cw invert and 3d thesis4 klm

104
Role of surface characteristics and mobile invertebrates on settlement choices and post- settlement survivorship of three scleractinian coral species Christopher Michael White Student #: 10406689 July 30 th , 2014 University of Amsterdam

Upload: christopher-white

Post on 17-Aug-2015

43 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: CW Invert and 3D Thesis4 KLM

Role of surface characteristics and mobile invertebrates on settlement choices and

post- settlement survivorship of three scleractinian coral species

Christopher Michael White Student #: 10406689

July 30th, 2014

University of Amsterdam

Faculty of Science: Master Biological Sciences

Track: Limnology and Oceanography

Supervisor: Dr. Kristen L. Marhaver

Examiner: Dr. Mark J.A. Vermeij

Page 2: CW Invert and 3D Thesis4 KLM

Table of Contents

Thesis ABSTRACT: 3

3D printing and settlement preference of two brooding scleractinian corals, a novel approach to rethinking artificial reef building

ABSTRACT: 4INTRODUCTION: 4METHODS: 7LARVAE COLLECTION 7SETTLEMENT TILES 8CORAL LARVAE COLOR CHOICE EXPERIMENT I 9CORAL LARVAE COMPLEXITY CHOICE EXPERIMENT II 10DATA ANALYSIS 10RESULTS: 11CORAL LARVAE COLOR CHOICE EXPERIMENT I 11COMPLEXITY CHOICE EXPERIMENT II 13DISCUSSION: 15COLOR PREFRENCE 15COMPLEXITY PREFERENCE 17IMPLICATIONS IN ARTIFICIAL REEF RESTORATION 18REFERENCES: 19

The ability of mobile reef invertebrates to influence coral settler survivorship

ABSTRACT: 22INTRODUCTION: 22METHODS: 25RESEARCH LOCATION 25STUDY SPECIES 26EXPERIMENTAL DESIGN 1-6 27-32DATA ANALYSIS 35RESULTS: 36EXPERIMENTS 1,2: M. FAVEOLATA RESPONSE TO CONDITIONED SEAWATER AND SURFACES 36-38EXPERIMENT 3: M. FAVEOLATA SURVIVORSHIP WITH VARYING INVERTEBRATES 39EXPERIMENT 4: F. FRAGUM SURVIVORSHIP WITH VARYING INVERTEBRATES 41EXPERIMENT 5: A. HUMILIS SURVIVORSHIP WITH VARYING INVERTEBRATES 42EXPERIMENT 6: FIELD STUDY ON INVERT. REEF TRAFFIC & F. FRAGUM SETTLER SURVIVORSHIP 44 DISCUSSION: 47SETTLEMENT OF CORAL LARVAE: 47INVERTEBRATE INTRODUCTIONS: 49

DECAPODS AND POLYCHAETES 49GASTROPODS 51ECHINODERMS 52

FIELD EXPERIMENT ON REEF: 53CONCLUSION: 56REFERENCE: 57ACKNOLOGEMENTS 62

2

Page 3: CW Invert and 3D Thesis4 KLM

APPENDIX PILOT STUDIES 7 & 8 62

Abstract:

Coral reefs, while occupying 1% of the oceans, provide habitat for a quarter of all

sea life. Scleractinian corals in particular build the foundation of the majority of structural

habitats found on reefs and are often referred to as ecosystem engineers (Williams et al.

2009). Their life histories, overall health, as well as their responses to biotic as well as

abiotic factors are therefore fundamentally important in our overall understanding of

coral reef systems.

Recently, coral reef health and recovery has been found to be particularly

dependent on coral larvae availability, settlement ecology, and post settlement

survivorship (Williams et al. 2009). Investigating the early life stages of various

scleractinian coral species, especially larval settlement preferences and post settlement

biological interactions, may therefore contribute to the foundation of our knowledge

regarding these systems. This report therefore focuses on two principle areas of research.

One, using innovative 3D technology to print various surface types in numerous colors,

surface and color settlement preferences of two scleractinian coral species, Favia fragum

and Agaricia humilis were investigated. The second investigation involved mapping the

interactions of various coral reef mesograzers with settlers of the coral species Favia

fragum, Agaricia humilis, and Montastraea faveolata, in an attempt to deduce the

invertebrate ecological influence on early life stages of those corals.

3

Page 4: CW Invert and 3D Thesis4 KLM

3D printing and settlement preference of two brooding scleractinian corals, a novel approach to rethinking artificial reef buildingAbstract:

Corals are subject to various pressures throughout their life histories that can

contribute to mortality, particularly during early life stages. Newly released coral larvae

handle these pressures by settling in favorable locations, taking into account factors such

as light, location, sedimentation, depth, and biotic competition. This study was conducted

in the southern Caribbean, on the leeward side of Curacao. Using 3D printing technology,

the settlement surface preferences for both color and surface complexity were determined

for larvae of two scleractinian coral species. Two brooding coral species were tested,

Favia fragum and Agaricia humilis. Both exhibited settlement preferences for white tiles

when presented color choices, however A. humilis showed a broader range of surface

preferences settling on both red and black tiles. Surface complexity played less of a role

in larvae settlement preferences for both species, while it is thought that larvae from both

species were more interested in settling relative to light exposure and not complexity

alone. These findings show that larvae from these two species choose surfaces that most

closely resemble the natural habitat in which they are prone to survive. A novel approach

for artificial reef building may be feasible by mimicking natural conditions for

scleractinian corals with the advent of 3D printing technology in future projects.

4

Page 5: CW Invert and 3D Thesis4 KLM

Introduction:

Coral larvae settlement cues and settlement surface preferences are the subject of

increasing research attention. Factors such as substrate orientation, chemical cues,

progeny distance from adults, and light have all been shown to influence larvae

settlement preferences and survivorship (Marhaver et al. 2013, Mason et al. 2011,

Harrington et al. 2004, Vermeij and Sandin 2004, Carlton and Sammarco 1987). In adult

corals, light is known to affect tentacle expansion and contraction, the regulation of

circadian clocks, and reproductive timing (Mason et al. 2012). Most recently,

photoreceptors in coral species were found to be a factor in coral larvae settlement

preferences. Mason et al. (2011) found using larvae of Porites astreoides and Acropora

palmata that chemical cues may be primarily responsible for settlement of larvae, but that

color also acts as a settlement cue. Notably, the color preference of Porites astreoides and

Acropora palmata was absent when the same experiment was conducted in the dark

(Mason et al. 2011). Mason et al. (2012) recently found genes in Acropora palmata that

code for acropsins, amino acid sequences similar to opsin proteins, which are photo

pigments known to occur in animals. These specific acropsins were found to align most

closely with rhodopsins of Homo sapiens and the veined squid Loligo forbesi. The ability

of juvenile corals to detect color could enable them to distinguish between different

species of crustose coralline algae (CCA), facilitating survival of coral settlers (Mason et

al. 2011).

Our knowledge of larval settlement surface preference and post-settlement

survivorship is important in understanding coral recruitment rates and patterns in general.

Additionally, as coastal development accelerates in the form of road systems, housing,

5

Page 6: CW Invert and 3D Thesis4 KLM

agriculture, and industry, disturbances such as sedimentation events, eutrophication

through waste effluent, and direct habitat destruction are likely to become more regular.

Disturbances like these may lead to changes in larval settlement patterns and overall

success. Due to reductions in overall reef system health and disproportionately large

disturbances on shallow reef systems, and given that shallow reef systems are

increasingly of interest for local tourism industries, understanding settlement preferences

of shallow water species may yield important insights in the functioning of reef

ecosystems.

Favia fragum is a scleractinian hermaphroditic coral that is limited in habitat

range, encountered mostly in the shallows of 3 m or less, attached to a variety of hard

surfaces and commonly found in the Caribbean (Goodbody and Putron 2009, Szmant-

Froelich et al. 1985). These shallow water habitats are often dominated by rubble and

prone to higher rates of disturbance from wave action and storms (Carlon 2002, Szmant-

Froelich et al. 1985).

Agaricia humilis, also a scleractinian coral commonly found in the Caribbean, can

be found in slightly deeper water, but is still limited for the most part to a depth of less

than 10 m, primarily occurring on shallow reefs and reef flats (Van Moorsel 1983).

Shallower depths of 10-20 m are known to have an increased level of disturbance in

comparison to similar landscapes at 30-40 m (Bak and Luckhurst 1980). This may have

contributed to the selective pressure for these two brooding coral species to planulate

throughout the year (Szmant-Froelich et al. 1985). Due to the fact that these coral species

recruit to an area with high levels of disturbance, it makes it that much more important

for larvae to settle on surfaces protected from disturbances.

6

Page 7: CW Invert and 3D Thesis4 KLM

Here, larvae from two species of scleractinian corals, F. fragum and A. humilis,

were given the opportunity to settle on artificial 3D-printed plastic surfaces of various

textures and colors. The aim was to test if larvae exhibited a preference for a specific

color and to determine which surface texture was preferred. 3D printing technology is

quickly being assimilated to support the activities and efforts of our daily lives. As costs

come down for printing of plastic parts, biological tissues, and even entire buildings,

technological developments are being realized more quickly. Therefore, in addition to

obtaining a better understanding of these two species’ settlement preferences, we show

how 3D printing technology may act as another tool in the near future for promoting

coral recruitment to artificial reefs within the shallow coral reef ecosystem.

Methods:

Larvae collection

38 colonies of Favia fragum and 23 colonies of Agaricia humilis were collected

from shallow reefs in Curacao at the Water Factory (12°6'37"N, 68°57'7"W) and

Piscaderabaai (12°7'20"N, 68°57'53"W) dive sites. Both collection sites are located on

the leeward, southwest coast of Curacao. Goodbody and Putron (2009) found over a four-

year period that F. fragum predictably begins releasing larvae three to four days after the

new moons between July and August, continuing for up to 19 days in total. Recorded

peaks in F. fragum planulation in Bermuda (Goodbody and Putron 2009) are correlated

with peaks in larval release in Puerto Rico and the US Virgin Islands (Carlon 2002,

Szmant-Froelich et al. 1985). In all three of these studies, larval release peaked six to

twelve days after the new moon. In this experimental time period for the current study,

peaks in F. fragum larval release in February and March of 2014 occurred on the 12th and

7

Page 8: CW Invert and 3D Thesis4 KLM

10th day following the new moon, respectively. Parent F. fragum colonies were collected

on February 5th 2014. A. humilis colonies were collected on March 5th 2014, with a larvae

release peak recorded the following day, five days after the new moon.

All corals were kept in the flow-through aquarium system at CARMABI. Each

coral was placed in an individual plastic beaker with a tube providing a continual flow of

fresh seawater. Water overflow ran onto a submerged 200 m filter (nylon mesh) held in

place using 6.35 cm diameter PVC piping. As larvae were released and floated to the

water surface, water flowed onto the submerged mesh for collection. Aquariums with

adult colonies were surrounded with shade cloth overnight to block out artificial light and

promote coral planulation. At sunrise, larvae were pipetted out of the filters into plastic

polystyrene containers filled with 0.45 m-filtered seawater (0.45 m FSW). Larvae

were pooled over multiple days of collection to obtain enough individuals for

experiments with sufficiently high replication. For each experiment, a mixed cohort of

one- and two-day-old larvae was used as daily larvae harvest rates varied. Larvae were

added to experimental treatments within two days of being collected.

Settlement tiles

Tiles were printed in white, black, and red plastic (Acrylonitrile butadiene

styrene; ABS) using an “UP!-Mini 3D printer” (Robotshop inc., Swanton/USA). Tiles

measured 4.15 cm in length by 4.15 cm in width with a thickness of 5 mm (Figure 1).

Features on the tiles added an additional maximum thickness of 1 cm. Tiles had a flat

surface faced upwards with four different surfaces on the underside: cones, bumps, flat,

and open cylinders. Tiles were designed using Blender software (Version 2.70, Blender

foundation). Four larger cones were located in the corners to act as support structures to

8

Page 9: CW Invert and 3D Thesis4 KLM

suspend the side with the complex printed

surfaces face down in the water column

(Figure 1).

