Crustacean Research 47 REPRODUCTIVE STRATEGIES of CRINOID-ASSOCIATED SHRIMPS

Home , Comatulidae, Cuapetes, Hippolyte (shrimp), Periclimenes rathbunae

Crustacean Research 2018 Vol.47: 89–99 ©Carcinological Society of Japan. doi: 10.18353/crustacea.47.0_89 Reproductive strategies of the crinoid-associated shrimps Cristimenes commensalis (Borradaile, 1915) and Pontoniopsis comanthi Borradaile, 1915 (Crustacea: Palaemonidae)

Sancia E. T. van der Meij, Magnus L. Johnson, Sammy De Grave

Abstract.̶ Tropical shallow-water crinoids offer shelter to a diversity of marine life, including various symbiotic palaemonid shrimp species. In this study, the reproductive features of the shrimps Cristimenes commensalis and Pontoniopsis comanthi inhabiting comatulid crinoids are studied and compared. Ovigerous females were collected from the crinoids Anneissia bennetti and Comaster nobilis from Iriomote-jima, Japan. All specimens were examined for the following traits: carapace length, fecundity, embryo volume, brood mass volume and dry weight of the females and the eggs. Cristimenes commensalis produced 15–117 eggs with a volume of 0.04 mm3±0.00, whereas the smaller-sized species P. comanthi produced 12–57 eggs with a volume of 0.02 mm3±0.01. The reproductive output (RO) of C. commensalis was 0.16±0.03, considerably lower than that of P. comanthi at 0.31±0.05. Data is beginning to emerge on the fecundity and reproductive output of symbiotic palaemonid shrimps, although comparisons remain hampered by the general lack of knowledge on their biology in general. Pontoniopsis comanthi has the highest recorded reproductive output so far, despite being the smallest species studied. Our results conﬁrm that inhabiting a sheltered habitat allows symbiotic shrimps to allocate more energy to embryo production than free-living species. Different reproductive strategies are, however, apparent for these species inhabiting the same host species, implying the existence of differential selective forces.

Key words: Caridea, co-occurrence, egg volume, fecundity, reproductive output, symbiosis

■ Introduction thus far (Feldman & Brett, 1998). The high diversity of crinoid dwellers is like- Coral reef ecosystems are teeming with sym- ly related to their complex external morpholo- biotic relationships. All major sessile reef taxa gy, which allows for co-occurrence of various have distinct associated faunas, and many of associated species. Crinoids are ﬁlter-feeders; these symbiont species are speciﬁcally adapted they concentrate food with their feather-like to their host organism. Crinoids host numerous arms, which they then transport from the pin- symbionts, de facto sustaining a microcosm nules to the mouth along the ambulacral ruled by crinoid-symbiont as well as symbiont- groove, making the food particles easily acces- symbiont interactions (Potts, 1915; Fishelson, sible to symbionts. The crinoid family Comatu- 1974; Fabricius & Dale, 1993). These associa- lidae Fleming, 1828 (previously Comasteridae tions are known from modern and fossil fau- A.H. Clark, 1908) lacks protective lappets over nas, providing evidence for one of the longest- the groove, and appears to be particularly at- ranging host-symbiont interactions documented tractive to symbiotic organisms (Deheyn et al.,

Received: 18 May 2018. Accepted: 25 July 2018. Published online: 24 Oct 2018. 89 SANCIA E. T. VAN DER MEIJ, MAGNUS L. JOHNSON, SAMMY DE GRAVE

