Colin L. McLay, Shirley S. L. Lim, and Peter K. L. Ng

Transcription

1 JOURNAL OF CRUSTACEAN BIOLOGY, 21(3): , 2001 ON THE FIRST ZOEA OF LAURIDROMIA INDICA (GRAY, 1831), WITH AN APPRAISAL OF THE GENERIC CLASSIFICATION OF THE DROMIIDAE (DECAPODA: BRACHYURA) USING LARVAL CHARACTERS Colin L. McLay, Shirley S. L. Lim, and Peter K. L. Ng (CLM, correspondence) Zoology Department, Canterbury University, PB 4800, Christchurch, New Zealand ( (SSLL) Biology Division, School of Science, Nanyang Technological University, 469 Bukit Timah Road, Singapore ( (PKLN) Department of Biological Sciences, National University of Singapore, Kent Ridge, Singapore ( ABSTRACT The first zoea of Lauridromia indica (Gray, 1831) is described and compared with other dromiid larvae. Forty-six characters of dromiid first zoea and megalopa larvae (seven genera and 11 species) are summarized, and their concordance with the taxonomy of recently revised genera, based on adult morphology, is tested. For the most part, larval characters support the new generic arrangement of McLay (1993). However, the larvae of Dromia wilsoni are very different from those of congeners, suggesting that it should be placed in a new genus. Almost half of the species whose larvae are known show various degrees of abbreviated development. All known dromiids with direct development occur in Australian waters. This mode of development is not confined to a monophyletic group of dromiids and seems to have evolved independently several times. The distinctiveness of the shell-carrying genera Conchoecetes and Hypoconcha is reinforced by larval characters and their similarity suggests descent from a common ancestor. On the basis of larval and adult characters, these two genera should eventually be placed in a separate new family. McLay (1993) substantially revised the generic classification of the Dromiidae, recognizing 99 species in 29 genera. Definitions of the new genera and emendations to the old genera can be found therein. The number of genera and species for which the larval development is known for certain, however, is rather small, with the larvae of only 10 genera and 13 species described. These species are (following McLay s 1993 generic system): Dromia personata (Linnaeus, 1758) by Lebour (1934), Rice et al. (1970); D. erythropus (George Edwards, 1771) by Laughlin et al. (1982); D. wilsoni (Fulton and Grant, 1902) by Wear (1970, 1977), Terada (1983); Lauridromia dehaani (Rathbun, 1923) by Terada (1983); Cryptodromiopsis antillensis (Stimpson, 1858) by Rice and Provenzano (1966); Dromidiopsis globosa (Lamarck, 1818) by Hale (1941); Stimdromia lateralis (Gray, 1831) by Montgomery (1922), Hale (1925); Paradromia japonica (Henderson, 1888) by Terada (1983), Hong and Williamson (1986); Austrodromidia octodentata (Haswell, 1882) by Hale (1925); Cryptodromia tuberculata (Stimpson, 1858) [= Cryptodromia pileifera Alcock, 1901] by Tan et al. (1986); Conchoecetes artificiosus (Fabricius, 1798) by Sankolli and Shenoy (1968); Hypoconcha parasitica (Linnaeus, 1763) [= H. sabulosa (Herbst, 1799)] by Lang and Young (1980); and H. arcuata (Stimpson, 1858) by Kircher (1970). Three of the above dromiid species, Dromidiopsis globosa, Stimdromia lateralis, and Austrodromidia octodentata, have direct development (to first crab), while one species, Cryptodromia tuberculata, is almost completely abbreviated (only one zoeal stage). One other species, Epipedodromia thomsoni (Fulton and Grant, 1902), is also believed to have direct development (Hale, 1925). The characters of the first stage zoea of Lauridromia indica (Gray, 1831) are described here for the first time. Comparisons are made with the only congener for which larvae are known, L. dehaani. We also take the opportunity to test the generic system of McLay (1993) utilizing the information available for the first zoeae. MATERIALS AND METHODS One ovigerous L. indica was collected by trawlers off Singapore in The first stage zoea that were obtained 733

2 734 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 21, NO. 3, 2001 from this female were preserved in Formalin. The adult and samples of the zoeae have been deposited in the Zoological Reference Collection of the Raffles Museum, National University of Singapore (ZRC ). Zoeas were dissected in glycerine, under an Olympus microscope, at 400, 1,000, phase contrast, equipped with a camera lucida, using fine entomological needles. Measurements and structural details of the specimens and appendages, viewed under a Nikon microscope, 100 oil immersion, were based on at least 10 specimens. The order of the zoeal description is based on the malacostracan somite plan and described from anterior to posterior. Setal armature on appendages is described from proximal to distal segments and in order of endopod to exopod as described by Clark et al. (1998). RESULTS Larval Description Lauridromia indica (Gray, 1831) Figs. 1 2 Carapace (Fig. 1A). Longer than high, ovate; 1 pair of posteriorly directed, gently recurved spines; rostrum longer than antennule; eyes sessile. Antennule (Fig. 2A). Uniramous, endopod absent; exopod unsegmented, with 1 subterminal plumose seta and 6 terminal aesthetascs (2 stout, 4 thin). Antenna (Fig. 2B). Endopod unsegmented, subequal to exopod (= scale), with 1 subterminal and 3 terminal plumose setae; exopod with 10 plumose setae on distal margin, fine setae on inner and outer margins. Mandible (Fig. 2C). With concave median surface, serrate lateral rim. Ventroposterior region with asymmetric group of teeth. Palp absent. Maxillule (Fig. 2D). Coxal endite with 12 setae (7 terminal denticulate, 5 subterminal plumose setae); basial endite with 9 setal processes (6 terminal denticulate, 3 subterminal plumose setae); endopod 2-segmented, proximal segment with 2 plumose setae, distal segment with 6 plumose setae (2 subterminal, 4 terminal). Maxilla (Fig. 2E). Coxal endite bilobed, with plumose setae; basial endite bilobed, with plumose setae; endopod bilobed, with (2 subterminal, 4 terminal) setae; exopod (scaphognathite) margin with 30 or 31 plumose setae. First Maxilliped (Fig. 1C). Coxal segment with 1 plumose seta; basis with 11 plumose setae arranged 2, 3, 3, 3; endopod 5-segmented, with 2, 3, 1, 2, 6 (1 basal, 1 subterminal, 4 terminal) plumose setae respectively; exopod with 4 long, terminal, plumose natatory setae. Second Maxilliped (Fig. 1D). Coxa with 1 plumose seta; basis with 4 plumose setae arranged 1, 1, 2; endopod 4-segmented, with 3, 3, 2, 6 (1 median, 2 subterminal, 3 terminal) plumose setae respectively; exopod 2-segmented, with 4 long, terminal, plumose natatory setae. Third Maxilliped (Fig. 1E). Uniramous; endopod with 2 terminal setae; exopod naked. Pereiopods (Fig. 2F). First pereiopod chelate; remaining pereiopods present as large, unsegmented rudiments extending beneath carapace. Abdomen (Fig. 1B). Six somites; somite 6 fused with telson; pleopods present on somites 2 5, uniramous, buds absent. Telson (Fig. 1B). Triangular, with distinct posterior median notch fringed with fine setae. Each half of telson with short, stout spine at posterolateral corner; second process present as short, plumose seta and 5 long, articulated setae. Comparison of Dromiid Genera Based on the First Zoeal Larvae and Megalopae Assuming that larval and adult characters have evolved in tandem, we can use the generic system of McLay (1993), which was based solely on adult characters, to make several predictions about how the larvae of these species should be grouped. Briefly, the generic system is based on two (not necessarily mutually exclusive) hypotheses: a) that the direction of evolution has been towards loss of the camouflage habit, so that advanced dromiids are non-camouflagers, and b) that there have been four theatres of dromiid evolution, the Atlantic Ocean (with many endemic species of the genus Dromia Weber, 1795), South African waters (with six endemic genera), southern Australia (with more than four endemic genera), and the remainder of the Indo-Pacific (which includes both the most primitive and advanced genera). Of the six predictions tested in Table 2 (see Table 1 for characters), we can find some larval evidence to support most of them to dif-

