Host selection and preferences of coral symbiotic crab Tetralia rubridactyla

Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

Contents lists available at ScienceDirect

Journal of Experimental Marine Biology and Ecology journal homepage: www.elsevier.com/locate/jembe

Host selection and preferences of coral symbiotic crab Tetralia rubridactyla Parinya Limviriyakul a,b, Li-Chun Tseng a, Tung-Wei Shih c,⁎⁎, Jiang-Shiou Hwang a,d,⁎ a

Institute of Marine Biology, College of Life Sciences, National Taiwan Ocean University, Keelung 20224, Taiwan Department of Marine Science, Faculty of Fisheries, Kasetsart University, Bangkok 10900, Thailand c National Museum of Marine Science and Technology, Keelung 20248, Taiwan d Department of Biomedical Science and Environmental Biology, Kaohsiung Medical University, Kaohsiung 80708, Taiwan b

a r t i c l e

i n f o

Article history: Received 7 January 2016 Received in revised form 21 July 2016 Accepted 1 August 2016 Available online xxxx Keywords: Tetralia rubridactyla Acropora Coral Host selection Symbiosis

a b s t r a c t Coral symbiotic crabs provide considerable benefits to their host corals. A comprehensive understanding of the association between these crabs and their hosts could help clarify the relationship, interaction, and importance of symbionts with coral reefs as hosts. In this study, the coral symbiotic crab Tetralia rubridactyla was test for host preference and fidelity. Five oceanic objects were provided to the crabs: common host corals (Acropora hyacinthus and A. digitifera), uncommon host corals (Pocillopora damicornis and Stylophora pistillata), and dead coral skeletons. The crabs were collected from the 2 source host corals A. hyacinthus and A. digitifera and subjected to an experiment comprising 7 treatments. Each treatment included 2 stages of no-choice and choice conditions to estimate the expected selection frequencies. The results revealed that the crabs chose any available object under the no-choice condition, and exhibited various preferences under the choice condition. Moreover, T. rubridactyla exhibited significantly higher frequencies to inhabit Acropora corals (p b 0.01, χ2 test), than dead coral skeletons and uncommon host corals. In all the treatments, the preferences of the crabs from the 2 source hosts were similar. Present results demonstrated T. rubridactyla host selection conditioning as follows: (1) Under the no-choice condition, inhabit any choice object for shelter; (2) under the choice condition, if without a common host, randomly inhabit any uncommon choice object as a host; and (3) under the choice condition, if a common host is available, selecting the common host is the first priority because it could provide food and space. This study revealed that T. rubridactyla express neither fidelity nor preference between A. hyacinthus and A. digitifera. Thus, these results also suggested that the distribution of T. rubridactyla on Acropora corals in the reef is affected by an abundance of corals rather than the preferences of coral species. © 2016 Elsevier B.V. All rights reserved.

1. Introduction A host is a crucial ecological factor for symbionts that, affecting their existence and distribution. Many obligate symbiotic decapods in coral reefs exhibit a high degree of habitat specialization (Stella et al., 2011b). A host provides crucial benefits to its symbionts including protection from predators (Baeza and Stotz, 2001; Bruce, 1972; Castro, 1978; Huang et al., 2005), direct or indirect food sources (e.g., tissue, mucus, eggs, fat bodies, and associated detritus) (Barry, 1965; Bruce, 1972; Castro, 1969; Fautin et al., 1995; Huebner and Chadwick, 2012; Patton, 1994; Stimson, 1990), a substratum for burrowing and gallforming decapods (Borradaile, 1921; Kropp, 1989; Wei et al., 2005,

⁎⁎ Corresponding author. ⁎ Correspondence to: J.-S. Hwang, Institute of Marine Biology, College of Life Sciences, National Taiwan Ocean University, Keelung 20224, Taiwan. E-mail addresses: [email protected] (T.-W. Shih), [email protected] (J.-S. Hwang).

http://dx.doi.org/10.1016/j.jembe.2016.08.001 0022-0981/© 2016 Elsevier B.V. All rights reserved.

2013), and a mating ground for seeking sexual partners (Huber, 1987; Ocampo et al., 2012). Coral reefs are among the most complex marine ecosystems, with the highest biodiversity (Veron, 2000), and are predominantly inhabited by invertebrates (Stella et al., 2011b). At least 310 decapod crustaceans live in the spaces of a coral structure, particularly on the branching corals of the genus Acropora and the Pocilloporidae family (Abele and Patton, 1976; Bruce, 1998; Castro et al., 2004; Patton, 1966, 1994; Stella et al., 2011b). Brachyuran crabs of the genus Tetralia belong to the Tetraliidae family and Trapezioidea superfamily; these crabs are well-known obligate coral symbionts distributed throughout the Indo-West Pacific region (Castro, 1988; Galil, 1988; Patton, 1994; Poupin, 2008). The coevolution of trapezioid crabs with corals occurred in the Eocene (Schweitzer, 2005). Tetraliid crabs are one of the most coral-dependent animals worldwide; they exhibit distinct habitat specialization with corals of only the genus Acropora (Castro et al., 2004; Galil, 1988; Patton, 1994). These crabs rely on their hosts for refuge, feed on coral mucus rather than on coral live tissue and eggs, and use coral space as a breeding ground (Castro, 1988; Knudsen, 1967;

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

Patton, 1994; Sin, 1999; Stimson, 1990). Few studies have reported an association between tetraliids and atypical hosts, such as Pocillopora, Seriatopora, and an alcyonarian (Castro, 2009; Chang et al., 1987; Garth, 1964, 1984; Knudsen, 1967). Symbiotic Tetralia crabs are ecologically crucial for their coral hosts (Glynn, 2013). A previous study observed that corals exhibited decreasing growth rates and increasing bleaching and mortality after symbiotic crabs were removed, whereas corals with crabs remained healthy (Stewart et al., 2006). Tetralia crabs could help host corals clean sediments and protect the host from predators such as starfish (Pratchett, 2001; Rouzé et al., 2014; Stewart et al., 2006). Most studies on the ecological perspectives of crabs symbiotic with corals have focused on larger crabs of the genus Trapezia (Glynn, 2013; Glynn and Enochs, 2011). To date, few studies have reported on small symbiotic Tetralia crabs. Choice experiments are commonly performed for examining the preference of animals exhibited for a particular resource (i.e., food and habitat). In a traditional experimental design, different choices have been presented to animals; however, with such an experimental design, preference can be confused with accessibility (Olabarria et al., 2002; Singer, 2000). Several recent researchers have demonstrated the appropriate methods and statistical analyses used for investigating the food and habitat preferences of animals (Jackson and Underwood, 2007; Olabarria et al., 2002; Underwood et al., 2004; Underwood and Clarke, 2005, 2006). Underwood and Clarke (2005) designed a protocol that includes a 2-stage test for determining the experimental preferences of animals. The 2-stage test design has an advantage over previous designs because it provides answers that are more accurate based on proper control rates of the Type I error, particularly with small samples (Underwood and Clarke, 2005). At Stage 1, a single type of habitat or food is provided to animals, followed by 2 or more choices at Stage 2. This analytical method has been widely applied in examining the preferences of various animals (Cacabelos et al., 2010; Hale et al., 2008; Mascaró et al., 2012; Pinna et al., 2012; Silva et al., 2010). In the coastal areas of northeast Taiwan, where the southern Ryukyu arc meets the Coral Triangle, the diversity of symbiotic crabs living with Acropora corals is abundant. Limviriyakul et al. (2016) reported that Acropora hyacinthus (Dana, 1846) (60.0%) and A. digitifera (Dana, 1846) (30.6%) were commonly found in this area. The most abundant symbiotic crab was Tetralia rubridactyla Garth, 1971, which was more abundant and frequently associated with A. hyacinthus (68.6%) than with A. digitifera (23.1%) (Limviriyakul et al., 2016). Therefore, this area can provide a rich resource of symbiotic crabs and host corals for determining the coral–crab relationship and selection preference of the crab. Relatively few studies have reported on the selection preferences of coral symbiotic crabs toward host corals. The present study adopted the methods by Underwood and Clarke (2005) and was conducted using a series of treatments involving T. rubridactyla and corals from northeast Taiwan. By examining the preferences for (1) an uncommon host coral and dead coral skeleton, (2) a common host coral and dead coral skeleton, (3) a common host coral and uncommon host coral, and (4) 2 host Acropora corals. The host selection behavior of T. rubridactyla in a reef environment was determined through a comprehensive statistical analysis. 2. Materials and methods 2.1. Collection of hosts and symbiotic crabs The experimental crab T. rubridactyla (Fig. 1A) and corals A. hyacinthus (Fig. 1B), A. digitifera (Fig. 1C), Pocillopora damicornis (Linnaeus, 1758) (Fig. 1D), and Stylophora pistillata Esper, 1797 (Fig. 1E) were collected by scuba diving at a 2–3-m depth in the coastal area of western Fan-Zai-Aou Bay near Keelung, northeastern Taiwan, from April to August 2014. The sampling area is composed of a patchy reef and coral

