biology and Ecology the worms inhabit, S. giganteus could act as 230:73-89. another component of the CIEE monitoring Hart, M. W. 1996. Deconstructing suspension program as a bioindicator of environmental feeders by analysis of film and video. stressors for coral reefs in Bonaire as well as Invertebrate Biology 115:185-190. the Caribbean. Hinrichsen, D. 2005. Coral reefs in crisis. BioScience 47:629-638.

64

Physis: Journal of Marine Science

Humann, P. and N. Deloach. 2003. Reef Creature Rini, A. 2008. Is #2 the number one problem in Identification. New World Publications, Inc., Bonaire? An examination of fecal Jacksonville, Florida. contamination and sedimentation from runoff. Hunte, W., B. E. Conlin, and J. R. Marsden. 1990. Physis 4:25-29. Habitat selection in the tropical polychaete Sower, S. A., K. L. Reed, and K. J. Babbitt. 2000. Spirobranchus giganteus. Marine Biology Limb malformations and abnormal sex 104:1. hormone concentrations in frogs. Lange, R. L., S. R. Scott, and M. Tanner. 1996. Environmental Health Perspectives 108:1085- Biomonitoring. Water Environment Federation 1090. 68:801-818. Smith, S. V. and R. W. Buddemeier. 1992. Global Marsden, J. R. 1987. Coral preference behavior by change and coral reef ecosystems. Annual planktotrophic larvae of Spirobranchus Reviews 23:89-118. giganteus corniculatus (Serpulidae: Szmant, A. M. 2002. Nutrient enrichment on coral Polychaeta). Coral Reefs 6:1-4. reefs: Is it a major cause of coral reef decline? McGlathery, K. J., R. Marino, and R. W. Howarth. Estuaries 25:743-766. 1994. Variable rates of phosphate uptake by Toonen, R. 2002. Aquarium invertebrates: shallow marine carbonate sediments: Christmas tree worms. Advanced Aquarist’s Mechanisms and ecological significance. Online Magazine 1:1-5. Biogeochemistry 25:127-146. Wu, R. S. S., W. H. L. Siu, and P. K. S. Shin. 2005. Pagliosa, P. R. 2005. Another diet of worms: The Introduction, adaptation and recovery of applicability of polychaete feeding guilds as a biological responses: Implications for useful conceptual framework and biological environmental monitoring. Marine Pollution variable. Marine Ecology 26:246-254. Bulletin 51:623-634 Patton, S. Personal communication. 18 September 2009. CIEE Research Station Bonaire, Bonaire, Netherlands Antilles. . Read, J. L. 1998. Are geckos useful bioindicators of air pollution? Oecologia 114:180-187.

65

Physis: Journal of Marine Science

Determining how coral reef habitat structure correlates with fish species richness at six dive sites in Bonaire, N.A.

Carolina Yanson Indiana University

Abstract Biodiversity of coral reef fish species is often related to the structural complexity and diversity of their habitats. This study explores the relationship between fish species richness, habitat diversity (substrate diversity) and habitat complexity (rugosity). Habitat diversity and topographic measures were used to predict reef fish diversity. It was hypothesized that high fish species diversity would show a positive correlation with greater habitat structure, which includes habitat diversity and topographic complexity. Fish species richness was determined at six dive sites in Bonaire, N.A. (Karpata, Andrea II, Cliff, Windsock, Angel City, and Red Slave) using data from 20 randomly chosen expert-level surveys provided by Reef Educational Environmental Foundation (REEF) for 2004 – 2009. Preliminary analysis of REEF data was used to select sites with high and relatively low fish species richness to make comparisons with the habitat structural complexity measurements (substrate diversity and rugosity). Substrate diversity and habitat complexity were measured using a 10 m transect randomly placed at 4 depths (2, 6, 12, and 18 m) at each site. Substrate diversity was determined by measuring the percent cover of the different substrates and then using the Shannon Diversity Index to determine H’. The rugosity of the sample area was measured by fitting a lead line to the reef at each of the determined depths. Overall results suggested that topographical complexity (rugosity) was not related to high fish species richness at dive sites on Bonaire. There was a weak positive correlation between H’ and fish species richness on the reef slope and a weak negative relationship between H’ and fish species richness on the reef flat. The results provide evidence that there are more factors to consider when explaining fish species richness on coral reefs than the structural complexity of the habitat at the scale of this study.