3 MCLAY ET AL.: FIRST ZOEA OF LAURIDROMIA INDICA 735 Fig. 1. Lauridromia indica (Gray, 1831), first zoea. A, lateral view of first zoea; B, dorsal view of abdomen and telson; C, first maxilliped; D, second maxilliped; E, third maxilliped.

4 736 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 21, NO. 3, 2001 Fig. 2. Lauridromia indica (Gray, 1831), first zoea. A, antennule; B, antenna; C, mandible; D, maxillule; E, maxilla; F, pereiopods.

5 MCLAY ET AL.: FIRST ZOEA OF LAURIDROMIA INDICA 737 Table 1. Summary of dromiid larval characters. Lauridromia Lauridromia Dromia Dromia Dromia Paradromia Cryptodromiopsis Cryptodromia Conchoecetes Hypoconcha Hypoconcha indica dehaani erythropus personata wilsoni japonica antillensis tuberculata artificiosus parasitica arcuata Number of zoel stages not known Zoea I: Carapace posterior margin smooth smooth smooth smooth smooth smooth serrated smooth smooth smooth smooth dorsal spine absent denticle absent denticle present denticle absent absent denticle absent absent posterolateral spines present present present present absent absent absent absent absent absent absent lateral spines absent absent absent absent present absent absent absent absent absent absent rostral length vs. antennule longer longer longer longer longer sub-equal sub-equal shorter sub-equal sub-equal sub-equal Telson median notch distinct distinct distinct indistinct indistinct indistinct distinct distinct distinct indistinct distinct length/width posterolateral fused spine present present present present present absent present present present present present second process plumose plumose naked, naked, plumose plumose plumose *vestigial plumose plumose plumose seta seta hair-like hair-like seta seta seta knob seta seta seta long setae (pairs) Antennule terminal aesthetascs *see footnote subterminal seta *see footnote Antenna scale setae (distal margin) / / (+ mesial (+ mesial margin) margin) endopod endopod length vs. scale length 1/1 3/4 2/3 7/10 1/1 3/4 7/10 1/1 4/5 3/5 3/4 terminal setae subterminal seta protopod spine absent absent present present absent absent present absent absent absent present Mandible palp absent not known not known not known absent small bud absent *equal in length to absent absent absent mandibular process Maxillule endopod distal segment setae 2, 2, 2 2, 2, 2 2, 2, 2?1, 1, 2 2, 2, 2 2, 2, 2 2, 2, 2 *segments 2, 2, 2 2, 2, 2 2, 2, 2 indistinct, naked proximal segment setae protopod basial endite: denticulate setae denticles basial endite: plumose setae *absent coxal endite setae ? denticles

6 738 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 21, NO. 3, 2001 Table 1. Continued. Lauridromia Lauridromia Dromia Dromia Dromia Paradromia Cryptodromiopsis Cryptodromia Conchoecetes Hypoconcha Hypoconcha indica dehaani erythropus personata wilsoni japonica antillensis tuberculata artificiosus parasitica arcuata Maxilla endopod: distal lobe setae 2, 2, 2 2, 2, 2 2, 2, 2?2, 1, 2 2, 2, 2 2, 2, 2 2, 2, 2 *segments 1, 2, 2 1, 2, 2 1, 2, 2 indistinct endopod: prox. lobe setae 4 2 3? denticles basial endite: distal lobe setae 5 5 4? denticles basial endite: prox. lobe setae 5 5 4? denticles coxal endite: distal lobe setae 4 4 3? denticles coxal endite: prox. lobe setae ? denticles scaphognathite setae First maxilliped exopod natatory setae endopod setae: prox. to distal 2, 3, 1, 2, 6 3, 3, 1, 2, 5 3, 3, 1, 2, 5 3, 3, 1, 2, 5 3, 3, 1, 2, 5 3, 3, 1, 2, 5 3, 3, 1, 2, 5 *0, 0, 0, 0, 3 3, 3, 1, 2, 5 3, 3, 1, 2, 5 3, 3, 1, 2, 5 basipod setae: prox. to distal 1, 2, 3, 3, 3 1, 2, 3, 3, 3 1, 2, 3, 3, 2 1, 2, 3, 3, 3 2, 2, 3, 3, 3, 2, 2, 3, 3, 3 1, 2, 3, 3, 3 *naked 2, 3, 3, 3, 3 3, 2, 3, 3, 3 2, 2, 3, 3, 3 Second maxilliped exopod natatory setae endopod setae: prox. to distal 3, 3, 2, 6 3, 3, 2, 5 3, 3, 2, 5 3, 3, 2, 5 3, 3, 2, 5 3, 3, 2, 5 3, 3, 2, 5 *0, 0, 1, 2 2, 3, 2, 5 2, 3, 2, 5 2, 3, 2, 5 basipod setae: prox. to distal 1, 1, 1, 2 1, 1, 1, 2 1, 1, 1, 1, 2, 3 1, 1, 1, 1, 2 1, 1, 1, 1 1, 1, 1, 1 1, 1, 1, 1, 2 *naked 1, 1, 1, 2 1, 1, 1, 1, 2 1, 1, 1, 1, 2 Third maxilliped exopod naked naked naked naked naked naked naked naked naked undeveloped absent endopod setae 2 (simple) 2 (simple) 2 (1 simple, 2 (simple) 3 (simple) 3 (simple) 2 (plumose) 1 (simple) 3 (plumose) undeveloped absent 1 denticulate) Megalopa: Carapace length/width not known not known not known greatest width not known not known nearer nearer nearer not known nearer nearer nearer nearer nearer posterior posterior posterior anterior posterior posterior posterior posterior margin margin margin margin margin margin margin margin Appendages fifth pereiopod subchelate not known not known yes yes yes yes yes yes no no no long setae on 5th pereiopod dactyl not known not known present present present present present absent present present present uropods not known not known biramous biramous uniramous uniramous biramous uniramous uniramous biramous biramous Telson posterior margin not known not known straight straight slightly strongly slightly strongly strongly straight straight concave concave concave concave concave marginal processes not known not known present present absent present present present present present present * Because of abbreviated development these structures are not comparable.? The larval data from Lebour (1934) may not be accurate.