2006; Britayev & Mekhova, 2011). Rates of up 1991), bivalve associates (Johnson & Liang, to 92% inhabitation of crinoids by symbionts 1966; Corey & Reid, 1991; Bruce, 2000), as have been reported, with the associated fauna well as several free-living species (Corey & mainly consisting of myzostomids, shrimps, Reid, 1991) of similar body size to the symbi- ophiuroids and polychaetes (Fabricius & Dale, otic species. All these studies have, however, 1993; Deheyn et al., 2006). only concentrated on a single species, with no The caridean family Palaemonidae encom- sympatric or host-sharing taxa documented, in passes species with a diverse range of life- contrast to studies in other families (Soledade styles, including free-living and symbiotic spe- et al., 2017). This is largely due to the fact that cies (De Grave, 2002). Studies on the most palaemonid species are highly host spe- symbiotic relationships between palaemonid cific and it is relatively rare that more than one shrimps and their invertebrate hosts are numer- shrimp species occurs in high population den- ous, and include research on cleaning behav- sities on a given host, with crinoid symbionts iour (Horká et al., 2018), host use (e.g., Guo et being a notable exception to this rule (Huang et al., 1996; Ďuriš et al., 2011; Levitt-Barmats & al., 2005). Shenkar, 2018), and host switching from an Here we report on the fecundity and repro- evolutionary viewpoint (Horká et al., 2016). ductive output of the symbiotic shrimps Cris- Although several studies have been carried timenes commensalis (Borradaile, 1915) and out detailing reproductive output, embryo vol- Pontoniopsis comanthi Borradaile, 1915, col- ume and mass of caridean species across sever- lected from the comatulid crinoid species An- al shrimp families (Palaemonidae, Hippolyti- neissia bennetti (Müller, 1841) (previously dae, Alpheidae; Corey & Reid, 1991; Kim & known as Oxycomanthus bennetti (Summers et Hong, 2004; Terossi et al., 2010; Soledade et al., 2014)) and Comaster nobilis (Carpenter, al., 2017), only two such studies have so far 1884). Cristimenes commensalis is widespread been carried out on symbiotic Palaemonidae. across the entire Indo-West Pacific, from Ke- The fecundity of the sea-anemone symbiont nya through to Japan and the Marshall Islands Periclimenes rathbunae Schmitt, 1924 was (De Grave, 2002), and is associated with a va- studied in Costa Rica (Azofeifa-Solano et al., riety of species in the crinoid families Comas- 2014), whilst the jelly-fish symbiont, Peri- teridae, Zygometridae and Himerometridae; climenes paivai Chace, 1969 was studied in generally occupying the distal part of the cri- Brazil (de Moraes et al., 2017). This level of noid arms and occurring in groups of several detail is also available for two free-living spe- heterosexual pairs and additional juveniles cies of a body size similar to the majority of (Zmarzly, 1984). Pontoniopsis comanthi is symbiotic species, Cuapetes americanus equally widespread (De Grave, 2002), primari- (Kingsley, 1878) (Negri et al., 2017) and Phy- ly associated with crinoids in the families Co- comenes siankaanensis (Martinez-Mayén & masteridae and Himerometridae; generally oc- Román-Contreras, 2006) (Martínez-Mayén & curring in relatively large groups, sharing the Román-Contreras, 2009), as Periclimenes sian- crinoid ecospace with other species, like said C. kaanensis). More information is available on commensalis. The co-occurrence on the same the number of eggs produced in symbiotic host species allows us to compare the repro- Palaemonidae, with this being available for ductive strategies of these crinoid-associated several sea-anemone symbionts (Omori et al., shrimps and study whether the same selective 1994; McKeon & O’Donnell, 2015), a fish- pressures govern their respective reproductive cleaner also inhabiting sea-anemones (Spotte, strategies. 1999), a gorgonian associate (Heard & Spotte,

90 Crustacean Research 47 REPRODUCTIVE STRATEGIES OF CRINOID-ASSOCIATED SHRIMPS

■ Material and Methods of the total energy budget diverted to reproduction. To obtain the RO values the brood dry Crinoids were collected during two consecu- weight was divided by the female’s dry weight. tive days (9–10 June 2014) from Baras Island, The statistical analyses were performed using Iriomote-jima, Yaeyama Islands, Japan (24°26′ R (R Core Team, 2018). Data were assessed for 25.9″N, 123°48′48.4″E). The collecting site normality using the Shapiro test. For non-para- was a relatively barren rock/canyon area, with metric data, analysis of covariance was exam- high current velocities and an approximate ined using the fANCOVA package (Wang, depth of 15–20 meters. Crinoids were individu- 2010). To test for any possible effects of host ally bagged into re-sealable zipper storage bags crinoid species on the fecundity of C. commen- (Umbrella Corp.) and transported in cooler salis a Wilcoxon signed-rank test was used boxes to the laboratory, where the individual (which in R is identical to the Mann-Whitney crinoids were placed in shallow trays. All visi- U test; Ennos & Johnson, 2018). A significance ble shrimp symbionts were removed from their level of 0.05 was adopted for the statistical host, aided by the occasional application of analyses and as the data were not normally dis- small quantities of diluted formalin by pipette. tributed, results are displayed as medians and After removal of the symbionts, the crinoids 95% confidence interval (CI) to cross compare were washed with jets of water over a 1 mm both species. CI’s were calculated using the sieve to dislodge remaining symbionts, and groupwise Median function (Mangiafico, 2018) subsequently preserved in 70% ethanol. The with the bootstrap level set to 5000. Although shrimp C. commensalis was collected from in order to compare with other species, means both A. bennetti and C. nobilis, whereas P. co- ±SD are also used, as only those values are manthi was only collected from A. bennetti. reported in literature. Female carapace length (CL) was measured from the post-orbital margin to the posterior ■ Results border of the carapace. The egg masses of gravid C. commensalis and P. comanthi were Inhabitation of crinoids by C. commensalis counted and classified into two stages (Zim- was 86.4%, compared to 50% by P. comanthi. mermann et al., 2015): stage I (early develop- The group size of C. commensalis ranged from ment, no eyespots visible) and stage II (late de- 2–13 specimens (median 6), and from 1–17 velopment, with visible eyespots). Fecundity specimens (median 5) for P. comanthi. The was determined as the total number of eggs at- groups consisted of a mix of male, female tached to the female’s abdomen. The methods (ovigerous and non-ovigerous) and juvenile for egg and brood volume measurements were specimens, of which only the ovigerous fe- adopted from literature (Zimmermann et al., males were examined herein. Of the 32 studied 2015). Egg samples and females were weighed females of C. commensalis, the majority (n＝ separately with a KERN ABT 120–5DM ana- 17) carried stage I eggs, with 15 individuals lytical balance, with an accuracy of 0.01 mg. carrying stage II eggs. Of the 25 P. comanthi Female weight was quantified after removing females, 13 carried stage I eggs and 12 carried all the eggs from the abdomen. All weight stage II eggs. There is no significant difference measurements are dry weight values after 24 in the number of eggs between stages I and II hours of oven drying at 60° C in a standard for C. commensalis (W＝159, df＝30, p＝0.24) drying cabinet. Brood weight was determined or P. comanthi (W＝60.5, df＝23, p＝0.35). as the total weight of the egg batch. The repro- The number of eggs ranged between 15 and ductive output (RO) is defined as the fraction 117 for C. commensalis and between 12 and 57