Introduction structurally complex habitats and the associated species are important to the overall The biodiversity of organisms is often health of the oceans, especially in Bonaire related to the structural complexity and where a major source of income from tourists diversity of the habitat. Habitat structure has and divers is very important to the island been defined as the physical and biological economy (Sandin et al. 2008). If reef habitats nature of the substratum (Jones 1991) and disappear, evidence suggests that associated includes both topography and substrate species will be lost as species show diversity. Topography is a measure of preferences towards particular environments complexity of the habitat structure, or a (Friedlander and Parrish 1998; Steele 1999; measure of the structure’s rugosity. Substrate Gratwicke and Speight 2005). In order to diversity is a measure of the diversity of the protect the rich fish diversity and economic habitat structure, or H’. Coral reefs are highly well-being of Bonaire, it is important to complex ecological systems, where multiple understand the relationship between habitat processes interact to create large assemblages diversity and complexity and the richness of of organisms (MacNeil et al. 2009). In this coral reef fish. study, fish species richness, or the number of Fish species, like terrestrial organisms, rely organisms that are found within a habitat was upon particular habitats for survival (Mumby calculated and compared to substrate diversity et al. 2003). Diverse characteristics of reef and rugosity to determine their relationship. habitats may directly influence the distribution Marine habitats of lower species richness, and size of species populations including such as those impacted by negatively by biological interactions such as predation and human interaction are rarely ever transformed competition (Hixon and Menge 1991; Jones back into high diversity habitats unless and Syms 1998). Diverse habitats are defined massive action is taken to salvage the damaged in this study as habitats with varying substrate ecosystem (Hewitt et al. 2008). Diverse and types. It is known that the loss of habitat [email protected] 66

Physis: Journal of Marine Science diversity leads to significant changes in species complexity and fish species richness has abundance and richness (Airoldi et al. 2008). received little attention in the past (Willis et al. Evidence suggests that diverse habitats 2005). The substrate diversity and increase reef species diversity (Abele 1974) as topographical complexity of the habitat in this fish species rely on diverse substrate and study is tested by measuring the types and habitats throughout different stages of their amounts of substrates at a dive site and the lives for reproduction, protection, and rugosity of the substrate. nourishment (Mumby et al. 2003). Structurally complex, or rugose reef Materials & Methods habitats also attract larger amounts of Fish Species Richness individual fish species and biomass in REEF is a non-for profit organization that comparison to reef flats where there is little collects surveys taken by experts and novices structural diversity (Friedlander and Parrish to monitor the population and diversity of fish 1998; Gratwicke and Speight 2005). Reefs species for specific dive sites globally (REEF with a variety of growth forms also increase 2009). Data is collected using a Roving Diver fish species richness and abundance Technique (RDT) wherein divers swim over (Gratwicke and Speight 2005). Fish species the reef at specific dive sites noting fish have shown higher assemblage rates where species and abundance observed. Surveys are reef substrate was more structurally and categorized by the experience of the surveyor; topographically complex (Friedlander and expert-level surveys were used exclusively for Parrish 1998), suggesting that fishes depend this study. Using expert-level surveys from upon a rugose environment for protection. For the REEF database Angel City, Red Slave, and example, gobies show higher recruitment and Karpata were selected as high fish richness survival rates when availability of shelter is sites whereas low levels of fish species increased through structural modifications richness occurred at Andrea II, Cliff, and (Steele 1999). As a result, local fish Windsock (Fig. 1). The high and low fish assemblages could be influenced by the diversity sites were originally selected based topography of their environment. In this study, the diversity and complexity of coral reefs along the leeward coast of Bonaire will be assessed to determine what effect they have on fish species diversity. The following hypotheses will be tested: H1: increases in fish diversity will be positively related to H’; and H2: increases in fish diversity will be positively related to topographic complexity. High and low fish species richness for six dive sites were determined using the Reef Environmental Educational Foundation (REEF) database by utilizing 20 individual surveys from 2004 – 2009 for each dive site. Analysis of how marine substrate diversity and topographical complexity affect fish species is lacking, primarily because these habitats are not inhabited by humans and are not easily accessible (Airoldi et al. 2008). This study is important as it will help to determine whether Figure 1. High and low fish diversity study sites on Bonaire, or not the loss of these diverse habitats could N.A., as estimated using REEF surveys. have a negative effect on their inhabitants. The purpose of this study is to better upon the following criteria: 1) there were understand if marine habitat diversity and between 100 – 150 expert level surveys total, complexity and fish species richness are 2) study sites were not adjacent to each other, positively correlated. This is ideal because the and 3) individual surveys were available from relationship between habitat structural the past decade. When more recent data were

