Cross Shelf Patterns in Habitat Selectivity of Hawkfish (Family

Cross Shelf Patterns in Habitat Selectivity of Hawkfish (Family:

Cirrhitidae) in the Red Sea; with a Special Case of Varying Color

Morphs in Paracirrhites forsteri.

Thesis by

Veronica Chaidez

In Partial Fulfillment of the Requirements

For the Degree of

Master of Science in Marine Science

King Abdullah University of Science and Technology, Thuwal,

Kingdom of Saudi Arabia

Approval Date: December 2014 2

The thesis of Veronica Chaidez is approved by the examination committee.

Committee Chairperson: Dr. Michael Berumen

Committee Member: Dr. Xabier Irigoien

Committee Member: Dr. Stein Kaartvedt

Veronica Chaidez

ABSTRACT

Cross Shelf Patterns in Habitat Selectivity of Hawkfish (Family: Cirrhitidae)

in the Red Sea; with a Special Case of Varying Color Morphs in

Paracirrhites forsteri.

Veronica Chaidez

Not much is known about hawkfish worldwide including those that occur in the understudied Red Sea reef system. Hawkfishes are small reef predators that perch in ambush-ready positions and shelter within or on various substrates including live and dead coral. The aim of this study was to look at the distribution and abundance patterns of Red Sea hawkfishes across an inshore and offshore gradient and to investigate the use of benthic habitats. This study was conducted on three inshore, four midshore, and two offshore reefs with surveys at 8 meters and along the reef crest. In total, three species were documented: Paracirrhites forsteri, Cirrhitichthys oxycephalus, and Cirrhitus spilotoceps. We found clear distinctions between depth zones and between continental shelf positions. Cirrhitichthys oxycephalus only occurs at the reef slope and Cirrhitus spilotoceps is only found on reef crests. Paracirrhites forsteri was the most abundant species across all reefs and was found in four varying color morphs. Morph 1 showed the most evidence of being a generalist as it utilized the greatest number of substrates. All three species were more abundant on midshore and offshore reefs which have healthier, 5 intact coral communities. Coral cover is a good indicator of hawkfish abundance even when the species in question does not utilize live coral directly.

Keywords: coral cover, habitat selectivity, hawkfish, Red Sea

ACKNOWLEDGMENTS

I would like to thank Drs. Michael Berumen and Darren Coker for their help in experimental design and data analysis. A special thank you to Dr. Darren Coker for joining me in the field and teaching me many tricks of the trade. Thank you also for teaching me the “whys” of the various techniques I used. Your insight and friendship were invaluable.

A warm thank you goes to the entire team of CMOR, especially my faithful skippers that took me out every day to any reef that I needed. I could not have completed this project without you. Skukran.

Thank you to Tane-Sinclair Taylor for the gorgeous close-ups of these beautiful fish.

Thank you Maha Khalil for creating the map for my sampling sites.

I’d like to thank the following people for joining me in the field: Alex Kattan, Alison

Monroe, Amr Gusti, Darren Coker, Ioannis Georgakakis, Maddie Emms, Manalle Al-

Salamah, May Roberts, Nora Kandler, Noura Ibrahim, Pedro De La Torre, Remy Gatins,

Rodrigo Villalobos, Royale Hardenstine, Song He, and Tullia Terraneo. You were excellent buddies.

Thank you to my roommate Marcela Herrera for being the domestic one and having been the one to brave IKEA. Thank you for being there for me until the very end of this thesis.

Thank you Manalle for all the love.

Thank you to my labmates for keeping me young with all the laughs. 7

TABLE OF CONTENTS

• Examination Committee Approvals Form………………….…………...…….p.2 • Copyright Page.…………………….………………….………………………p.3 • Abstract……………………………………....…………………...……..…….p.4 • Acknowledgments…………………………....……………………..…...... ….p.6 • Table of Contents……………………………..…………………….…....…....p.7 • List of Figures…………………………………………………………………p.8 • List of Tables.………………….....…………………………………………...p.9 • 1. Introduction…………..……………...…………………………..………....p.10

o 1.1 Habitat selectivity……………………………………………...... p.10 o 1.2 Habitat structure……………………………………….….…...... p.12 o 1.3 Research questions………………………………………...…...…p.13 • 2. Methods……………………………………………………..…..….…...... p.14 o 2.1 Study site………….………..……………………………..………p.14 o 2.2 Study species…...……………………………...……………….....p.14 o 2.3 Surveys of fish and microhabitat………………………….……....p.15 o 2.4 Data Analysis……………………………………………….…..…p.16 o 2.5 Microhabitat selection……………..……………………….……..p.16 • 3. Results………………………..……………………………………...... …p.18 o 3.1 Benthic profiles……………………...……………………...... p.18 o 3.2 Abundance……………………. …………………………….……p.19 o 3.3 Habitat selectivity ………...……………….………………….….p.20 o 3.4 Paracirrhites forsteri: four color morphs ……………………..….p.21 • 4. Discussion………………………..………………………………………...p.23 • 5. Conclusions………………………………………………………………...p.28 • References………………………………..………………………………..….p.30 • Appendices………………………………………………………………..…..p.33

LIST OF FIGURES

1. Figure 1. Sampling sites near Thuwal of the Central Red Sea. Green circles

indicate inshore reefs, red circles indicate midshore reefs, and white circles

indicate offshore reefs. Six sites were sampled at each of the three shelf

positions………………………………………………………………………..p.33

2. Figure 2. Various color morphs of the Red Sea freckled hawkfish (Paracirrhites

forsteri). a) Morph 1, b) Morph 2, c) Morph 3, d) Morph 4……………...……p.34

3. Figure 3. Substrate composition at three shelf positions: inshore, midshore, and

offshore; and at two zones: crest and slope. a) inshore crest, b) inshore slope, c)

midshore crest, d) midshore slope, e) offshore crest, f) offshore slope….…….p.35

4. Figure 4. Mean abundances of three species of hawkfish across two shelf

positions and two zones. a) midshore crest, b) midshore slope, c) offshore crest, d)

offshore slope Asterisks signify a significant difference between the two columns.