7 MCLAY ET AL.: FIRST ZOEA OF LAURIDROMIA INDICA 739 Table 2. Test of predicted versus observed larval groupings of dromiid crabs using data from Table 1. Only apomorphic characters are considered. [Note: when considering Dromia spp., D. wilsoni is excluded.] Predicted larval relationships based on adults Observed results in larvae 1) The species grouped under each genus should be more similar to each other than they are to species in other genera. 2) There should be a group of large sponge-camouflagers with a cheliped epipod, consisting of Tunedromia, Lauridromia, Dromidiopsis, and Dromia. 3) There should be a group of smaller dromiids, with a cheliped epipod, including Fultodromia, Stimdromia, with Petalomera, Paradromia, and Frodromia being the most derived in this group. 4) We would expect to find that the larvae of Conchoecetes and Hypoconcha, two shell-carrying genera, share some close similarities. 5) The genera without an epipod on the cheliped should fall into three groups: a) Cryptodromiopsis; b) Dromidia and Exodromidia; and c) Austrodromidia. 6) The species of Dromia should form a monophyletic group. Dromia (2 spp.) share: second telson process of zoea naked, hair-like; ten antennal scale setae extending onto mesial margin; single plumose seta on basal endite of maxillule; four setae on proximal lobe of maxilla basal endite. Lauridromia (2 spp.) share: no apomorphic characters. Hypoconcha (2 spp.) share: 4 pairs of long setae on telson; second maxilliped basipod setae are 1,1,1,1,2; exopod of third maxilliped absent or undeveloped. Lauridromia spp. and Dromia spp. share presence of: posterolateral spines on carapace; rostrum longer than antennule; third maxilliped endopod has 2 simple or denticulate setae. Larvae of Tunedromia are unknown. Only have data for P. japonica; this species differs from the rest in having: the posterolateral fused telson spine absent; mandibular palp represented by a small bud. No zoea known for the other genera. Conchoecetes (1 sp.) and Hypoconcha (2 spp.) share: maxilla endopod distal lobe setae are 1,2,2; second maxilliped endopod setae are 2,3,2,5. Cryptodromiopsis has serrated posterior carapace margin; there are 2 plumose setae on the third maxilliped endopod; and the greatest width of carapace is nearer the anterior margin. Larvae of the other 3 genera are unknown. The 4 shared unique characters of D. erythropus and D. personata are listed above (see 1), but D. wilsoni differs in every case. Additional differences are: presence of dorsal and lateral carapace spines in the zoea I and absence of telson marginal processes in the megalopa. ferent degrees. Dealing with each of them in turn we find that in relation to the first prediction, the species of Dromia (excluding D. wilsoni) share four unique characters, Lauridromia McLay, 1993, none, and Hypoconcha Guérin-Méneville, 1854, three characters. Amongst the species of Dromia we also have several characters which are only found in one or another genus: the presence of posterolateral carapace spines, and rostrum being longer than the antennule, are shared with Lauridromia; presence of an antennal protopod spine is shared with Cryptodromiopsis antillensis; and having a straight posterior telson margin in the megalopa is shared with Hypoconcha. The antennal protopod spine, shared with C. antillensis, may well prove to be an ancestral character. The telson margin character, shared with Hypoconcha, is hard to assess because, on the basis of adult characters, these two genera are only distantly related. The two distinctive characters (see above) of Lauridromia are shared with Dromia, so the larvae provide no evidence to support the generic concept of Lauridromia. This result is not unexpected because both genera form part of the group of larger camouflaging dromiids with a cheliped epipod (see second prediction). In addition to the three unique characters, Hypoconcha also has a distal lobe setal formula of 1, 2, 2 on the maxilla endopod; 2, 3, 2, 5 on the second maxilliped endopod; and in the megalopa the fifth pereiopod lacks a subchelate mechanism, all of which are shared with Conchoecetes Stimpson, These shared larval characters are part of our argument that these two genera should eventually be removed to a new family of their own. As noted above, the straight posterior telson margin in the megalopa of Hypoconcha and Conchoecetes is shared with Dromia.

