Mar Biol (2007) 151:1875-1886 DOI 10.1007/s00227-007-0618-z RESEARCH ARTICLE Gonopore sexing technique allows determination of sex ratios and helper composition in eusocial shrimps Eva Toth ? Raymond T. Bauer Received: 25 May 2006 / Accepted: 12 January 2007 / Published online: 8 February 2007 ? Springer-Verlag 2007 Abstract An evaluation of the social organization and sexual system of eusocial species of Synalpheus has been hindered because it has not been possible to determine the sexual composition of colony helpers (workers). The external sexual characters typically used to sex caridean shrimps are lacking in Synalpheus. We used SEM sexing technique to determine the sex- ual composition of helpers in colonies of Synalpheus regalis, S. rathbunae, S. chacei, S. rathbunae A (see Morrison et al. Mol Phylogen Evol 30:563-568, 2004), and S. filidigitus. Colonies consisted of both sexes and sex ratios of helpers generally conformed to 50:50 female to male. Females were characterized by gonop- ores with U-shaped slits on the coxae of the third pereopods (first walking legs) while males had oval gonopore openings on the coxae of the fifth pereopods (third or last walking legs). In S. chacei, S. filidigitus, and S. rathbunae A, a few helpers were found that had both male and female gonopores (intersexes). All three reproductive females (queens) of S. filidigitus examined were intersexes. Sexing of helpers allowed us to test some hypotheses about sexual differences in helper morphology that might indicate task specializa- tion (division of labor). Male helpers were not different from female ones in body size (except in S. regalis: Communicated by P.W. Sammarco. E. Toth (El) Smithsonian Tropical Research Institute, Unit 0948, APO AA, 34002-0948, USA e-mail: tothev@si.edu R. T. Bauer Department of Biology, University of Louisiana at Lafayette, Lafayette, LA, 70504-2451, USA males somewhat larger) and in fighting chela size. The lack of sexual dimorphism in these characters suggests no male-female specialization in colony tasks such as defense. The presence of male and female helpers sim- ilar in size suggests that the sexual system of these eusocial species is gonochoristic, although protandry of some sort in S. filidigitus can not be ruled out. The intersexuality observed in a few individuals may be due to developmental anomalies, protandry, or even simul- taneous hermaphroditism. Finally, the sexing tech- nique allowed us to establish that new colonizers of unoccupied sponges in S. rathbunae are a single male and female of helper size. Introduction Eusocial animals live in family groups with overlapping generations in which most colony members help to rear the progeny of one or a few queens (e.g. Wilson 1971; Michener 1974; Lacey and Sherman 1991). This repro- ductive skew is achieved because as a rule most non- reproductive colony members (called workers or help- ers) in eusocial colonies do not have functional gonads (e.g. Wilson 1971; Ratnieks 1988). Suppression of gonadal development may occur either by self restraint in which an individual chooses not to develop sexually to avoid fights with others, or through policing by fel- low helpers or the queen (e.g. Reeve and Sherman 1991; Faulkes and Bennett 2001; Hart and Ratnieks 2005). Instead of reproducing, helpers in social animal societies specialize in other tasks, such as nest defense, nest construction, and brood care, while reproductives (queens and kings) usually concentrate solely on repro- duction. Even though they refrain (at least temporarily) 4y Springer 1876 Mar Biol (2007) 151:1875-1886 from direct reproduction in many species, non-breeders in eusocial societies obtain indirect fitness benefits by helping to rear relatives (Hamilton 1964). Eusociality has evolved in several Hymenoptera (all ants, some bees and wasps), termites, some thrips, some aphids, one beetle, two species of mole rats and some snapping shrimps (Wilson 1971; Crespi 1992; Kent and Simpson 1992; Stern and Foster 1996; Jarvis and Bennett 1993; Duffy 1996a, b, 2003). In the Hyme- noptera, with haplo-diploid sex determination, in which males develop from unfertilized eggs and females from fertilized eggs (Wilson 1971), only females work in the colonies. Males are less related to their siblings than are females, and thus their incentive to help with colony activities is lower (e.g. Queller and Strassmann 1998). In termites and mole rats, however, with diplo-diploid sex determination (Wilson 1971; Alexander et al. 1991), both sexes are part of the work force. All well-studied eusocial societies are terrestrial. Snapping shrimps of the genus Synalpheus (Decapoda: Alpheidae) are the only known group of marine inver- tebrates with eusocial species. However, the sex of helpers has not previously been determined. Eusocial shrimps, like termites and mole rats, are diplo-diploid animals. Maximum colony sizes vary according to the species. S. filidigitus and S. chacei have small colonies with up to 120 individuals, while colonies of S. regalis and S. rathbunae may have more than 300 helpers (Duffy 2003; ET pers. obs). Colonies are headed by a single or a few queens, according to the species (Duffy 2003), and, as Duffy (1996a) showed for two species, queens mate with only one male (king). Colonies con- sist of animals with continuous range of sizes, and indi- viduals vary in body length from a few mm (juveniles) to about one cm (developed queens). Although euso- cial shrimps are small, like all snapping shrimps they are territorial and will defend their sponge from intrud- ers (e.g. Toth and Duffy 2005). They fight with their formidable weapon, the snapping claw (major or larger chela), which is also used in communication (Nolan and Salmon 1970; Duffy et al. 2002). This chela houses massive muscles that enable the fingers to close quickly and powerfully producing a forward-directed, high- velocity jet of water (Ritzmann 1974; Versluis