67

Physis: Journal of Marine Science available, fish species richness was determined marked with a colored zip tie and measured for 20 randomly chosen individual surveys later. The mean of the rugosity measures from from 2004 – 2009. The mean was calculated four depths was estimated by dividing the to determine the recent fish species richness length of the lead line by the length of the for each site. transect to determine the rugosity value of each site. Substrate Diversity / H’ H’ for each site was determined by placing Data Analysis 4 transect lines, 10 m in length, at depths of 2, Data were separated between the reef slope 6, 12 and 18 m at each dive site. The depth (18 and 12 m depths) and the reef flat (6 and 2 range covered the typical depths sampled by m depths). The mean H’ and rugosity (± SD) REEF volunteers. At each site the mooring was determined for the reef slope and the reef was used as the center point of the sampling flat to compare high and low fish species area. The sampling area was 100 m wide, and diversity sites to one another. A t-test was run the starting point for transects was determined to determine if the mean H’ and the mean prior to sampling using a random number rugosity on the reef slope and the reef flat was generator. A number between 0 and 100 was statistically different. H’ and rugosity chosen, with 0 being 50 m north of the measurements were then compared to fish mooring and 100 being 50 m south. Transects species richness on the reef slope and the reef then ran south of the random number. The flat. substrate found directly underneath the transect line was categorized as live coral (LC), coral Results rubble (CR), hard bottom (HB), sand (SA), algae (AL), invertebrates (IT), and other Fish Species Richness (OTH). Substrate type was estimated to the Mean fish species richness was highest at nearest centimeter and the percent of each Red Slave (93.1 ± 10.8), followed by Angel substrate was calculated for each transect by City, Cliff, Karpata, Windsock and Andrea II dividing the total substrate cover by the length (Table 1). Using recent REEF surveys, fish of the transect. The percent of each substrate species richness at Cliff was unexpectedly type was determined for each site and these higher than fish species richness found at percentages were used to calculate H’ using Karpata (Fig. 3). Karpata had higher fish the Shannon Diversity Index. A higher H’ species richness when calculated using value demonstrates a higher level of substrate diversity and vice versa.

Habitat Complexity Habitat complexity was determined using a rugosity measure similar to Willis (2005). A lead line was fitted to the reef along the same 10 m transect that was used to determine habitat diversity (Fig. 2). The lead line was

Figure 3. Twenty REEF surveys were randomly selected from 2004 – 2009 to determine the mean of fish species richness (± SD) at each study site.

REEF’s 20 year database that was used initially to select high and low fish diversity dive sites. Rather than grouping the sites into high and low diversity, diversity measures were used directly for statistical and graphical comparisons among sites. Figure 2. Diagram of rugosity measured by fitting a lead line to the reef over a given transect length (10 m).