Tiles were then secured, with textured

surfaces up, to 15 cm2 sections of white

plastic ceiling panels i.e., “egg crating”

(Plaskolite, Columbus/USA)

with plastic tie wraps. These panels were

then secured to the seabed in Piscaderabaai at a depth of five meters for five weeks. This

allowed for the development of biofilms and a small amount of colonization by CCA.

After five weeks, tiles were removed from the reef and lightly scrubbed with a test tube

cleaning brush to remove any large detritus and sediment.

Color choice (Experiment I)

Six clear plastic polystyrene containers with lids were filled with 180 ml of 200

m-filtered seawater (200 m FSW). 60 F. fragum coral larvae were added to each

container along with three tiles, one of each color, arranged haphazardly within each bin.

Larvae were released in the two days prior to beginning the experiment and pooled to

create an experimental cohort. The water in each container fully covered the 3D printed

tiles, which were negatively buoyant following light fouling by calcifying algae. Bins

were placed in the laboratory at CARMABI where temperature was maintained at 29˚C

with an alternating 12-hour light:dark regime. The light source consisted of overhead

fluorescent lights and diffused natural light from a strip of high windows. Tiles were

scored for settlers using a blue light and barrier filter system every 48 hr after initiation

Figure 1: Tiles design in Blender (View-Bottom face up)

9

Page 10: CW Invert and 3D Thesis4 KLM

until day 8. The blue lights used include a dive light with separate fluorescence excitation

filter, a BlueStar flashlight, or a Dual Fluorescent Protein Flashlight (all manufactured by

NightSea, Bedford/USA). Settlement was defined as the complete metamorphosis of

larvae into an oral-aborally flattened primary polyp with observable feeding tentacles and

secretion of a basal plate (Mason et al. 2011). Of the three colored tiles offered, the tile

color as well as specific location of each settler was scored as having settled on the top

flat surface, on the sides of tiles, or on the underside of the tile in one of the four sections

with different textures. The experiment was continued for eight days. Over the course of

the experimental period, the maximum number of settled larvae on each tile was

recorded. As larvae can only go through metamorphism once, are considered independent

from each other, and as some mortality occurred towards the end of the experimental time

period, this methodology was preferred as it gave each larva an equal choice opportunity.

The same experiment was repeated using the brooding species A. humilis

following the F. fragum experiment. Following obtaining results for F. fragum, juvenile

F. fragum settlers were scraped off the tiles using a razor blade and tiles were placed in

an aquarium where they were allowed to foul for an additional two weeks prior to the A.

humilis choice experiment. A. humilis colonies, collected on 5 March 2014, began

planulation immediately. Larvae collection methodology followed that for F. fragum.

Using the same tiles, the experiment was repeated with A. humilis using the same

methodology as for F. fragum.

Complexity choice (Experiment II)

Nine 235 ml clear plastic polystyrene cups with lids were filled with 60 ml 200

m FSW. A mixed cohort of larvae one to two days old was used. 30 F. fragum larvae

10

Page 11: CW Invert and 3D Thesis4 KLM

were added to each cup along with one white plastic tile, manufactured in the same

fashion as the other 3D settlement tiles. Tiles were scored for settlers using one of the

same blue light systems mentioned above every 48 hr for 6 days. The location of settlers

was recorded as having settled on the top flat surface, on the sides of tiles, or on the four

different textured areas on the underside of tiles. The methodology of taking the

maximum number of settlers over the experimental time period was maintained and the

experiment was repeated using A. humilis after tiles went through the same conditioning

process as the color choice experiment, following obtaining results for F. fragum.

Data analysis

After the eight day time period, the maximum number of settlers per tile was

recorded for each color and or complexity type. Aggregative coral settlement has been

observed to occur in P. astreoides and A. palmata, leading some researchers to consider

dishes as the smallest independent unit of measurement (Mason et al. 2011). However,

Gaines and Roughgarden (1985) described the settlement of barnacle larvae as a process

where settlement to a given area was proportional to the available space of that area. At

low densities, larvae settlement was determined to be independent of other larvae

settlement (Gaines and Roughgarden 1985). Thus, at low densities in this experiment

where we had a maximum density of one larva per 2 ml seawater, we assume settlement

of each larva to be independent.

For complexity and orientation, coral settlement location fell into four categories;

flat tops of tiles, sides of tiles, flat bottoms of tiles, and complex bottom areas of tiles.

Due to the varying size of each surface type available to larvae, total surface area of tiles

was calculated and sub-divided based on relative percentage of each respective area.

11

Page 12: CW Invert and 3D Thesis4 KLM

Results:

Color choice experiment (Experiment I)

In the color preference experiment, if the overall sum of settlers, aggregated

across all replicates is compared to the total number of coral larvae introduced (300),

only 16.5 percent of the Favia fragum introduced to settlement bins settled within the

eight-day allotted experimental time period. Of the 51 larvae that settled, settlement onto

white tiles was significantly higher than settlement on the other two colors when average

settlement across replicates were compared to expected settlement of a random

distribution across the three available tiles (Chi-square = 8.8748, DF = 2, P = 0.011827).

For A. humilis, a similar result was observed. The majority of settlement occurred on

white tiles (Chi-square = 7.0941, DF = 2, P = 0.028809). However, A. humilis exhibited a

more generalist behavior; a larger proportion of Agaricia humilis settlers were found on

red and black tiles as compared to F. fragum (Table 1).

12

Page 13: CW Invert and 3D Thesis4 KLM

1 2 3 4 5 60

10

20

30

40

50

60A.

Replicate

# of

set

tler

s

1 2 3 4 5 60

10

20

30

40

50

60B.

Replicate

# of

set

tler

s

White Black Red0

0.1

0.2

0.3

0.4C.

Color tiles

Pro

po

rtio

n o

f la

rvae

th

at s

et-tl

ed

Figure 2: A. A. humilis and B. F. fragum 3D tiles color choice settlement across

replicates. Key: White- settled on white tile, black- settled on black tiles, dark grey-

settled on red tiles, light grey- not settled. C. Average proportion of A. humilis (white)

and F. fragum (black) that settled each color tile. Standard error shown.

Table 1: Color settlement preference of two brooding scleractinian coral speciesAgaricia humilisTreatment White Black RedAverage 0.294444444 0.072222222 0.113888889Expected 0.33 0.33 0.33Standard deviation 0.144401703 0.0704483 0.0464479Standard error 0.058951748 0.028760398 0.018962276Chi-square = 7.0941, DF = 2, P = 0.028809.Favia fragumTreatment White Black RedAverage 0.119444444 0 0.022222222Expected 0.33 0.33 0.33Standard deviation 0.093936543 0 0.031031645Standard error 0.038349433 0 0.012668616Chi-square = 8.8748, DF = 2, P = 0.011827.

13

Page 14: CW Invert and 3D Thesis4 KLM

Complexity choice experiment (Experiment II)

Larval settlement choice for various surface complexity types followed similar

trends to what would be expected under a random distribution of the available space.

Although settlement was not significantly different than a random distribution,

differences between species are still visible. F. fragum preferred the light-exposed flat

surfaces of tiles to all other surfaces and showed minimal settlement on the flat bottom

(shaded) surfaces. A humilis preferred the cryptic shaded area over either of the flat light

exposed or flat shaded surfaces in relation to F. fragum (Figure 2).

Table 2: Color settlement preference of two brooding scleractinian coral species

Agaricia humilisTreatment Flat top Side Complex bottom Flat bottomAverage 0.159259259 0.1 0.207407407 0.02962963Expected 0.4 0.2 0.3 0.1Standard deviation 0.105116044 0.070710678 0.122222222 0.035136418Standard error 0.035038681 0.023570226 0.040740741 0.011712139Chi-square = 0.5426, DF = 3, P = 0.909427.Favia fragumTreatment Flat top Side Complex bottom Flat bottomAverage 0.27037037 0.122222222 0.140740741 0.003703704Expected 0.4 0.2 0.3 0.1Standard deviation 0.196104029 0.133333333 0.12886015 0.011111111Standard error 0.06536801 0.044444444 0.042953383 0.003703704Chi-square = 2.0523, DF = 3, P = 0.561621.

14

Page 15: CW Invert and 3D Thesis4 KLM

Flat top Side Complex bottom Flat bottom0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

Figure 3: A. humilis (black) and F. fragum (grey) 3D larvae settlement choice compared

to expected (white) normal distribution based off of available area of each complexity

type. A. humilis- Chi-square = 0.5426, DF = 3, P = 0.909427. F. fragum- Chi-square =

2.0523, DF = 3, P= 0.561621.

Discussion:

Color preference

The narrow ovulation of Favia fragum occurs within two to three days after the

new moon (Szmant-Froelich et al. 1985). This ability to time reproduction precisely using

available light yields insight into F. fragum’s capability as adult colonies to measure light

(Szmant-Froelich et al. 1985). Goodbody-Gringley (2010) additionally found that larvae

of F. fragum released right before sunrise exhibited a higher settlement rate than larvae

that were released just after sunset. This led Goodbody-Gringley to hypothesize that

newly released larvae with access to cues available under light conditions would be more

prone to locate suitable habitat for settlement quickly. This allowed them to conclude that

light plays a significant role in settlement behavior and success of F. fragum larvae

(Goodbody-Gringley 2010). Corals’ success in measuring light is not limited to F.

15

Page 16: CW Invert and 3D Thesis4 KLM

fragum. Miller and Mundy (2003) found similar results for broadcast spawning of

Platygyra daedalea with peaks in settlement corresponding to morning daylight hours.

F. fragum and A. humilis larvae had a distinct preference for white settlement

surfaces. However, the larvae of A. humilis exhibited a more generalist behavior, settling

on red and black tiles as well. Researchers have studied the connection between color

choice and settler survivorship in the past, with some attributing color preference for

settling on favorable surfaces such as CCA. Mason et al. (2011) found that Porites

astreoides and Acropora palmata, two species that often prefer to settle on CCA,

exhibited higher settlement on red surfaces than green or white surfaces. Vermeij and

Sandin (2004) found that higher survival of Siderastrea radians settlers was correlated

with proximity to CCA. Investigations of larval settlement and settler survival in the

Pacific coral species Acropora tenuis and A. millepora revealed that larvae preferred to

settle onto CCA, especially Titanoderma prototypum, and that this was correlated with

increased survivorship when compared to coral settlers on other species of CCA

(Harrington et al. 2004).

However, our results with two shallow water species F. fragum and A. humilis

show a significant preference for white tiles when given the choice between white, red, or

black (Figure 1). This may be explained by the typical habitat of F. fragum, which often

occurs in shallow reef flat habitats (Goodbody-Gringley 2010, Goodbody-Gringley and

Putron 2009, Reutter and Riggs 1985). These habitats are often composed of sand,

rubble, and solid limestone substrate, with minimal CCA coverage. Thus settlement

habitats of F. fragum are oftentimes white and not the pink-red of CCA that deeper reef

environments often harbor in greater amounts. As F. fragum has evolved to thrive in

shallow water environments, their larvae seem conditioned to use color as a tool for

16

Page 17: CW Invert and 3D Thesis4 KLM

seeking this habitat. Additional signaling cues may be at work when in-situ conditions are

present. When given the opportunity to settle on rubble from a similar depth habitat than

rubble from deeper origins, (Carlon 2002) found F. fragum preferred rubble from shallow

origins. Studies such as Carlon’s in conjunction with our color preference results show

how larvae play an active role in seeking ideal habitat, relying on a host of factors to

make settlement decisions.

As with F. fragum, A. humilis larvae also showed a preference for white tiles

when given the choice between white, red, and black. However, A. humilis exhibited a

broader settlement preference, settling on both black and red tiles. In previous studies, A.

humilis were shown to metamorphose at a higher rate when CCA was present, with

settlement almost entirely on the CCA surface when available (Morse et al. 1988). This

species has a greater depth range than F. fragum, a factor that could explain the broader

variety of surface colors to which A. humilis settled in our experiments. If we take into

consideration the previous color preference work of Mason et al. (2011) on Porites

astreoides and Acropora palmata, we observe how species show species-specific color

preferences relative to the characteristics of the habitats in which that species prefers to

settle and is most likely to survive.