et al. 2000). In crustaceans both closing force and speed increase with chela size (Taylor 2001; Levinton and Allen 2005), suggesting that the larger the claw, the better the fighting ability of the animal. Because euso- cial shrimps are "fortress defenders" (Queller and Strassmann 1998), in which young are independent and food does not need to be collectively gathered nor processed (Duffy 2003), the main colony task among colony members focuses on sponge host defense (fight- ing). As Duffy (1996a, b) stated, these shrimps qualify as eusocial because larger colony members take indi- rectly care of weaker, younger ones by defending the sponge from intruders and thus help them to survive until fully grown. To understand better these under- studied societies, knowledge of the sex and sex ratios of helpers, as well as sexual dimorphism in chela size, would be desirable. Consistent and accurate identification of the sexes, and thus the sexual composition of helpers, is not possible in Synalpheus using standard techniques of examination of abdominal appendages with light microscopy. In members of the same family, e.g. Alpheus species, males can be distinguished from females by the appendices masculinae on the second pleopods. These appendices have never been observed in Synalpheus spp. (Banner and Banner 1975; Felder 1982; Dardeau 1984). Other possible characteristics for sexing Synalpheus specimens, such as the shape of abdominal pleurae and the relative position of other pleopodal appendices, have been suggested (e.g., Ban- ner and Banner 1975) but have not proved reliable. Although some species of Synalpheus are eusocial, most sponge dwelling species live in heterosexual pairs (Duffy 2003). Mature reproductive females of such pairs are characterized by features such as developing ovaries with eggs (visible through the carapace), the modification of their pleopods and pleural plates to form an open abdominal brood pouch, the presence of embryos carried in that space, and by their larger body size. Identification of males in these pairs has usually been done by default: non-ovigerous individuals of adult size are usually considered males (e.g. Banner and Banner 1975; Dardeau 1984). This assumption may be valid when the shrimps live in pairs but not in euso- cial shrimps in which several to many non-breeding individuals live with a few or a single breeding female (queen). Ovarian development, incubated embryos, and modified pleopods are features found in breeding females of Synalpheus, and thus an absence of these characters might occur both in males and in non-repro- ductive females. In eusocial shrimps, as in the pair-forming species, the reproductive females (queens) based on the above mentioned characters are easily recognized. Other col- ony members have generally been assumed to be males and subadults (Chace 1972; Banner and Banner 1981a, b; Duffy 1996b, 1998, 2003; Duffy and Macdonald 1999; Duffy et al. 2002; Toth and Duffy 2005; Didderen et al. 2006), resulting in a 100% male sex ratio among help- ers. After eusociality in these shrimp was recognized, the excess number of males was suggested to occur 4y Springer Mar Biol (2007) 151:1875-1886 1877 because of a protandric sexual system in which the queen was derived by sex change from one of these males. The sex change would be environmentally (socially) mediated, which has been proposed in some of the approximately 40 species of protandric decapods (Bauer 2000), e.g. the alpheid genus Athanas (Suzuki 1970; Nakashima 1987; Gherardi and Calloni 1993) and Pandalus spp. (Pandalidae) (Charnov 1982; BregstrOm 2000), as well as in the protandric simultaneous her- maphroditic genus Lysmata (Baeza and Bauer 2004). In protandric species, the population may consist of some mixture of individuals that do not change sex (primary males or females) and protandric sex chang- ers (Bauer 2000). In eusocial Synalpheus spp., helpers might thus also be a mixture of males, sex changers and females with undeveloped gonads. Repressed ovarian development, however, is a very common helper fea- ture in most eusocial taxa (e.g. Hart and Ratnieks 2005). Assuming repressed ovarian development but no sex change, helpers in eusocial shrimp colonies could also consist of primary males and females in a 50:50 ratio. As noted above, identification of males in Synal- pheus by the external morphological characters nor- mally used for sexing shrimps is not reliable (Felder 1982; Dardeau 1984). The presence or absence of male gonadal characters, such as ejaculatory ducts with sperm (e.g. Bauer 1986), might be used to distinguish male and female individuals. However, because the ejaculatory ducts of Synalpheus males are not as large or well developed as those of other male caridean shrimps (see "Results"), routine observation of ejacu- latory ducts by dissection is fairly difficult (pers. obs.), especially in the eusocial species, in which helpers are quite small in body size. The position of gonopores on the third (female) or fifth (male) pereopods are a reli- able indicator of sex in caridean shrimps and all mala- costracan crustaceans (Bauer 2004). Male shrimps might be distinguished from undeveloped females by the presence of gonopores, through which sperms are ejaculated, on the coxae of their fifth pair of legs (pereopods). In female shrimps, eggs are spawned through gonopores located on the third pair of legs. In Synalpheus spp., however, male and female gonopores are very difficult to observe with light microscopy using dissecting microscopes (Dardeau 1984), the traditional and most convenient method for examination of exter- nal morphology in crustaceans. These gonopores, on the other hand, are readily apparent and observable using the scanning electron microscopy (SEM), as reported herein. Accurate sexing of colony members of known body size would allow evaluation of hypotheses about size, age, and sex-related polyethism ("division of labor") in eusocial colonies of Synalpheus, as well as the sexual system (gonochory, sequential or simulta- neous hermaphroditism) of these species. In this study, we tested the null hypothesis that help- ers are undifferentiated males (male-phase hermaphro- dites and/or primary males) in five eusocial species of Synalpheus, as well as the alternative hypothesis that the helper population also includes non-reproductive females. We tested these hypotheses by sexing helpers and queens of measured size using scanning electron observations on the location of gonopores (third pereopods, female; fifth pereopods, male). Moreover, we were able to determine the sex of dispersing indi- viduals colonizing unoccupied sponges and thus to test hypotheses about sexual composition in colony founders. In addition, accurate sexing of individuals also allowed us to compare possible differences in defensive roles between the sexes, as measured by the relative size of cheliped weapons used in agonistic interactions. Materials and methods Taxon sampling We collected in 2005 four eusocial species in Belize close to the Carrie Bow Cay marine station of the Smithsonian Institution (N16?46.46', W88?04.46'), Syn- alpheus "rathbunae A", S. filidigitus, S. regalis, and S. chacei, and another, S. rathbunae, in Panama at Bocas del Toro marine station (N09?21.10\ W82?15.54'). We collected whole sponges either by SCUBA or snorkel- ing. Each sponge was put in a very fine-meshed bag and closed immediately. At the surface sponges were broken into small pieces and all animals collected from them. Colonies were preserved in formalin or ethanol. Synalpheus filidigitus, S. regalis, S. rathbunae and S. "rathbunae A" are in one clade with a single origin of eusociality, whereas S. chacei has a separate, indepen- dent origin of eusociality (Morrison et al. 2004). Initially we sexed (using SEM, see later in this sec- tion) 7 and 80 helpers of various sizes (randomly selected) from two colonies, respectively, of S. rathbu- nae, along with six queens (of which four came from different colonies) with results clearly showing separate sexes. Based on these results, we decided to subsample and thus to select, measure, and sex 15 large, 15 mid- sized, and 15 small individuals all coming from the same colony of each species, or more when we needed clarification of the observed sex ratio. Because eusocial colonies have often only one sole queen and because we wanted to analyze three queens per species, we 4y Springer 1878 Mar Biol (2007) 151:1875-1886 added two extra queens originating from different colo- nies for S. rathbunae, S. "rathbunae A", and S. regalis. The S. filidigitus colony analyzed had one large and two smaller queens, and we examined those. For the examination of additional queens of the other species originating from different colonies we chose when pos- sible the smaller-sized individuals, reasoning that if there is sex change, the smaller queens might be still showing male gonopores. We thus sampled 45 helpers originating from one colony and three queens from S. filidigitus, S. rathbunae, and S. "rathbunae A". From S. regalis we sampled f 5 large, 25 midsized and 25 small non-reproductive individuals. For S. chacei, with a different origin of eusociality, which species is atypical in that there are often helpers with very slight ovarian development, we sampled two colonies. Because there is usually only one, or a few females with fully devel- oped ovaries and embryos in S. chacei colonies, we assumed that individuals with slight ovarian develop- ment (but otherwise with characters of helpers) either leave their colonies to start new colonies themselves, or remain in their colonies without further gonadal development. In these two S. chacei colonies we mea- sured all individuals (13 and 9, respectively) that showed some ovarian development as well as, from each colony, 10 large individuals without any indica- tions of ovarian development, 10 mid-sized and 10 small individuals, plus the single queen from each col- ony. Additionally, we included another queen from a different colony (Fig. 4b smaller queen). We also determined the sex of 34 dispersing individ- uals of S. rathbunae that colonized 17 unoccupied experimental sponges (Toth, submitted). We expelled all shrimp from small pieces of sponges by dipping them into fresh water and relocated them into the sea for 45 days. After this period we collected these experi- mental sponges by SCUBA, put the sponges individu- ally in a very fine-meshed bag and closed the bags. In the laboratory the sponges were destroyed and all colo- nizers collected. In some of these colonizers ovarian development was observable, but none of these indi- viduals were developed enough to carry embryos. Size measurements Most morphometric data were obtained using a dis- secting microscope, digital camera, and the program Image Pro Plus 4.1 for windows (Media Cybernetics L.P. 1999). Magnified images of animals were projected onto the computer screen, allowing us to draw straight lines between two points. For S. filidigitus queens, which change their large chela for a small one upon obtaining queen status (Duffy and Macdonald 1999), the largest of the two small chelae was measured. A number of specimens of S. chacei (30) and S. regalis (20) were measured with a dissecting microscope and an optical micrometer in the laboratory of the second author. We measured chela length and carapace length for each shrimp. We separated the major (fighting) chela of each individual shrimp and placed the shrimp on its left side for measurements. We recorded the com- monly used metric of body size in crustaceans, cara- pace length, as a straight line from the tip of rostrum to the posterior margin of the carapace. Chela length was measured as a straight line from the tip of the fixed (immovable) finger to the proximal edge of the chela. In individuals that were