…………………………………………………..………………………..…….p.36

5. Figure 5. Averages of the four P. forsteri color morphotypes at two shelf positions

and two reef zones. a) midshore crest, b) midshore slope, c) offshore crest, d)

offshore slope. Asteriks denote significant differences in abundances. In panel a),

double asterisks mean morph 2 and 3 are not different from each other but both

are different from morph 1……………………………………………………..p.36

LIST OF TABLES

1. Table 1. Percentage of live hard coral cover across shelf positions: inshore,

midshore, and offshore and across reef zones: crest and slope…………….….p.37

2. Table 2. Selectivity index of three species of hawkfish across an inshore to

offshore gradient for 14 categories of substrate. “=” denotes a category that was

used in proportion to its availability, “+” means the category was used in greater

proportion to its availability, “-“ means the category was used at a lower

proportion than its availability, and “U” means the category went unused. “NA”

means the substrate was not present. Subscripts represent Manly’s standardized

selection ratio (B)……………………………………………………….…...... p.38

3. Table 3. Selectivity index of the four color morphs of P. forsteri across an inshore

to offshore gradient for 14 categories of substrate. “=” denotes a category that was

used in proportion to its availability, “+” means the category was used in greater

proportion to its availability, “-“ means the category was used at a lower

proportion than its availability, and “U” means the category went unused. “NA”

means the substrate was not present. Subscripts represent Manly’s standardized

selection ratio (B)……………………………………………………..……….p.39

INTRODUCTION

1.1 Habitat Selectivity

Organisms use their environment in a variety of ways and become adapted to their

environment to different degrees. The degree of specialization for any organism lies on a

continuum with generalists using a variety of habitats or resources, specialists using a

narrower range of resources, and highly specialized organisms that optimize the use of

one or two resources. This type of partitioning, allows for a host of diverse organisms and

life strategies to co-exist and flourish in the same area (Morris 1996). The levels of biodiversity and speciation that we find in an ecosystem are functions of habitat selection strategies employed by members of the ecological community.

Habitat selection expressed by species and populations gives us basic ecological knowledge of a system. The manner in which resources are partitioned among organisms has direct effects on population densities, species interactions, and the assemblage of ecological communities (Fretwell and Lucas 1970, Fretwell 1972, Rosenzweig 1974,

Morris 2003). The various habitat selection strategies employed by organisms may also give us insight into their evolutionary trajectories (Morris 2003).

Knowing the distribution patterns of a particular resource or habitat, allows us to

better estimate abundance patterns of species that rely on them. For example, data on

resource selection, informed management’s decision in Southcentral Wyoming to remove

excess feral horses from the environment (Crane et al. 1997). Quantifying habitat and

resource use has been a common practice in terrestrial management, especially for large

mammals such as moose (Rounds 1981, Cederlund and Okarma 1988, Van Beest et al.

2010). In order to manage both moose populations and young pine seedlings which 11

moose like to eat, Van Beest et al. (2010) conducted a habitat use study to determine if

artificial feeding sites made a difference in foraging behavior. There are many habitat

selectivity studies that also look at seasonality trends (Kelt et al. 1994, Hancock and

Wilson 2003). Whittingham et al. (2005), assessed how much yellowhammers (Emberiza

citrinella) rely on the boundaries of farmland in order to inform conservation

management. These tend to be cost-effective projects when employing a mark and

recapture sampling technique.

Coral reefs are diverse ecosystems made up of complex interconnected

relationships. The diversity of reef species is likely due to the multitude of available

niches within these systems. These niches can be characterized based on an animal’s

habitat type and/or food source (Munday 2000, Bonin 2011, Lawton and Pratchett 2012,

Lawton et al. 2012). Quantifying habitat use helps us better understand whether an

organism is a specialist, generalist, or something in between (Gardiner and Jones 2005,

Pratchett et al. 2012). For example, Lawton et al. (2012) looked at the feeding behavior

of coralivorous butterflyfishes to assess their level of specialization and help predict how

these species would be affected in the case of disturbance. It is a common theme within

the ecological literature that specialists are more vulnerable to extinction when faced with

habitat destruction or alteration (McKinney 1997, Munday 2004). Essentially, the

resistance and resilience of specialized organisms is dependent on the stability of their critical resources (McKinney 1997, Vazquez and Simberloff 2002). This is especially concerning in light of climate change predictions with coral reefs being one of the most vulnerable ecosystems on Earth (Graham et al. 2006, Pratchett et al. 2012). Habitat 12 selectivity studies are important for predicting species vulnerabilities to environmental alterations.

1.2 Habitat structure

Biodiversity is often linked to ecosystem complexity. Coral reefs are made up of a plethora of reef-building organisms and these offer the rest of the reef fauna complex structures to inhabit and forage from. Branching corals are often highlighted in coral reef studies for their contribution to reef structure and complexity as well as their vulnerability to mechanical and chemical stressors (Brown and Suharsono 1990, Gleason

1993, Loya et al. 2001, Bonin 2011). Many small fish use branching corals to feed from, lay their eggs in, and shelter in (Shulman 1984, DeMartini 1996, Wilson et al. 2006).

Complex habitat is often crucial for the survival of small juvenile fish. We see evidence of this in kelp forests, seagrass beds, mangroves, and coral reefs (Dayton 1985, Alongi

2002, Duarte 2002, Steneck 2002, Graham et al. 2006). Bonin (2011) found that

Pomacentrid recruits seek out branching Acropora colonies and Wilson et al. (2006) showed that dead colonies are also important to this group. Such studies elucidate that structural complexity may be just as or more important than live coral.