25

community on rocks exposed to strong wave action. Among the collected corals, A. hyacinthus and A. digitifera are typical host corals (Castro et al., 2004; Patton, 1994), whereas P. damicornis and S. pistillata are atypical host corals (Garth, 1984; Knudsen, 1967). Colony shapes of A. hyacinthus in the sampling area is varied from corymbose to small platelike with slender branches, exhibiting mostly secondary branches (Fig. 1B), whereas A. digitifera is corymbose to digitate with shorter and thicker tapering vertical branches, rarely with secondary branches (Fig. 1C). Each coral colony was stored separately in a plastic zip-lock bag and immediately transported to the Chaojing Ocean Center (National Museum of Marine Science and Technology, Keelung, Taiwan). In the laboratory, the symbiotic crabs and all symbionts were gently removed from the corals by using flexible plastic rods. The total numbers of T. rubridactyla, including the adult and juvenile crabs in each colony, were recorded to evaluate the symbiont capacity of host corals. The taxonomic descriptions used for identifying tetraliid crabs and coral species were adapted from Castro et al. (2004), Wallace (1999), and Veron (2000). Furthermore, the coral colonies were wrapped tightly with Polyethylene food wrap film to measure the total volume by using the water displacement method (Vytopil and Willis, 2001). After the volume was recorded, the corals were stored in an aquarium for experimental use. The crabs and corals were isolated in 2 aquaria (90 cm × 60 cm × 60 cm; containing 324 L of seawater). Each aquarium was connected to 2 separate recirculation filtration systems to prevent any chemical contact between the host corals and crabs during acclimatization. The crabs were individually separated in small perforated plastic boxes, held in the aquarium, and fed shrimp meat once daily. Furthermore, the aquaria were maintained at 25 ± 1 °C, 35 ± 1 salinity, and under a 12-h light–dark cycle of artificial lighting. Seawater was partially exchanged weekly and measured twice weekly by using ammonia portable photometer (Hanna HI93700), nitrate portable photometer (Hanna HI96728), phosphate portable photometer (Hanna HI96713), pH meter (Hanna HI-2211), alkalinity colorimeters (Hanna HI772), Calcium test kit (Salifert) and Magnesium test kit (Salifert). The water quality was maintained at approximately constant concentrations of ammonia, 0 mg L−1; nitrates, 0 mg L−1; phosphates, 0.02–0.04 mg L−1; pH, 8.1–8.2; alkalinity, 8.0–8.3 dKH; calcium, 390– 410 ppm; and magnesium, 1290–1350 ppm. The experimental animals used for the host preference test were acclimated in a laboratory for at least 1 week. 2.2. Experimental design The experimental T. rubridactyla were divided into 2 groups according to their source host corals A. hyacinthus and A. digitifera. The “source host” refers to the species of acroporid coral from which the crab used in the experiment were collected; by contrast, the “non-source host” is another species of acroporid coral which the crab of that species has been reported to inhabit. The “common host” refers to the typical host corals of T. rubridactyla, commonly scleractinian corals of the genus Acropora, and the “uncommon host” refers to the reported atypical host corals of the crab, generally other genera than Acropora. Table 1 list of the hosts used in the experiments. The trial comprised 7 treatments (Table 2), involving 5 choice objects: A. hyacinthus, A. digitifera, P. damicornis, S. pistillata, and dead coral skeletons. Dead coral skeletons were obtained by killing acroporid corals, allowing their tissues to rot, washing them with freshwater, and drying them in an oven. The crabs in both groups were subjected to the 7 treatments, each of which involved 48 replicates. Treatments 1–4 were designed to examine whether host selection depends on food or shelter, whereas Treatments 5 and 6 were designed to investigate the preference of an uncommon host. The final treatment was designed to determine the preference between source and non-source hosts. Based on the design by Underwood and Clarke (2005), all treatments involved 2 test stages. At Stage 1 (comprising Stages 1-1 and 1-2), all crabs

26

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

Fig. 1. Experimental materials: T. rubridactyla (A), A. hyacinthus (B), A. digitifera (C), P. damicornis (D), S. pistillata (E), and dead coral skeleton (F). The scale bar represents 5 mm.

were treated with a no-choice treatment by providing them with 2 similar choice objects in order to collect information on the probability of choices made. At Stage 2, 2 choice objects were provided to investigate the preferences of the crabs (Stage 2; choice treatment; Table 2). All choice objects provided in the trials had a similar size (colony diameter, approximately 8–10 cm). Present study found that the body size of the smallest ovigerous T. rubridactyla was 3.98 mm in carapace width (CW); therefore, all crabs with a CW of N3.98 mm were considered adults. Only adult T. rubridactyla were used in the trial. The sex of the experimental crabs was not recorded; however, all crabs were non-ovigerous. All crabs were starved for 24 h before the experiments. Each coral colony and its inhabited crabs were tagged and never used in the same trial. In the same treatment, each crab was used only once, whereas the corals were used several times. All trials were conducted in the afternoon, at approximately 3 p.m., under artificial lighting. Before the beginning of each trial, a plastic

aquarium (38 cm × 24 cm × 16 cm) containing 9 L of seawater was aerated and maintained under the same conditions as during holding. An individual crab in an inverted glass tumbler was placed at the center, whereas the 2 choice objects were placed at opposite ends of the aquarium; the crab was acclimated for 15 min. Furthermore, the aeration was ceased to reduce the effects of the current. The position of the crab was observed and recorded 18 h after its release. A crab was considered to select a choice object when it stayed on or under the object throughout the inspection period of 5 min. After each trial, the used aquarium was cleaned thoroughly with freshwater to prevent any error in the next trial. All crabs were transported and corals were transplanted to the sampling area after the experiments.

Table 2 Experimental protocol for the treatments in assessing the host selection behavior of the crab T. rubridactyla, P.D. = P. damicornis, S.P. = S. pistillata, D.C.= dead coral skeleton, A.H. = A. hyacinthus, and A.D. = A. digitifera. Treatment

Table 1 Comparison and division of source hosts, non-source hosts, common hosts and uncommon hosts used in the experiments.

Source host Non-source host Common host Uncommon host

Crabs from A. hyacinthus

Crabs from A. digitifera

A. hyacinthus A. digitifera A. hyacinthus, A. digitifera P. damicornis, S. pistillata

A. digitifera A. hyacinthus A. hyacinthus, A. digitifera P. damicornis, S. pistillata

1 2 3 4 5 6 7

Stage 1 (no choice condition) Stage 1-1

Stage 1-2

P.D. + P.D. S.P. + S.P. A.H. + A.H. A.D. + A.D. A.H. + A.H. A.D. + A.D. A.H. + A.H.

D.C. + D.C. D.C. + D.C. D.C. + D.C. D.C. + D.C. P.D. + P.D. P.D. + P.D. A.D. + A.D.

Stage 2 (choice condition)

P.D. + D.C. S.P. + D.C. A.H. + D.C. A.D. + D.C. A.H. + P.D. A.D. + P.D. A.H. + A.D.