Crustacean Research 47 91 SANCIA E. T. VAN DER MEIJ, MAGNUS L. JOHNSON, SAMMY DE GRAVE

Table 1. Reproductive traits (median±95% CI) of Cristimenes commensalis and Pontoniopsis comanthi for all specimens combined, as well as separated according to embryo development stage.

CL Embryo Embryo volume Brood mass volume Dry weight-female Dry weight-eggs Means Embryo RO (mm) number (mm3) (mm3) (mg) (mg)

C. commensalis 2.03 49.5 0.04 1.98 2.83 0.47 0.17 15–117 (n＝32) (1.95–2.10) (38.0–58.0) (0.04–0.05) (1.57–2.50) (2.59–3.30) (0.33–0.56) (0.15–0.17)

Stage I 2.05 53 0.04 1.82 2.79 0.50 0.18 23–104 (n＝17) (2.00–2.25) (38–67) (0.03–0.04) (1.57–2.33) (2.34–3.83) (0.30–0.62) (0.16–0.19)

Stage II 2.00 45 0.05 2.59 2.86 0.43 0.14 15–117 (n＝15) (1.75–2.10) (28–60) (0.05–0.06) (1.35–2.81) (2.30–3.64) (0.28–0.58) (0.11–0.17)

P. comanthi 1.10 27.0 0.02 0.55 1.18 0.38 0.32 12–57 (n＝25) (1.10–1.20) (19.0–32.0) (0.02–0.03) (0.44–0.75) (1.04–1.29) (0.30–0.43) (0.30–0.35)

Stage I 1.10 21.0 0.02 0.44 1.04 0.36 0.35 12–57 (n＝13) (1.10–1.25) (18.0–31.0) (0.02–0.02) (0.28–0.45) (0.96–1.24) (0.29–0.41) (0.32–0.36)

Stage II 1.15 29.5 0.03 0.81 1.31 0.4 0.30 12–37 (n＝12) (1.10–1.20) (23.5–32.5) (0.02–0.03) (0.63–0.97) (1.18–1.40) (0.30–0.47) (0.24–0.34)

Fig. 1. Weight in relation to carapace length for a) C. commensalis and b) P. comanthi. ANCOVA indicated no significant difference in the relationships with egg stage as a factor (p＞0.05). Stage I eggs＝open circles, stage II eggs＝closed circles. for P. comanthi, with medians (95% CI) of nificantly larger than for P. comanthi (W＝800, 49.5 (38.0–58.0) and 27 (19.0–32.0), respec- df＝55, p＜0.001) as was female dry weight tively (Table 1). (W＝798, df＝55, p＜0.001) (Table 1). There Carapace length of C. commensalis was sig- is no significant difference in female dry

92 Crustacean Research 47 REPRODUCTIVE STRATEGIES OF CRINOID-ASSOCIATED SHRIMPS

Fig. 2. Fecundity in relation to carapace length in mm a) C. commensalis, and b) P. comanthi. ANCOVA indicated no signiﬁcant difference in the relationships with egg stage as a factor (p＞0.05). Stage I eggs＝open circles, stage II eggs＝closed circles.