8 740 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 21, NO. 3, 2001 Table 3. List of additional predictons about larval groupings and the data needed to test them. Predictions are numbered so as to follow on from Table 2. Predicted larval relationships based on adults Data needed to test the predictions 7) We would predict that Sphaerodromia or Eodromia should have the most primitive larvae. 8) The larvae of non-camouflaging dromiids, such as Epigodromia and Takedromia, should, in general, have more derived characters than those of the camouflaging dromiids, such as Dromia, Dromidiopsis, Cryptodromia, and Cryptodromiopsis. 9) Small camouflagers without a cheliped epipod: Cryptodromia and Homalodromia should be similar. 10) The South African dromiid species, including Pseudodromia, Exodromidia, Eudromidia, Dromidia, Barnardromia, and Speodromia, should form a monophyletic group. No larvae of these genera are known, but development is not likely to be direct because egg diameter is only mm. Larval data are only available for camouflagers and not the others, so we need, for example, the larvae of a common genus like Epigodromia. Also Cryptodromiopsis unidentata is very widespread and common and should be studied. Only have data for one species of Cryptodromia with abbreviated development. The species of these two genera are widespread and common in shallow waters so should be easy to study. No larvae are known. Many of these species are known to have large eggs, although none have been confirmed as direct developers. Species of Dromidia seem reasonably common. For testing the second prediction, we only have data for Lauridromia and Dromia (except D. wilsoni). For these two genera, the presence of posterolateral carapace spines in zoea I and two simple or denticulate setae on the endopod of the third maxilliped are unique characters. Shared characters include having a first maxilliped basipod setal formula of 1, 2, 3, 3, 3 (shared with C. antillensis) and having the rostrum longer than the antennule (shared with D. wilsoni). The third prediction cannot be adequately tested because data are only available for P. japonica, which has two unique characters: absence of a posterolateral fused spine on the telson and mandibular palp represented by a small bud. However, it is not known whether these are representative of the whole group of genera or only P. japonica. Stimdromia lateralis has direct development, and larvae of Fultodromia McLay, 1993, Petalomera Stimpson, 1858, and Frodromia McLay, 1993, are unknown. Testing the fourth prediction about the shell-carrying dromiids, two characters are unique to Hypoconcha (four pairs of telson long setae, and the third maxilliped exopod setae are undeveloped or absent), but only one to Conchoecetes (three plumose setae on third maxilliped endopod). Conchoecetes and Hypoconcha are united by three unique characters: they share a maxilla endopod distal lobe setal formula of 1, 2, 2; second maxilliped endopod setal formula of 2, 3, 2, 5; and the megalopal fifth pereiopods lack a subchelate mechanism. A carapace rostrum length subequal to the antennule is a character common to both Hypoconcha and Conchoecetes (but shared with C. antillensis and P. japonica); Hypoconcha has a second maxilliped basipod setal formula of 1, 1, 1, 1, 2 (but shared with D. wilsoni) while Conchoecetes has a formula of 1, 1, 1, 2. Conchoecetes has uniramous uropods in the megalopa (but this is shared with D. wilsoni and Cryptodromia tuberculata). Because the larvae of Dromidia Stimpson, 1858, and Exodromidia Stebbing, 1905, are unknown and Austrodromidia octodentata has direct development, we only have available data for Cryptodromiopsis antillensis to test the fifth prediction. There are three unique characters (Table 2). The serrated posterior carapace margin is also found in Homola barbata (see Rice and Provenzano, 1970). Cryptodromiopsis antillensis has a protopod spine on the antennae, but this is shared with both species of Dromia (excluding D. wilsoni); it also has a second maxilliped basipod setal formula of 1, 1, 1, 1, 2, but this is shared with both species of Hypoconcha. Most of the species of Dromia live in the Atlantic Ocean, and these, at least from the adult perspective, form a monophyletic group

9 MCLAY ET AL.: FIRST ZOEA OF LAURIDROMIA INDICA 741 (sixth prediction). While it is clearly apparent that Dromia erythropus and D. personata are closely related, D. wilsoni is not. Four characters are unique to the two Atlantic Dromia species (Table 2, prediction 2), but D. wilsoni does not share any of them. Of the four characters that the two Atlantic Dromia share with other genera, D. wilsoni again differs from all of them: posterolateral carapace spines are absent, lateral carapace spines are present, first maxilliped basipod setal formula is 2, 2, 3, 3, 3, and there are three simple setae on the third maxilliped endopod. Clearly, the larval data indicates that D. wilsoni does not belong in the genus Dromia. Most Dromia species are endemic to the Atlantic, probably resulting from a radiation that probably dates from the origin of the Atlantic Ocean. Dromia wilsoni is present in the Indo-Pacific as well as the Atlantic oceans and may well have been a much later colonist from the Indian Ocean, via the tip of South Africa (see Discussion). Other predictions can be made from the adult revision, but at present we lack the data to test them (Table 3). Amongst the adults, Sphaerodromia Alcock, 1899, and Eodromia McLay, 1993, clearly have many ancestral characters. We predict that their larvae should also have many primitive characteristics. The generic revision was based on the hypothesis that camouflaging using sponges, etc., is ancestral and that some advanced genera like Epigodromia McLay, 1993, and Takedromia McLay, 1993, do not camouflage. Thus, we would predict that larvae of these genera should have more derived characters. Also, we would expect Cryptodromia and Homalodromia Miers, 1884, two small camouflagers lacking a cheliped epipod, to be very similar. The 17 endemic South African dromiid species, in Pseudodromia Stimpson, 1858, Exodromidia Stebbing, 1905, Eudromidia Barnard, 1947, Dromidia Stimpson, 1858, Barnardromia, McLay, 1993, and Speodromia Barnard, 1947, should also form a monophyletic group. This group is predicted to provide a further example of a cool water stenothermic radiation in this area (Kensley, 1981). DISCUSSION The use of larval characters in trying to resolve adult taxonomy is still in its infancy and has not been used extensively for many groups. Perhaps the group for which larval characters have been most often used is the Xanthoidea (sensu Guinot, 1978). Fukuda (1979), Martin (1984), and Martin et al. (1985) provided synopses of the larvae of Xanthoidea in general, while more recently, larval information has been used in the generic taxonomy of the Xanthidae sensu stricto (e.g., Ng, 1993; Ng and Clark, 1994; Clark and Ng, 1998), Eumedonidae (Van Dover et al., 1986; Lim and Ng, 1988; Mori et al., 1991; Chia et al., 1993; Chia and Ng, 1995), and Pilumnidae (Lim et al., 1984, 1986; Ng and Clark, 2000). In many of these cases, larval characters have proved useful in understanding generic as well as familial relationships for various taxonomically difficult species and genera. It has also proved useful in affirming or refuting the various generic classifications proposed. The generic classification of the Dromiidae has always relied on adult characters. Prior to McLay (1993), the genera of the Dromiidae were largely based on the revision carried out by Borradaile (1903) who recognized 12 genera. The adult characters that he considered important were the presence or absence of an epipod on the cheliped, the distinctness of the furrows that delineate the regions of the carapace, the ratio of carapace width to carapace length, the shape of the legs, and the arrangement of the female sternal 7/8 sutures. The use of the cheliped epipod character was the major innovation introduced by Borradaile because it helped to resolve many of the problems with dromiid taxonomy. However, the sternal 7/8 sutures are only of limited value because they show ontogenetic change and can only be used for mature females. Carapace grooves seem to be of little value at the generic level. McLay (1993) recognized 29 genera and used a wider range of characters including presence or absence of epipods and podobranchs on the pereiopods, ratio of carapace width to length, texture of the carapace surface, development of the rostrum, sexual dimorphism of the chelipeds, tubercles of the first two pairs of walking legs, arrangement of spines on and around the dactyli of the legs, size of uropod plates on the abdomen, presence of vestigial pleopods on the male abdomen, and fusion of the last two segments of the abdomen. Variation in the structure of the male first pleopods, which has proved