With coral reefs suffering countless disturbances around the globe and resource managers trying to make the best decisions with limited funds, information on habitat selection becomes a good proxy for vulnerability risk. Modern disturbances altering coral reefs worldwide are storms, sediment dumping, eutrophication, boat traffic, rising temperatures and sea levels, as well as changes in pH (Pandolfi et al. 2005). Coral 13

colonies with high surface area such as branching corals, tend to be the ones with the

highest risk of bleaching and mechanical damage from boat anchors and storms.

1.3 Research questions:

Hawkfish are poorly studied worldwide, and this is particularly true in the understudied Red Sea reef system (Berumen et al. 2013). The Red Sea is home to many endemics including Cirrhitus spilotoceps (Gaither 2013), the stocky hawkfish.

Hawkfishes are small reef predators that perch in an ambush-ready position on various substrata including live and dead coral (Hobson 1974, Myers 1989, Hobson 1994). Their diet consists of small fish and crustaceans. The objective of this study is to document patterns of distribution, abundance, and habitat use of hawkfishes in the central Red Sea across the continental shelf and between two depths.

The aims of the study are as follows:

What is the abundance of hawkfish across an inshore to offshore shelf position gradient and across reef zones?

Are each species of hawkfish selecting for particular microhabitats?

Is each color morphotype of P. forsteri selecting for particular microhabitats?

METHODS

2.1 Study site

This study took place in the central Red Sea. In total nine reefs were surveyed: three inshore, four midshore, and two offshore (Figure 1). These reefs were chosen due to the ease of accessibility from KAUST and in order to look at the distributions of hawkfish across an inshore to offshore gradient.

2.2 Study species

The family Cirrhitiade has 34 species, four of them are found in the Central Red

Sea. Our study aimed to look at the selection patterns of Red Sea hawkfishes:

Paracirrhites forsteri (Schneider 1801) (freckled hawkfish), Cirrhitus oxycephalus

(Bleeker 1855) (pixie hawkfish), Cirrhitus spilotoceps (Schultz, L.P 1950, Randall 1963,

Gaither 2013) (stocky hawkfish), and Oxycirrhites typus (longnose hawkfish).

Paracirrhites forsteri is distributed across the Red Sea, South Oman, Hawai’i, South

Japan, South Africa, and New Caledonia and mainly feeds on small fishes and shrimp

(Hiatt and Strasburg 1960, Randall 1963, Hobson 1974, Randall 1985, Nakabo 2002).

Cirrhitus oxycephalus occurs in the Red Sea and southern Oman to Panama as well as from South Japan to South Africa and New Caledonia. It is also found on seaward reefs.

The range of C. spilotoceps is from the Red Sea to the Gulf of Oman (Randall 1963,

Gaither 2013). Oxycirrhites typus’ range includes the Red Sea to Panama and South

Japan to Mauritius and New Caledonia. 15

In the Red Sea, P. forsteri, exhibits at least four distinct color morphotypes. At present, all four color morphs are considered the same species. From here in, we have labeled the color morphs 1 through 4. Morph 1: distinguished by two prominent yellow horizontal stripes across the majority of its body, morph 2: white ventral side and the dorsal half may vary in a combination of yellow, red, and black, morph 3: half bright red and half dark grey, morph 4: completely brown except for a subdued orange tail (Figure

2).

2.3 Surveys of fish and microhabitats

Surveys were conducted at 18 sites with two zones at each site: the reef slope and the reef crest. Each site was randomly selected along the exposed side of each reef due to the higher coral cover found on the seaward side. Three 30 x 1 meter replicate transects were laid out per reef per site and per zone. Visual surveys were done on SCUBA at the reef slope at approximately 8 meters depth and on snorkel when depths were 3 meters or less at the crest. Differences in sampling technique are not expected to influence our results. For each individual fish, the substrate it was perched on and the fish’s length to the nearest 5 centimeters were noted. To document substrate composition and available habitat at each transect, substrate directly underneath each half meter mark was recorded.

Substrate was classified as one of fourteen categories including some of the most common genera of coral: (1) Acropora, (2) Pocillopora, (3) Stylophora, (4) Millepora,

(5) Porites, (6) other live hard coral, (7) Xeniidae, (8) other soft coral, (9) dead coral

(dead but structurally intact colonies), (10) rubble, (11) sand, (12) pavement, (13) turf or coralline algae, and (14) other sessile organisms. Pavement refers to clean slabs of rock, 16

usually partially covered in encrusting algae. Category 13 changes somewhat from the

crests to the slopes. On a reef crest “turf algae” refers to the leafy green and mossy algae

that tends to grow on dead coral. At the slopes of midshore and offshore reefs, this category referred to all sorts of algae including coralline and encrusting algaes.

2.4 Data analysis

Histograms depicting average abundances of substrate and fish were created in

Microsoft Excel©. The standard error was calculated for every descriptive statistic.

Hawkfish abundance data was log transformed and square root transformed in order to

test for normality. Normality was not found so we employed non-parametric tests (Mann-

Whitney U and Kruskall-Wallis) to find significant differences between groups.

2.5 Microhabitat selection

Our study followed Design I, sampling protocol A from Manly et al. 2002.