Complexity preference

F. fragum settled more on flat, light exposed surfaces, than sides or undersides of

tiles (Table 2). Goodbody-Gringley (2010) also found that F. fragum larvae did not settle

on underside of settlement tiles when given the chance. Conversely, A. humilis showed a

high affinity for complex undersides of tiles (Table 2). In previous research, Morse et al.

17

Page 18: CW Invert and 3D Thesis4 KLM

(1988) also found differences in settlement preferences for exposed and cryptic habitats

for two Agaricia species.

Babcock and Mundy (1996) found that larvae of Platygyra sinensis, a shallow-

reef scleractinian coral species, as well as Oxypora lacera, a deep-reef coral species with

preference for shaded habitats, both exhibited settlement responses that depended on light

regimes. Both species settled underneath tiles when exposed to high light, and both

species settled on the tops of tiles when exposed to low light. As F. fragum and A.

humilis showed preferences for either tops or bottoms of tiles and no distinction between

the species noted for settlement on the sides of tiles, light exposure is likely the dominant

factor for where larvae chose to settle vs. complexity alone in this study (Table 2). Future

efforts should focus on designing 3D tiles that take light into consideration when

providing clear complexity choice variables to coral larvae.

Potential for application in artificial reef restoration

It is interesting to note from this experiment the differences between two species’

preferences for color and possibly light intensity. Additional work shows our results do

not stand alone, with numerous coral species exhibiting preferences in the planular stage

for specific color regimes. Considering these results and with the dramatic progress in 3D

printing technology, the implications for artificial reef restoration are numerous. Spieler

et al. (2001) highlighted two goals for artificial reef restoration: 1) to define the surfaces

to be used to accomplish the set out goals, and 2) to identify the interactions of reef biota

with this surface given its texture and orientation, and given the current state of the

environment. They additionally stated that the choice of artificial surface should take into

account texture, color, chemistry, profile, shelter, shading, size, configuration, stability,

18

Page 19: CW Invert and 3D Thesis4 KLM

settlement attractants, economy, and availability (Spieler et al. 2001). Effective design

therefore depends on local ecology. Carlton and Sammarco (1987) found corals in the

Great Barrier Reef generally preferred vertically oriented substrates, which could differ

from corals in the Caribbean or elsewhere in the world. Thus, if one application is found

to work in a given area, it may not necessarily be as successful when implemented

somewhere else.

However, in order to provide different coral species, which occur at different

depth ranges, with their preferred surface color, irradiance level, and surface settlement

cues, a successful artificial reef substrate could involve a multi-layered structure,

designed to attract larvae of different coral species over different depth ranges. Materials

used in 3D printing to date include plastics, biotic tissues, minerals, metals, and even

concrete. Thus, 3D printing of surface coatings, pier wrappings, or pilings themselves

could include differently-colored sections at different depths, along with variable surface

textures to promote colonization of the entire piling by different species. This concept

could also perhaps be applied to exposed and sheltered sides of reef breaks. As our

knowledge of coral ecology continues to expand, a single solution approach becomes

increasingly difficult to support, and only in our continual creativity, do we poses the

capacity to enact positive, well-thought-out artificial reef restoration projects.

References:

Babcock, R., and C. Mundy. 1996. Coral recruitment: consequences of settlement choice

for early growth and survivorship in two scleractinians. Journal of Experimental

Marine Biology and Ecology 206:179–201.

Bak, R. P. M., and B. E. Luckhurst. 1980. Constancy and change in coral reef habitats

19

Page 20: CW Invert and 3D Thesis4 KLM

along depth gradients at Curaqao. Oecologia 47:145-155.

Carleton, J. H., and P. W. Sammarco. 1987. Effects of substratum irregularity on success

of coral settlement: quantification by comparative geomorphological techniques.

Bulletin of Marine Science 40:85–98.

Carlon, D. B. 2002. Production and supply of larvae as determinants of zonation in a

brooding tropical coral. Journal of Experimental Marine Biology and Ecology

268:33–46.

Gaines, S., and J. Roughgarden. 1985. Larval settlement rate: a leading determinant of

structure in an ecological community of the marine intertidal zone. Ecology

82:3707-3711.

Goodbody-Gringley, G. 2010. Diel planulation by the brooding coral Favia fragum

(Esper, 1797). Journal of Experimental Marine Biology and Ecology 389:70-74.

Goodbody-Gringley, G., and S. J. Putron. 2009. Planulation patterns of the brooding

coral Favia fragum (Esper) in Bermuda. Coral Reefs 28:959-963.

Harrington, L., K. Fabricius, G. De’ath, and A. Negri. 2004. Recognition and selection of

settlement substrata determine post-settlement survival in corals. Ecology

85:3428-3437.

Marhaver, K. L., M. J. Vermeij, F. Rohwer, and S. A. Sandin. 2013. Janzen-Connell

effects in a broadcast-spawning Caribbean coral: distance-dependent survival of

larvae and settlers. Ecology 94:146-60.

Mason, B., M. Beard, and M. W. Miller. 2011. Coral larvae settle at a higher frequency

on red surfaces. Coral Reefs 30:667-676.

20

Page 21: CW Invert and 3D Thesis4 KLM

Mason, B., M. Schmale, P. Gibbs, M. W. Miller, Q. Wang, K. Levay, V. Shestopalov,

and V. Z. Slepak. 2012. Evidence for multiple phototransduction pathways in a

reef-building coral. PLoS ONE 7:e50371.

Miller, K., and C. Mundy. 2003. Rapid settlement in broadcast spawning corals:

implications for larval dispersal. Coral Reefs 22:99–106.

Van Moorsel, G. W. N. M. 1983. Reproductive strategies in two closely related stony

corals (Agaricia, Scleractinia). Marine Ecology 13:273-283.

Morse, D. E., N. Hooker, A. N. C. Morse, and R. A. Jensen. 1988. Control of larval

metamorphosis and recruitment in sympatric agariciid corals. Journal of

Experimental Biology and Ecology 116:193-217.

Sammarco, P. W. 1985. The Great Barrier Reef vs. the Caribbean: comparisons of

grazers, coral recruitment patterns and reef recovery. Proceedings of the Fifth

International Coral Reef Congress 4:391-398.

Spieler, R. E., D. S. Gilliam, and R. L. Sherman. 2001. Artificial substrate and coral reef

restoration: what do we need to know to know what we need. Bulletin of Marine

Science 69:1013-1030.

Szmant-Froelich, A., M. Reutter, and L. Riggs. 1985. Sexual reproduction of Favia

fragum (Esper): lunar patterns of gametogenesis, embryogenesis and planulation

in Puerto Rico. Bulletin of Marine Science 37:880–892.

Vermeij, M. J. A., and S. A. Sandin. 2004. Density-dependent settlement and mortality

structure the earliest life phases of a coral population. Ecology 89:1994-2004.

Williams, R. R., S. N. Arnold, N. D. Fogarty, R. S. Steneck, M. J. A. Vermeij, and V. J.

Paul. 2009. New perspectives on ecological mechanisms affecting coral

recruitment on reefs. Smithsonian Contributions to Marine Sciences 38:437-452.

21

Page 22: CW Invert and 3D Thesis4 KLM

The ability of mobile reef invertebrates to influence coral settler survivorshipAbstract:

Coral settlers experience substantial mortality in the time between settlement and

adulthood. Although studies have shown that this high mortality rate may be due to

various abiotic factors and biotic competition with algae, less is known about interactions

with mesograzer invertebrates. The present study took place on the leeward side of

Curacao at the CARMABI marine research station in the 2013-2014 winter season. In

this study, various mesograzers including echinoderms, decapods, gastropods, and

predatory polychaetes were introduced to newly-settled polyps of three scleractinian coral

species. Coral survival and invertebrate behaviors were then observed. Decapods had an

overall negative influence on coral survivorship and size of individual coral settlers and

crabs may have been a contributing factor. A size effect was also observed for predatory

polychaetes, while other invertebrate coral interactions, influenced coral survivorship in

positive, negative, and neutral ways depending on the invertebrate and coral species in

question. In addition, a field experiment was conducted on a shallow reef near

CARMABI, revealing that reef traffic is highly structured in space, highly localized, and

increases at night. Traffic was comprised primarily of various decapods. Areas of

increased traffic were associated with higher levels of coral settler mortality. Both lab and

field experiments provide evidence that mobile reef invertebrates play a role in coral

ecology by influencing early life stage coral settler survivorship.

22

Page 23: CW Invert and 3D Thesis4 KLM

Introduction:

Coral larvae face a gauntlet of challenges before adulthood, including finding a

suitable place for settlement. When finally attached, coral settlers become particularly

vulnerable to abiotic threats such as sedimentation, wave action, storms, and

eutrophication, as well as biotic threats. Algae are a widely-recognized biotic source of

mortality for juvenile corals as they can hinder coral recruitment by preemptively taking

up space, reducing local water flow, releasing secondary metabolites, altering microbial

communities, and blocking sunlight from corals (Williams et al. 2009). While turf algae,

benthic cyanobacteria, and many macroalgae have a known ability to negatively affect

coral settler survivorship, mobile invertebrates are less well studied and their effects—if

any—on corals are often not known. (Williams et al. 2009, Birrell et al. 2008, Kuffner et

al. 2006, Birrell et al. 2005, Kuffner and Paul 2004).

Of the research that focuses on mobile invertebrate coral interactions, much

previous research focused on Indo-Pacific invertebrate species, especially the predatory

corallivorous crown-of-thorns starfish Acanthaster planci and gastropods in the genus

Drupella. (Carpenter 1997, Sammarco 1985). Coral reef systems in the Caribbean have

few invertebrate coral predators with such plainly visible negative effects. Caribbean

coral predators include the bearded fireworm Hermodice carunculata and the

corallivorous snail Coralliophila abbreviata. Other Caribbean invertebrate species known

to have an indirect effect on corals are the rock-boring urchin Echinometra lucunter, and

the long spined urchin Diadema antillarum (Wolf 2012, Baums et al. 2003, Bruckner et

al. 1997, Carpenter 1997, Hayes, 1990).

23

Page 24: CW Invert and 3D Thesis4 KLM

While keystone invertebrate species like A. planci in the Indo-Pacific are known

to have a direct predatory effect on corals, most invertebrates in the Caribbean, chiefly

the D. antillarum, have indirect effects on corals, such as increasing coral survivorship

through herbivory of algae. After the mass mortality of D. antillarum in 1983 and 1984,

resulting in the loss of up to 95% of the Caribbean population, a noted increase in algal

biomass took place throughout the Caribbean (Williams and Polunin 2001, Carpenter

1997). Sammarco (1985) identified echinoids as major grazers of the Caribbean up until

the D. antillarum die-off, whereas in recent years, fish have become significant algal

grazers (Williams and Polunin 2001).

Algal stressors are known to negatively influence settlement of coral larvae and

algae-dense habitats make for poor recruitment areas (Williams et al. 2009, Birrell et al.

2008, Kuffner et al. 2006, Birrell et al. 2005, Kuffner and Paul 2004). Therefore

herbivorous invertebrates are often assumed to be beneficial to coral recruitment and

survivorship in the Caribbean. The ability of mobile herbivores to influence algal

abundance through grazing—and thereby indirectly benefit scleractinian coral settlement

and survivorship—has been well studied (Coen 1988, Huston 1985, Lubchenco 1981). In

hermit crab exclusion tests, epibiont cover, specifically in the form of Dictyota spp., was

significantly higher than treatments with natural crab densities or crab inclusion

treatments, with algae increasing from 27% coverage to 77% within a single month in

exclusion treatments (Coen 1988). Edmunds and Carpenter (2001), found a significant

increase in the abundance of juvenile corals in areas dominated by D. antillarum. They

further noted a distinct decrease in macroalgae in areas with high D. antillarum densities

(Edmunds and Carpenter 2001).

24

Page 25: CW Invert and 3D Thesis4 KLM

Despite the general assumption that invertebrate herbivores should benefit corals

indirectly, invertebrate species with different anatomies, life histories, feeding strategies,

and food preferences will each have a unique effect on corals, especially when

encountering corals in their most vulnerable state as young settlers (Carpenter 1997).