10 742 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 21, NO. 3, 2001 useful in many brachyurans, is of little use in the Dromiidae because they are for the most part so uniform. An essential feature of the revision is the use of suites of characters rather than treating each character separately. For example the nature of the carapace surface, size of the last two pairs of legs, and the spines which they bear primarily reflect the way in which pieces of camouflage are carried. Another important suite involves the shape of the abdomen, fusion of abdominal segments, and the size and shape of the uropods in relation to their role in the abdominal locking mechanism. The assumption was made that comparison of these character suites amongst the genera would indicate the direction of evolution in the family as a whole. The genera Dromia Weber, 1795, Dromidiopsis Borradaile, 1900, and Lauridromia McLay, 1993, contain most of the larger dromiids that carry sponges. Other than their large size and smooth carapace surface, all three of the above genera are also related by having an epipod on the cheliped, and the uropods are small but visible externally, so we would expect them to have similar larvae. Dromia has 10 species, most of which are confined to the Atlantic, with only three species occurring in the Indo-Pacific. Larvae are known for two of the Atlantic species, D. personata (see Lebour, 1934; Rice et al., 1970) and D. erythropus (see Laughlin et al., 1982) and for D. wilsoni (see Wear, 1970, 1977; Terada, 1983) which occurs in both the Atlantic and Indo-Pacific oceans. Direct development in this genus (Dromia) has not been recorded, and, given their small egg size, it is not expected. While the larvae of both D. personata and D. erythropus agree very well in all major characters and support McLay s (1993) concept of Dromia, one species, D. wilsoni, poses some problems. The larval evidence does not support the transfer of Cryptodromia wilsoni Fulton and Grant, 1902, into Dromia. This species had been classified in Petalomera Stimpson, 1858, since Rathbun (1923) referred it to this genus, but McLay (1993) argued that the preponderance of adult characters suggests that it is a species of Dromia. However, it is now clear that D. wilsoni may be better classified in its own, as yet unnamed, genus. As presently constituted, Dromia is thus a paraphyletic group. During the revision of Dromia, McLay (1993) noted some adult features of D. wilsoni that did not quite fit the genus, in particular the presence of some spines on the inner margin of the dactyl of the third leg. More recently it was discovered that males of D. wilsoni have vestigial pleopods on the third to fifth abdominal segments (CLM, unpublished data). However, both of the abovementioned adult features are regarded as primitive dromiid characters found in such genera as Sphaerodromia Alcock, 1899, and Eodromia McLay, Vestigial male pleopods are also found in Exodromidia Stebbing, 1905, and on males of D. bollorei Forest, It is unwise to separate species into different genera using plesiomorphic characters. The only adult characters of D. wilsoni that might be used to separate it from the other Dromia spp. are the oval shape of the carapace and the peculiar undulating, sculptured appearance of the carapace tomentum. Compared to other dromiid genera, defined only using adult characters, these differences are not particularly significant and the only strong characters available are those of the larvae. This leaves things in a rather unsatisfactory state where one of the 30 dromiid genera (McLay, 1993, 1998) would be defined using mainly larval characters, while all the others are based on adults. The larvae of D. wilsoni suggest an origin from a different ancestor to that of the other Dromia spp., so that the adult features must be regarded as being convergent. On the basis of similar adult characters, D. foresti McLay, 1993, known only from New Caledonia, might also be included in the new genus. The generic affinities of Dromia wilsoni will be treated in depth separately by the first author in a subsequent paper. Dromidiopsis includes six species which occur in the Indo-Pacific. Dromidiopsis globosa has large eggs (2.0-mm diameter) and has direct development (McLay, 1993), but other species in this genus, e.g., D. dubia Lewinsohn, 1984, and D. lethrinusae (Takeda and Kurata, 1976) have much smaller eggs and would be not be expected to have direct development (McLay, 1993). Lauridromia contains three Indo-Pacific species. Larvae are known only for L. dehaani (see Terada, 1983) and L. indica (present study). The remaining species, L. intermedia (Laurie, 1906) also has small eggs (0.5-mm diameter) and therefore is unlikely to have direct development. The present first zoea of L. indica agree with those of L. de-