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

27

2.3. Statistical analysis

2.4. Field assessment

According to the methods by Underwood and Clarke (2005), the frequency of association from Stage 1 was used for estimating the null hypothesis of no preference. The expected frequencies were derived from maximum likelihood estimators by using information from both stages. When habitat selection is random, the proportion of crabs associated with the 2 choices at Stages 1 and 2 are similar and can be expressed as follows:

The benthic community was assessed in the study area by using the photoquadrat method modified by Hill and Wilkinson (2004). Fifty photographs of the 0.25 m2 (size 0.5 m × 0.5 m) quadrats were taken from 5 transect lines set parallel to the shore at a 2–3 m depth. The interval of each sampling point was 10 m along the 50-m-long transect lines. Two photographs of the left and right sides of the transect line were taken at each sampling point. An underwater digital camera was used to take horizontal plane photographs of the quadrats. At the laboratory, coverage of the benthic community was calculated according to 25 randomly generated points within each photograph by using Coral Point Count with Excel Extensions 4.1 software (Kohler and Gill, 2006).

H0 : q1 ¼ θp1 ; q2 ¼ θp2 ðor H0 : q1 =p1 ¼ q2 =p2 Þ

where q1 and q2 are the proportion of crabs associated with each choice object at Stage 1 (no-choice condition), p1 and p2 are the proportion of crabs associated with each choice object when they are presented at Stage 2 (choice condition), and θ is a constant unknown parameter. The χ2 test was conducted to determine whether the observed value differed from the expected value. To evaluate the symbiont capacity of a host coral, the Pearson product–moment correlation was used to estimate the correlation between crab abundance and the total volume of coral colonies.

3. Results 3.1. Laboratory experiment The results for the T. rubridactyla collected from the 2 acroporid source hosts were similar after all treatments. At Stage 1, almost all the crabs selected any object. At Stage 2, the crabs exhibited various preferences depending on the treatment (Table 3, Figs. 2 and 3).

Table 3 Observed numbers of T. rubridactyla that selected the provided choice objects in each treatment and χ2 and p values. The expected numbers of crabs (maximum likelihood) are shown in parentheses. The Stage 1 (choice was absent) and 2 (choice was available) treatments for all the experiments are explained in Table 2; 48 crabs were used in each treatment. Significant differences of the χ2 test (p b 0.05) are indicated in bold. The degree of freedom is 1. Source host of experimental crabs A. hyacinthus

A. digitifera

Treatment 1 Stage 1-1 Stage 1-2 Stage 2 χ2, p

P. damicornis 16 (16.00) – 6 (8.46) χ2 = 1.562, p = 0.212

Dead coral – 14 (14.27) 10 (7.54)

No selection 0 (0.00) 2 (1.73) 0 (0.00)

P. damicornis 13 (12.82) – 6 (6.90) χ2 = 0.233, p = 0.63

Dead coral – 15 (15.05) 9 (8.10)

No selection 3 (3.18) 1 (0.95) 1 (1.00)

Treatment 2 Stage 1-1 Stage 1-2 Stage 2 χ2, p

S. pistillata 13 (12.18) – 3 (6.43) χ2 = 3.693, p = 0.055

Dead coral – 14 (14.35) 11 (7.57)

No selection 3 (3.82) 2 (1.65) 2 (2.00)

S. pistillata 15 (14.73) – 4 (7.39) χ2 = 3.164, p = 0.075

Dead coral – 15 (15.17) 11 (7.61)

No selection 1 (1.27) 1 (0.83) 1 (1.00)

Treatment 3 Stage 1-1 Stage 1-2 Stage 2 χ2, p

A. hyacinthus 15 (15.28) – 15 (8.73) χ2 = 10.649, p = 0.001

Dead coral – 14 (12.71) 1 (7.27)

No selection 1 (0.72) 2 (3.29) 0 (0.00)

A. hyacinthus 16 (16.00) – 15 (8.13) χ2 = 12.329, p b 0.001

Dead coral – 15 (15.50) 1 (7.87)

No selection 0 (0.00) 1 (0.50) 0 (0.00)

Treatment 4 Stage 1-1 Stage 1-2 Stage 2 χ2, p

A. digitifera 15 (15.25) – 14 (8.64) χ2 = 7.733, p = 0.005

Dead coral – 14 (12.99) 2 (7.36)

No selection 1 (0.75) 2 (3.01) 0 (0.00)

A. digitifera 16 (16.00) – 16 (8.13) χ2 = 16.016, p b 0.001

Dead coral – 15 (15.50) 0 (7.87)

No selection 0 (0.00) 1 (0.50) 0 (0.00)

Treatment 5 Stage 1-1 Stage 1-2 Stage 2 χ2, p

A. hyacinthus 15 (15.34) – 16 (7.83) χ2 = 16.871, p b 0.001

P. damicornis – 16 (16.00) 0 (8.17)

No selection 1 (0.66) 0 (0.00) 0 (0.00)

A. hyacinthus 16 (16.00) – 15 (8.39) χ2 = 12.593, p b 0.001

P. damicornis – 13 (14.50) 1 (7.61)

No selection 0 (0.00) 3 (1.50) 0 (0.00)

Treatment 6 Stage 1-1 Stage 1-2 Stage 2 χ2, p

A. digitifera 15 (15.34) – 16 (7.83) χ2 = 16.871, p b 0.001

P. damicornis – 16 (16.00) 0 (8.17)

No selection 1 (0.66) 0 (0.00) 0 (0.00)

A. digitifera 16 (16.00) – 16 (8.39) χ2 = 16.155, p b 0.001

P. damicornis – 13 (14.50) 0 (7.61)

No selection 0 (0.00) 3 (1.50) 0 (0.00)

Treatment 7 Stage 1-1 Stage 1-2 Stage 2 χ2, p

A. hyacinthus 15 (14.94) – 7 (7.97) χ2 = 0.242, p = 0.623

A. digitifera – 15 (15.06) 9 (8.03)

No selection 1 (1.06) 1 (0.94) 0 (0.00)

A. hyacinthus 16 (16.00) – 9 (8.00) χ2 = 0.250, p = 0.617

A. digitifera – 16 (16.00) 7 (8.00)

No selection 0 (0.00) 0 (0.00) 0 (0.00)

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

Selection ratio (%)

100

100

A

P.D D.C.

80 60 40 20

Selection ratio (%)

28

0

A.H. D.C.

Stage 2

40 20

D Selection ratio (%)

Selection ratio (%)

C

A.D. D.C.

80 60 40 20 0

Stage 1

Stage 2

Stage 1 100

E

A.H. P.D.

80 60 40 20 0

Selection ratio (%)

Selection ratio (%)

20

Stage 1

0

Stage 2

F

A.D. P.D.

80 60 40 20 0

Stage 1

Selection ratio (%)

40

100

60

100

60

Stage 2

80

100

S.P D.C.

0 Stage 1

100

B

80

Stage 2

G

Stage 1

Stage 2

A.H. A.D.

80 60 40 20 0 Stage 1

Stage 2

Fig. 2. Percentage of T. rubridactyla collected from the source host A. hyacinthus that selected the choices in Treatments 1 (A), 2 (B), 3 (C), 4 (D), 5 (E), 6 (F), and 7 (G). P.D. = P. damicornis, S.P. = S. pistillata, D.C. = dead coral skeleton, A.H. = A. hyacinthus, and A.D. = A. digitifera.