missing the fighting chela, we measured only the carapace length and plotted chela length as 0 (Fig. 4). Sexing of individuals For SEM, specimens of S. chacei that were initially pre- served in seawater formalin and later changed to 70% ethanol (ETOH) for storage, or 70-95% ETOH and were later taken through the solution series: 95% ETOH to 100% ETOH, then 50:50 ETOH: hexamethyldisilaz- ane (HMDS), and finally 100% HMDS, with at least 1 h in each solution. Other specimens preserved in 95% ETOH were taken through the solution series: 95% ETOH to 100% ETOH, then 50:50 ETOH: HMDS, and finally 100% HMDS, specimens were drained of HMDS and air dried. Several dried specimens (up to 26, depend- ing on size) were attached with adhesive to round stubs of 25 mm in diameter for sputter-coating with 40-80 nm gold prior to observation with a Hitachi S-3000-N scan- ning electron microscope. The presence or absence of gonopores on the coxae of the third and fifth pereopods was recorded for each specimen. Reproductive females in colonies (queens) were identified as individuals carry- ing embryos and/or with many eggs clearly visible from dorsal view through the translucent exoskeleton of the cephalothorax and abdomen. Data analysis In most individuals, we observed only male or female gonopores, but both types of gonopores were observed in some individuals ("intersexes"). Data from these lat- ter were grouped with that of males or females or treated apart in different tests. We used X2 to test the null hypotheses of no difference in observed versus expected sex ratios. Expected sex ratios proposed in different species were 50:50 females: males. We used log transformed data and used f-tests and ANCOVA to 4y Springer Mar Biol (2007) 151:1875-1886 1879 Fig. 1 Male and female go- nopores in Synalpheus spp. a Ventral view of posterior tho- rax and pereopods of a male S. rathbunae helper, showing go- nopores {arrow) on the coxae of pereopods 5 and their ab- sence from the coxae of the third pereopods; b higher magnification of male P5 cox- ae and gonopores (arrow) from (a); c male gonopore; d ventral view of posterior tho- rax of a female S. rathbunae A helper, showing gonopore slits (arrows) on P3 coxae and an absence of gonopores on P5 coxae; e higher magnification of female gonopores in (d) showing gonopore slits (white arrows), presumed egg guid- ing setae (black arrows marked "g"), and setae on the distal spur of coxae (un- marked black arrows); f fe- male gonopore from the P3 coxa of a S. chacei female helper. 3, coxa of pereopod 3; 5, coxa of pereopod 5; g, egg- guiding setae test the null hypothesis of no differences in carapace and chela length between male and female helpers. Results Sexing Helpers and queens of all Synalpheus species examined by SEM had gonopores typical of decapod crustaceans in appearance and position. Males were identified by gonopores on the basal segments (coxae) of the fifth pereopods (last pair of walking legs) and females by gonopores on the coxae of the third pereopods (=first pair of walking legs). Well developed male gonopores appear as protuberances with a conspicuous opening on the medial sides of the peropod 5 coxae (Fig. la-c). Female gonopores are broadly U-shaped slits on the proximomedial sides of the third pereopod coxae (Fig. ld-f). Just distal to the female gonopore are sev- eral posteriorly directed setae (Fig. ld-f) unique to both helper and reproductive females. This somewhat fan-shaped row of presumptive egg-guiding setae is not to be confused with other setae, which may be found on the coxae of both males and females, especially a line of setae along a distal medially pointing spur (Fig. le, f), which may also include short stout spines. All reproductive females (queens) examined had female gonopores. Observation of ejaculatory ducts by dissection, although not consistent enough for reliable sexing, did reveal information about these structures (Fig. 2). This structure in Synalpheus spp. widens from a narrow juncture with the distal vas deferens into a thin sac that extends into the coxa of the last walking leg (Fig. 2a-c). Sperm cells appear as discoid refractile structures within the ejaculatory duct (Fig. 2c). Sexual composition of eusocial colonies In all species, the helper population had a mixed sexual composition. We observed helpers that had 4y Springer 1880 Mar Biol (2007) 151:1875-1886 Fig. 2 Synalpheus spp. ejacu- latory ducts, a Entire pereo- pod 5 of S. regalis with ejaculatory duct (arrow) at- tached to limb coxa; b higher magnification of (a) showing proximal articles of pereopod 5 with ejaculatory duct (ar- row) attached just inside limb coxa; c ejaculatory duct of S. rathbunae filled with sperm cells; arrow shows broken dis- tal end of vas deferens. c coxa, m muscle tissue, s mass of sperm cells gonopores either on the fifth pair of pereopods (males) or the third pair of pereopods (females). Additionally, there were a small number of individuals that had both sets of gonopores (termed here "intersexes"). We observed no gonopores in a few of the smallest individ- uals, which we term "juveniles." We found such indi- viduals of small size in S. chacei, (one in colony 1, and two in colony 2), S. "rathbunae A" (1), S. filidigitus (6) and S. rathbuanae (2) without gonopores. In S. rathbu- nae and S. regalis, and S. chacei colony 1, helpers were males, females, and juveniles, whereas one helper in S. chacei colony 2, four in S." rathbunae A", and nine individuals in S. filidigitus appeared to be intersexes (Fig. 3a, b). The male gonopores of intersexes (Fig. 3c) appeared relatively smaller and rudimentary com- pared to male gonopores of purely male individuals, or only one of the male gonopores was present while the other was not (Fig. 3a). The reproductive females (queens) in all species had only female gonopores except in S. filidigitus, in which all three queens exam- ined, also had male gonopores (Fig. 