Selection ratios were calculated in order to determine if fish were choosing a microhabitat in higher, lower, or equal proportion to what is available (for equation see

Manly et al. 2002). Selection ratios were calculated for each substrate category at each continental shelf position and reef zone for each species of hawkfish and color morphotype of the freckled hawkfish. Selection ratios for each reef site were standardized so that they equal 1 using Manly’s standardized selection ratio (B). Alpha values were adjusted for pairwise comparisons using Bonferroni-corrected 95% confidence intervals 17

(Miller 1981). A confidence interval that encompasses 1 means the microhabitat is being used in proportion to its availability. An interval that is less than 1 means the microhabitat is underused or avoided and an interval greater than 1 means the microhabitat is chosen or selected for.

RESULTS

3.1 Benthic profiles

Across the continental shelf, the benthic profiles of the reef crests change. Almost half of the substrate on inshore crests is turf algae (45.7% SE +/- 4.4). The next most common substrate is pavement (21.8% SE +/- 2.6). Of the coral categories, Porites and other genera of live coral are the most common (6.0% SE +/- 1.6, 4.5% SE +/- 1.1 respectively) (Figure 3a). At midshore crests we see closer proportions of turf algae

(17.9% SE +/- 4.2) and several genera of coral (Acropora: 9.5% SE +/- 1.3, Pocillopora:

12.5% SE +/- 1.7, Stylophora: 11.2% SE +/- 2.2) including the soft coral family Xeniidae

(15.1% SE +/- 3.8) (Figure 3c). The offshore reefs have the highest percentage of coral cover spread out among: Acropora 4.4% SE +/- 0.9, Pocillopora 19.0% SE +/- 2.7,

Stylophora 11.3% SE +/- 1.2, Millepora 7.4% SE +/- 2.4, and other live hard coral: 4.6%

SE +/- 1.0. Pavement (19.4% SE +/- 2.3) and turf algae (13.9% SE +/- 2.2) are also prominent (Figure 3e).

Inshore slopes were mainly composed of sand (35.6% SE +/- 4.0) and rubble

(24.8% SE +/- 2.3). The next abundant categories were pavement (11.4% SE +/- 2.3) and turf algae (13.0% SE +/- 1.9). Live coral cover was 4.0% SE +/- 1.1 (Figure 3b).

Midshore and offshore slopes showed similar profiles with Pocillopora being the dominant coral group (18.0% SE +/- 2.3 at midshore positions and 16.0% SE +/- 2.3 at offshore positions). Midshore reefs had about twice as much Acropora cover than offshore reefs (9.1% SE +/- 1.5 and 4.6% SE +/- 1.3 respectively). Overall, live hard coral cover (47.4% at midshore and 42.0% at offshore) (Table 1) and soft coral cover 19

(13.6% at midshore and 13.2% at offshore) were similar across midshore and offshore

positions (Figures 3d and 3f).

The habitat composition of the reefs differs when moving between the crest and

the slope. On inshore reefs, one finds about 30% more turf algae at the crest whereas the

slopes are dominated by sand and rubble. The crests and slopes of midshore reefs and

offshore reefs share similar profiles with offshore reefs having more substrate covered in

coralline, encrusting, and turf algae (Figure 3). The crests and slopes of offshore reefs

share similar coral cover (47.2% and 42.0%) while the crests have twice as much

Xeniidae than is found at deeper depths (14.0% SE +/- 3.5, 7.1% SE +/- 1.3 respectively)

(Figure 3e and 3f).

3.2 Abundance

A total of 295 fish were recorded across all locations. Of the four species of hawkfish found in the Red Sea, three were documented. Overall, P. forsteri was the most abundant species with an average of 2 (at reef crest) and 4 (at reef slope) individuals per transect on mid and offshore reefs (Figure 4). Paracirrhites forsteri was completely absent from inshore reefs. Of the other two species, only 5 fish (2 C. oxycephalus and 3

C. spilotoceps) were spotted on inshore sites. Cirrhitus spilotoceps was only recorded on the reef crest while C. oxycephalus was only recorded on the reef slope (Figure 4). In contrast, P. forsteri was recorded in both zones but was twice as abundant on reef slopes

(Figure 4).

As normality was not found in the abundance data even after it was log

transformed and square root transformed (Kolgomorov-Smirnov d = 0.20319, p < 0.01),

non-parametric tests were used. To test differences among shelf positions and reef zones,

we used a t-test to compare the two species found at each zone. The results from the

Mann-Whitney U test are as follows: midshore crest (U = 88.500, p = 0.033); midshore slope (U = 68.500, p = 0.002); offshore crest (U = 120.500, p = 0.189); and offshore slope (U = 23.500, p = 0.000).

3.3 Habitat Selectivity

Table 2 shows the selectivity ratios for three hawkfish species on 14 substrate categories. Manly’s standardized selection ratio (B) is attached to each selectivity value as an index of how likely a category would be utilized if all categories were available at equal capacities. At the midshore crest zones, P. forsteri were seen utilizing Acropora,

Pocillopora, Stylophora, and other live hard corals as well as structurally intact dead coral colonies (Table 2). At the midshore slope zones, P. forsteri positively selected for

Acropora corals and negatively, or avoided, soft coral and turf algae (Table 2). According to the B value, an individual is twice as likely to use an Acropora colony than a

Pocillopora colony. Colonies of Pocillopora, Stylophora, Porites, and other live hard coral were used in expected proportions as were dead colonies and other sessile organisms. As we move offshore, the number of utilized categories increases. At reef crests, P. forsteri uses all the live corals in equal proportions except Pocillopora although this selectivity value is close to encompassing 1 and so the negative selection should be 21

interpreted with caution (Table 2). At the offshore crest, they also seem to avoid

pavement. All of the live coral categories are utilized at the reef slope with Pocillopora

being positively selected for and Porites being underutilized or avoided (Table 2). Turf

algae is also underutilized.

Cirrhitus oxycephalus was seen perched on Acropora, Millepora, other live hard

coral, other soft coral, dead coral colonies, and turf algae at the midshore reefs (Table 2).