Fig. 3. Boxplot of embryo volume in mm3 at stages I and II for C. commensalis and P. comanthi. The embryo volume of stage II eggs is signiﬁcantly higher than those of stage I (W＝10, p＜0.01 and W＝20, p＜0.01 for each species respectively). weight between egg stages I and II for C. com- between female dry weight and carapace length mensalis (W＝117, df＝30, p＝0.706), but (Fig. 1). There was also a positive relationship there is for P. comanthi (W＝32.5, df＝23, p＜ between the number of eggs carried by an indi- 0.05), with stage II bearing females being vidual shrimp and its carapace length (Fig. 2). heavier. A signiﬁcant relationship was found The embryo volume (Table 1, Fig. 3) of C.

Crustacean Research 47 93 SANCIA E. T. VAN DER MEIJ, MAGNUS L. JOHNSON, SAMMY DE GRAVE

Table 2. Reproductive biology values of symbiotic Palaemonidae, only studies with measurements on reproductive output are included (Brood mass volumes for Periclimenes paivai and P. rathbunae were calculated as the mean embryo number x the mean embryo volume in mm3).

Embryo Embryo Brood mass average min–max Shrimp species Host species n CL number volume mm3 volume mm3 RO comment Reference RO (mean±SD) (SD) (SD)* (SD)

Anneissia bennetti; All stage Cristimenes 0.04 0.16 Comaster nobilis 32 1.55–2.60 15–117 2.15±0.53 0.09–0.23 females, but This study commensalis (±0.00) (±0.03) (crinoid) see Table 1

20–517 0.1038 0.0073– stage I Lychnorhiza lucerna 65 0.0237 5.43 Periclimenes (229±120) (±0.056) 0.2905 females de Moraes (scyphozoan 3.42–5.62 paivai 32–540 stage II et al., 2017 jellyﬁsh) 18 0.0307 5.86 n/a n/a (191±115) females

Stichodactyla Azofeifa- Periclimenes 80–605 0.24 only stage I helianthus 70 2.25–5.25 0.038 10.98 0.10–0.50 Solano et al., rathbunae (289±120) (±0.094) females (sea anemone) 2014

All stage Pontoniopsis Anneissia bennetti 0.02 0.31 25 1.00–1.50 12–57 0.59±0.27 0.23–0.40 females, but This study comanthi (crinoid) (±0.01) (±0.05) see Table 1

Fig. 4. Reproductive output vs carapace length. Closed circles represent C. commensalis, open circles represent P. comanthi. There was no significant relationship between carapace length and reproductive output for P. comanthi (p＞0.05) and no difference in reproductive output by stage for C. commensalis (ANCOVA, p＞0.05). commensalis was significantly larger than that df＝30, p＞0.05). of P. comanthi (W＝766.5, df＝55, p＜0.001), The mean reproductive output (RO) of C. with a significant increase from stage I to stage commensalis over the two stages combined had II embryos in P. comanthi (W＝22, d.f＝23, a median value of 17%, with the values per re- p＜0.05) but not for C. commensalis (W＝91, productive event ranging from 9% (stage II