3d), i.e., were intersexes. The 34 individuals of S. rathbunae colonizing unoc- cupied sponges consisted of 17 male-female pairs. The females of such pairs had female gonopores and egg- guiding setae like those of other female helpers in this species (Fig. 3e, f). Sex ratios of helpers were calculated for all species. Exclusion of intersex individuals from the calculations results in a significant deviation from a 50:50 sex ratio Table 1 Results of X2 (1 df) for five social Synalpheus species testing the null hypothesis of a 50:50 female-male sex ratio in colony helpers (juveniles are excluded from the analysis) Species Total A* value P S. chacei colony 1 39 0.421 0.516 S. chacei colony 2 40 4.333 0.037* S. filidigitus 30 6.533 0.011* S. rathbunae A 40 1.600 0.206 S. rathbunae 43 0.023 0.879 S. regalis 65 3.461 0.063 Represents significant P values in colony 2 in S. chacei, resulting in a female biased sex ratio, and in S. filidigitus, providing a male biased sex ratio (Table 1). Body and fighting-chela size of helpers Body size (carapace length) did not differ significantly between males and females in S. chacei col 2, S. "rath- bunae A", S. filidigitus, and S. rathbunae (Fig. 4; Table 2). Among S. chacei colony 1, helper females were slightly larger than males, and among S. regalis helpers, males were larger than females (Table 2). Intersex individuals were scattered among males and females in S. "rathbunae A" and S. filidigitus (Fig. 4). The size of the major (fighting) chela did not differ among the sexual types of helpers (males, females) in any of the species examined (Fig. 4; Tables 3,4). 4y Springer Mar Biol (2007) 151:1875-1886 1881 Fig. 3 Gonopores of intersex- es and colonizing individuals in Synalpheus spp. a S. rathbu- nae A helper, b S. filidigitus helper, c rudimentary male gonopores of S. filidigitus helper, d male pores of S. fili- digitus queen, e S. rathbunae colonizing female, f S. rathbu- nae helper from established colony, for comparison with (e). Black arrows indicate male gonopores on pereopod 5 coxae, white arrows show fe- male gonopores on pereopod 3 coxae Discussion Our results show that SEM observation of gonopores is a reliable method, the only one to date, for sexing many helpers of social Synalpheus that can be done rel- atively quickly, and likely will prove essential for sex- ing males and non-reproductive females in other Synalpheus species. The small size of helpers in euso- cial species makes this practical, i.e., several to numer- ous specimens can be mounted and examined on a single SEM observation stub. In all five species studied in this report, females show a fan-shaped row of setae just below (distal to) the female gonopores. Hoglund (1943) hypothesized that in another caridean shrimp, Palaemon squilla (Palaemonidae), these "egg-guid- ing" setae (his "Group I" setae) direct the eggs posteri- orly towards the abdomen, under which they will be attached during spawning. Because these setae show up consistently, they might be useful for sexing individ- uals of Synalpheus spp. of larger body size using light microscopy to save money and time or when sex deter- mination of living animals is needed. Presence (female) or absence (male) of these egg-guiding setae, which are present in non-reproductive females observed in this study, could be determined using light microscopic observation of the coxa of the third pereopod (first walking leg) dissected from a specimen. However, for each species, the combination of (1) the presence of these setae above female gonopores on pereopods 3 plus the absence of gonopores on pereopods 5 (=female) and (2) the absence of these setae and pereo- pod 3 gonopores plus the presence of gonopores on pereopods 5 (=male) should be verified initially on some ovigerous and non-ovigerous specimens using SEM. The currently accepted assumption that in eusocial shrimps helpers are males or juvenile males is not sup- ported by our study. Our results show that the helper populations mainly consist of males, females, and a few smaller individuals which have not yet developed gonopores (juveniles). In some species, there were individuals with both male and female gonopores (intersexes). Our results generally confirm a 50:50 male: female sex ratio (except for S. chacei colony 2 and S. filidigitus), suggesting that the basic sexual sys- tem for S. chacei, S. rathbunae and S. regalis, is gonoch- ory (separate sexes) rather than protandric hermaphroditism. Observed deviations may be due to non-random sampling in S. chacei (see "Materials and methods") and thus except for S. filidigitus, the 50:50 sex ratio among helpers, with males and non-reproduc- tive females of similar size, is not indicative of protandry (Bauer 2004). Moreover, in S filidigitus, if intersex 4y Springer 1882 Mar Biol (2007) 151:1875-1886 00 o o o 5 ; a & chacei 1 4 - 3 ; A?& * 2 : O 1 i n - v OCX) A b 5. chacei 2 A 44 8 A O A s ^ - &A O %" ? : o-o-o ?o ?0 0 ': d S. filidigitus 9?V 5; e & rathbunae 4 : 3 -. 2 : o o ACS,^ OA ny o Ac#& AO ^ 0% ?A ? A* 1 i A V n - f S. regalis ?<0(# A A A a )QOOO o AA o ' A OOA' 3 4 5 1 1.5 2 Carapace length (mm) 2.5 4 3 2 1 0 5 * & 4 ? 