68

Physis: Journal of Marine Science

Table 1. Fish species richness, H’, and rugosity on the reef slope and the reef flat of six dive sites in Bonaire, NA. Reef Slope Reef Flat Mean Mean Mean Mean Richness H' Rugosity Mean H' Rugosity n = 20 SD n = 2 SD n = 2 SD n = 2 SD n = 2 SD Red Slave 93.05 10.82 1.41 0.06 1.31 2.58 1.15 0 1 0 Angel City 91.7 10.04 1.4 0.08 1.59 0.73 1.34 0.09 1.25 1.37 Karpata 84.8 10.78 1.37 0.06 1.82 0.73 1.43 0.07 1.3 3.13 Cliff 86.15 9.19 1.37 0.03 1.56 0.23 1.14 0.14 1.04 1.78 Windsock 77.75 18.36 1.38 0.03 1.56 0.11 1.37 0.15 1.18 0.32 Andrea II 72.2 12.61 1.31 0.07 1.4 0.19 1.3 0.29 1.14 0.54 .

Fish Species Richness and H’ 100 H’ was highest at Karpata on the reef flat (1.4 Reef Slope ± 0.1) and was lowest at Cliff on the reef flat Reef Flat (1.1 ± 0.1; Fig. 4). A t-test showed there was 90 no significant difference in mean H’ between the reef slope (1.4 ± 0.0) and the reef flat (1.3 ± 0.1; p = 0.061; Fig. 5). There was a weak negative relationship between H’ and fish 80 species richness on the reef flat (R2 = 0.013), Richness Species Fish and there was a weak positive relationship between H’ and fish species richness on the 70 0.511.52

2 Reef Slope H' Reef Flat Figure 6. Mean H’ on the reef flat (R2 = 0.085) and the reef 2 1.5 slope (R = 0.013) versus mean fish species richness at 6 dive sites in Bonaire, N.A.

1 reef slope (R2 = 0.085; Fig. 6). Mean H' 0.5 Fish Species Richness and Rugosity Karpata had the highest rugosity on both 0 the reef slope (1.8 ± 0.7) and the reef flat (1.3

Cliff ± 3.1); however, Karpata also had lower mean

Karpata fish species richness (84.8 ± 10.8) than Cliff Andrea II Andrea Windsock Red Slave Angel City Angel (86.2 ± 9.2), Angel City (91.7 ± 10.0), and Figure 4. Mean H’ (± SD) at each dive site on of the reef flat (2 Red Slave, the highest ranking site in fish m, 6 m) and reef slope (12 m, 18 m). species richness (93.1 ± 10.8), showed the 2 lowest substrate diversity and rugosity (1.0 ± 0.0) on the reef flat as well as the lowest 1.5 rugosity on the reef slope (1.3 ± 2.6). Cliff and Windsock had equivalent mean rugosity 1 (1.6 ± 0.2; ± 0.1) on the reef slope but

Mean H' different fish species richness, 86.2 ± 9.2 and 0.5 77.8 ± 18.4 respectively. The reef slope had significantly higher 0 rugosity than the reef flat (p < 0.001; Fig. 7). Reef Slope Reef Flat There was no relationship between rugosity and fish species richness on either the reef Figure 5. Mean H’ (± SD) of the reef flat (2 m, 6 m) and reef slope (R2 = 0.001) or the reef flat (R2 = 0.002; slope (12 m, 18 m)(t-test, p = 0.061). Fig. 8).

69

Physis: Journal of Marine Science

reef flat and a weak positive relationship 2 between H’ and fish species richness on the reef slope. There was no relationship between 1.5 fish species richness and rugosity of the reef flat or reef slope. This study provides evidence that fish species richness, when 1 estimated using REEF surveys, is not related

Mean Rugosity Mean to habitat diversity or rugosity at the scale 0.5 used in this study. REEF surveys contain information about all fish on the reef including 0 territorial and non-territorial species. Reef Slope Reef Flat Studies of the relationship between fish Figure 7. Mean rugosity (± SD) of reef slope and the reef flat species richness and H’ show varying results. habitats at 6 dive sites in Bonaire (t-test, p ˃ 0.001). For example, MacNeil et al. (2009) found that corallivores were most strongly associated

Reef Slope with higher H’, and the bulk of the analysis 100 found that habitat composition was the