Additionally, invertebrates vary in their foraging times, digestive abilities, size, speed,

nutrient requirements, reactions to altering temperature, and responses to various toxins.

These fundamental differences along with invertebrates being furthermore subject to both

bottom up and top down controls, yields different relative abundances of micro-

invertebrates with specific pressures enacted on coral settlers (Lubchenco and Gaines

1981). The complex and sensitive micro-invertebrate guild therefore holds numerous

ecologically important questions. How do various invertebrates affect coral settlement

and the survivorship of new coral settlers? Which invertebrate species hurt and which

help coral settlers?

In the present study, we investigated the impact of various invertebrate species

when introduced to coral settlers. Coral reef invertebrates tested within this experiment

included coral predators, opportunistic feeders, as well as herbivores. Two pilot studies

(see appendix) and five lab experiments were conducted with various invertebrate species

and coral settlers of the species Montastraea faveolata, Favia fragum, and Agaricia

humilis to deduce various ecological relationships. Of the five lab experiments, the first

two investigated M. faveolata larvae settlement behavior based on different invertebrate-

conditioned water and surface signals. The remaining three lab experiments tested direct

interactions of various invertebrates with settlers of each of the three coral species.

Additionally, a field experiment was carried out to examine overall invertebrate reef

traffic and the effect of varying reef traffic on newly settled F. fragum polyps.

25

Page 26: CW Invert and 3D Thesis4 KLM

Methods:

Research location

All experiments were conducted on the island of Curacao, located in the southern

Caribbean, just north of Venezuela within the chain of islands commonly referred to as

the Netherlands Antilles. Experiments were conducted between 1 September 2013 and 31

March 2014. Locations for invertebrate collections, coral collections, coral larvae

harvesting, incubations, and field experiments are shown in Figure 1.

Figure 1: Site locations. 1. Buoy Zero (12°7′25"N, 68°58′50"W) 2. Piscaderabaai

(12°7'20"N, 68°57'53"W) 3. Water factory (12°6'37"N 68°57'7"W) 4. Sea Aquarium

(12°5'3"N, 68°53'51"W) 5. CARMABI marine research station

Study species

Seven mobile, benthic coral reef invertebrate species were included in this

investigation, although not all invertebrate species were included in all experiments. The

invertebrate species of interest were: the red legged hermit crab (also known as the red

2 miles

Curacao

26

Page 27: CW Invert and 3D Thesis4 KLM

reef hermit crab or scarlet reef hermit crab) Paguristes cadenati, the red striped hermit

crab Phimochirus holthuisi, the bearded fireworm Hermodice carunculata, the yellow

corallivorous snail Coralliophila abbreviata, the rock-boring urchin Echinometra

lucunter, the long-spined urchin Diadema antillarum, and the sea lettuce slug Elysia

crispata. The three scleractinian coral species included within this investigation were the

broadcast spawning species Montastraea faveolata and the brooding species Favia

fragum and Agaricia humilis.

Experiment 1 & 2: M. faveolata settlement in response to seawater and surfaces pre-

conditioned by invertebrates:

36 polypropylene plastic beakers were cleaned with 10% dilute bleach for 1 hr,

moved into a separate rinse tank with 10% dilute sodium thiosulfate pentahydrate to

neutralize the bleach, and then rinsed with 100 m-filtered seawater (100 m FSW;

Millipore, Billerica MA/USA). A peristaltic pump was then used to prepare 0.45 m-

filtered seawater (0.45 m FSW). Each beaker was filled with 0.65 L of 0.45 m FSW.

Two microscope slides, pre-cured for four days in the CARMABI flow-through seawater

system, were then added to each polypropylene plastic beaker. The flow-through

seawater system was comprised of aquaria with 200 m 0.45 m FSW flow-through

from Piscaderabaai, pumped from a depth of 5 m.

Incubating invertebrates in the beakers with 0.45 m FSW and glass microscope

slides created conditioned seawater and conditioned surfaces for use in two separate

experiments. Two specimens of each invertebrate study species were added to each

beaker. For this experiment, four beakers were prepared per treatment. Invertebrates were

collected at 1-5 m depth from the dive sites at Sea Aquarium, Water Factory,

Piscaderabaai, Snake Bay, and Buoy Zero (Figure 1). In addition, eight clean glass

27

Page 28: CW Invert and 3D Thesis4 KLM

microscope slides were rinsed with 0.45 m FSW and placed into four beakers for a

sterile or “negative” control, and eight microscope slides pre-treated for four days in the

aquarium wet lab were placed into four beakers for a biofilm “positive” control. Both

controls had no invertebrate species present. All beakers were then placed in aquaria to

use as a water bath, stabilizing temperature at 29°C for a 24 hr incubation. In the aquaria,

beakers were assigned random locations relative to one another in space to ensure

interspersion of treatments and replicates. Beakers were placed within one of two

aquariums. In previous studies of settlement cues in pelagic marine invertebrates, a 24-

hour conditioning period was used to prepare seawater (Lau and Qian 2001, Griffith and

Frost 1976). We followed this precedent and incubated invertebrates in the beakers for 24

hr. Following this conditioning period, invertebrates were removed and the conditioned

water and conditioned glass slides were then used immediately for experiments with coral

larvae.

Experiment 1: M. faveolata settlement in invertebrate conditioned seawater

Water was pooled from each set of four polypropylene plastic beakers containing

the same invertebrate species and filtered through a 0.45 m filter. 40 ml was then added

to each of seven replicate petri dishes per treatment. There were eight treatments in total,

including one control treatment (0.45 m FSW from control with biofilm conditioned

slides and no invertebrates). Petri dishes were randomly assigned locations on the lab

bench and kept in a 12:12 night:day light regime. Lab temperature was maintained at

29°C. 30 M. faveolata larvae, collected six days prior as gamete bundles on 24 September

2013 at the Water Factory (Figure 1), were added

to each petri dish. Petri dishes were scored every 48 hr for settlers until day 16.

Experiment 2: M. faveolata settlement on invertebrate conditioned surfaces

28

Page 29: CW Invert and 3D Thesis4 KLM

Glass microscope slides were removed from polypropylene plastic beakers with

invertebrates after a 24-hour incubation period. One slide per sample was then placed in a

petri dish yielding a treatment value of nine with a positive biofilm control and separate

negative biofilm control. All treatments were replicated eight times. 35 ml of single batch

0.45 m FSW obtained fresh the same morning from the CARMABI flow-through

seawater system was added to each petri dish. Petri dishes were randomly assigned

locations on the lab bench and kept in a 12:12 night:day light regime. Lab temperature

was maintained at 29°C. 30 M. faveolata larvae, again collected six days prior as gamete

bundles on 24 September 2013, at the Water Factory (Figure 1), were added to each petri

dish. Petri dishes were scored every 48 hr for settlers until day 16.

Experiment 3: M. faveolata survivorship with varying invertebrate species:

M. faveolata larvae were reared from gamete bundles collected on 24 September

2013, from the Water Factory dive site (Figure 1). Larvae were allowed to settle on the

bottom of polystyrene plastic deli bins with lids, each filled with approximately 150 ml of

0.45 m FSW. Water changes were performed every second day to maintain newly

settled polyps. After approximately two months, the flat bottoms of bins were cut into

4x4 cm square sections while being submerged underwater to keep settlers alive. The

number of settlers on each plastic square was counted using a dissection microscope.

Squares were then incubated for 24 hr in a flow-through seawater system at CARMABI.

The number of settlers per plastic section varied from 8 to 142. Plastic sections were

grouped into groups of seven by settler density, then one plastic section from each group

was randomly assigned to each of eight experimental treatments. Species used in this

experiment were: the red legged hermit crab P. cadenati, red striped hermit crab P.

holthuisi, bearded fireworm H. carunculata, yellow corallivorous snail C. abbreviata,

29

Page 30: CW Invert and 3D Thesis4 KLM

rock-boring urchin E. lucunter, long-spined urchin D. antillarum, and the sea lettuce slug

E. crispata. Individual invertebrates were added to clear circular plastic containers of 0.5

L volume with lids. Containers were 13 cm in diameter, 6 cm in depth, with a 4x4 cm

square hole cut into the lid for water exchange. These holes were then covered with black

shade cloth (plastic mesh with 2 mm diameter holes). Controls consisted of a container

without the presence of an invertebrate. In each replicate, one 4x4 cm square of

polystyrene plastic, with settlers attached, was secured to the bottom of the container with

a plastic-coated paper clip. Paper clips were held in place with double-sided, 0.3 cm thick

poster board adhesive. Replicate containers were then submerged in aquaria with flowing

seawater for nine days. In each replicate, settler survivorship was scored 24 hr after

incubation began and then every 48 hr thereafter until the final measurement on day 9.

Experiment 4: F. fragum survivorship with varying invertebrate species:

F. fragum colonies were housed in a coral nursery that was set up on the reef at

CARMABI in Piscaderabaai. These colonies were originally collected from the Sea

Aquarium reef in December 2013. On 10 January 2014, fifteen of these colonies were

transported to the flow-through aquarium system at CARMABI.

For both Experiments 4 and 5, larvae harvesting and settlement techniques were

the same. Each coral was placed in an individual cup with a tube providing a continual

flow of a few ml of fresh seawater per second. Water overflow ran down the handle of

the cup onto a submerged 200 m plastic mesh filter, surrounded with PVC piping. As

corals planulated, larvae would rise to the surface, flow over the handle and into the

submerged mesh filter for collection. Aquariums with adult colonies were surrounded

with shade cloth overnight to block artificial light and promote coral planulation. At

sunrise, larvae were pipetted out of mesh catchers into polystyrene bins filled with 0.45

30

Page 31: CW Invert and 3D Thesis4 KLM

m FSW. Larvae were pooled over multiple days of collection to obtain enough

individuals for high replication. A mixed cohort of one- and two-day-old larvae were

introduced to settlement surfaces as daily larvae harvest rates varied. Larvae were

collected over six-days, starting eight days after the new moon. Every time 60 larvae

were collected, they were placed into polystyrene plastic bins filled with 200 mL of 0.45

m FSW. Terracotta tiles measuring 3x4 cm were incubated underwater in Piscaderabaai

for three weeks prior to the experiment, allowing the development of CCA and biofilms.

Detritus was brushed off tiles before starting the experiment. Four tiles were added to

each settlement bin.

Settlement of F. fragum varied between 2 and 16 settlers per tile. Tiles were

grouped into groups of seven by settler density, then one tile from each group was

randomly assigned to each of three experimental treatments. Species used in this

experiment were: the red-legged hermit crab P. cadenati, and the red striped hermit crab

P. holthuisi. The control treatment had no invertebrates. The experiment was conducted

in the same plastic containers used in Experiment 3. One individual hermit crab was

added to each container. One terracotta tile was then added to each container. Containers

were then submerged in aquariums with flow-through seawater for 13 days. Survival of

settlers was scored at 24 and 48 hr, and then every subsequent 48 hr until day 13.

Experiment 5: A. humilis settler survivorship with varying invertebrate species:

Nine A. humilis colonies were collected from Water Factory and Piscaderabaai

between 21_24 January 2014. The methods and approach used in Experiment 4 were

repeated for this experiment. Differences included testing the bearded fireworm

31

Page 32: CW Invert and 3D Thesis4 KLM

H. carunculata, in addition to the hermit crabs P. cadenati, P. holthuisi, and control.

Settlement of A. humilis larvae on terracotta tiles ranged from 5 to 16. Tiles were divided

among treatments using the same grouping method as in Experiment 4, resulting in nine

replicates for four treatments, control included. The tiles were scored 24 hr after initiation

and every 48 hr after that for 11 days at which point the experiment was terminated.

Experiment 6A: Time-lapse reef surveys on invertebrate reef traffic and composition:

To quantify invertebrate reef traffic and species composition, time-lapse

photography was used. Cameras used included the Canon G9, Canon G10, Canon

PowerShot S110 (all manufactured by Canon, Tokyo/Japan), and a GoProHero3+

(GoPro, San Mateo/USA). GoPro cameras include time-lapse capabilities as a

programmed feature. For the Canon cameras, publicly-available time-lapse software

known as CHDK or “Canon Hack Development Kit,” was used to program time-lapse

specifications onto memory cards, allowing for time-lapse photography with these

cameras. Diving at Buoy Zero, starting on 18 February through 29 March 2014, 40 dives

in total using SCUBA were conducted in which cameras were placed and collected.