At the offshore reefs, it seemed to avoid or underuse Pocillopora. It used Millepora,

Porites, other live coral, rubble, and turf algae at expected proportions based on substrate

availability (Table 2).

Cirrhitus spilotoceps was present on turf algae at inshore, midshore, and offshore

crest zones. Only at the offshore sites was there positive selection for this substrate

(Table 2). At offshore positions it underused pavement. It also utilized more categories

than at the midshore and inshore positions including Millepora and other live coral. All

zones that contained only one individual were omitted from Tables 2.

3.4 Paracirrhites forsteri: four color morphs

A Kruskall-Wallis test was done to compare the abundance among color morphs

(Kruskall-Wallis, p = 0.0035). A posterior comparison of p values showed significant

differences between morph 1 and morph 4 for the midshore (p = 0.0032) and offshore (p

= 0.0001) slope habitats. In the midshore crest habitat there was a difference when morph

1 was compared with morphs 2 and 3 (p < 0.0105) but no difference between morphs 2

and 3 (Figure 5). 22

In the habitats that did not exhibit all four morphotypes, a more appropriate

analysis was conducted where the non-present color morphs were removed. A Mann-

Whitney test showed differences in the abundances at the offshore crest habitat where

only two morphotypes (morph 1 and morph 4) were found (U = 2.000, p = 0.000).

Morph 1 is the most abundant morph across the continental shelf and is found on both crests and slopes (Figure 5). Morphs 2, 3, and 4 are more common on the slope than the crest (Figure 5). None of the morphs were found on inshore reefs.

All zones that contained only one individual were omitted from Table 3. These included most of the inshore zones and one offshore crest zone for Morph 4. All four morphotypes utilized at least one live coral category at each of the zones. Morph 1 used

Acropora, Pocillopora, Stylophora, and dead colonies in both reef zones on midshore and offshore reefs (Table 3). Everywhere, these were equally selected for except the offshore crest where Stylophora was positively selected for and Pocillopora was negatively selected for (Table 3). More live coral categories were utilized at the offshore slope habitat including Porites which was underutilized.

The other color morphs used less live coral categories but also showed evidence of using more of them at the offshore sites (Table 3). Morph 3 seemed to avoid turf algae at the slopes of midshore and offshore reefs. Morph 4 utilized the least number of substrate categories but also utilized Acropora, Pocillopora, and turf algae (Table 3) DISCUSSION

In our sampling area, three of the four Central Red Sea hawkfish species were observed. Oxycirrhites typus was not recorded, most likely because its preferred habitat of gorgonians and black corals occur at greater depths than those explored in this study.

Of the other three species, all were rare or absent from inshore reefs which lack live coral. Midshore and offshore shelf positions share similar substrate profiles as well as similar abundances of hawkfish. Midshore and offshore reefs contain a much higher percentage of live coral cover than inshore reefs, suggesting that abundances of hawkfish may be related to abundance of coral. How directly linked this association is, we can not say but it does seem to be clear that hawkfish prefer habitats where live coral is an important part of the reef mosaic. This is true even for species that were not frequently observed perched on live coral (C. oxycephalus and C. spilotoceps). There also seems to be niche partitioning as C. spilotoceps is only found at the reef crest, while C. oxycephalus is only found on reef slopes. Paracirrhites forsteri seems to be more of a generalist occupying all four habitats. On a macro-habitat scale, this family of fish seem to occupy distinct niches.

Cirrhitus spilotoceps most likely occupies the reef crest because its preferred microhabitat occurs there. Cirrhitus spilotoceps was most often found on turf algae made

up leafy green macroalgae. This substrate needs adequate sunlight and nutrients, it likely grows well on the shallow seaward side of reefs. This area is also subject to waves and surge which C. spilotoceps is well adapted to as they are able to cling to their substrate.

The wave action may also help to disturb small invertebrates that live in the algae, making the choice of turf algae an optimal perching location to catch a meal. Although 24

not empirically recorded, it was noted that this species was often seen underneath

overhangs of hard substrate (coral or pavement). In addition to its cryptic coloration, the

behavior of perching on turf algae as well as lying under overhangs, may add protection

from other visual predators as well as its prey.

The selectivity index showed that P. forsteri positively selected for Acropora and

Pocillopora at midshore and offshore sites respectively. At these sites, they also avoided slope turf algae and Porites, based on availability. If all options were available in equal proportion, the B value tells us that P. forsteri would occupy branching corals first.

Interestingly, dead coral colonies would be occupied second. This gives us more evidence for the importance of coral, specifically, structurally complex coral. Not only does P. forsteri prefer the major branching coral genera to sit on but it will also choose structurally intact dead colonies over other live massive corals. The structure of the coral head must provide a useful niche for P. forsteri. Perhaps it serves as an effective shelter from predators as evidenced by the many species of small reef fish that use corals in this manner. Branching corals also require more three-dimensional space and perhaps P. forsteri like to perch on these for the variety of ambush angles such a structure provides.

If all substrates were equally available, C. oxycephalus would most likely be found on Millepora. This is different from the other two species, again reiterating the niche partitioning that seems to be evident in this group. The second choice of substrate would be dead, structurally intact corals. The structure of the habitat seems to also matter to C. oxycephalus. On average, C. oxycephalus is the smallest of the three hawkfish species. It is found in the same zones as P. forsteri which on average has a more robust body. Perhaps there is competition for branching corals. In order to test such a 25 hypothesis, more complex behavioral experiments would have to be conducted but it might explain why C. oxycephalus is not often seen on Acropora and Pocillopora and instead occupies its own unique niche.

Of the four color morphotypes of P. forsteri, morph 1 is by far the most abundant.