Reef Flat primary determinant of fish species assemblages. This data suggests that particular species of fish may be more 90 influenced by the composition of their environment than others. Risk (1971) found that H’ was strongly correlated with fish species richness, and fish species richness 80 increased with depth in Roberts and Ormond Fish Species Richness (1987). While there is still conflicting findings among studies, the depths chosen in a study can impact the results. For example, the high fish species richness site, Red Slave, had 70 extremely low mean H’ at all four depths. 0.511.52 Mean H’ on the reef slope was relatively Rugosity average in comparison to other sites, whereas Figure 8. Mean rugosity of reef flat (R2 = 0.002) and reef slope H’ at depths of 2 and 6 m were very low. In (R2 = 0.001) habitats versus fish species richness at 6 dive sites between 2 and 6 m there is a large reef flat in Bonaire. zone with greater H’ in particular areas that were not sampled. Perhaps if data were Discussion collected at eight depths instead of four, smaller differences in H’ between habitats The purpose of this research was to may be detected. examine two aspects of habitat complexity, H’ While several studies have shown evidence and rugosity, to determine how habitat that topographic complexity correlates with complexity correlates to coral reef fish species fish species abundance and diversity richness. A strong positive relationship was (Friedlander and Parrish 1998; Steele 1999; expected based on other studies (Abele 1974; Gratwicke and Speight 2005), not all of these Friedlander and Parrish 1998). Overall, there studies have used the same measurement of was no statistically significant relationship rugosity. For example, in Gratwicke and between fish species richness and H’ or Speight’s (2005) study rugosity is measured between fish species richness and rugosity. on a simulated reef with simulated structural However, when sites were examined by complexity. As the area studied was a separating reef flat from reef slope simulation, one must take into consideration measurements, there was a significant all the possible factors that influence fish difference in H’ and rugosity when comparing species abundance on natural coral reefs such the reef slope to the reef flat. Additionally, as the abundance of predators, invasive there was a weak negative relationship species, and competitors that may not be as between H’ and fish species richness on the abundant on a simulated plot. In Willis et al.