Cameras were either placed just after sunrise so that the first shot occurred between

08:35-09:35, or just after sunset with the first shot beginning between 18:25-19:00. Time-

lapses were set to take one photo per minute for GoPro cameras and one photo per three

minutes for the remainder of cameras, due to unique camera specifications. The duration

of each time-lapse series ranged between 03:45 hr and 07:48 hr. Cameras were secured in

place on the reef using a 2.54 cm diameter PVC stand. Two monofilament lines were

secured to the PVC stand and hooked to reef rubble to keep cameras secure. Cameras

were placed at a depth of 5-6 m, distance to the reef face was between 10-20 m, and

camera height was between 40-50 cm from the frame objective. Two day and one

32

Page 33: CW Invert and 3D Thesis4 KLM

evening time series of the same location were compared over six sites for overall

invertebrate reef traffic as well as composition of reef traffic. Time-lapse series were

analyzed using ImageJ, version 1.47 with the plugin designed by Biomedical Imaging

Group Rotterdam of Erasmus MC – University Medical Center, MTrackJ, version 1.3.0.

Using MTrackJ, each photo series measurement was standardized using a ruler

photographed in the first photo of each time-lapse. All visible tracks were recorded,

regardless of when in the time series they began or ended. Tracks across time series were

standardized by taking the average number of tracks per hour for each time series.

Experiment 6B: Field survivorship of F. fragum settlers on coral reef at Buoy Zero:

This experiment was conducted to quantify the level of invertebrate traffic on reefs and

determine the effect of traffic on coral settler survivorship. 38 F. fragum colonies were

transferred from the CARMABI coral nursery located in Piscaderabaai. Colonies were

initially collected from the Sea Aquarium reef, the Water Factory reef, and Piscaderabaai.

Larvae were collected from 7-12 March 2014. Once 100 larvae were collected, they were

placed into polystyrene plastic bins filled with 200 ml of 0.45 m FSW. Prior to the

experiment, terracotta tiles measuring 3x4 cm were incubated for one month underwater

on the reef in Piscaderabaai at a depth of 5 m, allowing for the development of CCA and

biofilms. Detritus was brushed off tiles before adding four of the tiles to each settlement

bin. Tiles were scored using a blue light system. These ranged from a dive light used in

conjunction with a Fluorescence Excitation Filter, a BlueStar flashlight, or a Dual

Fluorescent Protein Flashlight combined with the matched barrier filter glasses

(NightSea, Bedford/USA). Settlers per tile ranged from 8 to 37 settlers. Tiles were

assigned treatments based on the same grouping methodology observed in Experiments 4

and 5. Treatment sites included three vertical surfaces and three horizontal surfaces. In

33

Page 34: CW Invert and 3D Thesis4 KLM

addition, one vertical control surface, and one horizontal control surface were used. Only

one control per orientation was used, as the number of tiles with settlers was a limiting

factor in the number of controls possible. For each site, six terra cotta tiles were secured

in place. Tiles were secured by placing three nails in the rock surface and placing each

tile against the rock face, wedged between the nails. This allowed for removal of tiles for

scoring purposes but also helped to keep tiles in place. Controls consisted of pieces of

steel reinforcement bar, placed in the same area but in sand instead of rock. White plastic

ceiling panels i.e., “egg crating” (Plaskolite, Columbus/USA) measuring 25x25cm was

then attached either vertically or horizontally at approximately 0.5 m above the sandy

bottom. Six tiles were then attached to each egg crate in the same fashion as other

treatments with nails. Nails were kept in the egg crating by wrapping the tip of each nail

with tape, until each nail could be wedged into a single 1x1 cm hole of egg crating. Steel

reinforcement bars for controls were used to prevent invertebrates from accessing the

tiles by keeping the egg crate surfaces off the sea floor while keeping tiles exposed to

similar abiotic conditions as other treatments. Tiles were installed on the reef at Buoy

Zero on 13 March 2014. Treatment sites with similar conditions were chosen

haphazardly. Treatment sites had a consistent depth of 5-6 m, distance to reef face 10-20

m, and with rocky bottoms except for controls. These conditions were chosen based off

the assumption that coral settlers can only settle and begin their life cycle on open and

solid substrate. Solid limestone areas are therefore the primary areas of interest for the

study as coral larvae are most capably of settling and recruiting to these surfaces.

Tiles were placed on designated treatment surfaces haphazardly and secured with

three nails hammered into the substrate. Tiles were scored at night after sunset using

SCUBA and a blue light system at day 0, 1, 3, 5, 11, and 16. Time-lapse cameras were set

34

Page 35: CW Invert and 3D Thesis4 KLM

up for all treatment sites except controls, and were the same as those mentioned in

experiment 6A. Two day time-lapses (one with tiles present, one without tiles present)

and one night time-lapse per location were recorded.

Data Analysis

For Experiments 1 through 5, a maximum likelihood approach was used. Data

were analyzed by grouping experimental treatments in various combinations and then

comparing the models to one another to find which distribution or model best described

the data (Hartmann et al. 2013, Vermeij et al. 2009, Vermeij and Sandin 2004). A

binomial error distribution was calculated across all treatments, between time zero and

the final time point. Success was defined as settlement in Experiments 1 and 2 and as

survivorship in Experiments 3 through 5. Models were synthesized with varying numbers

of parameters and groupings within parameters in an attempt to deduce the best-fit mean

values. A Chi-squared distribution was calculated between models of varying parameters

on the sum of the natural logarithm of the binomial distribution to yield the best-fit

model, assuming a single tail test. Samples throughout experiments were considered

independent and it was assumed there was no density-dependent effect of coral larvae or

coral settlers within samples. In closed system samples (Experiments 1 and 2), larvae

density did not exceed one larva per ml. In all other experiments, continuous water flow

and exchange was provided throughout the trial period, reducing microbial “bottle

effects” and larval density effects.

For experiment 6A, day time-lapses of treatment sites were compared to alternate

day photo series as well as an evening photo series of the same treatment site using a one-

way ANOVA to test whether there was a change in invertebrate reef traffic volume at

35

Page 36: CW Invert and 3D Thesis4 KLM

night. For experiment 6B, reef traffic for each time series was calculated as number of

invertebrate tracks per hour. Each area’s invertebrate traffic per hour and settler

survivorship was compared using a Kruskal-Wallis Test to determine whether settler

survivorship varied across treatment areas in relation to invertebrate reef traffic.

RESULTS:

Experiment 1- M. faveolata settlement response to conditioned invertebrate seawater

Montastraea faveolata larvae exhibited settlement over time for all treatments of

conditioned invertebrate seawater. Final settlement ranged from an average of four

settlers per sample in water conditioned with bearded fire worms, Hermodice carunculata

to nine settlers in control samples with no invertebrates (Figure 2). When applying the

maximum likelihood model, grouping the two hermit crab species Paguristes cadenati

and Phimochirus holthuisi, the control separate, and all other treatments together yielded

a significantly more accurate model than any other model P=0.002618225 (Table 1).

Additionally, there was a higher initial rate of settlement within the control treatment,

followed by that of the two hermit crab treatments (Figure 2). It should be noted that

prior to filtering invertebrate conditioned sea-water for this experiment, invertebrate

excrement and in the case of gastropods, slime, was generated within the 24-hour

incubation period. Varying levels of solubility regarding these compounds could have

disproportionately altered water chemical cues. The solubility of invertebrate excrement

was not measured within this study.

Table 1: Maximum likelihood model for invertebrate-conditioned seawater and

settlement of M. faveolata

36

Page 37: CW Invert and 3D Thesis4 KLM

Experiment 1 (Maximum likelihood model)3-parameter model is preferred to 2-parameter model Sum-LN(BINOMDIST) # ParametersLikelihood for Treatment Hermits (5,6) vs. Control (8), vs. rest (1,2,3,4,7) 119.640798 3Likelihood for Treatment Control (8) vs. rest (1,2,3,4,5,6,7) 124.1688386 2Difference 4.528040618Difference*2 (for one-tail test) 9.056081235Degrees of Freedom 1Chi-test P-value 0.002618225

0 48 96 144 192 240 288 336 384

0

1

2

3

4

5

6

7

8

9

10

Long spined urchinSea lettuce slugBoring rock urchinBearded firewormRed legged hermit crabRed striped hermit crab

Time (hours)

Ave

rage

num

ber

of s

ettle

rs o

ver

each

tr

eatm

ent

Figure 2: Settlement of M. faveolata over time within invertebrate conditioned seawater.

Experiment 2- M. faveolata settlement response to conditioned invertebrate glass

surfaces

M. faveolata larvae exhibited a more narrow range of settlement over time for

conditioned invertebrate surfaces in comparison to that of conditioned seawater observed

in Experiment 1. Besides an initial high settlement of M. faveolata on conditioned

37

Page 38: CW Invert and 3D Thesis4 KLM

surfaces consisting of microscope slides from red striped hermit crabs P. holthuisi, all

other treatments followed similar rates of settlement, leveling off after six days and

resuming again after eight days including both biofilm negative and biofilm positive

controls. The likelihood model that was preferred was a two-parameter model in which

all treatments were grouped separate from the rock-boring urchin Echinometra lucunter,

with the next-best-fit three-parameter model not able to represent the data better

P=0.13638.

Table 2: Maximum likelihood model for invertebrate conditioned surfaces and settlement

of M. faveolata

Experiment 2 (Maximum likelihood model)3-parameter model is not preferred to 2-parameter model Sum-LN(BINOMDIST) # ParametersLikelihood for Treatment long spine urchin (1) vs. Rest (2,5,6,7,8,9) vs. fireworms & rock-boring (3,4) 169.8667831 3Likelihood for Treatment rock-boring urchin (7) vs. rest (1,2,3,4,5,6,8,9) 171.5390522 2Difference 1.109182883Difference*2 (for one-tail test) 2.218365766Degrees of Freedom 1Chi-test P-value 0.13638

38

Page 39: CW Invert and 3D Thesis4 KLM

0 48 96 144 192 240 288 336 384

0

1

2

3

4

5

6

7

8

Long spined urchinSea lettuce slugBoring rock urchinBearded firewormRed legged hermit crab

Time (Hours)

Ave

rage

# o

f S

ettle

rs o

ver

each

tr

eatm

ent

Figure 3: Settlement of M. faveolata over time on various conditioned invertebrate

surfaces.

Experiment 3- M. faveolata survivorship with varying invertebrate species

Introduction of invertebrate species to M. faveolata settlers yielded variable

results. In most treatments, mortality increased except for H. carunculata and E. lucunter

where mortality rates were similar to that of the control. Additionally, in the presence of

sea lettuce slugs Elysia crispata, M. faveolata mortality actually decreased relative to the

control. Mortality of coral settlers was greatest in presence of both species of hermit

crabs (Figure 4). The exact cause in which hermit crabs were contributing to coral

mortality was not observed during the experimental period with possible mechanisms

including predation and trampling. Analysis of video recordings of separate introductions

of hermit crabs and M. faveolata did reveal incidental removal of M. faveolata by red-

legged hermit crabs P. cadenati. Additionally, a fine layer of sediment present on all

sample surfaces during the experiment was visibly absent on portions of both hermit crab

samples. These “cleaned” areas were void of previously established settlers. The long

39

Page 40: CW Invert and 3D Thesis4 KLM

spine urchins Diadema antillarum, experienced mortality during the experiment and

could have contributed to the observed mortality of M. faveolata settlers. Urchin

mortality may have been a result of habitat constraints or insufficient water flow. The

maximum likelihood model favored each treatment kept separately as the model

describing eight parameters was significantly more accurate than that of a seven

parameter model P=0.009430735 (Table 3).

Table 3: Maximum likelihood model for M. faveolata settler survivorship with variable

invertebrate introductions.