It shows the most evidence of being a generalist as it is found at every zone and position except the inshore. The low abundances of morph 2 and 3 at the midshore crest compared to the slope also points to niche partitioning. This is interesting as the reef crest is a narrower zone and a smaller area will induce more competition for the best corals.

Perhaps they can not compete for habitat or food with morph 1. Also, on average, morph

1 is the largest of the 4 morphs which might make it more suitable for the rougher waters of the crest like the more robust bodied C. spilotoceps.

Like the general pattern of the species, the morphs use branching corals:

Acropora, Pocillopora, and Stylophora in addition to dead coral colonies. The selectivity index for the color morphs is not as straight forward as the one for all the species. This is due to a much lower number of fish in three of the categories. The most robust data is what we have for morph 1, it follows similar patterns as P. forsteri. Pocillopora, dead colonies, and Stylophora are the most important substrates to morph 1 although it also used other substrate categories with similar B ratios. This is evidence that morph 1 is a generalist. Not only is observed on branching corals but also other live but flat corals.

While the other color morphs seem to be more niche specific, morph 1 is able to exploit several perch configurations as well as reef zones. 26

Like morph 1, Pocillopora and Acropora are most important to morph 2 but

unlike morph 1, it does not use as many substrate categories. The coloration of morph 2 is

highly conspicuous, perhaps it prefers equally colorful corals. The preference for dead

coral colonies is most prominent for morph 3 according to the B ratio. On the offshore slope, Stylophora is also a first choice. This also gives evidence for niche partitioning among the morphs. Morph 2 prefers live corals from the genera Pocillopora and

Acropora while morph 3 is more niche specific utilizing most often dead coral colonies.

The specializations of morph 2 may have to do with the type of prey that they ambush and so they might rely on live Pocillopora corals for the small crustaceans that inhabit them. Where morph 4 lies on the specialist spectrum is hard to say. According to the selectivity index, morph 4 would choose other live, non-branching corals although this interpretation should be taken with caution as morph 4 was found in small numbers.

None of the morphs showed positive selection for a substrate except morph 1.

Details on what would be selected for if all substrate were available in equal proportions, should be interpreted with caution. There are limitations to this analysis, as habitats that are used in higher frequencies might not be able to be pulled out. Still, we can see an overall pattern for the heavier use of branching corals. For all four morphs, coral from the genus Stylophora seemed to play an more important role. Again, the complexity of their

microhabitat is important as it provides food and shelter.

Different color morphotypes are a common phenomenon in reef fish species

(Randall and Randall 1960, Thresher 1978, DeMartini and Donaldson 1996).

Paracirrhites arcatus (Arceye hawkfish) of the Indo-Pacific has been described in two distinct color morphotypes (DeMartini and Donaldson 1996). The two morphs occur at 27

different depth ranges but still overlap enough to interbreed (DeMartini and Donaldson

1996). They have also been known to have diverging ecologies which raises the question

of whether with enough time, these morphotypes will diverge to form two separate

species. (DeMartini and Donaldson 1996). Whether the color morphs of P. forsteri

represent different life stages or different species altogether requires further investigation

including genetic work.

Worldwide, we know little about the associations between fish and their coral

habitats. To better understand reef fish assemblages, it is important to better understand

these relationships although it is not an easy task as associations vary greatly among

species. For example, some studies show reef fish to be coral-obligates as is the case with most species of Gobidae and Pomacentridae (Patton 1994, Munday et al. 1997, Munday

2000, Nadler et al. 2013). Our study shows that P. forsteri prefer to sit on branching corals but not necessarily live ones. Live coral cover but also habitat complexity are important drivers of distribution and abundance patterns. And although a fish might not use live coral directly as is the case with C. spilotoceps, sufficient coral cover seems to be

a requirement for the presence of hawkfish.

The most abundant species of hawkfish, P. forsteri, most often sits on branching

corals. These are susceptible to a number of disturbances including mechanical stress

from storms and biological stress from changing water parameters. Understanding habitat

associations is important because organisms that rely on vulnerable habitat or scarce

resources, may themselves become vulnerable or endangered if these resources are

altered. 28

CONCLUSION

The distribution of hawkfish across the continental shelf shows that they favor reefs with high coral cover. Although we do not understand the larval dispersal patterns of these fish, it is probable that the lack of live coral and structurally complex coral at the inshore positions do not support adult hawkfish populations. Perhaps the productivity of the reef is not able to sustain diverse assemblages of reef fish or the lack of live corals decreases the abundance of prey. The abundance patterns of each species among reef zones may be evidence that these species occupy separate ecological niches. Each species of hawkfish may prey on different organisms and this determines what substrates are most effective.

Of the species recorded in this study, all three showed evidence of niche partitioning as they occupy different reef zones although P. forsteri occupies all but only individuals of morphotype 1. Morph 1 is a generalist. Morph 2 and 3 are more specialists.

The different color morphs may separate themselves out among certain genera of coral as a way to exploit different food sources and not overlap with each other too much. There might be some competition among the morphs as well. It would be interesting to know which color morph would dominate if all morphs were set out on a habitat at once.

Although our selectivity index gives us some indicator of what each morph prefers, we do not know which morph had the first choice on any given swath of reef.

Corals are part of healthy reef systems. Without them, biodiversity goes down.

We see that coral cover has a large impact on hawkfish abundances. Where there is little or no live coral, as in the inshore reefs, we find close to zero hawkfish. What is also 29 important is the structural complexity of the habitat. The importance of branching corals to this group of fish can not be underestimated.

REFERENCES

Alongi DM (2002) Present state and future of the world’s mangrove forests. Environmental Conservation 29, 331-349.