3.1.1. Treatments 1 and 2 Treatments were designed for investigating the preferences between the uncommon host corals (P. damicornis and S. pistillata) and dead coral skeletons; 100% and 81.25% of T. rubridactyla from A. hyacinthus selected P. damicornis and S. pistillata at Stage 1-1 of Treatments 1 and 2, respectively. Moreover, 81.25% and 93.75% of the crabs from A. digitifera selected P. damicornis and S. pistillata at Stage 1-1 of Treatments 1 and 2, respectively (Table 3, Figs. 2A, B, 3A, and B). In both groups, the frequency of the crabs that selected dead coral skeletons at Stage 1 in both treatments was 87.5%–93.75%. At Stage 2, N 56.25% of the crabs opted to stay with dead coral skeletons. On average, 37.5% and 21.88% of the crabs inhabited P. damicornis and S. pistillata, respectively. Nonetheless, the results revealed no significant differences in the frequencies of the crabs to select uncommon host corals and dead coral skeletons (p N 0.05, χ2 test, Table 3). These results indicated that the crabs passively selected any encountered object when the main host acroporid corals were absent; thus, unknown

harmful stress from an uncommon host coral may have discouraged the crabs. Consequently, the crabs selected dead coral skeletons at a higher frequency solely for the basic shelter provided. 3.1.2. Treatments 3 and 4 Treatments 3 and 4 involved using common acroporid host corals and dead coral skeletons to investigate whether the selection behavior of crabs depends on the food resource or shelter function. The results of Treatments 3 and 4 revealed that T. rubridactyla from A. hyacinthus and A. hyacinthus preferred to associate with a common host coral at Stage 1-1 (93.75%–100%); the selection proportion of dead coral skeletons in Stage 1-2 ranged 87.5%–93.75%. Moreover, at Stage 2, most crabs opted to stay with host acroporid corals rather than with dead coral skeletons (Figs. 2C, D, 3C, and D). Overall, both tetraliid crabs from A. hyacinthus and A. digitifera revealed significantly higher association ratios with live acroporid corals than with dead coral skeletons (p b 0.001, χ2 test, Table 3). Thus, this observation showed that the

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

100

A

P.D D.C.

80 60 40 20

Selection ratio (%)

Selection ratio (%)

100

0

A.H. D.C.

40 20

Selection ratio (%)

Selection ratio (%)

C

Stage 2

D

A.D. D.C

80 60 40 20 0

Stage 1

Stage 2

Stage 1 100

E

A.H. P.D.

80 60 40 20 0

Stage 1

Stage 2

F

A.D. P.D.

80 60 40 20 0

Stage 2

G

Selection ratio (%)

Selection ratio (%)

20

Stage 1

0

Selection ratio (%)

40

100

60

100

60

Stage 2

80

100

S.P D.C.

80

0 Stage 1

100

B

29

Stage 1

Stage 2

A.H. A.D.

80 60 40 20 0

Stage 1

Stage 2

Fig. 3. Percentage of T. rubridactyla collected from the source host A. digitifera that selected the choices in Treatments 1 (A), 2 (B), 3 (C), 4 (D), 5 (E), 6 (F), and 7 (G). P.D. = P. damicornis, S.P. = S. pistillata, D.C. = dead coral skeleton, A.H. = A. hyacinthus, and A.D. = A. digitifera.

crabs exhibited a preference for common host corals because, in addition to the shelter function, live corals provided food resource.

3.1.3. Treatments 5 and 6 At treatment 5 and 6, common and uncommon host corals were provided to the crabs to examine their selection behavior. The results of Treatments 5 and 6 revealed that T. rubridactyla from A. hyacinthus and A. digitifera preferred to stay with their common hosts at Stage 11 (93.75–100%), whereas the percentage of crabs that selected P. damicornis at Stage 1-2 was higher than 81.25%. At Stage 2, most crabs inhabited the host acroporid coral, rather than the uncommon coral P. damicornis (Figs. 2E, F, 3E, and F). The statistical results revealed that acroporid host corals were preferred over the uncommon coral P. damicornis by both crabs from A. hyacinthus (p b 0.001, χ2 test) and A. digitifera (P b 0.001, χ2 test; Table 3).

3.1.4. Treatment 7 The main purpose of treatment 7 was to examine host selection between the source and non-source hosts of the crabs from the 2 groups. T. rubridactyla from A. hyacinthus selected A. hyacinthus at Stages 1-1 and 2 for 93.75% and 43.75%, while selected A. digitifera at Stages 1-2 and 2 for 93.75% and 56.25%, respectively (Fig. 2G). Similarly, the crabs from A. digitifera selected A. hyacinthus at Stages 1-1 and 2 for 100% and 55.0%, respectively, but preferred to stay with A. digitifera at Stages 1-2 and 2 for 100% and 45.0%, respectively (Fig. 3G). The statistical results revealed no significant differences between the proportion of T. rubridactyla from A. hyacinthus that selected A. hyacinthus and A. digitifera (χ2 = 0.242, p = 0.623) and the crabs from A. digitifera (χ2 = 0.250, p = 0.617; Table 3). These results indicated that the crabs preferred the 2 host coral species through random selection. They also suggested that T. rubridactyla had no fidelity to host species of congeners.

30

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

3.2. Field assessment A total of 50 quadrats and 1250 random points were analyzed for coverage of the benthic community. The highest coverage percentage of benthic taxa was scleractinian corals (29.6 ± 18.5%, mean ± standard deviation), followed by macro algae (25.9 ± 16.5%) and coralline algae (19.1 ± 12.1%). Coverage of dead corals with algae was 16.7 ± 11.5%, and none of them were branching corals. The present study found no diseased or bleached corals from the investigated area (Fig. 4A). Among scleractinian corals, 13 genera were identified. The most abundant genus was Favites (25.71%), followed by Favia (20.06%), and Platygyra (9.89%) (Fig. 4B). The proportions of branching corals in the genera Acropora and Stylophora were 2.82% and 3.11% of scleractinians, respectively. Among the collected corals, T. rubridactyla crabs were found only with Acropora corals. Most of the coral colonies were inhabited by adult and juvenile crabs, whereas the megalopa stage crabs were

60

recorded in only a few colonies. In a colony, 2 larger crabs were generally an adult male and female, whereas the remaining individuals were often juveniles of smaller size. From 21 colonies of A. digitifera, the number of T. rubridactyla varied from 1 to 8 individuals per colony (ind. colony − 1) (Fig. 4C). The colony sizes of collected A. digitifera were 45 to 395 cm3 in volume. The number of 2 ind. colony− 1 was the most frequently recorded (47.6%) and found in corals with the total volume ranging from 91 to 395 cm 3 (Fig. 4D). Among 30 heads of A. hyacinthus with the volume ranging from 54 to 494 cm3, the numbers of the crab found varied from 1 to 5 ind. colony− 1 (Fig. 4E). The number of 2 ind. colony− 1 was the most frequently recorded (50%) (Fig. 4F) in corals with the total volume ranging from 47 to 494 cm3. The Pearson product–moment correlation revealed no significant correlation between crab abundance and the total volume of the coral colonies (p N 0.05). The pattern of symbiont capacity of both host corals with T. rubridactyla crabs in the study area was unclear.

Favites Favia Platygyra Echinopora Hydnophora Porites Cyphastrea Goniopora Oxypora Stylophora Acropora Mycedium Pavona

B

A

Coverage (%)

50 40 30 20 10

Sp Sc o .C Al .C G o Zo r Ma a C c o. A O D t. e. L C -A Sa Di.C -P -R U nk

0

Proportion of scleractinian genera

Major taxa categories

Numbers of crab -1 (individual colony )

D

C

8

Individual

7

1 2 3 4 5 6 7 8

6 5 4 3 2 1 0

100

200

300

400

500

Sizes of A. digitifera (cm3)

F

E Numbers of crab -1 (individual colony )

Proportion of numbers recorded

5

Individual 1 2 3 4 5

4 3 2 1 0

100

200

300

400

500

600

Proportion of numbers recorded

Sizes of A. hyacinthus (cm3) Fig. 4. Coverage (%) of benthic fauna (A), Spo = sponge, Sc. C = scleractinian corals, Al. C = alcyonarian corals, Gor = gorgonians, Zoa = zoanthids, Mac = macro algae, Co. A = coralline algae, Ot. L = other living, De. C-A = dead coral with algae, Di. C = diseased corals, Sa-p-r = sand-pavement-rubble, Unk = unknowns; proportion of scleractinian genera (B); numbers of T. rubridactyla recorded in various sizes of A. digitifera (C); proportion of numbers of crab per A. digitifera colony (D); numbers of T. rubridactyla recorded in various sizes of A. hyacinthus (E); and proportion of numbers of crab per A. hyacinthus colony (F).