Experiment 3 (Maximum likelihood model)8-parameter model is preferred to 7-parameter model Sum-LN(BINOMDIST) # ParametersLikelihood for Treatment Each vs. all 1 vs. 2 vs. 3 vs. 4 vs. 5 vs. 6 vs. 7 vs. 8 448.5718432 8Likelihood for Treatment1 vs. 2 vs. 3 vs. 4 vs. hermits (5,6) vs. 7 vs. 8 451.9415337 7Difference 3.369690473Difference*2 (for one-tail test) 6.739380947Degrees of Freedom 1Chi-test P-value 0.009430735

40

Page 41: CW Invert and 3D Thesis4 KLM

0 24 48 72 96 120

144

168

192

216

00.10.20.30.40.50.60.70.80.9

1A.

ControlRed legged hermit crabRed striped hermit

Time (hours)

Sur

vivo

rshi

p

0

24

48

72

96

12

0

14

4

16

8

19

2

21

6

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1B.

Control Fire Worm

Time (hours)

Sur

vivo

rshi

p

0

24

48

72

96

12

0

14

4

16

8

19

2

21

6

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1C.

Control

Sea lettuce slug

Time (hours)

Sur

vivo

rshi

p

Figure 4: Survivorship of M. faveolata settlers, incubated with various invertebrates with

standard error shown. A. Hermit crabs B. Fireworms C. Gastropods.

Experiment 4- F. fragum survivorship with varying invertebrate species

Mortality of F. fragum increased when either of the hermit crab species was

introduced compared to the control. F. fragum, attached to terracotta tiles experienced a

higher mortality with the larger P. cadenati hermit crabs than with the smaller P.

holthuisi hermit crabs (Figure 5). Overall, settler mortality rates were lower (25 % and

below) than that experience by M. faveolata in experiment five (as high as 100%). In this

experiment, a two parameter maximum likelihood model in which the hermit crab

treatments were aggregated, did not describe the data better than when all treatments

were grouped together P=0.148493661.

41

Page 42: CW Invert and 3D Thesis4 KLM

Table 4: Maximum likelihood model for F. fragum settler survivorship with variable

invertebrate introductions.

Experiment 4: Maximum likelihood model

2-parameter is not preferred to 1-parameter Sum-LN(BINOMDIST) # ParametersLikelihood for Treatment control (1) vs. hermit crabs (2,3) 27.32113134 2Likelihood for Treatment all aggregated (1,2,3) 28.36495982 1Difference 1.043828483Difference*2 (for one-tail test) 2.087656965Degrees of Freedom 1Chi-test P-value 0.148493661

0 1 2 3 4 5 6 7 8 9 10 11 12 13 140.7

0.8

0.9

1

ControlRed striped hermit crabRed legged hermit crab

Days

Cor

al s

ettle

r su

rviv

orsh

ip

Figure 5: Survivorship of F. fragum incubated with hermit crabs.

Experiment 5- A. humilis settler survivorship with varying invertebrate species

Agaricia humilis experienced increased rates of mortality under introduced

invertebrate treatments of H. carunculata and P. cadenati in relation to the control, and

interestingly less mortality when smaller P. holthuisi crabs were present (Figure 6).

42

Page 43: CW Invert and 3D Thesis4 KLM

Overall, mortality rates of A. humilis were similar to that of F. fragum, below 25%. The

maximum likelihood model describes the data best as a two parameter model where

controls and P. holthuisi crabs were grouped and compared to H. carunculata and P.

cadenati crabs P=0.045143297.

Table 5: Maximum likelihood model for Agaricia humilis settler survivorship with

variable invertebrate introductions.

Experiment 5: (Maximum likelihood model)2-parameter is preferred to 1-parameter Sum-LN(BINOMDIST) # ParametersLikelihood for Treatment control and red striped hermit (1,4) vs. fireworm and red-legged hermit (2,3) 83.25943908 2Likelihood for Treatment all aggregated (1,2,3,4) 85.26607815 1Difference 2.006639065Difference*2 (for one-tail test) 4.013278131Degrees of Freedom 1Chi-test P-value 0.045143297

0 2 4 6 8 10

0.7

0.8

0.9

1

ControlBearded firewormRed legged hermit crab

Days

Ave

rage

Sur

vivo

rshi

p ov

er e

ach

trea

tmen

t

Figure 6: Survivorship of A. humilis introduced to hermit crabs and the bearded

fireworm.

43

Page 44: CW Invert and 3D Thesis4 KLM

Experiment 6A- Reef time-lapse surveys on invertebrate reef traffic and composition

The in-situ reef monitoring experiment took place at Buoy Zero on horizontal

surfaces. Vertical surfaces were initially recorded as well, but as tiles of coral settlers

discussed in Experiment 6B were prone to excessive movement, analysis of vertical

surfaces were not included. Horizontal time series revealed a stark increase in

invertebrate reef traffic at night when compared to time-lapse series of the same locations

studied during the day (ANOVA, One-way randomized, DF = 2, F=5.513, P=0.0438,

Figure 7). Invertebrate traffic increased in areas with more hiding places, i.e., crevices,

algae, rock complexity, and sediment, and was visibly less on exposed rocky surfaces

(personal observations). The most prevalent invertebrates found during each time-lapse

series were various species of hermit crabs. The second most common invertebrates were

non-hermit crab, crab morphotypes (Figure 8).

Low Med/High High0

1

2

3

4

5

6

7

8

9

Traffic levels at each photo quadrant location

Num

ber

of tr

acks

per

tim

e se

ries

pe

r hr

Figure 7: Invertebrate reef traffic at three horizontal locations at Buoy Zero, standardized

as average number of tracks per hr. Control not shown as invertebrate traffic was

44

Page 45: CW Invert and 3D Thesis4 KLM

restricted to zero. Key: Black-Day time-lapse, Grey-Day time-lapse alternate, White-

Evening time-lapse. ANOVA, One-way randomized, DF = 2, F=5.513, P=0.0438.

(AM)#1 (AM)#1 (PM)#1 (AM)#2 (AM)#2 (PM)#2 (AM)#3 (AM)#3 (PM)#30123456789

(timelapse time) site location

Num

ber

of in

vert

ebtr

ate

trac

ks p

er ti

mel

apse

sta

ndar

d-iz

ed f

or 1

hr

Figure 8: Composition of invertebrate reef traffic on exposed rock faces for the nine time

series shown in Figure 7. Note that the nine series take place across three locations and

either after sunrise or after sunset each day (AM/PM) Key: Black- hermit crabs, Grey-

crabs, White-other invertebrates.

Experiment 6B- Reef survivorship of F. fragum settlers at Buoy Zero

Horizontally-oriented F. fragum tiles placed in heavily trafficked areas tended to

have increased levels of mortality compared to their counterparts placed in either a low

trafficked area or the control area with no traffic (Figure 9). The results were not

statistically significant when comparing each of four levels of traffic independently using

a Kruskal-Wallis Test. Grouping the low traffic site with the control together and

comparing them to the grouped data from the moderate to heavy and heavy traffic sites

still failed to yield statistical significance of P ≤ 0.05 using a one-tail Mann-Whitney U

test, however the visible difference was very close to being considered statistically

significant (U value = 43.5, P-value = 0.05262).

45

Page 46: CW Invert and 3D Thesis4 KLM

Unfortunately, tiles placed on vertical surfaces were prone to falling and were

found numerous times lying in the sediment upon returning to the experimental area in

subsequent days. A few times, damselfish could be observed knocking the vertical tiles

off the reef face. As this certainly affected survivorship of settlers on these tiles, it is

impossible to deduce what portion of settler mortality was attributed to invertebrate

traffic for vertical oriented treatment sites. Therefore, data from vertical treatment areas

were removed from analysis and only horizontally-oriented treatment sites were

considered.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

0

1

2

3

4

5

6

7

8

9

Settler survivorship

Inve

rteb

rate

traf

fic

per

tile

Figure 9: Survivorship of F. fragum settlers on tiles located within four treatment areas

with invertebrate traffic standardized to tracks per area per hour. Key: X- Control with

zero traffic, Square- low traffic, Diamond- med/high traffic, Triangle- high traffic, Mann-

Whitney U test on site med/high and high traffic sites vs. low and zero traffic sites

yielded Z-Score =-1.6166, U value=43.5, p-value of 0.05262.

46

Page 47: CW Invert and 3D Thesis4 KLM

Discussion:

Settlement of coral larvae

Coral larvae are known to respond to various settlement cues including substrate

orientation, chemical cues, progeny distance from adults, and light (Marhaver et al. 2013,

Mason et al. 2011, Harrington et al. 2004, Vermeij and Sandin 2004, Carlton and

Sammarco 1987). As it has been found that a five-to-sevenfold increase in recruitment of

coral spat occurs under reduced grazing pressure (Sammarco 1985), an attempt was made

to deduce if Montastraea faveolata larvae responded differently in regards to settlement

patterns when presented with invertebrate conditioned seawater, as seen in Experiment 1.

Settlement of M. faveolata was found to increase in rate in the control treatment

(Figure 2). The likelihood model best described the data by grouping the hermit crabs and

control separately from the rest of the invertebrate treatments. This grouping could

indicate treatments with fewer chemical signals in the seawater were preferred over that

of other treatments, which were observed to have higher excrement (echinoderms and

polychaetes) and slime (gastropods) prior to particulate filtration (personal observations).

Due to the lack of clear responses to the various invertebrate conditioned seawater

treatments, our data shows that it is unlikely that M. faveolata larvae are settling based

off of unique invertebrate signatures and most likely simply choosing cleaner water in

this case.

A narrower settlement range of M. faveolata was observed under our conditioned

surface experiment and settlement began generally later in time when compared to the

conditioned water experiment (Figure 3). The maximum likelihood model favored a

single parameter model, indicating minimal differences between the various conditioned

47

Page 48: CW Invert and 3D Thesis4 KLM

surfaces. As similar settlement rates were reflected across all but the Phimochirus

holthuisi treatment, it looks as if M. faveolata does not respond directly to surface cues

left by different invertebrates. From these two experiments, these results give evidence

that this coral species may respond positively to cleaner saltwater, but does not exhibit

clear positive or negative settlement cues to these specific invertebrates found on the reef,

despite the fact that these invertebrates fulfill unique roles as predators, herbivores, and

omnivores.

Invertebrate introductions

Decapods and polychaetes

In Experiments 3 through 5, we tested various invertebrate introductions with

three coral species: M. faveolata, F. fragum, and A. humilis. In all experiments we saw a

marked decrease in survivorship of coral settlers when either of the hermit crabs was

introduced except for when P. holthuisi crabs were introduced to A. humilis, yielding an

increase in survivorship in comparison to the control (Experiment 5, Figure 6).

Although the literature regarding the effect of hermit crabs on coral settlers is

scarce, various research efforts shed light into the mechanisms in which various crabs can

contribute to coral settler death. Glynn (1997) found that bioerosion by the hermit crabs

Trizopagurus spp. on a fringing reef in Panama were responsible for the removal of about

1 metric ton of coral sediment per ha, per yr. Additionally, Glynn (1997) goes on to

discuss how hermit crabs and other crustaceans are strong enough to erode reef rock. It is

known that herbivores may contribute to coral recruitment mortality indirectly by

scraping of algae off rocks (Huston 1985), and that decapod crustaceans exhibit various

feeding behaviors including predation, detritivory, and suspension feeding (Wadell and

48

Page 49: CW Invert and 3D Thesis4 KLM

Pawlik 2000). This provides the clearest mechanism from the literature into how hermit

crabs contribute to coral settler death.

In a few instances throughout the experimental time periods, hermit crab

exoskeletons were observed after a few days. Although speculative, these observations

point to a level of malnourishment amongst the hermit crabs during the experiments. As

crabs were observed to incidentally dislodge settlers through video observations, the

mechanism of “knock off” seems to be the chief pathway in which these two species of

hermit crabs contribute to settler death of these three coral species, over the alternative of

direct predation.