70

Physis: Journal of Marine Science

(2005) rugosity was measured with a chain of H’ and rugosity on fish species richness. instead of a lead line. Though measurements This research is important because diversity is were only taken in 1 m sections instead of positively related to reef resilience. entire 10 m transects, results still indicated a positive correlation between fish species Acknowledgements assemblages during day and night and rugosity. However Roberts and Ormond I would like to thank my advisors Dr. Rita (1987) found that topographical complexity Peachey and Kate Jirik for their advice and has no significant effect on fish species unending support, Alyssa Adler and Grant richness. Karpata, the study site with the Frank for their patience and support for my highest rugosity on both the reef slope and the project, and of course Caren Eckrich for reef flat, had the fourth lowest mean fish always allowing me to dive and training me so species richness overall. This site is highly well. I would also like to thank REEF and topographically complex because it has a Christy Semmens for providing me with the shallow coral reef on the reef flat and highly fish survey data and all of the expert REEF complex coral structures on the reef slope. surveyors who made this study possible. Karpata’s low fish species richness in surveys collected from the past 5 years could have References been influenced by hurricane Omar, a large hurricane that came within 100 miles of the Abele, L. G. 1974. Species diversity of decapods leeward coast of Bonaire in 2008. Some dive crustaceans in marine habitats. Ecology sites are located leeward of Klein Bonaire and 55:156-161. may have been more protected from the wave Airoldi, L., D. Balata and M. W. Beck. 2008. The energy whereas Karpata was exposed to the Gray Zone: Relationships between habitat loss and marine diversity and their applications in waves from the south west, however Aronson conservation. Journal of Experimental Marine (1992) found that hurricanes had little effect Biology and Ecology 366:8-15. on tropical reef fish species and Red Slave had Alvarez-Filip, L., N. K. Dulvy, J. A. Gill, I. M. the highest fish diversity and was also exposed Cote and A. R. Watkinson. 2009. Flattening of to the storm waves. Windsock had high Caribbean coral reefs: region-wide declines in rugosity on the reef slope and the reef flat, but architectural complexity. Proceedings of the had the fifth lowest mean fish species richness, Royal Society of London 276:3019-3025. a possible result of a lack of fish species site Aronson, R. B. 1992. The effects of geography and fidelity, or low habitat substrate diversity. hurricane disturbance on a tropical predator- While this study focused on a small number prey interaction. Journal of Experimental Marine Biology and Ecology. 162:15-33. of sites, the result suggests that H’ and Friedlander, A. M. and J.D. Parrish 1998. Habitat rugosity does not influence fish species characteristics affecting fish assemblages on a richness. Perhaps fish species richness at Hawaiian coral reef. Journal of Experimental varying depths should be taken into Marine Biology and Ecology 224:1-30. consideration, and measured directly instead Gratwicke, B. and M. R. Speight. 2005. Effects of of using the REEF database. Fish species habitat complexity on Caribbean marine fish richness has been shown to increase with assemblages. Marine Ecology Progress Series depth (Roberts and Ormond 1987), and may 292:301-310. be higher on the reef slope than the reef flat. Hewitt, J. E., S. F. Thrush and P. D. Dayton. 2008. This data suggests that focusing on one depth Habitat variation, species diversity and ecological functioning in a marine system. range, the reef flat or the reef slope would be Journal of Experimental Marine Biology and beneficial instead of finding a mean over a Ecology 366:116-122. large range of depths. Fish species richness Hixon, M. A. and B. A. Menge. 1991. Species can also be greatly influenced by the presence diversity: prey refuges modify the interactive of pelagic predators with few if any effects of predation and competition. boundaries or permanent territories, as well as Theoretical Population Biology 39:178-200. site specific species such as corallivores Jones G. P. 1991. Postrecruitment processes in the (MacNeil et al. 2009). Perhaps focusing on ecology of coral reef fish populations; a fish that have greater site fidelity would have a multifactorial perspective. The Ecology of different outcome. Modifications of methods Fishes on Gorat Reefs 294-328. may be necessary to estimate the importance

71

Physis: Journal of Marine Science

Jones, G. P. and C. Syms. 1998. Disturbance, Risk, M. J. 1971. Fish diversity on a coral reef in habitat structure and the ecology of fishes on the Virgin Islands. Department of Biological coral reefs. Australian Journal of Ecology Sciences 1-6. 23:287-297. Roberts C. M., and R. F. G. Ormond. 1987. Habitat MacNeil, M. A., N. A. Grahm, A. Polunin, V. C. complexity and coral reef fish diversity and Nicholas, M. Kulbicki, R. Glazin, M. abundance on Red Sea fringing reefs. Marine Harmelin-Vivien and S. P. Rushton. 2009. Ecology Progress Series 41:1-8. Hierarchial drivers of reef-fish Sandin, S. A., E. M. Sampayo and M. J. Vermeij. metacommunity structure. Ecology 90:252- 2008. Coral reef fish and benthic community 264. structure of Bonaire and Curacao, Netherlands Mumby, P. J., A. J. Edwards, J. Arias-Gonzalez, K. Antilles. Caribbean Journal of Science C. Lindeman, P. G. Blackwell, A. Gall, M. L. 44:137-144. Gorczynska, A. R. Harborne, C. L. Pescod, H. Steele, M. A. 1999. Effects of shelter and predators Renkmen, C. C. Wabnitz and G. Llewellyn. on reef fishes. Journal of Experimental Marine 2003. Mangroves enhance the biomass of coral Biology and Ecology 233:65-79. reef fish communities in the Caribbean. Nature Willis, S. C., K. O. Winemiller and H. Lopez- 427:533-536. Fernandez. 2005. Habitat structural REEF. 2009. Reef Environmental Education complexity and morphological diversity of fish Foundation. World Wide Web electronic assemblages in a neotropical floodplainriver. publication. www.reef.org, 29 October 2009. Oecologia 142:284-295.

