Host specificity and microhabitat preference of symbiotic copepods (Cyclopoida: Clausiididae) associated with ghost shrimps (Decapoda: Callichiridae, Callianideidae)

Abstract We examined the host specificity of two ectosymbiotic Clausidium Kossman, 1874 copepods (Cyclopoida: Clausiididae) on two co‐occurrence species of host ghost shrimps. Our results revealed that both species of symbiotic copepod demonstrated extremely high host specificity. Moreover, within a single host shrimp species, each symbiont species displayed strong spatial patterns in microhabitat selection on their hosts’ bodies. Clausidium persiaensis Sepahvand & Kihara, 2017, was only found on the host Callianidea typa Milne Edwards, 1837 and almost exclusively within the host shrimp gill chamber, while C. iranensis Sepahvand, Kihara, & Boxshall, 2019 was only found on the host Neocallichirus jousseaumei (Nobili, 1904) and showed extremely strong preferences for the chelae and anterior walking legs. We also found that while the number of symbionts tends to increase with the host size, the two host species differed in the degree of symbiont infestation, with large C. typa hosting approximately 7× as many symbionts as the similarly sized N. jousseaumeia. The mechanisms resulting in the observed differences in infestation levels and microhabitat preferences of clausidium copepods among their hosts, including differences in physiology, burrowing pattern, and host grooming behavior should be further investigated.

type of site specificity, the specific physicochemical microhabitat is the most commonly invoked explanation for this phenomenon (e.g., Bychowsky, 1961;Wootten, 1974). Despite the ubiquity of this phenomenon, the mechanisms underlying host specificity are largely understudied.
From a symbiont's perspective, a population of potential hosts is a heterogeneous landscape. Hosts frequently vary in quality across species (Brown & Creed, 2004;Farrell, Creed, & Brown, 2014;Rohde, 1994) and there may even be significant variation in habitat quality across individuals within the same host species (Lie, 1973).
Even at the within-host level, microhabitats or specific tissues may vary with respect to the resources they offer, or the risk of mortality within each microhabitat patch (Mestre, Mesquita-Joanes, Proctor, & Monrós, 2011;Skelton, Creed, & Brown, 2014;Skelton, Geyer, Lennon, Creed, & Brown, 2017). Moreover, each host and each microhabitat presents a limited pool of resources, creating the possibility of strong inter-and intra-specific interactions among symbionts (Baker, Andras, Jordán-Garza, & Fogel, 2013;Råberg et al., 2006;Ulrich & Schmid-Hempel, 2012). Recently, Ivanenko et al. (2018) revealed that there was a lack of host specificity of associated copepods with mushroom corals in the red sea. The authors suggested that the association between copepods and their host corals is not strict, and not phylogenetically constrained. To address interaction among symbionts in one specific system, we investigated the host specificity and microhabitat preferences of two cyclopoid copepod associate with ghost shrimps.
Most clausidiid copepods live attached to the marine invertebrates host, and species of Clausidium Kossman, 1874 are known to live close association with burrowing shrimps (Boxshall & Halsey, 2004). The information on the behavior of these copepods, on their interactions with their host, with the environment, is very scarce.
per host was dependent upon host habitat type and recruitment. Corsetti and Strasser (2003)

| MATERIAL AND ME THODS
We conducted a field survey to collect ghost shrimps and their associated Clausidium copepods to assess host specificity, microhabitat selection, and relationships between symbiont abundance to host size. All organisms included in this study were collected at the Oli Total length (TL, measured from the tip of the rostrum to the posterior end of the telson) and carapace length (CL, measured from the tip of the rostrum to the posterior end of the carapace) were recorded for each ghost shrimp. We identified ghost shrimps to species and transported each species to the laboratory in separate collection falcon tube to prevent interspecific transfer of clausidium copepods. We collected copepod symbionts from submerged hosts in the laboratory using a dissecting microscope. In order to map the distribution of copepods on their host, we divided the exoskeleton of the ghost shrimp into four regions based on natural morphological divisions, as illustrated in (Figure 3). We removed copepods by hand and recorded the location of each copepods on the exoskeleton.
We analyzed the distribution and intensity of symbiotic copepods using multiple methods that allowed us examine both the pre- when analyzing community data, in our case, the copepod symbiont community was censused on each host rather than sampled, so observations of zero at a site are meaningful and Euclidean distance was more appropriate. We also examined the multivariate dispersion of symbionts using permutational analysis of dispersion (Anderson, 2006; function betadisper ( ) in the R package vegan).