Relative size of these organisms may play a role in survivorship and behavior as

well. P. holthuisi hermit crabs were smaller in size than that of the P. cadenati hermit

crabs. Measuring shell length and width in Experiments 5 and 7 yielded shell area

averages of 2.42 cm2 and 1.58 cm2 for P. cadenati in Experiments 5 and 7 respectively,

while P. holthuisi had average shell areas of 2.05 cm2 and 0.97 cm2 in Experiments 5 and

7, respectively When encountering M. faveolata settlers, P. holthuisi caused a greater

level of mortality than that of the P. cadenati (Figure 4). Conversely, when P. holthuisi

were introduced to larger settlers of F. fragum and A. humilis, a reversal was noted where

P. cadenati had a more negative effect than P. holthuisi (Figures 5, 6). P. holthuisi, in

fact, improved the survivorship of A. humilis in relation to the control, which may be in

part to their ability to navigate around these larger settlers while still maintaining a

herbivore grazing capacity. As M. faveolata settlers are tiny and smaller crabs are certain

to make shorter more frequent steps, this size effect could be a product of crab decision-

making of what to avoid on the larger end of settler size and biological mechanics on the

smaller end.

49

Page 50: CW Invert and 3D Thesis4 KLM

Looking at the results for the fireworms may also contribute insight for this

theory. While fireworms of various sizes did not exhibit a more negative effect on M.

faveolata settler survivorship when compared to the control (Figure 4), they did have a

greater negative effect when presented with the larger settlers of A. humilis in Experiment

5 (Figure 6). As H. carunculata is a known corallivorous polychaete, there may be a

threshold initial size at which their prey choice becomes worth consuming (Wolf 2012).

Our likelihood model in Experiment 5 supports this evidence as there was a notable

difference between treatments of P. cadenati and H. carunculata compared to that of the

control and red striped hermit crabs when introduced to A. humilis (P = 0.045143297).

Gastropods

While Corallophilia abbreviata had a negative effect on the survivorship of M.

faveolata coral settlers, the sea lettuce slug Elysia crispata, improved survivorship. As E.

crispata is a true herbivore and may promote coral settler health as it controls algae while

avoiding coral spat, C. abbreviata is a known predator of adult corals, feeds on 14

species of scleractinian corals (Miller 1981). Studies on the rate at which C.

abbreviata feed on corals found a mean coral tissue consumption rate of 1.9

cm2/snail/day with a maximum of 6.5 cm2/snail/day (Baums et al. 2003, Bruckner et

al. 1997). This species is also prevalent throughout various habitats within the

Caribbean. C. abbreviata was found on 64% of coral colonies examined on a reef in

Panama (Carpenter 1997), and associated with coral mortality when corals were stressed

(Hayes 1990). Other studies conducted on Florida Keys reefs found 50% of Montastraea

spp. and 20% of Acropora palmata colonies were infested with C. abbreviata (Baums et

al. 2003). Given the fact that C. abbreviata is a significant predator of adult corals,

50

Page 51: CW Invert and 3D Thesis4 KLM

and with our results pointing to predation of M. faveolata settlers by C. abbreviata

(Figure 4), this species may be of greater ecological importance than previously noted

as it is found widely throughout the Caribbean and has the capacity to target

scleractinian corals throughout their life history.

Echinoids

One of the most widely studied invertebrates within the Caribbean is the long

spined urchin Diadema antillarum. Although in the experiment reported here, the

presence of D. antillarum led to increased mortality of M. faveolata, D. antillarum

performed poorly in the experimental setup. Individuals in some cases lost spines or died.

Results in this study concerning D. antillarum are therefore unreliable. However the

literature on these organisms is important to discuss as it covers the trade off to coral

species that can both benefit and hinder the process of coral recruitment. Research carried

out by Sammarco (1980) suggested that echinoid density can affect the success of

different coral species. Agaricia spp. was found to increase in abundance and become

dominant in areas where D. antillarum was removed (Sammarco 1980). In the same

study, the increase in Agaricia spp. in the absence of all echinoids led to a competitive

advantage over F. fragum, which saw a marked reduction in abundance. Coral diversity

has been found to be highest in locations inhabited with echinoids and lowest in areas

absent of echinoids (Sammarco 1982). F. fragum disappeared entirely from experimental

areas subjected to high urchin densities and intense grazing favored corals predominately

geared for survival in cryptic habitats (Sammarco 1980). Thus, while echinoids can create

competitive advantages for various coral species, and in some cases lead to increases in

51

Page 52: CW Invert and 3D Thesis4 KLM

biodiversity, at high densities they can become detrimental to the coral recruitment

process.

Similarly, it has been shown numerous times that low densities of echinoids leads

to favorable conditions for algae, able to then outcompete corals. D. antillarum at high

densities will exhibit strong grazing pressure, reducing both turf algae and coral spat,

while at low densities, coral spat will be overgrown by turf algae. A moderate grazing

pressure is therefore thought to be optimum for coral recruitment. (Edmunds and

Carpenter 2001, Carpenter 1997, Huston 1985, Sammarco 1980).

While our results for D. antillarum may not be reliable due to apparent stress of

individuals in the experimental chambers, mortality of M. faveolata was similar to that of

the control when introduced to a second urchin, the rock-boring urchin Echinometra

lucunter (Experiment 3). During the experiment, the urchins were observed to be very

sedentary, which agrees with literature; E. lucunter, which graze primarily on drift algae,

are a sedentary crevice dweller found in shallows and thought to only have local effects

within their burrows (Carpenter 1997, Ogden 1977).

Field experiment on reef

Invertebrate roles as predators, grazers, and tramplers may yield various benefits

and hindrances to coral settlers depending on the species of invertebrate and of the coral

in question. Although specific relationships can be observed in the lab setting, under

natural conditions, various other factors come into play such as space, algae, and top

down predation, likely to alter overall interactions from those observed under lab

conditions. On reefs for example, C. abbreviata can most often be found between the

coral columns of M. annularis, where there might be some level of protection

52

Page 53: CW Invert and 3D Thesis4 KLM

(personal observations). Predators of C. abbreviata are numerous and include

snapping shrimp, Caribbean spiny lobster, pufferfish, hogfish, filefish, and octopuses

(Baums et al. 2003).

The time-lapse reef monitoring done in Experiments 6B showed variable

survivorship of F. fragum settlers on tiles within four different treatment areas with a

range of levels of invertebrate traffic. The field experiment also shows that there is a

marked increase of invertebrate reef traffic at night when compared to the same areas

monitored during the day (Figure 7). The two treatment areas with moderate to heavy

traffic yielded more variable survivorship with low survivorship of coral spat on

certain tiles when compared to that of the control site with no invertebrate traffic or

site with low invertebrate traffic (Figure 9). Although six to nine tracks per hour may

seem low, invertebrates were often observed moving extensively within a small area,

counting towards a single track. If these results are extrapolated to entire nights,

weeks, and years, the movement of invertebrates presents a tangible pressure on coral

settlers.

In addition, of all the recorded reef traffic, hermit crabs made up the largest

proportion of recorded invertebrates, followed by other crabs and then gastropods

(Figure 8). Previous research on Curacao reef systems found both P. cadenati and P.

holthuisi to be the dominant hermit crab species present on the reefs (Jorissen 2013).

Invertebrate diversity at the species level in the present study was difficult to deduce

as the camera angle gave a close to 90-degree shot of each treatment site. In surveys of

corals in the bay of Panama, 859 individuals representing up to 55 species of decapods

were recorded, demonstrating the diversity present within this taxon (Abele 1974).

53

Page 54: CW Invert and 3D Thesis4 KLM

Despite the overwhelming representation of decapods in this reef experiment, it was

conducted on sites characterized by exposed rocky surfaces. If surveys were done to

include cryptic habitats, other invertebrate species would most likely be found in

higher densities.

However, given that the dominant form of traffic on exposed rock surfaces is

from hermit crabs, and that coral settlers in heavily trafficked areas are more prone to

mortality than settlers in less trafficked areas, the results of the field experiment

reflect the findings found within both lab Experiments 4 and 5. When taken together,

these experiments provide strong evidence in support of the ability of hermit crabs to

contribute a consistent negative pressure on coral spat survivorship. The adverse

response to hermit crabs is highlighted in research on reef sponges, finding that of 30

species of sponges, 87 % produced organic extracts found to deter P. punticeps, a hermit

crab (Wadell and Pawlik 2000).

This negative pressure from various species of crabs could indicate that what is

typically considered habitat for coral settlers, is in fact not equal, based off of variable

densities of these invertebrates. Previous studies have found higher densities of decapods

in lagoons and reef slopes when compared to reef flats on three southern Cuban reefs, as

well as a strong correlation of species to hard substrate types and increasing complexity

(Martínez and Garcia 1999). Caribbean reefs have entered into an algae-dominated

phase shift and it is thought that coral recruitment strategies have changed from favorable

long-lived broadcast spawners like as Acropora sp. and M. annularis to brooding species

such as Agaricia spp. (Williams et al. 2009). By increasing complexity and habitat for

various invertebrates with algae species such as Halimeda spp., rises in invertebrate

54

Page 55: CW Invert and 3D Thesis4 KLM

densities may further reduce coral settler survivorship in those areas and provide an

additional selection mechanism to favor larger spat of brooding corals over small

spawning species.

Despite convincing results pointing to a negative short-term interaction, it is

critical to take into consideration long-term interactions and the implications thereof.

It is known that majid crabs and echinoids preferentially feed on turf algae, known itself

to reduce coral settler survivorship (Carpenter 1997). Additionally, some Trapezia spp.

aid corals as they maintain the ability to deter coral predators (Carpenter 1997, Glynn

1983a, b). Thus, while there may be short-term pressures exhibited by invertebrates on

corals, their long-term predation on algae could help promote coral recruitment.

Future studies should focus on the long-term trade-off of herbivory regarding these

invertebrate species, to that of the short-term trampling effect, shown here to have a

negative impact on coral settler survivorship.

Conclusion:

While small hermit crabs, gastropods, and polychaetes were thought to

have minimal impact on corals due to small body size, limited numbers, and short ranges

(Carpenter 1997), both lab and reef experiments showed that invertebrates, particularly

hermit crabs, exhibited a negative pressure on coral settler survivorship. Additionally,

larger initial size of coral settlers may allow them to avoid trampling, but places them at

risk for becoming of interest to predatory polychaetes. There may therefore be a tradeoff

regarding initial size of coral settlers with invertebrate composition able to shift this

balance, thereby influencing coral health and species diversity on reefs. Additional

research on other invertebrate species, and those found within cryptic habitats, would add

55

Page 56: CW Invert and 3D Thesis4 KLM

to this picture. Micro-invertebrate traffic at Buoy Zero, although specific, highlights the

daily variability and intensity on exposed reef faces. Although micro-invertebrates

remain scarcely studied, and have yet to be considered as an important factor in coral

recruitment ecology, evidence continues to emerge on species-specific behavior and the

implications thereof.

Reference:

Abele, L. G. 1974. Species diversity of decapod crustaceans in marine habitats. Ecology

55:156-161.

Baums, I. B., M. W. Miller, and A. M. Szmant. 2003. Ecology of a corallivorous

gastropod, Coralliophila abbreviata, on two scleractinian hosts. I: population

structure of snails and corals. Marine Biology 142:1083-1091.

Birrell, C. L., L. J. McCook, and B. L. Willis. 2005. Effects of algal turfs and sediment

on coral settlement. Marine Pollution Bulletin 51:408–414.

Birrell, C. L., L. J. McCook, B. L. Willis, and L. Harrington. 2008. Chemical effects of

macroalgae on larval settlement of the broad- cast spawning coral Acropora

millepora. Marine Ecology Progress Series 362:129–137.

Bruckner R. J., A. W. Bruckner, and E. H. Jr. Williams. 1997. Life history strategies of

Coralliophila abbreviata lamarck (Gastropoda: Coralliophilidae) on the southeast

coast of Puerto Rico. Proceedings of 8th International Coral Reef Symposium

1:627–632.

Carleton, J. H., and P. W. Sammarco. 1987. Effects of substratum irregularity on success

of coral settlement: quantification by comparative geomorphological techniques.

Bulletin of Marine Science 40:85–98.

56

Page 57: CW Invert and 3D Thesis4 KLM

Carpenter, R. C. 1986. Partitioning herbivory and its effects on coral reef algal

communities. Ecological Monographs 56:343–363.

Carpenter, R. C. 1986. Invertebrate predators and grazers. Life and death of coral reefs.