11 MCLAY ET AL.: FIRST ZOEA OF LAURIDROMIA INDICA 743 haani very well in all main features, but they do not share any unique larval characters. The presence of posterolateral carapace spines and a rostrum that is longer than the antennule are both shared with Dromia. This result is not unexpected because both genera form part of the group of larger camouflaging dromiids with a cheliped epipod. The absence of apomorphic larval characters in Lauridromia larvae may simply indicate that their evolutionary separation from Dromia is more recent. The first zoea of L. indica, however, differs from that of L. dehaani in that its abdominal pleopods are very well developed (absent in the first zoea of L. dehaani). This suggests that the larval development of L. indica is probably also semi-abbreviated. Lauridromia dehaani is reported to have four zoeal stages (Terada, 1983), and on the basis of its abdominal pleopods, L. indica probably has only two or three zoeal stages (i.e., development is semi-abbreviated). Most of the eight species in Cryptodromiopsis were previously placed in Dromidia Stimpson, 1858, but it is clear that this genus is restricted to South African waters. Only the larva of C. antillensis is known (Rice and Provenzano, 1966). This Atlantic species has a sister species, C. sarraburei Rathbun, 1910 (sic C. larraburei, see Boyko, 1998) in the eastern Pacific separated by the isthmus of Panama. The adults of these two species are scarcely distinguishable, so we would predict that the larvae of C. sarraburei and C. antillensis would also be difficult to separate. The transfer of Dromidia antillensis into Cryptodromiopsis by McLay (1993) is well supported by the possession of several features that we interpret as apomorphic, viz., serrated posterior carapace margin of the first zoea, presence of two plumose setae on the third maxilliped endopod, and greatest width of the megalopal carapace is nearer the anterior margin. The serrated posterior carapace margin of C. antillensis bears a remarkable resemblance to that found in homolid zoeae (see Rice, 1980; Konishi et al., 1995). The larvae of the members of these two families, however, differ markedly in many other ways (see Rice, 1980). It is possible, however, that such serrated carapace margins represent a plesiomorphic character shared by dromiids and homolids, but with only one dromiid known to possess this feature, nothing much can be said. Abbreviated and direct development are features of dromiid life histories (47% of species whose development is known have two or fewer zoeal stages). Direct development is known in Austrodromidia octodentata, Dromidiopsis globosa, Stimdromia lateralis, and may also occur in Epipedodromia thomsoni. These genera are not necessarily closely related, and the only things that they have in common are that they all have large eggs and come from southern Australia. This mode of development seems to have evolved independently in each case. Furthermore, most of the South African dromiids, Pseudodromia, Exodromidia, Eudromidia, Dromidia, Barnardromia, and Speodromia, also have large eggs (around 1.6-mm diameter or larger), but direct development has not yet been reported for these species. McLay (1993) argued that the South African dromiids form a monophyletic group, so it is possible that, in contrast with the southern Australian dromiids, their reproductive characteristics are derived from the common ancestor. Other dromiid species, Cryptodromia tuberculata, Dromia wilsoni, Paradromia japonica, and Lauridromia indica have abbreviated larval development. Many dromiids are associated with sponges, which they use for camouflage and food. Shortened or direct development may be a response to exploiting a patchy resource, like sponges, because it enhances local recruitment. Finally, we have the shell-carrying dromiids Conchoecetes and Hypoconcha, both of which have an epipod on the cheliped and the carapace subpentagonal, flattened, and without anterolateral teeth. Also, the last two pairs of legs are shorter than the first two pairs (fourth pair shortest), and the female sternal sutures 7/8 end apart between or behind the first pair of legs. Uropods are well developed in Conchoecetes but vestigial or absent in Hypoconcha, and the two genera differ in the way they carry their bivalve shell. In Conchoecetes the dactyl of the third legs is talonlike and opposed by a stout proximal extension of the propodus while the dactyls of the fourth legs are each fashioned as a tiny lunate hook, twisted to point anterovertically and not opposed by any spines. Conchoecetes mainly uses the talon on the third legs to grasp the bivalve hinge margin posteriorly, and the fourth legs hold and support the shell edges. There are three species in this Indo-Pacific

12 744 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 21, NO. 3, 2001 genus, but the larval development is known only for C. artificiosus (see Sankolli and Shenoy, 1968). The genus Conchoedromia Chopra, 1934, has recently been shown to be a synonym of Conchoecetes Stimpson, 1858 (Guinot and Tavares, 2000). Hypoconcha lacks the talons on the third legs so that both of the last two pairs of legs must be used to elevate the shell allowing the crab to move. These legs have the dactyli fashioned into tiny lunate hooks (which can fit into a recess on the propodus) twisted to point anteriorly and posteriorly respectively, and the dactyli not opposed by any spines. As in the case of Cryptodromiopsis antillensis and C. sarraburei, the species of Hypoconcha are separated into two groups of sister species by the isthmus of Panama: on the Atlantic side we have H. arcuata, H. parasitica, and H. spinosissima Rathbun, 1933, while on the Pacific side we have H. californiensis Bouvier, 1898, H. lowei Rathbun, 1933, and H. panamensis Smith, 1869, respectively. The larvae are only known for two of the Atlantic species group, H. parasitica (see Lang and Young, 1980) and H. arcuata (see Kircher, 1970). The most striking observation is that the larvae of Conchoecetes artificiosus, Hypoconcha parasitica, and H. arcuata form one very coherent group with the following diagnostic characters: setation pattern of the endopod of the maxilla is 1, 2, 2; setation pattern of the endopod of the second maxilliped is 2, 3, 2, 5; setation on the second maxilliped basipod is 1, 1, 1, 1, 2 (but shared with C. antillensis); first zoeal rostrum subequal in length to antennule in situ (shared with C. antillensis and P. japonica); possession of four pairs of long telson setae (in both Hypoconcha species but Conchoecetes artificiosus has five); third maxilliped exopod and endopod setae are undeveloped or absent (in both Hypoconcha but developed in Conchoecetes); and the posterior margin of megalopa telson is straight (but shared with Dromia). In addition, all have only two or three zoeal stages (although this is shared with D. wilsoni and P. japonica). In the megalopae of both Conchoecetes and Hypoconcha, the uropods are well developed, with only the exopod present in the first, but they are biramous in the second genus. However, well-developed uropods are not present in postlarval Hypoconcha. The larval characters thus appear to support the separation of these two genera from other dromiids, suggesting that they should perhaps be referred to their own subfamily or family. McLay (1993) suggested that Hypoconcha could be removed from the Dromiidae, and now that the larval evidence has been examined, it is clear that Conchoecetes should be united with Hypoconcha. Differences between the adults of these two genera in the method of carrying the shell (see Guinot and Tavares, 2000) indicate that there is more than one way to grasp the shell. The separation of shell-carrying from sponge-carrying dromiids is also supported by the analysis of 18sRNA and rdna sequences (Spears et al., 1992) who showed that H. arcuata and Cryptodromiopsis antillensis are not closely related. Rice (1980) was the last to review the usefulness of larval characters in determining the higher classification of the Brachyura. No clear patterns, however, were discerned that could prove useful in determining the generic limits of the Dromiidae. The chief arguments presented by Rice (1980) centered on discussing the affinities of the Dromiidae as a family and its place in the overall classification of the Brachyura. Much has been said about the so-called anomuran hair on the telson of the larvae, but the actual value of this character in deciding the phylogenetic position of the Dromiidae is still not fully established. The other larval features, used by Rice (1980) and others to argue for the anomuran relationships of dromiids, could merely be plesiomorphic characters and thus useless in a phylogenetic analysis. Discussion about the origin of crabs using larval information has been complicated by the lack of a clearly stated hypothesis that states which groups shared a common ancestor. There have been endless arguments about whether or not the Dromiidae should be included or excluded from the Brachyura. These discussions have not really addressed the question of sister-group relationships between the podotremes and the eubrachyurans. So for the adults we can ask, Are the Podotremata the sister group of the Eubrachyura (Heterotremata + Thoracotremata)? We would answer yes to this question on the basis of adult characters (see McLay, 1999). We believe that the hypothesis of Saint Laurent (1980a, b), that the Podotremata and Eubrachyura are sister groups, is primarily supported by the fact that