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

4. Discussion 4.1. Host preference of 2 Acropora species Preference is a result of active behavioral selection expressed by an organism (Singer, 2000). Observed patterns of association and distribution in nature might not correspond with an actual preference for a particular resource. The outcome of associations may be influenced by ecological factors (biotic and abiotic) that force symbionts to alter their preferences to less favorable choices (Mascaró et al., 2012; Underwood et al., 2004). Several factors influence these associations: resource limitations, the size and space of available habitats, intra- and interspecific competition, predation, and oceanographic conditions (i.e., light, depth, and waves) (Baeza et al., 2002; Castro, 1978; Huber, 1987; Knudsen, 1967; Sin, 1999; Vannini, 1985). The mechanism underlying host selection by T. rubridactyla remains unclear yet. The present study is the first to report the multipurpose of choice tests on tetraliid crabs through laboratory experiments. The findings reveal that T. rubridactyla exhibited no significant difference in host selection between A. hyacinthus and A. digitifera. Records obtained from observing the reef confirmed that the crab T. rubridactyla is commonly associated with acroporid corals (Limviriyakul et al., 2016). By contrast, the density of the available host could influence the distribution and abundance of its obligate symbiotic decapods (Mascaró et al., 2012; Sin and Lee, 2002; Thiel et al., 2003b). The density of A. hyacinthus was higher than that of A. digitifera in the study area. The opportunity for host coral encounters is a potential factor influencing the association reported by Limviriyakul et al. (2016). The morphology and complexity of habitats affect the distribution and abundance of associated invertebrates in marine environments (Chabanet et al., 1997; McCoy and Bell, 1991; Untersteggaber et al., 2014; Vytopil and Willis, 2001). The complexity of coral structures provides microhabitats characterized by a reduced water flow, dissolved oxygen, and the delivery of organic matter (Fabricius, 2006; Schiller and Herndl, 1989; Untersteggaber et al., 2014). Vytopil and Willis (2001) reported that Tetralia crabs selected their host corals according to the space between branches, and the inter-branch space volume could affect the size of the crabs. A tighter branch of Acropora provides a more complex structure, providing the most favorable habitat for epifauna (Vytopil and Willis, 2001). In the present study, the branches of A. digitifera were shorter than those of A. hyacinthus; thus, symbiotic crabs were exposed more easily to their predators from the top view. The results reveal no preference between these 2 corals, possibly because of the lack of predators in the experimental environment. Therefore, the crabs were free to select any coral as a host without survival pressure. Coral crabs of the genus Tetralia is strictly associated with host of the genus Acropora, but without a preference for any congeneric species (Castro, 1976; Knudsen, 1967; Patton, 1994). By contrast, another report showed that the dark phase of T. nigrolineata inhabited A. gemmifera more frequently whereas the light phase inhabited A. tenuis (Sin, 1999). Furthermore, several studies have reported the specific host preferences of various marine invertebrates in situ (Baeza and Stotz, 2001; Mascaró et al., 2012; Pinna et al., 2012); however, the results of these studies do not reflect an actual preference for a host because of the limitation of a host resource. The present study confirmed these results by comparing the laboratory results with in situ records. 4.2. Fidelity of the crabs on host corals This study used 2 groups of Tetralia crabs collected from 2 source hosts. The results reveal that the crabs randomly selected an alternative coral host of congeners. Present outcomes are similar to those of previous reports indicating that Tetralia crabs exhibit ecological specialization to Acropora corals (Castro, 1976; Castro et al., 2004; Knudsen, 1967; Patton, 1994). According to another perspective, Acropora corals are abundant in healthy reefs worldwide (Veron, 2000). Using various

31

congeneric Acropora corals as hosts is beneficial for expanding the population and existence, because dietary resources and habitat space would be relatively abundant. A high degree of host fidelity has been recorded in various marine decapods, particularly in Pinnotheridae (Baeza and Stotz, 2003; Derby and Atema, 1980; Góngora-Gómez et al., 2015; Guo et al., 1996; Hamel et al., 1999; Ocampo et al., 2012; Silbiger and Childress, 2008; Stevens, 1990). For instance, a parasitic pea crab (Calyptraeotheres garthi) exhibited high host fidelity in experiments (Ocampo et al., 2012). Evidently, female pea crabs were attracted to the host species from which they were collected, not to the alternative host species that may have provided a greater fecundity and reproductive output. By contrast, low host fidelity has been reported for the sea anemone shrimp Ancylomenes pedersoni, which randomly selected the host anemone (Mascaró et al., 2012); the authors inferred that A. pedersoni could not recognize its previous host because of its short-term memory. Reports have revealed that symbiotic crustaceans frequently exhibit movement among hosts of the same species (Castro, 1978; Mascaró et al., 2012; Patton et al., 1985; Thiel et al., 2003a,b). Trapezia ferruginea randomly moved among P. damicornis colonies at night under laboratory conditions (Castro, 1978). Castro (1978) assumed that the movement of T. ferruginea depends on the quantity of mucus secreted by the host coral. Previous studies have reported that symbionts migrate to another host when food is lacking to avoid harming their host and to fulfill the food requirement (Castro, 1969, 1978; Thiel et al., 2003b). Mucus, a major food source for crabs, is insufficiently produced by bleached corals (Bythell and Wild, 2011; Tsuchiya et al., 1993). Consequently, unhealthy corals become an unfavorable habitat for the crabs. A coral bleaching event may drive coral symbiotic crabs to wander (Tsuchiya et al., 1993). Field surveys and manipulative experiments have shown that crabs on bleached corals attempt to move and occupy other healthy corals (Stella et al., 2011a). Both species richness and abundance of coral symbionts were lacking in bleached colonies compared with healthy ones (Jeng, 2000; Glynn et al., 1985; Tsuchiya, 1999; Tsuchiya et al., 1993). Trapezia crabs inhabiting bleached corals have shown a decline in defensive behavior, an increase in mortality (Glynn et al., 1985), and a decrease in fecundity and the number of ovigerous females (Stella et al., 2011a; Tsuchiya, 1999). Based on previous studies, the present results indicate that T. rubridactyla expressed no fidelity toward the host colony because of their propensity to seek a healthy host and suitable food. Symbiont movement is affected by host characteristics such as size (Castro, 1978), color (Baeza and Stotz, 2003), available space (Abele and Patton, 1976), and density (Thiel et al., 2003b). Numerous studies have reported an evident correlation between the number and size of coral-associated crustaceans with the size of the host (Abele, 1976; Abele and Patton, 1976; Castro, 1978; Glynn, 1976; Gotelli and Abele, 1983; Patton, 1994; Vytopil and Willis, 2001). Trapeziid and tetraliid crabs tend to inhabit new coral heads when they grow larger or the habitat space is insufficient (Castro, 1978; Vytopil and Willis, 2001). The host color also affects the selection behavior of symbionts. The reddish-green crab Allopetrolisthes spinifrons commonly lives with the sea anemone Phymactis clematis, which exhibits various colors. In laboratory experiments, the crabs frequently acclimated to hosts with colors similar to their own and avoided staying on the blue sea anemone, where homochromy decreased (Baeza and Stotz, 2003). The present study observed low fidelity of T. rubridactyla toward Acropora species, thus suggesting that the crabs frequently migrate among coral colonies in nature when host corals cannot provide sufficient food or habitat space, as well as after coral bleaching. 4.3. Crabs with an uncommon habitat The results reveal that T. rubridactyla preferred a common host (acroporid corals) under the choice condition, but they selected an

