Zoeal stages of Pseudomicippe varians Miers, 1879 (Decapoda: Brachyura: Majoidea: Majidae) and a comparison with other Majidae larvae

Pseudomicippe varians Miers, 1879 is a majid crab recorded from Western Australia (Shark Bay) and northern Queensland. The zoeal stages are described from laboratory reared material. The zoeal stages of P. varians can be easily distinguished by the absence of carapace spines and extremely large mandibles. These characters are likely diagnostic among majoideans in general. Additionally, recent phylogenetic studies of majoids using larval characters showed the Majidae as one of the few families for which there is larval support for its monophyly. Furthermore, based on the monophyly of Majidae and the morphology of P. varians, a set of characters is established that could be used as a diagnostic for majids in general.

Within this family, the genus Pseudomicippe Heller, 1861 is represented by 11 species distributed throughout the Indo-west Pacific region. Pseudomicippe varians Miers, 1879 occurs in Western Australia (Shark Bay), and northern Queensland (Griffin & Tranter 1986).
The purpose of this is to describe the zoeal stages of P. varians, which comprise the first larval description for the genus. In addition, the zoeal information available in the literature is used to compare the findings of this study with those of other species previously described for this family.

Larval development and description
A single ovigerous specimen of Pseudomicippe varians was collected in August 2001 on seaweed beds, in Heron Island, Queensland, Australia (23u279S, 151u559E). The specimen was held in an aquarium in a temperature-controlled room (24¡1uC) until hatching, which occurred at night. After hatching, 50 of the most active, positively phototactic larvae were placed individually into 100 ml acrylic jars containing 50 ml of filtered seawater. The remaining larvae were kept in mass culture as extra specimens to be used for morphological description.
Newly hatched larvae were fed ad libitum with Artemia nauplii. Sea water was changed, and larvae were inspected and fed daily. All acrylic jars were washed in fresh water and airdried before re-use with fresh seawater on the following day. Average salinity was 32 psu. A natural photoperiod was maintained (%14L:10D).
Whenever possible, a minimum of five specimens of each stage were dissected for morphological description. For slide preparations polyvinyl lactophenol was used as the mounting medium with acid fuchsin and/or chlorazol black stains. The description of setae follows Pohle and Telford (1981), but here includes only analysis by light microscopy (LM), using an Olympus BX-51 microscope with Differential Interference Contrast and drawing tube. Some of the setae designated as plumose herein may be plumodenticulate setae because of the lower resolution limits of LM as compared to scanning electron microscopy (SEM). Description guidelines of Clark et al. (1998) were generally followed.
Specimens of larval stages and a spent female crab have been deposited at the Museu de Zoologia da Universidade de São Paulo, São Paulo, Brazil, accession number 17281.

Larval development and description
Zoeal development of Pseudomicippe varians consists of two zoeal stages. The duration of the first zoeal stage was 4-11 days (3.9¡0.3), 5-12 days (5.0¡0.6) for the second stage. It was not possible to obtain megalop stages because of fungus infection in our culture that arrested the development of the second zoeal stage leading to death. Only morphological changes are described for the second zoea. Carapace ( Figure 1A). Dorsal, rostral, and lateral spines absent. Ventral margin posterior to scaphognathite notch with small plumodenticulate seta preceding densely plumose ''anterior seta'', followed by five plumose setae. Eyes sessile. Small indistinct median ridge frontally and a small median tubercle on posterodorsal margin, each bearing cuticular dorsal organ (sensu Martin and Laverack 1992). Pair of simple setae present on posterodorsal margin.
Antenna ( Figure 1C). Biramous, spinous process of the protopod pointed, bearing two rows of sharp spinules; endopod bud present with half of the exopod length; onesegmented exopod shorter than protopod with two subterminal setae with similar size.
Mandible ( Figure 1D). With medial toothed molar process and much enlarged lateral incisor processes, as long as antenna. Palp absent.
Telson ( Figure 1J). Bifurcated, distinct median notch, three pairs of plumodenticulate setae on inner margin; each furcal shaft proximally bearing lateral spine, furcal shafts and spines covered in rows of spinules to just below tips. Grouped denticulettes present.