Berumen ML, Hoey AS, Bass WH, Bouwmeester J, Catania D, Cochran JEM, Khalil MT, Miyake S, Mughal MR, Spaet JLY, Saenz-Agudelo P (2013) The status of coral reef ecology research in the Red Sea. Coral Reefs doi: 10.1007/s00338-013-1055-8.

Bonin MC (2011) Specializing on vulnerable habitat: Acropora selectivity among damselfish recruits and the risk of bleaching-induced habitat loss. Coral Reefs doi: 10.1007/s00338-011-0843-2.

Brown BE, Suharsono (1990) Damage and recovery of coral reefs affected by El Nino related seawater warming in the Thousand Islands, Indonesia. Coral Reefs 8, 163-170.

Cederlund GN, Okarma H (1988) Home range and habitat use of adult female moose. The Journal of Wildlife Management 52(2), 336-343.

Crane KK, Smith MA, Reynolds D (1997) Habitat selection patterns of feral horses in southcentral Wyoming. Journal of Range Management 50, 374-380.

Dayton PK (1985) The ecology of kelp communities. Annual Review of Ecology and Systematics 16, 215- 245.

DeMartini EE (1996) Sheltering and foraging substrate uses of the arc-eye hawkfish Paracirrhites arcatus (Pisces: Cirrhitidae). Bulletin of Marine Science 58, 826-837.

DeMartini EE, Donaldson TJ (1996) Color morph-habitat relations in the arc-eye hawkfish Paracirrhites arcatus (Pisces: Cirrhitidae). Copeia 2, 362-371.

Duarte CM (2002) The future of seagrass meadows. Environmental Conservation 29, 192-206.

Fretwell SD (1972) Populations in a seasonal environment. Princeton University Press, Princeton

Fretwell SD, Lucas Jr HL (1970) On territorial behavior and other factors influencing habitat distribution in birds. Acta Biotheoretica 14, 16-36.

Gaither MR, Randall JE (2013) Reclassification of the Indo-Pacific hawkfish Cirrhitus pinnulatus (Forster) Zootaxa 3599(2), 189-196.

Gardiner NM, Jones GP (2005) Habitat specialisation and overlap in a guild of coral reef cardinalfishes (Apogonidae). Marine Ecology Progress Series 305, 163-175.

Gleason MG (1993) Effects of disturbance on coral communities: bleaching in Moorea, French Polynesia. Coral Reefs 12, 193-201.

Graham NAJ, Wilson SK, Jennings S, Polunin NVC, Bijoux JP, Robinson J (2006) Dynamic fragility of oceanic coral reef ecosystems. Proceedings of the National Academy of Sciences 103(22) 8425- 8429.

Hancock MH, Wilson JD (2003) Winter habitat associations of seed-eating passerines on Scottish farmland. Bird Study 50, 116-130.

Hiatt RW, Strasburg DW (1960) Ecological relationships of the fish fauna on coral reefs of the Marshall Islands Ecological Monographs 30, 65-127. 31

Hobson ES (1974) Feeding relationships of teleostean fishes on coral reefs in Kona, Hawai’i. Fishery Bulletin U.S. 72, 915-1031.

Hobson ES (1994) Ecological relations in the evolution of acanthopterygian fishes in warm-temperate communities of the northeastern Pacific. Environmental Biology of Fishes 40, 49-90.

Kelt DA, Meserve PL, Lang BK (1994) Quantitative habitat associations of small mammals in a temperate rainforest in Southern Chile: empirical patterns and the importance of ecological scale. Journal of Mammalogy 75(4), 890-904.

Lawton RJ, Pratchett MS (2012) Influence of dietary specialization and resource availability on geographical variation in abundance of butterflyfish. Ecology and Evolution 2(7), 1347-1361.

Lawton RJ, Pratchett MS, Berumen ML (2012) The use of specialisation indices to predict vulnerability of coral-feeding butterflyfishes to environmental change. Oikos 121, 191-200.

Loya Y, Sakai K, Yamazato K, Nakano Y, Sambali H, van Woesik R (2001) Coral bleaching: the winners and the losers. Ecology Letters 4, 122-131.

Manly BFJ, McDonald LL, Thomas DL, McDonald TL, Erickson WP (2002) Resource Selection by Animals: Statistical Design and Analysis for Field Studies Springer Netherlands ISBN: 978-1- 4020-0677-7.

McKinney ML (1997) Extinction vulnerability and selectivity: combining ecological and paleontological views. Annual Review of Ecology Evolution and Systematics 28, 495-516.

Miller Jr. RG (1981) Simultaneous Statistical Inference Second Edition Springer-Verlag ISBN-13: 978-1- 4613-8124-2.

Morris DW (1996) Coexistence of specialist and generalist rodents via habitat selection. Ecology 77, 2352- 2364.

Morris DW (2003) Toward an ecological synthesis: a case for habitat selection. Oecologia 136, 1-13.

Munday PL, Jones GP, Caley MJ (1997) Habitat specialisation and the distribution and abundance of coral- dwelling gobies. Marine Ecology Progress Series 152, 227-239.

Munday PL (2000) Interactions between habitat use and patterns of abundance in coral-dwelling fishes of the genus Gobiodon. Environmental Biology of Fishes 58, 355-369.

Munday PL (2004) Habitat loss, resource specialization, and extinction on coral reefs. Global Change Biology 10, 1642-1647.

Myers RF (1989) Micronesian reef fishes: a practical guide to the identification of the coral reef fishes of the tropical central and western Pacific. Coral Graphics Ter. Guam p. 298.

Nadler LE, McNeill DC, Alwany MA, Bailey DM (2013) Effect of habitat characteristics on the distribution and abundance of damselfish within a Red Sea reef. Environmental Biology of Fishes doi: 10.1007/s10641-013-0212-9.

Nakabo T (2002) Fishes of Japan with pictorial keys to the species Tokai University Press, Tokyo.