| RE SULTS
We discovered two co-occurrence species of ghost shrimp, Neocallichirus jousseaumei (Nobili, 1904) and Callianidea typa, Milne Edwards, 1837 that served as hosts to copepod symbionts. with approximately 85% of symbiotic copepods present at that microhabitat, followed by the chelipeds (Chel, 5%), thoracic legs (legs, 3%), and Abdomen and telson (Abtel, 2%; Figure 5). Copepods were most frequently attached to N. jousseumei at the chelipeds (Chel, ca 80%) with the carapace and abdomen/telson microhabitats infrequently occupied (8% and 2%, respectively, Figure 5). For both symbiont species, microsite preference was highly nonrandom. On C. typa, intensity on the dominant site, carapace, was strongly predicted by both host sex and host size, though there was no interaction between these factors (Table 1). For the nondominant sites (chelae, legs, and abdomen/telson), host sex was the only significant predictor. However, for all sites, the significant sex effect was largely driven by the inclusion of juveniles whose sex could not be identified (mean symbionts ± SD, female = 74.5 ± 38.6, male = 68.3 ± 39.1, juvenile = 9.5 ± 4.2). There was also no significant relationship in intensity between the dominant site, carapace, and nondominant sites (Table 1). For N. jousseaumei, intensity on the dominant site, chelae, was significantly related to both host size and host sex, and there was a significant interaction between the predictors (Table 1).
For N. jousseaumei, the host sex effect was not simply driven by lower intensities on juveniles. There was a significant difference be-  (Table 2). PERMANOVA showed strong effects of host species, host size, and host sex, as well as all 2-way interactions (Table 3).
There was no significant difference in multivariate dispersion between the 2 hosts (p = .084). The data analysis showed that host-size adjusted density of clausidium copepods was affected by the host species and the host sex. While the two species of ghost shrimp hosts are similar in size, the symbiont density × host size relationship differed markedly between hosts. While both hosts had few symbionts at sizes < 35 mm, the density of symbionts on C. typa increased far more rapidly with host size than on N.jousseaumei. Symbiont densities on the largest sized C. typa were 4× greater than comparably sized N. jousseaumei.