2 'gj i : o JT 4 f- 3 2 1 0 3.5 Fig. 4 Synalpheus spp. Relationship between carapace length squares) in a 5. chacei colony l,b S. chacei colony 2, c S. rathbunae and major chela length in helpers (males open circles, females A, A S. filidigitus, e S. rathbunae, f S. regalis. Note that the dimen- open triangles, intersexes open squares, and juveniles upside down sions on the x-axis are different for the left and right column triangles) and queens (female filled triangles, intersexes filled helpers are considered as females along with the queens, there is no deviation from a 50:50 sex ratio (TV = 38, X2 = 0.641, P = 0.423). It seems more likely, given the mixed sexual composition of the helper pop- ulation, that queens are derived by reproductive matu- ration of a female helper (except in S. filidigitus, see below). Supporting this assumption are also the results from S. chacei helpers with slight ovarian development having only female gonopores, as well as from the female dispersers of S. rathbunae (future queens) colo- nizing empty sponges showing only female gonopores. Lastly, female and male helpers were similar in body size except in S. regalis, in which male helpers were slightly but significantly larger than female helpers, and not as one would expect in protandric sex change, in which females are the larger sex. The hypothesis that some individuals may mature sequentially from male to female (protandry), suggested as a possibility in Duffy (2003), is not supported by our study. Instead, a gonochoristic sexual system is indicated for these spe- cies in which reproductive males are derived from male helpers and queens from female helpers. 4y Springer Mar Biol (2007) 151:1875-1886 1883 Table 2 Mean carapace length (CL) of male and female helpers and the results of (-tests of the null hypothesis: no difference in log carapace length between male and female helpers of 5 social Syn- alpheus species (juveniles and intersexes are excluded from the analysis) Species Mean CL females Mean CL males df T-value P S. chacei colony 1 2.87 2.46 38 3.781 0.001* S. chacei colony 2 2.63 239 39 1.760 0.085 S. filidigitus 2.1 2.0 29 0.652 0.520 S. rathbunae A 3.34 3.43 39 0.633 0.531 S. rathbunae 2.44 2.56 42 0.720 0.473 S. regalis 2.5 2.8 64 3.483 0.0001* * Represents significant P values In S. chacei colony 2, S. rathbunae A and S. filidigitus (Fig. 4), we also found intersex individuals having both types of gonopores. Intersexes have been found in a variety of crustacean species, with different causes pro- posed, both non-functional (developmental anomalies; e.g. Sagi et al. 1997; Ford et al. 2004) and functional (protandric hermaphroditism, Rudolph 1995). In par- astacid crayfishes, in which the presence of a double set of gonopores is common, intersexes have been shown by histological studies to be male to female sex chang- ers, i.e. protandric hermaphrodites (Rudolph 1995; Almeida and Buckup 2000). In the caridean genus Lys- mata, individuals have both pairs of gonopores but the sexual system is protandric simultaneous hermaphro- ditism, in which individuals mature first in a male phase and later acquire female characters, and thereafter are functional simultaneous hermaphrodites (Bauer 2000). In our study we found that in intersexes, the male gonopores of these helpers were smaller than those of pure males or, in a few individuals, only one male gon- opore was present (in addition to the female ones). In S. filidigitus 20% of the helpers had rudimentary male gonopores as well as female gonopores. In this species only, all three queens examined (coming from the same colony), which clearly reproduce as females, were also intersexes. Although helpers do not function sexu- ally, it seems likely that the sexually reproductive males and queens must first pass through a helper (non-reproductive) phase when smaller (younger). In this case, intersexes may have the potential to become a reproductive male or female. It is possible that in some species (S. filidigitus) individuals change sex from male to female (Fig. 4) as a sex change hypothesis would predict. However, intersexes in this species were not smaller than females and larger than males, on the contrary, the largest and oldest individual (queen) in the colony was an intersex. It is also difficult to explain why not all animals in the colony with female gonop- ores possessed also male gonopores. What would be the function of pure female helpers if the queens (breeding females) were intersex? Because our species results are based on individuals coming from single col- onies, and because it is not known how many genes or chromosomes are required to determine their sex, or whether factors such as sponge secondary metabolites can affect sexual development, we need to be careful with drawing conclusions. Developmental anomalies occur quite frequently in social shrimps, such as major chela missing the pollex, or animals missing their ros- trum etc. (Didderen et al. 2006, pers obs). Such devel- opmental anomalies, possibly including double sets of gonopores (intersexes) might be due to genetic events such as population bottleneck events or inbreeding. Inbreeding could occur in some colonies if the original queen dies and she is replaced by a daughter. In S. fili- digitus all three queens analyzed came from the same colony and this makes generalizations invalid. Even if the underlying mechanism for intersexuality were known, interpretation of this phenomenon would be difficult without further behavioral observations and experiments. Unfortunately at this stage experimental studies can not be carried out because individuals need to be sacrificed to sex them by gonopore examination with SEM. In summary thus, the intersex characteristics of S. filidigitus queens may be due to (1) a developmental Table 3 Mean chela length (ChL) of male and female helpers, the slopes of the regression lines of log transformed carapace ver- sus chela length, and the results of ANCOVA of the null hypoth- esis: no difference in log chela length between male and female helpers of 5 social Synalpheus species (juveniles and intersexes are excluded from the analysis). Parasitized individuals were excluded from the analysis Species Mean ChL Mean ChL Slope females Slope males SS N F-value P females males (a=) (af) S. chacei colony 1 3.58 3.10 1.45 (0.71) 1.54 (0.95) 0.000 30 0.225 0.639 S. chacei colony 2 3.47 3.09 1.32 (0.61) 1.27 (0.61) 0.004 34 1.298 0.262 S. filidigitus 2.50 2.50 2.53 (0.83) 2.50 (0.80) 0.050 30 1.012 0.324 S. rathbunae A 3.43 3.45 1.71 (0.53) 1.80 (0.80) 0.003 38 2.065 0.160 S. rathbunae 2.53 2.77 1.90 (0.59) 1.77 (0.87) 0.002 38 0.628 0.434 S. regalis 2.29 2.72 1.68 (0.77) 1.84 (0.84) 0.006 63 1.007 0.320 4y Springer 1884 Mar Biol (2007) 151:1875-1886 00 o 3 Q E o o -^ ^0 X u is >>:5 _a o 8 1 '3 .5 & o ^# 3 o ^ N o PS " ? > u .a i I 3 ^? o o? ? ON -a o, ^2 a c '?? S3 t XI 3 o Q a-a " o ? B ?i o H5 o X 11 CJ o 3 _a a- PJ a u "3 E u o 11 o "3 E CO -G "5 = ?l) X ?l) 1) 3 CO CT" 03 V C 11 > 3 i-i X u 11 c 1-1 03 1) a 2 05 1) CJ E u PH 73 u >. a c < * "a c >^ 6 3 c ?1) E u CL 0 u CJ > _a > 0 ?1) 73 .3 - c 1) 11 6 3 'CT 9* * C >, a 73 _o u 1 3 ?i u >, 60 a C a 3 u o V Cu CO o a> (U (U i a >, >, >, r A 8 13 13 s s u u x y u u u g 13 13 13 "ESS H O (J O HhhUi w r4 m m n n rH C4 ^ H M O ON Tj" O O (N rH CS H M ^ rH T^ 00 00 00 00 ?