94 Crustacean Research 47 REPRODUCTIVE STRATEGIES OF CRINOID-ASSOCIATED SHRIMPS eggs, CL 1.75 mm) to 23% (stage I eggs, CL duced in both C. commensalis and P. comanthi 2.10 mm) of egg weight relative to female is broadly comparable to other symbiotic weight. The RO of P. comanthi was higher Palaemonidae (Azofeifa-Solano et al., 2014), with a median value of 32% over both stages the maximum number is generally lower. Only combined, with the values per reproductive Periclimenes patae Heard & Spotte, 1991, a event ranging from 23% (stage II eggs, gorgonian commensal, produces fewer eggs CL1.20 mm) to 40% (stage II eggs, CL (10–35 (Heard & Spotte, 1991)), with the ma- 1.15 mm) (Table 2, Fig. 4). jority of species having maximal values of 200 Possible effects of host crinoid species on the or more. A meaningful comparison in this re- fecundity or reproductive output of C. com- gard is, however, hampered by the fact that our mensalis. which occurred on two host species samples originated from a single event, and (n＝20 on A. bennetti, n＝12 on C. nobilis) seasonal variation has been noted in Actin- were tested using a Wilcoxon signed-rank test, imenes ornatus (Bruce, 1969) [(Omori et al., but no significant differences were found be- 1994), as Periclimenes ornatus]. Nevertheless, tween the two host species (p＞0.05). low fecundity in these crinoid-inhabiting species is likely offset by successive spawning, ■ Discussion once sexual maturity is reached, often in relatively quick succession (Bauer, 1991). The crinoids in this study show high levels The embryo volume (mm3) of C. commensa- of inhabitation by the shrimps C. commensalis lis is approximately in the same range as the and P. comanthi, which is in line with earlier other species of symbiotic Palaemonidae for studies (Deheyn et al., 2006). Both species which this is known (Azofeifa-Solano et al., show a lack of embryo loss between egg stages 2014). In contrast, the embryo volume of P. co- (Table 1).The jellyfish symbiont P. paivai had a manthi is the smallest recorded to date (Table mean percentage of embryo loss between egg 2). Although this is clearly linked to the size of stages of 16.6% (de Moraes et al., 2017), and the species, which is the smallest studied so for the sea-anemone symbiont, P. rathbunae far, it can be speculated that combined with the embryo mortality was reported to be 24% lower number of eggs produced that the spe- (Azofeifa-Solano et al., 2014). This species ex- cies would undergo a form of abbreviated lar- hibited similar levels of brood loss to free liv- val development compared to C. commensalis ing species, e.g. the similar sized C. america- and other species. Although in shrimps, abbre- nus (Negri et al., 2017). Notwithstanding the viated larval development is generally linked relatively small sample size in the present to a larger egg size (Strathmann, 1977; Wowor study, we speculate that this difference is et al., 2009), we postulate that the physical size caused by the more enclosed and thus protected limitations of the pleon in P. comanthi do not ecospace in crinoids, enabling higher levels of allow for enlarged eggs. In common with the parental care and potentially less predation, as majority of symbiotic Palaemonidae, no infor- opposed to the more open ecospace in sea mation is available on larval biology, so this anemones. remains somewhat speculative. As is generally known for caridean shrimps The area available for egg attachment physi- (Corey & Reid, 1991), fecundity is closely re- cally limits the brood mass volume (mm3) of a lated to female size (Azofeifa et al., 2014, and species, and is strongly linked to carapace vol- references therein). This trend is herein con- ume (Corey & Reid, 1991). When looking at firmed once again for the two species studied. the average brood mass volume, C. commensa- Although the minimum number of eggs pro- lis has a brood mass volume of 2.15±

Crustacean Research 47 95 SANCIA E. T. VAN DER MEIJ, MAGNUS L. JOHNSON, SAMMY DE GRAVE

0.53 mm3, averaged over both egg stages (stage chiridae and Pinnotheridae) are able to produce I: 1.80±0.68 mm3, stage II: 2.54±1.60 mm3), proportionately large broods by specific mor- compared to 0.59±0.27 mm3 for P. comanthi phological features (extension of ovaries into (stage I: 0.42±0.19 mm3, stage II: 0.76± the abdomen and reduced calcification of their 0.24 mm3). These numbers are much lower exoskeleton) (Vehof et al., 2016). Their RO than the estimated brood mass volumes of P. values are even higher than those of the symbi- paivai and P. rathbunae. In general, ovigerous otic shrimp, with investment values of 59% for females of Palaemonidae display a linear rela- the cryptochirids and 66% and 97% for the tionship between carapace length and the size pinnotherids. The pea crabs (Zaops ostreus of the second abdominal pleura (Baeza et (Say, 1817) [as Pinnotheres ostreum] and Fa- al.,2013, 2015), the main structural component bia subquadrata Dana, 1851) have short repro- of the brood pouch. It follows that larger spe- ductive seasons bringing their annual invest- cies, such as the two Periclimenes have a larger ment up to 99% and 146%, respectively (Hines, brood pouch than the herein studied species, 1992). The tropical cryptochirid crab species which are smaller in body size. Logically this Hapalocarcinus marsupialis Stimpson, 1859 allows for fewer eggs (see above) and thus a has an estimated annual investment of 472%, smaller total brood mass volume. based on a conservative estimate of 8 broods Reproductive Output (RO) is a widely used per year (Kotb & Hartnoll, 2002). A study on value to compare inter- and intraspecific varia- the temperate sea anemone-associated species tion in the fraction of the total energy budget A. ornatus indicates year-round breeding, al- diverted to reproduction, allowing for a cross though with a drop in the number of ovigerous comparison of the reproductive strategies be- females in the colder months. In the warmest tween species, although this value has only month the mean interval of two successive egg been reported for two species of symbiotic layings was 10 days (Omori et al., 1994). In Palaemonidae studies so far (Table 2; Azofeifa- tropical species year-round breeding has been Solano et al., 2014; de Moraes et al., 2017). reported for symbiotic Palaemonidae (Johnson These studies and the values reported herein & Liang, 1966; Bruce, 2000), with embryo in- show that the RO values of symbiont species cubation periods lasting 5–10 days in several are higher than those of free-living species. sea grass inhabiting caridean shrimp (Bauer, Commensal caridean shrimp species are thus 1991). Such short embryo incubation periods seemingly able to allocate more energy in in tropical species may be a life history adapta- brood production than free-living species, po- tion for increasing the number of broods tentially linked to shorter larval duration. De- spawned per unit time, or a metabolic response spite being the smallest species, the RO of P. to elevated water temperatures (Bauer, 1991). comanthi is the highest recorded so far for The annual investment of the tropical species symbiotic palaemonids (Table 2). A large RO C. commensalis and P. comanthi could there- may be adaptive to a commensal lifestyle, as fore be as high as 384% and 744%, respective- larvae may suffer greater mortality during their ly, based on an estimated 24 broods per year search for a specific settlement site (Hines, (estimated embryo incubation of 10 days, plus 1992). Moreover, low survival rates have been ca. 5 days for moulting and spawning of new reported for juvenile symbiotic palaemonids eggs (derived from Bauer, (1991) and Omori et (Omori et al., 1994). An increased RO will al., (1994)). These annual investment rates, in- help to compensate for larval and juvenile cluding those of the cryptochirid and pinnoth- mortality. erids, do not take female maturation time and The smallest of symbiotic crabs (Crypto- longevity into account and true values are