| D ISCUSS I ON
While the explanation for this disparity is not readily apparent, one explanation is that the symbiotic copepod C. persiaensis preferentially attaches to C. typa, or that N. jousseaumei exhibits selective host defense and repels the symbiont, possibly through removal by grooming behavior.
Another possibility is that the feeding mechanism of C. typa and N. jousseaumei may influence the abundance of copepods. Ghost shrimp feed in a variety of ways including filtration of plankton and deposit feeding, and generally consume microalgae and other diatoms (Felder & Griffis, 1994). Most callianassid ghost shrimp feed by sifting sand for microscopic organisms using their mouthpart to remove food particles from setae of the maxillipeds (Pohl, 1946). Hayes (1976), showed that recruitment of the planktonic host-seeking stage of copepods, Copepodid I in clausidium copepods may depend on the host's feeding mechanism, hence the trophic mode of We also found that there was a significant difference in the number of copepods selecting each host sex. This result could signify that copepods base their selection of host in a hierarchical fashion, with host species forming a first hierarchy, and host size a second.
Since females are generally larger in size than male higher copepod colonization on female specimens may be a result of size and not sex.
Explanations for host selection may also be evolutionary in origin and be related to contact time among these copepods and ghost shrimps in the Persian Gulf. The Persian Gulf is relatively young with coastlines that formed only in the past 3,000-6,000 years (Riegl & Purkis, 2012) and the evolutionary process within it may have been affected by historical events such as glaciation and sea level fluctuation (Sheppard et al., 2010). Regarding the short history of the region, contact time between ghost shrimps and Clausidium copepods has been limited, as has the time during which clausidium copepods have adopted new species of ghost shrimp hosts. Although some studies suggest that coevolution between symbiont and host can occur rapidly when parasitism is high (Soler, Martinez, Soler, & Møller, 1994;Takasu, Kawasaki, Nakamura, Cohen, & Shigesada, 1993), it is not clear whether the intensity of selection, either positive or negative, and time since the clausidium copepod introduction have been sufficient for ghost shrimps to have evolved responses.
Each symbiont species displayed strong patterns of microhabitat selection. This site selection was apparent in two ways: (a) within a single symbiont species, there was clear site selection on a host; (b) there were also strong differences in site selection between the two hosts. Clausidium persiaensis, was only found on C. typa and almost exclusively on the carapace and within the gill chamber, while C. iranensis was only found on N. jousseaumei and showed extremely strong preferences for the chelae and anterior walking legs.
Our result showed that for both symbiont species, microsite preference was highly nonrandom. Possible explanations for site preferences are minimizing risks from host defensive grooming behaviors, or constraint from environmental parameters such as current strength.
We consider three hypotheses that may explain microhabitat preference in two clausidium copepod species on their hosts: first, grooming behavior (GB) of the ghost shrimp (Bauer, 1981) and copepods occupying protected zones making them inaccessible to " grooming "; second, burrowing behavior (BB) of hosts determines the current velocity experienced by symbionts (Dworschak, Felder, & Tudge, 2012) and copepods choose microhabitats that minimize the force of "flow"; third, niche partitioning (NP) could be a strategy to increase the possibility of finding mates. Diverse grooming structures and behaviors have evolved in decapod crustaceans in response to the selective pressure of fouling (Bauer, 1981). General body grooming of decapods, performed by serrate setal brushes on chelipedes and/ or posterior pereiopods (Bauer, 1981). Fifth pereiopods as a main appendage for grooming of the carapace and gills of C. typa and N.jousseaumei are almost morphologically similar, while these shrimps are different in maxiliped 3 and in another appendages (major and minor chelipeds, pereiopods 2-4).
The (BB) and the (NP) hypotheses explain the microhabitat selection advantage in Clausidium.
Two ghost shrimp hosts do differ in behavior in several ways including their patterns of burrowing, and habitat selection (Griffis & Suchanek, 1991). Burrowing patterns determine the current velocity of water in burrows and consequently on the body of the ghost shrimps (Griffis & Suchanek, 1991). Possible differences in water current strength in the burrow of hosts may also explain the differences of microhabitat preferences in clausidium copepods.
The flattened body shape of the genus Clausidium (Figures 1b,   2b) is probably an adaptation against stress experienced at their habitats. Marin and Nascimento (1993)  were found coupling during the mating process (Figures 1b, 2b), providing evidence that mating may benefit from microsite partitioning on the host. However, Timi (2003) suggested that microhabitat restriction of Lernanthropus cynoscicola (parasitic copepod) is not due to facilitation of mating. Additionally, he showed that aggregation among individuals of the same sex was stronger than among males and females, and the co-occurrence of both sexes did not depart from that expected by chance. While this present study was not designed to explore the ecological significance of TA B L E 3 Results of PERMANOVA testing the effects of host species, host size, and host sex on the multivariate distribution of symbiotic copepods on ghost shrimp hosts host site preference, we suggest that some evidence does support the burrowing and mating hypotheses. However, the discussed hypotheses should be further studied to illuminate the evolutionary and adaptive advantages of niche differentiation in copepods and their hosts. Rohde (1979) reviewed intrinsic and extrinsic factors those are responsible for niche restriction in parasites. The author emphasized that intrinsic (intraspecific) factors are largely responsible for niche restriction. Rohde (1979) argued that intrinsic factors play some roles in determining niches in monogen species and suggested that narrow microhabitats may function to enhance mate-locating chances. On the other hand, site selection within the host also may relate to the physicochemical environment (e.g., Bychowsky, 1961;Wootten, 1974).
Positive correlations between host size and symbiont density or biomass are frequently reported, especially in parasite systems (e.g., Arneberg, Skorping, Grenfell, & Read, 1998;Grutter & Poulin, 1998;Mohr, 1961;Poulin, 2007;Saad-Fares & Combes, 1992). However, some authors have also observed that lower levels of parasitism may occur in the largest hosts (Shotter, 1973;Kabata, 1959;Etchegoin & Sardella, 1990). Our result revealed that N. jousseaumei, despite a larger body size, was host to fewer copepods on average (mean 15.4) when compared to C.typa (mean 64.41). These results do not agree with a previous study in which the abundance of clausidium copepods was directly related to the size of host (Marin & Nascimento, 1993). It is possible that the difference in densities level of clausidium on the body of hosts relate to the host's physiology, since member of the Clausidium genus is same in biology and attachments mechanisms. Corsetti and Strasser (2003) showed that copepod densities are correlated to the host species and host size, but host sex was unimportant.
In conclusion, there is strong evidence from field surveys that

ACK N OWLED G M ENTS
The authors would like to thank the Iranian National Institute for Oceanography and Atmospheric Science for cooperation during this study and Dr. Terue C. Kihara from Senckenberg am Meer, German Center for Marine Biodiversity Research, Wilhelmshaven for help and encouragement during this study. We are also very grateful to Dr. Francisco Neptalí Morales-Serna for his helpful and interesting comments and suggestions to improve the quality of the manuscript.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The raw data are available at Dryad with this https://doi.org/10.5061/ dryad.4qrfj 6q6p.