32

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

uncommon host coral at a high frequency (P. damicornis and S. pistillata) under the no-choice condition. Previous studies have reported that tetraliids are occasionally associated with P. damicornis (Garth, 1964, 1984), P. verrucosa (Chang et al., 1987), Seriatopora sp. (Knudsen, 1967), and an unidentified alcyonarian (Castro, 2009). Similar studies have reported that trapeziids are associated with the uncommon host corals A. divaricate (Patton, 1966), A. digitifera (Patton, 1994), A. solitaryensis (Tsuchiya and Nojima, 2002), and A. hyacinthus (Limviriyakul et al., 2016). Most relevant studies have examined a single coral symbiotic crab. Notably, a small population of the crab Trapezia cymodoce (a heterosexual pair and several females are ovigerous) was observed on the uncommon host coral A. solitaryensis in southern Japan (Tsuchiya and Nojima, 2002); the authors suggested that T. cymodoce may complete the post-larval life cycle on the coral. Although Trapezia and Tetralia crabs strongly depend on food from the host coral, they use zooplanktons as additional food sources (Patton, 1994). This observation suggested that the flexibility in obtaining food from various resources may increase the ability of the crabs to adapt to an uncommon host coral. Trapezioid crabs inhabit uncommon host corals during immature stages (Castro, 2009; Chang et al., 1987; Garth, 1984; Patton, 1994). Habitat selection during larval settlement is a common process among decapods (Baeza, 2015; Bruce, 1972; Hamel et al., 1999; Patton et al., 1985). For instance, the adult symbiotic crab A. spinifrons live only on sea anemones, whereas juveniles inhabit various hosts during ontogeny (Baeza and Stotz, 2001). Thus, Castro (2009) suggested that Tetralia larvae are not necessarily restricted to Acropora, because they might settle on an alternative host and seek their common host after reaching the post-larval stage (Baeza, 2015; Castro, 2009). The present study derived evidence of host selection preference by the coral symbiotic crab Tetralia; however, the reason for the specific symbiotic relationship between the Tetralia crab and acroporid coral remains unclear. In all 7 treatments, Tetralia crabs exhibited high percentages of selection rates for each choice object at Stage 1 (81.25%–100%), and even dead coral skeletons were found to have a selection rate of 87.5%. This observation indicates that the shelter function is a priority for the crabs. Any object that could be used as shelter by the crabs is instantly selected. In reefs, dead coral skeletons are the most valuable microhabitat for crustaceans in terms of abundance, biomass, and productivity (Kramer et al., 2014). A high occurrence of trapeziid crabs living on dead colonies of branching corals was recorded in the Central Indian Ocean (Head et al., 2015). Their results suggested that decapods actively inhabit dead coral skeletons, particularly when their host has died. Present results reveal that symbiotic crabs use dead coral skeletons for shelter, and might change to another host if a more suitable choice is available. 4.4. Field assessment In the study area, the symbiotic crab T. rubridactyla inhabited only Acropora corals; neither P. damicornis nor S. pistillata was inhibited, thus corresponding to the laboratory experimental results. Unexpectedly, no dead branching coral skeletons were found in the area, thus, there were no field records on T. rubridactyla inhabiting dead coral skeletons. The field investigation results also revealed that Acropora had low coverage (0.8%) and all the Acropora colonies were occupied by symbiotic crabs. This suggests that the low density of Acropora might not have been conducive to the settlement of post-larval stage crabs because such a condition result in fewer opportunities to meet common host corals. Furthermore, juvenile and post-larval crabs might be at a high risk of encountering predators without adequate shelter while searching for the common host. Gotelli et al. (1985) reported that densities of Trapezia spp. and other common decapods were significantly correlated with environmental factors such as spatial, temporal, and density of host corals. In addition, studies have indicated that tetraliid crabs are specialized to inhabit genus Acropora corals (Castro et al.,

2004; Galil, 1988; Patton, 1994). The present results suggest that the density of Acropora corals might have been a crucial factor affecting the population size of T. rubridactyla in the study area. In this study, the correlation of symbiont capacity between T. rubridactyla crab density and the total volume of host-corals was non-significant. Studies have demonstrated that the size of a host coral colony has a significantly positive correlation and influences the population of symbiotic crab Trapezia spp. (Abele and Patton, 1976; Castro, 1978; Glynn, 1976). By contrast, Vytopil and Willis (2001) found that an abundance of Tetralia crabs was not affected by either the colony size or live surface area of Acropora corals. Gotelli et al. (1985) reported that the occurrence frequency of symbiotic Trapezia spp. was not affected by coral head size or season, but by reef location. Present results confirm those of Vytopil and Willis (2001) and Gotelli et al. (1985). Furthermore, territory behavior might be a reason why only a few individuals (1–2 crabs) were found in a colony (Castro, 1976; Baeza et al., 2002; Vytopil and Willis, 2001). On the other hand, the present study proved that the density of T. rubridactyla and the size of Acropora exhibited no pattern of a linear regression relationship. 5. Conclusion Present results reveal that when provided with various choice objects, T. rubridactyla crabs select their host according to the following factors: (1) Under the no-choice condition, they use any choice object as a host for shelter; (2) under the choice condition without a common host, they randomly select any uncommon choice object as a host; and (3) under the choice condition, if a common host is present, the selection of a common host coral is based on priority because it could provide adequate food and space. The crabs express neither fidelity nor preference between A. hyacinthus and A. digitifera. Thus, these results also suggest that the distribution of T. rubridactyla on Acropora corals in the reef is affected by an abundance of corals and the territory behavior of the crabs rather than the preferences of coral species. Furthermore, the results comply with Schweitzer (2005) who indicated that coevolution exist between corals and crabs; developing the optimal selection model for crabs that beyond food and space provided from uncommon hosts and dead coral skeletons. The data represent the multidimensional aspects of the host selection behavior of the symbiotic crab T. rubridactyla; therefore, its life history and ecology must be comprehensively examined in future studies. Acknowledgments We are grateful for the financial support from the Ministry of Science and Technology (MOST), Taiwan, through grants MOST 103-2611-M019-002, MOST 104-2621-M-019-002, and MOST 104-2611-M-019-004 provided to J.S. Hwang, grant NSC-1021324062 and 103AS-14.3.2-FAF1(5-1) provided to T.W. Shih, as well as grant MOST 104-2811-M-019005 provided to L.C. Tseng. We thank Dr. Yu-Rong Cheng (Institute of Oceanography, National Taiwan University, Taiwan), who helped and advised on the taxonomy work of corals. [SS] References Abele, L.G., 1976. Comparative species richness in fluctuating and constant environments: coral-associated decapod crustaceans. Science 192, 461–463. Abele, L.G., Patton, W.K., 1976. The size of coral heads and the community biology of associated decapod crustaceans. J. Biogeogr. 3, 35–47. Baeza, J.A., 2015. Crustaceans as symbionts: an overview of their diversity, host use and life styles. In: Thiel, M., Watling, L. (Eds.), The Life Styles and Feeding Biology. Oxford University Press, New York, pp. 163–189. Baeza, J.A., Stotz, W.B., 2001. Host-use pattern and host-selection during ontogeny of the commensal crab Allopetrolisthes spinifrons (H. Milne Edwards, 1837) (Decapoda: Anomura: Porcellanidae). J. Nat. Hist. 35, 341–355. Baeza, J.A., Stotz, W.B., 2003. Host-use and selection of differently colored sea anemones by the symbiotic crab Allopetrolisthes spinifrons. J. Exp. Mar. Biol. Ecol. 284, 25–39. Baeza, J.A., Stotz, W.B., Thiel, M., 2002. Agonistic behaviour and development of territoriality during ontogeny of the sea anemone dwelling crab Allopetrolisthes spinifrons (H.