96 Crustacean Research 47 REPRODUCTIVE STRATEGIES OF CRINOID-ASSOCIATED SHRIMPS therefore likely somewhat lower. ■ Literature Cited A study on Periclimenes yucatanicus (Ives, 1891) collected from two morphologically dif- Azofeifa-Solano, J. C., Elizondo-Coto, M., & ferent sea anemone hosts (Stichodactyla heli- Wehrtmann, I., 2014. Reproductive biology anthus (Ellis, 1768) and Bartholomea annulata of the sea anemone shrimp Periclimenes (Le Sueur, 1817)) found that larger sized hosts rathbunae (Caridea, Palaemonidae, Ponto- were capable of supporting larger and more fe- niinae), from the Caribbean coast of Costa cund shrimp irrespective of the host species Rica. ZooKeys, 457: 211–225. (McKeon & O’Donnell, 2015). The authors Baeza, J. A., Ritson-Williams, R., & Fuentes, M. suggested that host preference might be deter- S., 2013. Sexual and mating system in a car- mined by host size, via its effect on fecundity idean shrimp symbiotic with the winged in P. yucatanicus. In this study, C. commensalis pearl oyster in the Coral Triangle. Journal of Zoology, 289: 172–181. was collected from two morphologically differ- Baeza, J. Antonio, Hemphill, C. A., & Ritson- ent crinoid species; A. bennetti is generally Williams, R., 2015. The sexual and mating larger in size and has more arms than C. nobil- system of the shrimp Odontonia katoi is. We found, however, no significant effects of (Palaemonidae, Pontoniinae), a symbiotic host species on fecundity or reproductive out- guest of the ascidian Polycarpa aurata in put of C commensalis. the Coral Triangle. PloS ONE, 10: e0121120. Comparison of the reproductive strategies of Bauer, R. T., 1991. Analysis of embryo produc- C. commensalis and P. comanthi reveals a dif- tion in a caridean shrimp guild from a tropi- ferent reproductive approach for each species. cal seagrass meadow. In: A. Wenner & A. Cristimenes commensalis specimens are larger Kuris, (eds.), Crustacean Egg Production, in size and produce a higher number of eggs, AA Balkema, Rotterdam, pp. 181–191. yet the RO of P. comanthi is higher than that of Britayev, T. A., & Mekhova, E. S., 2011. Assess- C. commensalis. Our results indicate there are ment of hidden diversity of crinoids and different reproductive strategies between sym- their symbionts in the Bay of Nhatrang, biotic species inhabiting the same host species, Vietnam. Organisms Diversity & Evolution, linked to differential evolutionary pressures 11: 275–285. despite equal environmental forces. In the Bruce, A. J., 2000. Biological observations on the present case, this is clearly linked to the size commensal shrimp Paranchistus armatus (H. difference between both symbionts. Milne Edwards) (Crustacea: Decapoda: Pon- toniinae). Beagle: Records of the Museums ■ Acknowledgements and Art Galleries of the Northern Territory, 16: 91–96. Funding for the fieldwork was provided by Corey, S., & Reid, D. M., 1991. Comparative fe- Oxford University’s John Fell Fund and hosted cundity of decapod crustaceans I. the fecun- by the Iriomote Station of the Tropical Bio- dity of thirty-three species of nine families of caridean shrimp. Crustaceana, 60: 270– sphere Research Center (University of the 294. Ryukyus), facilitated by T. Naruse. Crinoid De Grave, S. 2002. Biogeography of Indo-Pacific species were identified by Y. Fujita for which Pontoniinae (Crustacea, Decapoda): a PAE we are grateful. We thank Fabrício Carvalho analysis: Indo-Pacific pontoniid biogeogra- and an anonymous reviewer for their construc- phy. Journal of Biogeography, 28: 1239– tive comments. 1253. de Moraes, I. R. R., Wolf, M. R., Gonçalves, G.