* Tt ?+ ^ ?* vo ? u a c O c "3 a c % % % ?11 > ?ii > aj aj aj ii ?ii I I o o o o o o rH ON 00 NO (NI GO m in t 1 (Nl" rH m m rH S g 3 -S s s ^ g "5 ? be ^ ^ c>i oo co 05 anomaly, (2) a protandric origin from a male or inter- sex helper, or (3) simultaneous hermaphroditism. Acceptance or rejection of these or other sexual sce- narios is not possible at the moment. Future studies need to involve more colonies to obtain additional data including larger sample sizes, behavioral experiments, genetic methods, and histological studies to help to elu- cidate the sexual functionality in these animals. Colony founders, i.e. dispersing individuals coloniz- ing unoccupied sponges, were examined in S. rathbu- nae, the only species in which the sponge hosts survive experimental manipulation. Colonizing individuals occurred in heterosexual pairs, suggesting that dispers- ing males and females similar in size to helpers give rise to the reproductive males and queens in a new colony (Toth, submitted). It is thus possible that colony founders originate from male or female helpers that develop into reproductives by release from the inhibi- tion of gonadal development that occurs within their mother colony. One characteristic of eusocial animals is specializa- tion in tasks among helpers. In eusocial shrimps the most important task is colony defense (Duffy 2003). In social Hymenoptera, thrips, and beetles, mostly females perform defensive tasks, while in termites and mole rats both sexes help equally (Wilson 1971; Lacey and Sher- man 1991; Choe and Crespi 1997). We investigated the possibility that helper females and males differ in defen- sive roles by measuring of the size of the major (fight- ing) chela, which is used in colony defense in snapping shrimps. However, we found no significant differences among males and female helpers in major chela size, suggesting no such specialization among the sexes. The reliable method for sexing Synalpheus individu- als reported herein has allowed us to investigate some important characteristics of the sexual composition and social organization of these eusocial shrimps, and the results can be used as a stepping stone for further investigations. Our results show that the sexual organi- zation of these diploid shrimps resemble that of other diplo-diploid animals societies, such as mole-rats (Mammalia) and termites (Isoptera), in which workers are composed of both males and females and not just a single sex. Most other aspects of social organization and sexual biology of eusocial shrimps, e.g. the devel- opmental origin of colony queens and reproductive males, and the mode of helper gonad repression remain still unknown and open to further study. Acknowledgments We gratefully acknowledge use of facilities at the University of Louisiana at Lafayette's (ULL) Electron Microscopy Center and help from its Director, Dr. Thomas Pesa- creta and his assistant Anne Hume. We also thank J. Emmett Duffy for providing us with individuals of colony 1 of S. chacei to 4y Springer Mar Biol (2007) 151:1875-1886 1885 analyze, and the Fisheries Department of the Virginia Institute of Marine Science for letting us use their measuring equipment. This research was supported by the Marine Science Network Fellow- ship of Smithsonian Institute and the Smithsonian Tropical Re- search Institute (to ET). We thank JE Duffy, M Thiel and a third anonymous reviewer for their helpful comments and suggestions. This is contribution no. 791 from Caribbean Coral Reefs Ecosys- tem program (supported in part by the Hunterton Oceonographic Research Fund), and no. 113 of the ULL Laboratory of Crusta- cean Research. Sample collection complies with the current laws of both Belize and Panama, and collection permits were obtained from the Belizean department of Fisheries and AN AM in Panama. References Alexander RD, Noonan KM, Crespi BJ (1991) The evolution of eusociality. In: Sherman PW, Jarvis JUM, Alexander RD (eds) The biology of the naked mole-rat. Princeton Univer- sity Press, Princeton, pp 3-44 Almeida AO de, Buckup L (2000) Occurrence of protandric her- maphroditism in a population of the neotropical freshwater crayfish Parastacus brasiliensis (Parastacidae). J Crust Biol 224-230 Baeza JA, Bauer RT (2004) Experimental test of socially medi- ated sex change in a protandric simultaneous hermaphrodite, the marine shrimp Lysmata wurdemanni (Caridea: Hippolyt- idae). Behav Ecol Sociobiol 55:544-550 Banner DM, Banner AH (1975) The alpheid shrimp of Australia. Part 2: The genus Synalpheus. Rec Aust Mus 29:267-389 Banner DM, Banner, AH (1981a) An annotated checklist of the alpheid shrimp from the western Indian ocean. Traveaux et Document de L'Orstrom 157:92-96 Banner DM, Banner, AH (1981b) An annotated checklist of the alpheid shrimp of the Red Sea and Gulf of Aden. Zoologi- sche Verhandelingen 190:68-74 Bauer RT (1986) Sex change and life history pattern in the shrimps Thor mannitigi (Decapoda: Caridea): a novel case of partial protandric hermaphroditism. Biol Bull 170:11-31 Bauer RT (2000) Simultaneous hermaphroditism in caridean shrimps: a unique and puzzling sexual system in the Deca- poda. J Crust Biol 20(spec. no. 2):116-128 Bauer RT (2004) Remarkable shrimps: natural history and adap- tations of the Carideans. University of Okalahoma