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34 Milne Edwards, 1837) (Decapoda: Anomura: Porcellanidae). Mar. Freshw. Behav. Physiol. 35, 189–202. Barry, C.K., 1965. Ecological Study of the Decapod Crustaceans Commensal With the Branching Coral Pocillopora meandrina var. nobilis Verrill Master thesis University of Hawaii, Honolulu. Borradaile, L.A., 1921. On the coral-gall prawn Paratypton. Mem. Proc. Manch. Lit. Philos. Soc. 65, 1–11. Bruce, A.J., 1972. A review of information upon the coral hosts of commensal shrimps of the subfamily Pontoniinae, Kingsley, 1878 (Crustacea, Decapoda, Palaemonidae). In: Mukundan, G., Pillai, C.S.G. (Eds.), Proceedings of the Symposium on Corals and Coral Reefs, India, pp. 399–418. Bruce, A.J., 1998. New keys for the identification of Indo-West Pacific coral associated pontoniine shrimps, with observations on their ecology (Crustacea: Decapoda: Palaemonidae). Ophelia 49, 29–46. Bythell, J.C., Wild, C., 2011. Biology and ecology of coral mucus release. J. Exp. Mar. Biol. Ecol. 408, 88–93. Cacabelos, E., Olabarria, C., Incera, M., Troncoso, J.S., 2010. Do grazers prefer invasive seaweeds? J. Exp. Mar. Biol. Ecol. 393, 182–187. Castro, P., 1969. Symbiosis Between Echinoecus pentagonus (Crustacea, Brachyura) and Its Host in Hawaii, Echinothrix calamaris (Echinoidea) Ph.D. thesis University of Hawaii, Honolulu. Castro, P., 1976. Brachyuran crabs symbiotic with scleractinian corals: a review of their biology. Micronesica 12, 99–110. Castro, P., 1978. Movements between coral colonies in Trapezia ferruginea (Crustacea: Brachyura), an obligate symbiont of scleractinian corals. Mar. Biol. 46, 237–245. Castro, P., 1988. Animal symbioses in coral reef communities: a review. Symbiosis 5, 161–184. Castro, P., 2009. Shallow-water Trapeziidae and Tetraliidae (Crustacea: Brachyura) of the Philippines (Panglao 2004 expedition), New Guinea, and Vanuatu (Santo 2006 expedition). Raffles Bull. Zool. 20, 271–281. Castro, P., Ng, P.K.L., Ahyong, S.T., 2004. Phylogeny and systematics of the Trapeziidae Miers, 1886 (Crustacea: Brachyura), with the description of a new family. Zootaxa 643, 1–70. Chabanet, P., Ralambondrainy, H., Amanieu, M., Faure, G., Galzin, R., 1997. Relationships between coral reef substrata and fish. Coral Reefs 16, 93–102. Chang, K.H., Chen, Y.S., Chen, C.P., 1987. Xanthid crabs in the corals Pocillopora damicornis and P. verrucosa of southern Taiwan. Bull. Mar. Sci. 41, 214–220. Derby, C.D., Atema, J., 1980. Induced host odor attraction in the pea crab Pinnotheres maculatus. Biol. Bull. 158, 26–33. Fabricius, K.E., 2006. Effects of irradiance, flow, and colony pigmentation on the temperature microenvironment around corals: implications for coral bleaching? Limnol. Oceanogr. 51, 30–37. Fautin, D.G., Guo, C.C., Hwang, J.S., 1995. Costs and benefits of the symbiosis between the anemoneshrimp Periclimenes brevicarpalis and its host Entacmaea quadricolor. Mar. Ecol. Prog. Ser. 129, 77–84. Galil, B.S., 1988. Further notes on species of Tetralia (Decapoda, Trapeziidae). Crustaceana 54, 57–68. Garth, J.S., 1964. The Crustacea decapoda (Brachyura and Anomura) of Eniwetok Atoll, Marshall Islands, with special reference to the obligate commensals of branching corals. Micronesica 1, 137–144. Garth, J.S., 1984. Brachyuran decapod crustaceans of coral reef communities of the Seychelles and Amirante Islands. In: Stoddart, D.R. (Ed.), Biogeography and Ecology of the Seychelles Islands. Dr. W. Junk Publishers, The Hague, pp. 103–122. Glynn, P.W., 1976. Some physical and biological determinants of coral community structure in the eastern Pacific. Ecol. Monogr. 431–456. Glynn, P.W., 2013. Fine-scale interspecific interactions of coral reef: functional roles of small and cryptic metazoans. In: Lang, M.A., Marinelli, R.L., Roberts, S.J., Taylor, P.R. (Eds.), Research and Discoveries: The Revolution of Science Through Scuba. Smithsonian Institution Scholarly Press, Washington D.C., pp. 229–248. Glynn, P.W., Enochs, I.C., 2011. Invertebrates and their roles in coral reef ecosystems. In: Dubinsky, Z., Stambler, N. (Eds.), Coral Reefs: An Ecosystem in Transition. Springer, New York, pp. 273–325. Glynn, P.W., Perez, M., Gilchrist, S.L., 1985. Lipid decline in stressed corals and their crustacean symbionts. Biol. Bull. 168, 276–284. Góngora-Gómez, A.M., Muñoz-Sevilla, N.P., Hernández-Sepúlveda, J.A., García-Ulloa II, M., García-Ulloa, M., 2015. Association between the pen shell Atrina tuberculosa and the shrimp Pontonia margarita. Symbiosis 66, 107–110. Gotelli, N.J., Abele, L.G., 1983. Community patterns of coral-associated decapods. Mar. Ecol. Prog. Ser. 13, 131–139. Gotelli, N.J., Gilchrist, S.L., Abele, L.G., 1985. Population biology of Trapezia spp. and other coral-associated decapods. Mar. Ecol. Prog. Ser. 21, 89–98. Guo, C.C., Hwang, J.S., Fautin, D.G., 1996. Host selection by shrimps symbiotic with sea anemones: a field survey and experimental laboratory analysis. J. Exp. Mar. Biol. Ecol. 202, 165–176. Hale, R., Downes, B.J., Swearer, S.E., 2008. Habitat selection as a source of inter-specific differences in recruitment of two diadromous fish species. Freshw. Biol. 53, 2145–2157. Hamel, J.F., Ng, P.K.L., Mercier, A., 1999. Life cycle of the pea crab Pinnotheres halingi sp. nov., an obligate symbiont of the sea cucumber Holothuria scabra Jaeger. Ophelia 50, 149–175. Head, C.I., Bonsall, M.B., Koldewey, H., Pratchett, M.S., Speight, M., Rogers, A.D., 2015. High prevalence of obligate coral-dwelling decapods on dead corals in the Chagos Archipelago, central Indian Ocean. Coral Reefs 34, 905–915. Hill, J., Wilkinson, C., 2004. Methods for Ecological Monitoring of Coral Reefs. Australian Institute of Marine Science, Townsville. Huang, H.D., Rittschof, D., Jeng, M.S., 2005. Visual orientation of the symbiotic snapping shrimp Synalpheus demani. J. Exp. Mar. Biol. Ecol. 326, 56–66.