Crustacean Research 47 97 SANCIA E. T. VAN DER MEIJ, MAGNUS L. JOHNSON, SAMMY DE GRAVE

R. L., & Castilho, A. L., 2017. Fecundity the rule. American Zoologist, 32: 503–511. and reproductive output of the caridean Horká, I., Grave, S. D., Fransen, C. H. J. M., Pe- shrimp Periclimenes paivai associated with trusek, A., & Ďuriš, Z., 2016. Multiple host scyphozoan jellyfish. Invertebrate Reproduc- switching events shape the evolution of tion & Development, 61: 71–77. symbiotic palaemonid shrimps (Crustacea: Deheyn, D. D., Lyskin, S., & Eeckhaut, I., 2006. Decapoda). Scientific Reports, 6: 26486. Assemblages of symbionts in tropical shal- Huang, H. D., Rittschof, D., & MingShiou, J., low-water crinoids and assessment of sym- 2005. Multispecies associations of macro- bionts’ host-specificity. Symbiosis, 42: 161– symbionts on the comatulid crinoid Coman- 168. thina schlegeli (Carpenter) in southern Tai- Ďuriš, Z., Horká, I., Juračka, P. J., Petrusek, A., wan. Symbiosis, 39: 47–51. Sandford, F., 2011. These squatters are not Johnson, D. S., & Liang, M., 1966. On the biolo- innocent: the evidence of parasitism in gy of the watchman prawn, Anchistus custos sponge-inhabiting shrimps. PloS One 6: (Crustacea; Decapoda; Palaemonidae), an e21987. Indo-West Pacific commensal of the bivalve Ennos, R,, & Johnson, M. L., 2018. Statistical Pinna. Journal of Zoology, 150: 433–455. and Data Handling Skills in Biology. Pear- Kim, S., & Hong, S., 2004. Reproductive biology son, 258 pp of Palaemon gravieri (Decapoda: Caridea: Fabricius, K. E., & Dale, M. B., 1993. Multispe- Palaemonidae). Journal of Crustacean Biol- cies associations of symbionts on shallow ogy, 24: 121–130. water crinoids of the central Great Barrier Kotb, M. M. A., & Hartnoll, R. G., 2002. Aspects Reef. Coenoses, 8: 41–52. of the growth and reproduction of the coral Feldman, H. R., & Brett, C. E., 1998. Epi- and gall crab Hapalocarcinus marsupialis. Jour- endobiontic organisms on Late Jurassic cri- nal of Crustacean Biology, 22: 558–566. noid columns from the Negev Desert, Israel: Levitt-Barmats, Y., Shenkar, N., 2018. Observa- Implications for co-evolution. Lethaia, 31: tions on the symbiotic relationship between 57–71. the caridean shrimp Odontonia sibogae Fishelson, L., 1974. Ecology of the northern Red (Bruce, 1972) and its ascidian host Herdma- Sea crinoids and their Epi- and Endozoic nia momus (Savigny, 1816). PloS ONE, 13: fauna. Marine Biology, 26: 183–192. e0192045. Guo, C. C., Hwang, J. S., & Fautin, D. G., 1996. Mangiafico, S., 2018. Package ‘rcompanion’, ver- Host selection by shrimps symbiotic with sion 1.13.2. CRAN Repository, 1–93. http:// sea anemones: A field survey and experi- rcompanion.org/ mental laboratory analysis. Journal of Ex- Martínez-Mayén, M., & Román-Contreras, R., perimental Marine Biology and Ecology, 2009. Reproduction of Periclimenes sian- 202: 165–176. kaanensis (Decapoda, Caridea, Palaemoni- Heard, R. W., & Spotte, S., 1991. Pontoniine dae) in Bahía De La Ascensión, Quintana shrimps (Decapoda: Caridea: Palaemonidae) Roo, Mexico. Crustaceana, 82: 27–37. of the northwest Atlantic. II. Periclimenes McKeon, C. S., & O’Donnell, J. L., 2015. Varia- patae, new species, a gorgonian associate tion in partner benefits in a shrimp̶sea from shallow reef areas off the Turks and anemone symbiosis. PeerJ, 3: e1409. Caicos Islands and Florida Keys. Proceed- Negri, M., Magalhães, T., Rossi, N., Felder, D. L., ings of the Biological Society of Washing- & Mantelatto, F. L., 2017. Reproductive as- ton, 104: 40–48. pects of the shrimp Cuapetes americanus Hines, A. H., 1992. Constraint on reproductive (Kingsley, 1878) (Caridea, Palaemonidae) output in brachyuran crabs: pinnotherids test from Bocas del Toro, Panama. Crustaceana,