Pandolfi JM, Jackson JBC, Baron N, Bradbury RH, Guzman HM, Hughes TP, Kappel CV, Micheli F, Ogden JC, Possingham HP, Sala E (2005) Are U.S. coral reefs on the slippery slope to slime? Science Mag 307, 1725-1726.

Patton WK (1994) Distribution and ecology of animals associated with branching corals (Acropora spp.) from the Great Barrier Reef, Australia. Bulletin of Marine Science 55(1), 193-211. 32

Pratchett MS, Coker DJ, Jones GP, Munday PL (2012) Specialization in habitat use by coral reef damselfishes and their susceptibility to habitat loss. Ecology and Evolution doi: 10.1002/ece3.321.

Randall JE (1963) Review of the hawkfishes (family Cirrhitidae) Proceedings of the United States National Museum 114(3472), 389-451.

Randall JE (1985) Guide to Hawaiian Reef Fishes Harrowood Books, Newtown Square, PA 70.

Rosenzweig ML (1974) On the evolution of habitat selection. Proceedings of the first international congress of ecology. Centre for Agricultural Publishing and Documentation, The Hague 401-404.

Rounds RC (1981) First approximation of habitat selectivity of ungulates on extensive winter ranges. The Journal of Wildlife Management 45(1), 187-196.

Schultz LP (1950) Three new species of fishes of the genus Cirrhitus (family Cirrhitidae) from the Indo- Pacific. Proceedings of the United States National Museum 100(3270) 547-552.

Shulman MJ (1984) Resource limitation and recruitment patterns in a coral reef fish assemblage. Journal of Experimental Marine Biology and Ecology 74, 85-109.

Steneck RS, Graham MH, Bourque BJ, Corbett D, Erlandson JM, Estes JA, Tegner MJ (2002) Kelp forest ecosystems: biodiversity, stability, resilience and future. Environmental Conservation 29(4), 436- 459.

Thresher RE (1978) Polymorphism, mimicry and the evolution of the hamlets (Hypoplectrus, Serranidae). Bulletin of Marine Science 28, 345-353.

Van Beest FM, Loe LE, Mysterud A, Milner JM (2010) Comparative space use and habitat selection of moose around feeding stations. The Journal of Wildlife Management 74(2), 219-227.

Vazquez DP, Simberloff D (2002) Ecological specialization and susceptibility to disturbance: conjectures and refutations. American Naturalist 159, 606-623.

Whittingham MJ, Swetnam RD, Wilson JD, Chamberlain DE, Freckleton RP (2005) Habitat selection by yellowhammers Emberiza citronella on lowland farmland at two spatial scales: implications for conservation management. Journal of Applied Ecology 42, 270-280.

Wilson SK, Graham NAJ, Pratchett MS, Jones GP, Polunin NVC (2006) Multiple disturbances and the global degradation of coral reefs: are reef fishes at risk or resilient? Global Change Biology 12, 2220-2234.

APPENDICES

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Table 1.

Percentage of live Crest Slope coral cover

Inshore 14.8 7.3

Midshore 37.9 47.4

Offshore 47.2 42.0

Table 2.

Species /

Zone

lgae Shelf Position olony Acropora Pocillopora Stylophora Millepora Porites Other live hard coral Xeniidae Other soft coral Dead c oral c Rubble Sand Pavement Turf a Other sessile organisms

P. forsteri

Midshore Crest = .129 = .355 = .132 U U = .147 U U = .237 U U U U U

Slope + .232 = .133 = .193 U = .031 = .053 U - .015 = .188 U U U - .011 = .144

Offshore Crest = .118 - .024 = .226 = .123 = .302 U U U = .155 U NA - 0.00 = .053 NA

Slope = .121 + .146 = .358 = .062 - .012 = .069 U = .023 = .177 U NA U - .032 NA

C. oxycephalus

Midshore Slope = .036 U U = .349 U = .202 U = .030 = .258 U U U = .125 U

Offshore Slope U - .041 U = .265 = .057 = .132 U U U = .299 NA U = .206 NA

C. spilotoceps

Inshore Crest U U U NA U U U U U U U U = 1.0 U

Midshore Crest U U = .059 U U U U U = .132 = .255 U U = .554 U

Offshore Crest U U U = .274 U = .101 U U U U NA - .016 + .608 NA

Table 3.

Morphotype /

Zone

lgae

Shelf Position olony Acropora Pocillopora Stylophora Millepora Porites Other live hard coral Xeniidae Other soft coral Dead c oral c Rubble Sand Pavement Turf a Other sessile organisms Morph 1

Midshore Crest = .139 = .287 = .172 U U = .229 U U = .173 U U U U U

Slope = .236 = .091 = .275 U = .032 = .069 U U = .299 U U U U U

Offshore Crest = .116 - .023 + .240 = .120 = .296 U U U = .152 U NA - 0.00 = .052 NA

Slope = .107 = .124 = .234 = .119 - .013 = .108 U = .027 = .209 U NA U = .060 NA

Morph 2

Midshore Crest = .304 = .696 U U U U U U U U U U U U

Midshore Slope = .426 = .260 U U = .155 U U U = .159 U U U U U

Offshore Slope = .203 = .141 = .508 U = .018 = .014 U = .092 = .024 U NA U U U

Morph 3

Midshore Crest = .100 = .153 = .171 U U U U U = .575 U U U U U

Midshore Slope = .108 = .091 U U U U U = .025 = .174 U U U - .013 = .589

Offshore Slope = .165 = .159 = .434 U U = .038 U U = .202 U NA U - .003 NA

Morph 4 40

Midshore Slope = .322 = .109 U U U = .493 U U U U U U = .076 U

Offshore U = .644 U U U U U U U U U NA U = .356 NA