Press, Norman Bergrstom B (2000) The biology of Pandalus. Adv Mar Biol 38:1- 245 Charnov E (1982) The theory of sex allocation. Princeton Univer- sity Press, Princeton Chace FA (1972) The shrimps of the Smithsonian-Bredin Carib- bean expeditions with a summary of the West Indian shal- low-water species (Crustacea: Decapoda:Natantia). Smithsonian Cotrib Zool 98:103 Choe JC, Crespi BJ (1997) Social behavior in insects and arach- nids. Harvard University Press, Cambridge Crespi BJ (1992) Sociality in Australian gall thrips. Nature 359:724-726 Dardeau MR (1984) Synalpheus shrimps (Crustacea: Decapoda: Alpheidae). I. The gambarelloides group, with a description of a new species. Mem Hourglass Cruises 7, part 2:1-125 Didderen K, Fransen CHJM, Voogd de NJ (2006) Observations on sponge-dwelling colonies of Synalpheus (Decapoda, Alp- heidae) of Sulawesi, Indonesia. Crust 79:961-975 Duffy JE (1996a) Eusociality in coral reef shrimp. Nature 381:512-514 Duffy JE (1996b) Synalpheus regalis, new species, a sponge-dwell- ing shrimp from the Belize barrier reef, with comments on host specificity in Synalpheus. J Crust Biol 16:564-573 Duffy JE (1998) On the frequency of eusociality in snapping shrimps (Decapoda: Alpheidae), with description of a sec- ond eusocial species. Bull Mar Sci 63:387-400 Duffy JE (2003) The ecology and evolution of eusociality in sponge-dwelling snapping shrimp. In: Kikuchi T, Azuma N, Higashi S (eds) Genes, behaviors and evolution in Social In- sects. Hokkaido University Press, Sapporo, pp 217-252 Duffy JE, Macdonald KS (1999) Colony structure of the social snapping shrimp Synalpheus filidigitus in Belize. J Crust Biol 19:283-292 Duffy JE, Morrison CL, Macdonald KS (2002) Colony defense and behavioral differentiation in the eusocial shrimp Synal- pheus regalis. Behav Ecol Sociobiol 51:488-495 Faulkes CG, Bennett NC (2001) Family values: group dynamics and social control of reproduction in African mole-rats. Trends Ecol Evol 16:184-190 Felder DF (1982) Reproduction of the snapping shrimps Synal- pheus fitzmuelleri and S. apioceros (Crustacea: Decapoda: Alpheidae) on a sublittoral reef off Texas. J Crust Biol 2:535- 543 Ford AC, Fernandes TF, Read PA, Robinson CD, Davies IM (2004) The costs of intersexuality: a crustacean perspective. Mar Biol 145:951-957 Gherardi F, Calloni C (1993) Protandrous hermaphroditism in the tropical shrimp Athanas indicus (Decapoda: Caridea), a symbiont of sea urchins. J Crust Biol 13:675-89 HamiltonWD (1964) The genetical evolution of social behaviour I, II. J Theor Biol 7:1-52 Hart AG, Ratnieks FLW (2005) Crossing the taxonomic divide: conflict and its resolution in societies of reproductively toti- potent individuals. J Evol Biol 18(2):383-95 Hoglund H (1943) On the biology of larval development of Lean- der squilla (L.) forma typica De Man. Sven. Hydrogr. Biol. Komm. Skr. Ny. Ser. Biol. Band III, No 6, p 44 Jarvis JUM, Bennett NC (1993) Eusociality has evolved indepen- dently in 2 genera of bathergid mole-rat-but occurs in no oth- er subterrean mammal. Behav Eco Soc 33:253-260 Kent DS, Simpson JA (1992) Eusociality in the beetle Austro- platypus incompretus (Coleoptera, Curculionidae). Natur- wissenschaften 79:86-87 Lacey EA, Sherman PW (1991) Social organization of naked mole-rat colonies: evidence for division of labor. In: Sher- man PW, Jarvis JUM, Alexander RD (eds) The biology of the naked mole-rat. Princeton University Press, Princeton pp 275-336 Levinton JS, Allen BJ (2005) The paradox of the weakening com- batant: trade-off between closing force and gripping speed in a sexually selected combat structure. Funct Ecol 19:159-165 Michener CD (1974) The social behaviour of the bees. Harvard University Press, Cambridge Morrison C, Rios R, Duffy JE (2004) Phylogenetic evidence for an ancient rapid radiation of Caribbean sponge-dwelling snapping shrimps (Synalpheus). Mol Phylogen Evol 30:563- 568 ^ Nakashima Y (1987) Reproductive strategies in a partially pro- tandric shrimp, Athanas kominatoensis (Decapoda: Alphei- dae): sex change as the best of a bad situation for subordinates. J Ethol 5:145-159 Nolan BA, Salmon M (1970) The behavior and ecology of snap- ping shrimp (Crustacea: Alpheus heterochaelis and Alpheus normanni). Form Funct 2:289-335 Queller DC, Strassmann JE (1998) Kin selection and social in- sects. Bioscience 48:165-175 4y Springer 1886 Mar Biol (2007) 151:1875-1886 Ratnieks FLW (1988) Reproductive harmony via mutual policing by workers in eusocial Hymenoptera. Am Nat 132:165-175 Reeve HK, Sherman PW (1991) Intracolonial aggression and nepo- tism by the breeding female naked mole-rat. In: Sherman PW, Jarvis JUM, Alexander RD (eds) The biology of the naked mole-rat. Princeton University Press, Princeton pp 337-357 Ritzmann R (1974) Mechanisms for the snapping behavior of the two alpheid shrimps, Alpheus californiensis and Alpheus het- erochaelis. J Comp Physiol 95:217-236 Rudolph E (1995) Partial protrandric hermaphroditism in the burrowing crayfish Parastacus nicoletti (philippi 1982) (Decapoda: Parastacidae). J Crust Biol 15:720-732 Sagi A, Snir E, Khalaila I (1997) Sexual differentiation in decapod crustaceans: role of the androgenic glad. Invertebr Reprod Dev 31:55-61 Stern DL, Foster WA (1996) The evolution of soldiers in aphids. Biol Rev 77:27-79 Suzuki H (1970) Taxonomic review of four alpheid shrimps belonging to the genus Athanas with reference to their sexual phenomena. Sci RepYokohama Natl Univ Sect 2 17:1-37 Taylor GM (2001) The evolution of armament strength: evidence for a constraint on the biting performance of claws of duro- phaguous decapods. Evolution 55:550-560 Toth E, Duffy JE (2005) Coordinated group response to nest intruders in social shrimp. Biol Lett 1:49-52 Versluis M, Schmitz B, von der Heydt A, Lohse D (2000) How snapping shrimp snap: through cavitation bubbles. Science 289:2114-2117 Wilson EO (1971) The insect societies. Harvard University Press, Cambridge 4y Springer