98 Crustacean Research 47 REPRODUCTIVE STRATEGIES OF CRINOID-ASSOCIATED SHRIMPS

90: 1061–1078. 0.5–1. https://CRAN.R-project.org/package= Omori, K., Yanagisawa, Y., & Hori, N., 1994. fANCOVA Life history of the caridean shrimp Peri- Wowor, D., Muthu, V., Meier, R., Balke, M., Cai, climenes ornatus Bruce associated with a Y., & Ng, P. K. L., 2009. Evolution of life sea anemone in southwest Japan. Journal of history traits in Asian freshwater prawns of Crustacean Biology, 14: 132–145. the genus Macrobrachium (Crustacea: De- Potts, F. A., 1915. The fauna associated with the capoda: Palaemonidae) based on multilocus crinoids of a tropical coral reef: with espe- molecular phylogenetic analysis. Molecular cial reference to its colour variations. Carne- Phylogenetics and Evolution, 52: 340–350. gie Institution of Washington Publication, 8: Zimmermann, U., Carvalho, F. L., & Mantelatto, 71–96. F. L., 2015. The reproductive performance R Core Team., 2018. R: A language and environ- of the red-algae shrimp Leander paulensis ment for statistical computing. R Foundation (Ortmann, 1897) (Decapoda, Palaemonidae) for Statistic Computing, Vienna, Austria. and the effect of post-spawning female https://www.R-project.org/ weight gain on weight-dependent parame- Soledade, G. O., Santos, P. S., Pescinelli, R. A., ters. Brazilian Journal of Oceanography, 63: Costa, R. C., & Almeida, A. O., 2017. Em- 207–216. bryo production of two sympatric snapping Zmarzly, D. L., 1984. Distribution and ecology of shrimps of the genus Alpheus (Caridea: Al- shallow-water crinoids at Enewetak Atoll, pheidae) from Northeastern Brazil: a com- Marshall Islands, with an annotated check- parative approach. Thalassas: An Interna- list of their symbionts. Paciﬁc Science, 38: tional Journal of Marine Sciences, 33: 105–122. 151–158. Strathmann, R., 1977. Egg size, larval development, and juvenile size in benthic marine in- Addresses vertebrates. The American Naturalist, 111: (SETVDM) Groningen Institute for Evolutionary 373–376. Life Sciences, University of Groningen, PO Box Summers, M. M., Messing, C. G., & Rouse, G. 11103, 9700 CC Groningen, the Netherlands; W., 2014. Phylogeny of Comatulidae (Echi- Naturalis Biodiversity Center, PO Box 9517, nodermata: Crinoidea: Comatulida): A new 2300 RA Leiden, the Netherlands classiﬁcation and an assessment of morpho- (MLJ) School of Environmental Sciences, logical characters for crinoid taxonomy. Mo- Faculty of Science and Engineering, University lecular Phylogenetics and Evolution, 80: of Hull, HU6 7RX, United Kingdom 319–339. (SDG) Oxford University Museum of Natural Terossi, M., Wehrtmann, I. S., & Mantelatto, F. History, University of Oxford, Parks Road, L., 2010. Interpopulation comparison of re- Oxford OX1 3PW, United Kingdom production of the Atlantic shrimp Hippolyte obliquimanus (Caridea: Hippolytidae). Jour- nal of Crustacean Biology, 30: 571–579. E-mail addresses Vehof, J., van der Meij, S. E. T, Türkay, M., & (SETVDM)* [email protected] Becker, C., 2016. Female reproductive mor- phology of coral-inhabiting gall crabs (Crus- (MLJ) [email protected] tacea: Decapoda: Brachyura: Cryptochiri- (SDG) [email protected] * dae). Acta Zoologica, 97: 117–126. Corresponding author Wang XF., 2010. fANCOVA: Nonparametric Analysis of Covariance. R package version

Crustacean Research 47 99