13 MCLAY ET AL.: FIRST ZOEA OF LAURIDROMIA INDICA 745 podotreme females all have spermathecae between sternal sutures 7/8 (and external fertilization) whereas eubrachyuran females all have sternal gonopores through which the spermathecae are accessed (and internal fertilization). Some crab-like features were present in the common ancestor (which lacked a means of sperm storage), but, apart from these, the two groups of crabs have evolved independently. So we must ask the same question using larval information. Again we would answer yes, because the podotreme larvae that are known have many characters intermediate between anomurans and eubrachyurans. Rice (1980), interestingly, reconstructed the ancestral zoea of the Eubrachyura. The question of whether the Podotremata are in or out of the Brachyura is really a red herring, because in cladistic terms it does not matter. The podotremes can be either a subgroup of the Brachyura or they can be a separate group outside the Brachyura, because the status of the group does not alter the way in which the question is posed or how it must be answered. Much discussion has centered around the Dromiacean Paradox wherein it is pointed out that although the adult Dromiidae (one of the three dromiacean families) have many eubrachyuran features, the larvae have many anomuran features. The suggestion of larval or horizontal gene transfer (Williamson, 1988, 1992; Williamson and Rice, 1996) whereby the life cycles of an anomuran and a crab-like ancestor were somehow spliced together to produce a dromiid seems to defy current scientific wisdom. In other words, the horizontal gene transfer hypothesis implies that dromiids did not evolve but were the result of an unfortunate accident, in this case between a crab ancestor with direct development and a wayward male hermit crab with a free-living larval phase! This is what comes from taking phylogenies, based on molecular data from a few distantly related representatives of each group (see Spears et al., 1992), too seriously! ACKNOWLEDGEMENTS We thank Professor Daniele Guinot and Paul Clark for their valuable comments on an earlier draft of this manuscript. LITERATURE CITED Alcock, A An account of the deep-sea Brachyura collected by the marine survey ship Investigator. Trustees of the Indian Museum, Calcutta. 85 pp., 4 pl Catalogue of the Indian decapod Crustacea in the collection of the Indian Museum. Part I. Brachyura. Fasc. I Introduction and Dromides or Dromiacea (Brachyura Primigenia). Trustees of the Indian Museum, Calcutta. 80 pp., pl Barnard, K. H Descriptions of new species of South African decapod Crustacea, with notes on synonymy and new records. Annals and Magazine of Natural History, (11), 13 (102), 1946 (1947): Borradaile, L. A On some crustaceans from the South Pacific. Part IV. The crabs. Proceedings of the Zoological Society of London 1900: , pl On the genera of the Dromiidae. Annals and Magazine of Natural History (7)11: Bouvier, E.-L Sur quelques Crustacés anomoures et brachyures recueillis par M. Diguet en Basse Californie. Bulletin du Museum d Histoire naturelle, Paris 4: Boyko, C. B The correct spelling of Cryptodromiopsis sarraburei (Rathbun, 1910) not C. larraburei (Schmitt, 1921) (Decapoda, Brachyura, Dromiidae). Crustaceana 71: 234, 235. Chia, D. G. B., and P. K. L. Ng A revision of the genus Rhabdonotus A. Milne Edwards, 1879, with descriptions of two new species and the first zoeal stage of R. pictus A. Milne Edwards, 1879, (Brachyura: Eumedonidae). Crustacean Research 24: ,, and D. Vandenspiegel The identities of two crinoid symbionts, Harrovia albolineata Adams and White, 1849, and H. longipes Lanchester, 1900 (Decapoda, Brachyura, Eumedonidae). Crustaceana 64: Chopra, B Further notes on Crustacea Decapoda in the Indian Museum. VI. On a new dromiid and a rare oxystomous crab from the sandheads, off the mouth of the Hooghly River. Records of the Indian Museum 36: Clark, P. F., D. K. Calazans, and G. W. Pohle Accuracy and standardization of brachyuran larval descriptions. Invertebrate Reproduction and Development 33: , and P. K. L. Ng The complete larval development of the poisonous mosaic crab, Lophozozymus pictor (Fabricius, 1798) (Crustacea: Decapoda: Brachyura: Xanthidae: Zosiminae), with comments on familial characters for first stage zoeas. Zoosystema 20: Edwards, G Catalogue of the animals and plants represented in Catesby s Natural History of Carolina. With the Linnean names. Appended to Edward s (1771) edition of The Natural History of Carolina, Florida and the Bahamas Islands by Mark Catesby. Volume I, 100 pp., 100 pl.; volume II, 100 pp., 100 pl.; Appendix, 230 pp., 20 pl. Fabricius, J. C Supplementum Entomologiae Systematicae. Hafniae, Proft et Storch. 572 pp. Forest, J Les dromies de l Atlantique oriental. Description de Sternodromia gen. nov. et deux especes nouvelles du genre Dromia Weber (Crustacea Decapoda Dromiidae). Annales de I Institut Oceanographie, Paris 50: , fig. 1 7, pl Fukuda, Y Larvae of brachyuran crab: a review. Benthos Research 17/18: Fulton, S. W., and F. E. Grant Some little known Victorian decapod Crustacea with description of new species, no. 2. Proceedings of the Royal Society of Victoria 15: 59 68, pl