Taxonomic grouping
The zoeae of Pseudomiccipe varians conform to previous characterizations of this phase for Majoidea (Rice 1980(Rice , 1983 in having nine or more setae on the scaphognatite of the maxilla, and well-developed pleopods in the second zoeal stage. However, P. varians does not share the characters proposed by Ingle (1979) for the family Majidae, which he divided into two groups according to morphological resemblance. Ingle's (1979) larval classification of the Majidae (i.e. his Majinae) included species of the genera Maja and Schyzophrys.
The zoeae of the group I taxa are diagnosed by the presence of lateral carapace spines; a dorsal spine that is often well developed and usually of moderate length; a prominent rostral spine; dorso-lateral processes on abdominal segments two and three; and posterolateral processes on abdominal somites 3-5 that are prominent and sometimes long. The group II, which comprises the genera Leptomithrax and Acanthophrys, is characterized in zoeal stages by the absence of lateral carapace spines; dorsal spine sometimes reduced or absent; rostral spine sometimes reduced; dorso-lateral processes only on abdominal somite two; and posterolateral processes on abdominal somites 3-5 that are not prominent and sometimes short. The zoeal stages of P. varians agree with Group II in having no dorsal or lateral carapace spines; a dorsolateral process that is restricted to abdominal somite 2; posterolateral processes on abdominal somites 3-5 being short; the basal segment of maxilliped 2 with no more than three marginal setae; and the antennal exopod having terminal setae. However, P. varians differs substantially in lacking a rostral spine and in having only one spine on each telson fork (Tables I and II). Among the Majidae the zoeal stages of Pseudomiccipe varians can be easily distinguished by lacking rostral, dorsal, and lateral spines, the relative size of the mandible that is as long as the antenna, and the number of furcal spines (Tables I and II). The absence of carapace spines and very large mandible are especially distinctive and likely diagnostic among majoideans in general.

Larval comparison and taxonomic affinities among species of Majidae
Within Majidae, a considerable number of workers have addressed the larval descriptions for species of Maja, Schizophrys, Jacquinotia, Notomithrax, and Leptomithrax. However, some of these accounts are not suitable for comparisons because they lack descriptions of appendages or stages, as is the case for the descriptions of Schizophrys aspera (Milne-Edwards, 1834) (cf. Kurata 1969;Tirmizi & Kazmi 1987), Leptomithrax edwardsi (de Haan, 1839) and L. bifidus Ortmann, 1893(cf. Kurata 1969, and L. longimanus Miers, 1876(cf. Webber & Wear 1981. Descriptions of the three species of majids by Kurata (1969) lack zoeal information such as on the maxillule, maxilla and other possibly distinctive appendages. The description of Tirmizi and Kazmi (1987) for S. aspera was included in Tables I and II, but the data should be viewed with caution because of considerable disparities between illustrations and descriptions. Due to the inaccurate or incomplete nature of these older descriptions, the comparisons we make are based on the most recent larval descriptions or those we consider to provide adequate morphological information.
The genus Jacquinotia, represented by J. edwardsii (Jacquinot 1853), was considered as a member of the family Mithracidae by Webber and Wear (1981). However, Griffin and Tranter (1986) placed this monotypic genus within the family Majidae based on several adult characters, and we followed this taxonomic placement here since the larval characters for this genus match the pattern observed for Majidae. Additionally, the genus Eurynome Leach, 1814, traditionally viewed as a member of the family Pisidae (Sakai 1965(Sakai , 1976Griffin 1966;Ingle 1980Ingle , 1991Salman 1982;Aiyun & Siliang 1991;Santana et al. 2004), was placed in the family Majidae by Hale (1927Hale ( -1929 and by Griffin and Tranter (1986). In this case we do not follow Hale's or Griffin and Tranter's approach, leaving the genus as a member of Pisidae based on the differences with the majid larval characters. This is largely based on the setation of the maxilla, first maxilliped, and other appendages (Tables I  and II). The genus Acanthophrys A. Milne-Edwards, 1865, traditionally considered as a majid (Griffin 1966;Kurata, 1969;Ingle 1979;Clark 1986), was placed in the family Pisidae by Griffin and Tranter (1986). However, due to the lack of larval information we could not make any assertion on the position of the genus, and we do not include Acanthophrys in the comparisons.
Based on morphological comparisons (Tables I and II), several characters appear to be diagnostic for some majid genera. For species of Maja, the presence of rostral, dorsal and lateral spines (also in Leptomithrax longipes (Thomson 1902)), and dorsolateral processes on the second and third abdominal somites in both zoeal stages are diagnostic for this genus. For Schizophrys, the setation of the basial endite of the maxillule in zoea I, and the setation of the basis of the first maxilliped, and basial endite of the maxilla in both zoeal stages are

Phylogenetic relationships
In recent phylogenetic analyses of majoids using larval characters (Marques & Pohle 1998;Pohle & Marques 2000), the family Majidae appears to be one of the few families that have larval support for its monophyly. The authors base the monophyly of Majidae on the exopod of the antenna bearing a well-developed terminal spine, half or more the length of apical setae but not extending beyond the tip of the setae in the zoea; proximal coxal lobe of the maxilla in zoea II bearing three setae; scaphognathite of maxilla bearing 21-28 setae in zoea II; and the fork of the telson bearing three lateral spines (Tables I and II). Except for the presence of one rather than three telson furcal spines in P. varians and S. aspera, and the number of setae on the scaphognathite of the maxilla of zoea II in N. minor (Filhol 1885), the other majid species are in agreement with the synapomorphies determined by Pohle (1998, 2003), and Pohle and Marques (2000) for Majidae. So we believe that those characters cited above, based on a cladistic analysis, could better define this family, and should be used instead of the Ingle's (1979) groups.