Chapman and Hall, New York, New York, USA.

Coen, L. D., 1988a. Herbivory by crabs and the control of algal epibionts on

Caribbean host corals. Oecologia 75:198-203.

Downing N., and C. R. El-Zahr. 1987. Gut evacuation and filling rates in the rock-boring

sea urchin, Echinometra mathaei. Bulletin of Marine Science 41:579-584.

Edmunds, P. J., and R. C. Carpenter. 2001. Recovery of Diadema antillarum reduces

macroalgal cover and increases abundance of juvenile corals on a Caribbean reef.

Proceedings of the National Academy of Sciences 98:5067-5071.

Glynn, P. W. 1983a. Crustacean symbionts and the defense of corals: coevolution on the

reef? Pages 111-178 in M. H. Nitecki, editor: Coevolution. University of Chicago

Press, Chicago, Illinois, USA.

Glynn, P. W. 1983b. Increased survivorship in corals harboring crustacean symbionts.

Marine Biology Letters 4:105-111.

Glynn, P. W. 1997. Bioerosion and coral-reef growth: a dynamic balance. Pages 68-95 in

C. Birkeland, editor: Life and Death of Coral Reefs. Chapman and Hill, New

York, New York, USA.

Griffiths, A. M., and B. W. Frost. 1976. Chemical communication in the marine

planktonic copepods Calanus pacificus and Pseudocalanus sp. Crustaceana

30:1-8.

Harrington, L., K. Fabricius, G. De’ath, and A. Negri. 2004. Recognition and

57

Page 58: CW Invert and 3D Thesis4 KLM

selection of settlement substrata determine post-settlement survival in corals.

Ecology 85:3428-3437.

Hartmann, A. C., K. L. Marhaver, V. F. Chamberland, S. A. Sandin, and M. J. A.

Vermeij. 2013. Large birth size does not reduce negative latent effects of harsh

environments across life stages in two coral species. Ecology 94:1966-1976.

Hayes, J. A. 1990. Distribution, movement and impact of the corallivorous gastropod

Coralliophila abbreviata (Lamarck) on a Panamanian patch reef. Journal of

Experimental Marine Biology and Ecology 142:25-42.

Huston, M. A. 1985. Patterns of species diversity on coral reefs. Annual Review of

Ecological Systems 16:149-177.

Iglesias, J. C. M., and J. E. G. Raso. 1999. The crustacean decapod communities of

three coral reefs from the southwestern Caribbean sea of Cuba: species

composition, abundance and structure of the communities. Bulletin of Marine

Science 65:539–557.

Jorissen, H. 2013. The abundance, grazing rates and feeding preference of mesograzers

on Curacaoan coral reefs. Master thesis. University of Wageningen, Wageningen,

The Netherlands.

Kuffner, I. B., and V. J. Paul. 2004. Effects of the benthic cyanobacterium Lyngbya

majuscula on the larval settlement of the reef corals Acropora surculosa and

Pocillopora damicornis. Coral Reefs 23:455–458.

Kuffner, I. B., L. J. Walters, M. A. Becerro, V. J. Paul, R. Ritson-Williams, and K.

Beach. 2006. Inhibition of coral recruitment by macroalgae and cyanobacteria.

Marine Ecology Progress Series 323:107–117.

58

Page 59: CW Invert and 3D Thesis4 KLM

Lau, S. C. K., and P. Y. Qian. 2001. Larval settlement in the serpulid polychaete

Hydroides elegans in response to bacterial films: an investigation of the nature of

putative larval settlement cue. Marine Biology 138:231-328.

Lubchenco, J., and S. D. Gaines. 1981. A unified approach to marine plant-herbivore

interactions. I. Populations and communities. Annual Review of Ecological

Systems 12:405-437.

Marhaver, K. L., M. J. A. Vermeij, F. Rohwer, and S. A. Sandin. 2013. Janzen-Connell

effects in a broadcast-spawning Caribbean coral: distance-dependent survival of

larvae and settlers. Ecology 94:146-60.

Mason, B., M. Beard, and M. W. Miller. 2011. Coral larvae settle at a higher frequency

on red surfaces. Coral Reefs 30:667-676.

Miller, A. C. 1981. Cnidarian prey of the snails Coralliophila abbreviata and C. caribaea

(Gastropoda: Muricidae) in Discovery Bay, Jamaica. Bulletin of Marine Science

31:932-934.

Ogden, J. C. 1977. Carbonate-sediment production by parrotfish and sea urchins on

Caribbean reefs: reef biota. Pages 281-288 in H. Stanley et al. editors: Reefs and

Related Carbonate-Ecology and Sedimentology No 4. American Association

Petrol Geology, Tulsa, Oklahoma, USA.

Sammarco, P. W. 1980. Diadema and its relationship to coral spat mortality: grazing,

competition, and biological disturbance. Journal of Experimental Marine Biology

and Ecology 45:245–272.

Sammarco, P. W. 1982. Echinoid grazing as a structuring force in coral communities:

whole reef manipulations. Journal Experimental Marine Biology and Ecology

61:31-55.

59

Page 60: CW Invert and 3D Thesis4 KLM

Sammarco, P. W. 1985 The Great Barrier Reef versus the Caribbean: comparisons of

grazers, coral recruitment patterns and reef recovery. Proceedings of the 5th

International Coral Reef Symposium 4:391–398.

Sandin, S.A., J. E. Smith, E. E. DeMartini, E. A. Dinsdale, S. D. Donner, A. M.

Friedlander, T. Konotchick, M. Malay, J. E. Maragos, D. Obura, O. Pantos, G.

Paulay, M. Richie, F. Rohwer, R. E. Schroeder, S. Walsh, J. B. C. Jackson, N.

Knowlton, and E. S. mail. 2008. Baselines and degradation of coral reefs in the

Northern Line Islands. PLoS ONE 3:e1548.

Steneck, R. S., and M. N. Dethier. 1994. A functional group approach to the structure of

algal-dominated communities. Oikos 69:476–498.

Steneck, R. S., and J. C. Lang. 2003. Rapid assessment of Mexico’s Yucatan reef in 1997

and 1999: pre- and post-mass bleaching and hurricane Mitch (stony corals, algae

and fish). Atoll Research Bulletin 496:294–317.

Waddell, B., and J. R. Pawlik. 2000. Defenses of Caribbean sponges against invertebrate

predators. I. Assays with hermit crabs. Marine Ecology Progress Series 195:125-

132.

Vermeij, M. J. A., and S. A. Sandin 2004. Density-dependent settlement and mortality

structure the earliest life phases of a coral population. Ecology 89:1994-2004.

Vermeij, M. J. A., J. E. Smith, C. M. Smith, R. V. Thurber, and S. A. Sandin 2009.

Survival and settlement success of coral planulae: independent and synergistic

effects of macroalgae and microbes. Oecologia 159:325-336.

Williams, R. R., S. N. Arnold, N. D. Fogarty, R. S. Steneck, M. J. A. Vermeij, and V. J.

Paul. 2009. New perspectives on ecological mechanisms affecting coral

60

Page 61: CW Invert and 3D Thesis4 KLM

recruitment on reefs. Smithsonian Contributions to the Marine Sciences 38:437-

452.

Williams, I., and N. Polunin. 2001. Large-scale associations between macroalgal cover

and grazer biomass on mid-depth reefs in the Caribbean. Coral Reefs 19:358-366.

Wolf, A. 2012. The role of macroalgae on the corallivorous fireworm Hermodice

carunculata on coral reef resilience in the Caribbean. Dissertation. University of

Bremen, Bremen, Germany.

Acknowledgements:

I would like to foremost thank Kristen Marhaver for her consistent dedication and

support helping to not only realize this project and research effort, but her continual drive

to further my fundamental education of ecological principles and techniques found within

the scientific field of ecology at large. In addition, Mark Vermeij and Kristen Marhaver

were easily accessible, yielding a wealth of information ranging from field methods to

avenues of new research. I can’t thank them enough for making CARMABI feel like a

second home during this internship. I would further like to thank CARMABI for use of

their research facilities as well as StimulIT for taking the time to support this

investigation. Finally, I would like to thank CARMABI and DiveVersity for their help

with the logistical organization of this project.

61

Page 62: CW Invert and 3D Thesis4 KLM

Appendix:

During the fieldwork portion of this internship that occurred between 1 September

2013 and 31 March 2014, additional undertakings and projects that are not mentioned in

the above report were carried out with either my assistance or leadership. These efforts

are listed below.

-Revamped and secured additional medical supplies to bring CARMABI marine

field station up to par in dealing with emergency situations should they arise.

-Set up recycling for CARMABI marine field station so that the center could act

as a local leader in sustainable living.

-Created a coral reef husbandry protocol in conjunction with Ari Muskat for

future researcher’s efforts in maintaining coral health within the CARMABI wet

lab.

-Established a coral nursery in conjunction with Robin Hogenboom to create the

ability to research brooding coral colonies over extended periods of time and

promote the sustainable harvest of corals.

-Assisting MSc student Cas Grupstra in cage exclusion setup experiment (Dec.

2013).

-Assisting Valerie Chamberland in Diploria spawning monitoring dives (Sep.

2013).

-Setting up materials and assisting on research dives for the UvA master student

class project by Petra Nobs and Michiel Klaassen (Jan. 2014).

Pilot projects not included in above report

62

Page 63: CW Invert and 3D Thesis4 KLM

- Experiment 7 M. Faveolata Conditioned water pilot

- Experiment 8 M. Faveolata settlement and interaction test pilot

- Experiment 9 M. Faveolata 3D tile choice experiment

Experiment 7 (Pilot): Invertebrates were added to 0.45 m filtered seawater. Incubated

for 24 hr before taken out. Conditioned water was then added to 250 mL plastic bins with

M. faveolata larvae and incubated in the lab.

Expectations: Different invertebrate conditioned seawater would have an effect on the

rate of M. faveolata settlers over time.

Experimental Power: replication seven, 30 larvae per settlement bin/sample. Measured

at 36 hr and every 48 hr after that; duration eight days.

Results: Non-significant results in comparison to control except for D. antillarum

treatment. This could indicate that larvae do not make settlement choices based on water

column invertebrate signals.

18 36 54 72 90 108 126 144 162 180 198 216 234

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

Long spined urchin

Red striped hermit

Boring rock urchin

Control

Bearded fireworm

Sea lettuce slug

Time (hours)

Ave

rage

# o

f Set

tler

s ov

er e

ach

trea

tmen

t

63

Page 64: CW Invert and 3D Thesis4 KLM

Figure 1a: Settlement of M. faveolata over time within conditioned seawater. Standard

error of control shown.

Experiment 8 (Pilot): M. faveolata larvae settled on fresh terracotta tiles. Lower

settlement than expected. Tiles with settlers were introduced to circular containers with

single micro-invertebrates and incubated in aquariums.

Expectations: Significant mortality for fireworms, exploratory interaction for other

organisms.

Experimental Power: Replication four, three replicates with one settler per tile, one

replicate with 2-24 settlers per tile; measured at 18 hr, and every 48 hr following that for

six days.

Results: Low replication and power made experiment exploratory. However, noticeable

decrease in survivorship with red striped hermit crabs gave the initial signal that pushed

us to explore this interaction further. Furthermore, fireworm treatment did not show

expected mass mortality, which was examined in more depth with follow-up

experiments.

0 20 40 60 80 100 120 140 1600

0.2

0.4

0.6

0.8

1Long spined urchinRed striped hermitBoring rock urchinControlBearded firewormSea lettuce slug

Time (hours)

Ave

rage

% o

f sur

vivo

rshi

p ac

ross

tr

eatm

ents

64

Page 65: CW Invert and 3D Thesis4 KLM

Figure 2a: M. faveolata settler survivorship over time with various introductions of

invertebrate species. Standard error of control shown.

Experiment 9 (Pilot): M. faveolata larvae introduced to various 3D tiles of color and

shape.

Expectations: Significant settlement preference for color and complexity.

Experimental Power: Replication nine for complexity, three for color preference;

measured every 24 hr for 10 days.

Results: Low settlement on all tiles due to newly printed 3D surfaces, which were not

pre-conditioned in seawater before beginning experiment.

65