72

Physis: Journal of Marine Science

STUDENT PROFILES

ALLYSA MOLLIE ADLER SINNOTT Oregon State Wake Forest University University Biology Biology Sisters, OR Pittsburg, PA

GRANT MAGGIE MARSHALL THOMAS Trinity College FRANK Environmental Science Colorado College & Biology Biology Sacramento, CA Middleton, WI

NOELLE CHELSEY HAWTHORNE WEATHERSBEE Bucknell University Wofford College Biology Biology Oxford, CT Florence, SC

ALISON PAMELA MASYR WILLIAMS Oberlin College University of Colorado Biology at Boulder Brooklyn, NY Ecology & Evolutionary Biology Menlo Park, CA

AURORA CAROLINA SCHRAMM YANSON Eckerd College Indiana University Marine Science Biology Franklinville, NJ Ogden Dunes, IN

73

Physis: Journal of Marine Science

CIEE FACULTY

Dr. Rita Peachey is the co-instructor for Independent Research and the Resident Director of CIEE Research Station Bonaire. She earned her B.S. in Biology and M.S. in Zoology from the University of South Florida and her Ph.D. in Marine Science from the University of South Alabama. Her research interests include coral biology and how UV affects the early stages of life in the ocean. In addition, she has studied how pollution can enhance the detrimental effects of sunlight on larval crabs, corals and oysters. RITA PEACHEY, PH.D. Primary advisees: Alyssa Adler, Alison Masyr, Aurora Resident Director Schramm, Chelsey Weathersbee, and Carolina Yanson

Dr. Amanda Hollebone is the co-instructor of Independent Research and Marine Conservation Biology faculty at CIEE Bonaire. She received her B.S. in Biology from the UNC Chapel Hill, a Ph.D. in Marine Ecology from Georgia Tech and taught in the Biology Department at Georgia Southern University. Amanda’s research interests include reef community ecology and invasive species. Her dissertation research focused on the population dynamics and pre- and post- settlement ecology of a non-native porcelain crab in the oyster AMANDA HOLLEBONE, PH.D. reefs of Georgia, USA. Tropical Marine Conservation Primary advisees: Grant Frank, Noelle Hawthorne, Mollie Biology faculty Sinnott, Maggie Thomas, and Pamela Williams

Caren Eckrich is the Assistant Resident Director and the Dive Safety Officer for CIEE. She holds a B.S. in Wildlife and Fisheries from Texas A&M University and a M.S. in Biological Oceanography at the University of Puerto Rico in Mayaguez. Caren manages dive planning for the student independent research projects and has a wealth of local experience on the reefs that contributes to the success of student projects. Caren’s research interests include fish behavior, seagrass and algal ecology, sea turtle ecology, and CAREN ECKRICH coral disease. Assistant Resident Director

Anouschka van de Ven is the Technical Coordinator for CIEE. She is a PADI Dive Instructor and underwater photographer/videographer. She has a B.A., First Class Honours Degree in Communications Studies, from the London Metropolitan University and worked in television and advertising in Amsterdam before moving to Bonaire. Anouschka provides administrative support for student projects and participates in the CIEE long-term research program in Bonaire. ANOUSCHKA VAN DE VEN Technical Coordinator

74

Physis: Journal of Marine Science

CIEE INTERNS

Lauren Saulino assisted Dr. Amanda Hollebone’s advisees with their Independent Research projects. She holds a M.A. in Environmental Sciences from Miami University and a B.A. in Biological Sciences and Spanish from Clemson University.

LAUREN SAULINO Conservation Biology Intern

Kate Jirik assisted Dr. Rita Peachey’s advisees with their Independent Research projects. She holds a M.S. in Biology from California State University, Long Beach, and a B.A. in Biology from the University of California, Santa Barbara.

KATE JIRIK Coral Reef Ecology Intern

Jill Lau assisted with two of Dr. Amanda Hollebone’s advisees with their Independent Research projects. She holds a B.S. in Biology from the University of Maryland.

JILL LAU Volunteer Intern

75

DROB