33

Huber, M.E., 1987. Aggressive behavior of Trapezia intermedia Miers and T. digitalis Latreille (Brachyura: Xanthidae). J. Crustac. Biol. 7, 238–248. Huebner, L.K., Chadwick, N.E., 2012. Reef fishes use sea anemones as visual cues for cleaning interactions with shrimp. J. Exp. Mar. Biol. Ecol. 416–417, 237–242. Jackson, A.C., Underwood, A.J., 2007. Application of new techniques for the accurate analysis of choice of prey. J. Exp. Mar. Biol. Ecol. 341, 1–9. Jeng, M.S., 2000. Decapod crustaceans associated with unbleached and bleached colonies of the coral, Seriatopora hystrix from near the thermal discharge of a nuclear power plant in Taiwan. Crustac. Issues 12, 77–84. Knudsen, J.W., 1967. Trapezia and Tetralia (Decapoda, Brachyura, Xanthidae) as obligate ectoparasites of pocilloporid and acroporid corals. Pac. Sci. 21, 51–57. Kohler, K.E., Gill, S.M., 2006. Coral Point Count with Excel extensions (CPCe): a visual basic program for the determination of coral and substrate coverage using random point count methodology. Comput. Geosci. 32, 1259–1269. Kramer, M.J., Bellwood, D.R., Bellwood, O., 2014. Benthic Crustacea on coral reefs: a quantitative survey. Mar. Ecol. Prog. Ser. 511, 105–116. Kropp, R.K., 1989. A revision of the Pacific species of gall crabs, genus Opecarcinus (Crustacea: Cryptochiridae). Bull. Mar. Sci. 45, 98–129. Limviriyakul, P., Tseng, L.C., Hwang, J.S., Shih, T.W., 2016. Anomuran and brachyuran symbiotic crabs in coastal areas between the southern Ryukyu arc and the Coral Triangle. Zool. Stud. 55, 7. Mascaró, M., Rodríguez-Pestaña, L., Chiappa-Carrara, X., Simões, N., 2012. Host selection by the cleaner shrimp Ancylomenes pedersoni: do anemone host species, prior experience or the presence of conspecific shrimp matter? J. Exp. Mar. Biol. Ecol. 413, 87–93. McCoy, E.D., Bell, S.S., 1991. Habitat structure: the evolution and diversification of a complex topic. In: Belll, S.S., McCoy, E.D., Mushinsky, H.R. (Eds.), Habitat Structure: The Physical Arrangement of Objects in Space. Chapman and Hall, New York, pp. 3–27. Ocampo, E.H., Nuñez, J.D., Cledón, M., Baeza, J.A., 2012. Host-specific reproductive benefits, host selection behavior and host use pattern of the pinnotherid crab Calyptraeotheres garthi. J. Exp. Mar. Biol. Ecol. 429, 36–46. Olabarria, C., Underwood, A.J., Chapman, M.G., 2002. Appropriate experimental design to evaluate preferences for microhabitat: an example of preferences by species of microgastropods. Oecologia 132, 159–166. Patton, W.K., 1966. Decapod Crustacea commensal with Queensland branching corals. Crustaceana 10, 271–295. Patton, W.K., 1994. Distribution and ecology of animals associated with branching corals (Acropora spp.) from the Great Barrier Reef, Australia. Bull. Mar. Sci. 55, 193–211. Patton, W.K., Patton, R.J., Barnes, A., 1985. On the biology of Gnathophylloides mineri, a shrimp inhabiting the sea urchin Tripneustes ventricosus. J. Crustac. Biol. 5, 616–626. Pinna, S., Pais, A., Campus, P., Sechi, N., Ceccherelli, G., 2012. Habitat preferences of the sea urchin Paracentrotus lividus. Mar. Ecol. Prog. Ser. 445, 173–180. Poupin, J., 2008. Biogeography of the decapod and stomatopod Crustacea of the tropical Pacific: issues and prospects. Pac. Sci. 62, 377–383. Pratchett, M.S., 2001. Influence of coral symbionts on feeding preferences of crown-ofthorns starfish Acanthaster planci in the western Pacific. Mar. Ecol. Prog. Ser. 214, 111–119. Rouzé, H., Lecellier, G., Mills, S.C., Planes, S., Berteaux-Lecellier, V., Stewart, H., 2014. Juvenile Trapezia spp. crabs can increase juvenile host coral survival by protection from predation. Mar. Ecol. Prog. Ser. 515, 151–159. Schiller, C., Herndl, G.J., 1989. Evidence of enhanced microbial activity in the interstitial space of branched corals: possible implications for coral metabolism. Coral Reefs 7, 179–184. Schweitzer, C.E., 2005. The Trapeziidae and Domeciidae (Decapoda: Brachyura: Xanthoidea) in the fossil record and a new Eocene genus from Baja California Sur, Mexico. J. Crustac. Biol. 25, 625–636. Silbiger, N.J., Childress, M.J., 2008. Interspecific variation in anemone shrimp distribution and host selection in the Florida Keys (USA): implications for marine conservation. Bull. Mar. Sci. 83, 329–345. Silva, A.C., Hawkins, S.J., Clarke, K.R., Boaventura, D.M., Thompson, R.C., 2010. Preferential feeding by the crab Necora puber on differing sizes of the intertidal limpet Patella vulgata. Mar. Ecol. Prog. Ser. 416, 179–188. Sin, T., 1999. Distribution and host specialization in Tetralia crabs (Crustacea: Brachyura) symbiotic with corals in the Great Barrier Reef, Australia. Bull. Mar. Sci. 65, 839–850. Sin, T.M., Lee, A.C., 2002. Host specialisation in trapeziid crabs: consequences for rarity at local scales. In: Moosa, M.K., Soemodihardjo, S., Soegiarto, A., Romimohtarto, K., Nontji, A., Suharsono, S. (Eds.), Proceedings of the Ninth International Coral Reef Symposium, Bali, pp. 533–536. Singer, M.C., 2000. Reducing ambiguity in describing plant-insect interactions: “preference”, “acceptability” and “electivity”. Ecol. Lett. 3, 159–162. Stella, J.S., Munday, P.L., Jones, G.P., 2011a. Effects of coral bleaching on the obligate coraldwelling crab Trapezia cymodoce. Coral Reefs 30, 719–727. Stella, J.S., Pratchett, M.S., Hutchings, P.A., Jones, G.P., 2011b. Coral-associated invertebrates: density, ecological importance and vulnerability to disturbance. Oceanogr. Mar. Biol. Annu. Rev. 49, 43–104. Stevens, P.M., 1990. Specificity of host recognition of individuals from different host races of symbiotic pea crabs (Decapoda: Pinnotheridae). J. Exp. Mar. Biol. Ecol. 143, 193–207. Stewart, H.L., Holbrook, S.J., Schmitt, R.J., Brooks, A.J., 2006. Symbiotic crabs maintain coral health by clearing sediments. Coral Reefs 25, 609–615. Stimson, J., 1990. Stimulation of fat-body production in the polyps of the coral Pocillopora damicornis by the presence of mutualistic crabs of the genus Trapezia. Mar. Biol. 106, 211–218. Thiel, M., Zander, A., Baeza, J.A., 2003a. Movements of the symbiotic crab Liopetrolisthes mitra between its host sea urchin Tetrapygus niger. Bull. Mar. Sci. 72, 89–101.

34

P. Limviriyakul et al. / Journal of Experimental Marine Biology and Ecology 485 (2016) 24–34

Thiel, M., Zander, A., Valdivia, N., Baeza, J.A., Rueffler, C., 2003b. Host fidelity of a symbiotic porcellanid crab: the importance of host characteristics. J. Zool. 261, 353–362. Tsuchiya, M., 1999. Effect of mass coral bleaching on the community structure of small animals associated with the hermatypic coral Pocillopora damicornis. Galaxea 1, 65–72. Tsuchiya, M., Nojima, S., 2002. Occurrence of Trapezia associated with Acropora: on the “wrong” host coral? Coral Reefs 21, 160. Tsuchiya, M., Yamauchi, Y., Moretzsohn, F., Tsukiji, M., 1993. Species composition and some population traits of obligate symbiotic xanthid crabs, Trapezia and Tetralia, associated with bleached corals. In: Richmond, R.H. (Ed.), Proceedings of the Seventh Coral Reef International Symposium, Guam, pp. 56–63. Underwood, A.J., Clarke, K.R., 2005. Solving some statistical problems in analyses of experiments on choices of food and on associations with habitat. J. Exp. Mar. Biol. Ecol. 318, 227–237. Underwood, A.J., Clarke, K.R., 2006. Response on a proposed method for analysing experiments on food choice. J. Exp. Mar. Biol. Ecol. 335, 151–153. Underwood, A.J., Chapman, M.G., Crowe, T.P., 2004. Identifying and understanding ecological preferences for habitat or prey. J. Exp. Mar. Biol. Ecol. 300, 161–187.

Untersteggaber, L., Mitteroecker, P., Herler, J., 2014. Coral architecture affects the habitat choice and form of associated gobiid fishes. Mar. Biol. 161, 521–530. Vannini, M., 1985. A shrimp that speaks crab-ese. J. Crustac. Biol. 5, 160–167. Veron, J.E.N., 2000. Corals of the World. Australian Institute of Marine Science, Townsville. Vytopil, E., Willis, B., 2001. Epifaunal community structure in Acropora spp. (Scleractinia) on the Great Barrier Reef: implications of coral morphology and habitat complexity. Coral Reefs 20, 281–288. Wallace, C.C., 1999. Staghorn Corals of the World. CSIRO Publishing, Melbourne. Wei, T.P., Hwang, J.S., Tsai, M.L., Fang, L.S., 2005. New records of gall crabs (Decapoda, Cryptochiridae) from Orchid Island, Taiwan, northwestern Pacific. Crustaceana 78, 1063–1077. Wei, T.P., Chen, H.C., Lee, Y.C., Tsai, M.L., Hwang, J.S., Peng, S.H., Chiu, Y.W., 2013. Gall polymorphism of coral-inhabiting crabs (Decapoda, Cryptochiridae): a new perspective. J. Mar. Sci. Technol. 21, 304–307.