14 746 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 21, NO. 3, 2001 Gray, J. E Description of a new genus, and some undescribed species of Crustacea. Zoological Miscellany 1: 39, 40. Guérin-Méneville, F. E Description du genre Hypoconcha, nouveaux crabes, faux Bernard l Hermite, qui protègent leur corps avec la moitié d une coquille bivalve. Revue du Magazine Zoologie pure et appliquee (2)6: Guinot, D Principes d une classification evolutive des Crustacés Décapodes Brachyoures. Bulletin biologie de France et de Belgique, nouvelle serie 112: , and M. Tavares Conchoedromia alcocki Chopra, 1934: megalopa of Conchoecetes artificiosus (Fabricius, 1798) (Crustacea, Brachyura, Dromiidae). Journal of Crustacean Biology 20 (special number 2): Hale, H. M The development of two Australian sponge-crabs. Proceedings of the Linnean Society of New South Wales 50: , 2 pl British, Australian and New Zealand Antarctic research expeditions: decapod Crustacea. British, Australian and New Zealand Antarctic Research Expedition Rep., (B) 4(9): , fig. 1 16, pl. 3. Haswell, W. A Description of some new species of Australian Decapoda. Proceedings of the Linnean Society of New South Wales 6, pt 4: Henderson, J. R Report on the Anomura collected by HMS Challenger during the years Report of the Scientific Research of the Voyage of the Challenger 27: 1 221, 21 pl. Herbst, J. F. W Versuch einer Naturgeschichte der Krabben und Krebse, nebst einer systematischen Beschreibung ihrer verschiedenen Arten. Vols Berlin und Straslund. 515 pp., 62 pl. Hong, S. Y., and D. I. Williamson The larval development of Petalomera japonica (Henderson) (Decapoda, Dromiidae) reared in the laboratory. Journal of Natural History 20: Kensley, B On the zoogeography of southern African decapod Crustacea, with a distributional checklist of the species. Smithsonian Contributions to Zoology 338: Kircher, A. B The zoeae and glaucothoë of Hypoconcha arcuata Stimpson (Decapoda: Dromiidae) reared in the laboratory. Bulletin of Marine Science 20: Konishi, K., H. Takeoka, and H. Taishaku Description of the first zoea of Paromola macrochira Sakai (Brachyura: Homolidae) with notes on larval characters of archeobrachyuran families. Crustacean Research 24: Lamarck, J. B. P. A. de Histoire naturelle des Animaux sans vertebres, presentant les caracteres generaux et particuliers de ces animaux, leur distribution, leurs classes, leurs familles, leurs genres, et la citation des principales espece qui s y rapportent; precedee d une Introduction offrant la determination des caracteres essentiels de l Animal, sa distinction du vegetal et des autres corps naturels, enfin, l Exposition des principes fondamentaux de la Zoologie. Vol. 5: Lang, W. H., and A. M. Young Larval development of Hypoconcha sabulosa (Decapoda: Dromiidae). Fisheries Bulletin, U.S. 77: Laurie, D Report on the Brachyura collected by Professor Herdman, at Ceylon, in Pp , 2 pl. in W. A. Herdman, Report to the Government of Ceylon on the Pearl Oyster Fisheries of the Gulf of Manaar, Part 5, suppl. Rep. 40. Laughlin, R. A., P. J. Rodriguez, and J. A. Marval The complete larval development of the sponge crab Dromia erythropus (George Edwards, 1771) (Brachyura: Dromiidae) from the Archipiélago de los Roques, Venezuela. Journal of Crustacean Biology 2: Lebour, M. V The life history of Dromia vulgaris. Proceedings of the Zoological Society of London 1934: Lewinsohn, C Dromiidae from Madagascar and the Seychelles (Crustacea Decapoda Brachyura). Bulletin du Museum national d Histoire naturelle, Paris (4), 6, Section A, (1): , fig. 1 4, pl Lim, G. S. Y., and P. K. L. Ng The first zoeal stage of Harrovia albolineata Adams and White, 1848 (Crustacea: Brachyura: Pilumnidae), with a note on eumedonine systematics. Journal of Natural History 22: ,, and W. H. Tan The larval development of Heteropanope glabra Stimpson, 1858 (Decapoda: Xanthidae) in the laboratory. Crustaceana 47: 1 16., L. W. H. Tan, and P. K. L. Ng The complete larval development of Pilumnopeus eucratoides Stimpson, 1858 (Decapoda, Brachyura, Pilumnidae) in the laboratory. Crustaceana 50: Linnaeus, C Systema Naturae per Regna Tria Naturae, Secundum Classes, Ordines, Genera, Species, cum Characteribus, Differentiis, Synonymis, Locis. Tomus I. Edit. Decima Reformata Stockholm, Salvius: i iii, Centuria Insectorum, Quam, Praesidae D.D Car. von Linne, Proposuit Boas Johansson, Calmariensis. Pp in C. Linnaeus, Amoenitates Academicae; seu Dissertations variae, physicae, medicae, botanicae, Antenae seorsim editae, nunc collectae and auctae. Vol. 6. Martin, J. W Notes and bibliography on the larvae of xanthid crabs, with a key to the known xanthid zoeas of the western Atlantic and Gulf of Mexico. Bulletin of Marine Science 34: , F. M. Truesdale, and D. L. Felder Larval development of Panopeus bermudensis Benedict and Rathbun, 1891 (Brachyura, Xanthidae) with notes on zoeal characters in xanthid crabs. Journal of Crustacean Biology 5: McLay, C. L Crustacea: Decapoda: The Sponge Crabs (Dromiidae) of New Caledonia and the Philippines with a review of the genera. Pp in A. Crosnier, ed. Résultats des campagnes MUSORSTOM, 10. Mémoires du Museum national d Histoire naturelle A new genus and species of dromiid crab (Brachyura, Dromiidae) from the Timor Sea, Northwest Australia with records of other species from the China Sea. Zoosystema 20: Crustacea Decapoda: revision of the family Dynomenidae. Pp in A. Crosnier, ed. Resultats des Campagnes MUSORSTOM, 20. Memoires du Museum national d Histoire naturelle 180. Miers, E. J Crustacea. Pp , in Report of the zoological collections made in the Indo- Pacific during the voyage of H.M.S. Alert, Part I. The collections from Melanesia. Part II. The collections from the Western Indian Ocean. London. Trustees of the British Museum.