Contributions to Zoology, 73 (3) 165-205 (2004) SPB Academic Publishing bv, The Hague Carcinization in the Anomura - fact or fiction? II. Evidence from larval, megalopal and early juvenile morphology Patsy+A. McLaughlin Rafael Lemaitre² & Christopher+C. Tudge² ¹, 1 Shannon Point Marine Center, Western Washington University, 1900 Shannon Point Road, Anacortes, 2 Washington 98221-908IB, U.S.A; Department ofSystematic Biology, NationalMuseum ofNatural History, Smithsonian Institution, P.O. Box 37012, Washington, D.C. 20013-7012, U.S.A. Keywords: Carcinization, Anomura, Paguroidea, Lithodidae, Paguridae, Lomisidae, Porcellanidae, larval, megalopal and early juvenile morphology, pleonal tergites Abstract Existing hypotheses 169 Developmental data 170 Results 177 In this second carcinization in the Anomura ofa two-part series, From hermit to king, or king to hermit? 179 has been reviewed from early juvenile, megalopal, and larval Analysis by Richter & Scholtz 179 perspectives. Data from megalopal and early juvenile develop- Questions of asymmetry- 180 ment in ten ofthe Lithodidae have genera provided unequivo- Pleopod loss and gain 18! cal evidence that earlier hypotheses regarding evolution ofthe Uropod loss and transformation 182 king crab erroneous. of and pleon were A pattern sundering, - Polarity or what constitutes a primitive character decalcification has been traced from the megalopal stage through state? 182 several early crabs stages in species ofLithodes and Paralomis, Semaphoronts 184 with evidence from in other supplemental species eight genera. Megalopa/early juvenile characters and character Of major significance has been the attention directed to the states 185 inmarginallithodidsplatesareofnotthehomologoussecond pleomere,with thewhichadult whenso-calledseparated“mar- Cladistic analyses 189 Lomisoidea 192 ginal plates” ofthe three megalopal following tergites. Auxiliary Porcellanidae 193 and early juvenile lithodid data, aswell as equivalent data from Megalopal and early juvenile phase 193 other the direction indicated paguroids, support evolutionary Zoeal phase 194 by lithodid while car- pleonal plate development. Therefore, Conclusions 194 cinization, or of a crab-like form, has oc- development body Acknowledgements 196 curred the it has from hermit in Lithodidae, not proceeded a References 196 crab the data the thus ancestor. Rather suggest reverse, effec- tively refutingthe “hermit to king” myth. Brief reviews of data available from the Lomisidae and Porcellanidae support the Introduction Proposition of independent anomuran carcinization events in these taxa as well. Results of cladistic analysis of megalopal and juvenile somewhat do not In Part I of review of carcinization data, although unconventional, a (the process support the lithodid claim ofa sister-group relation of the gen- of developing a crab-like body form) in the Ano- era Lithodes and with the Paralithodes pagurid genus Pagurus. mura, McLaughlin & Lemaitre (1997) provided a Attempts to subject larval phase data to similar analysis were historical review of instances of this thwarted reported phe- by the tendency in paguroids, including lithodids, for nomenon andexamined evidence provided by adult lecithotrophic development.Additionally, presumed initial and terminal the stage deletions disallow the ontogenetic stage homologies morphology throughout Galatheoidea, Hippoi- required for meaningful phylogenetic results. dea, Lomisoidea, and Paguroidea (classification of Martin & Davis 2001), with particular emphasis & on the Paguroidea. McLaughlin Lemaitre (1997) Contents rejected the traditional concept ofpagurid carciniza- tion, which had consistently maintained that the Introduction 165 lithodid crab-like body form evolved from a typical Carcinization: an evolutionary reality? 169 shell-dwelling hermit crab ancestor (Boas 1880a, Downloaded from Brill.com10/08/2021 10:07:18PM via free access 166 P.A. McLaughlin et al. - Carcinization in Anomura, Part II Bouvier Borradaile I The is followed series b, 1924; 1894a, b, 1895, 1897; c, g). megalopa directly by a 1916; Wolff 1961; Cunningham et al. 1992; Gould of juvenile stages (Figs. Id, h). 1992; Richter & Scholtz 1994). The conclusion There is ongoing debate among developmentalists the reached by McLaughlin & Lemaitre (1997) was that as to whether megalopa is correctly considered the ofthe larval although the acquisition of a crab-like body form last stage phase (Williamson 1969, had to be acknowledged, the hypothesis that the 1982; Gould 1992; Harvey 1998; Clark & Calazans such in had & Provenzano evolution of a form the Paguroidea pro- 1999), a single postlarval stage (Goy the ceeded from a shell-dwelling hermit was fallacious. 1978, 1979), or first stage of the postlarval phase, As an alternative hypothesis, they argued just the i.e., preceding the juvenile stages (Lebour 1934; reverse, i.e., the 1 ithodid crab-like body form had Gurney 1942; Gore 1985; Kaestner 1970; Felder given rise to the simple hermit crab body form et al. 1985; Paul et al. 1993). Zoeal morphology is calcium habitat and through loss, change conse- dramatically transformed with the molt to the me- quential morphological adaptations. Although their galopa. The animal assumes a body shape adapted based characters original hypothesis was on derived more for walking or crawling than for swimming, least still from adult morphology, they suggested that devel- although many megalopae at initially are opmental data would support their view. Borradaile quite capable of swimming. However, locomotor the and/or the (1916) had theorized that carcinization was not functions are transferred to pleopods simply “happen chance” but rather that certain de- thoracic ambulatory legs, while the maxillipeds, the converted capod crustaceans possessed the genetic constitu- swimming appendages of the zoea, are carcinizationwhile others did into ents for not. Similarly, feeding appendages. Morphologically, progres- al. evidence sion from to in- Cunningham et (1992), referring to put megalopa young adult, although of forward by MacDonald et al. (1957) stated that volving a series molts, is usually a relatively and molt heterochronic modifications did not appear until gradual, not always a controlled, transi- metamorphosis. McLaughlin & Lemaitre (1997) tion. rationalized that if such the For the how- was really case, one many viewing megalopa as larval, might expect that those anomurans destined to ex- ever, the metamorphic molt is interpreted as oc- hibit aspects of carcinization would exhibit corol- curring between the megalopa and the first juvenile lary larval, megalopal and early juvenile attributes stage (Dawirs et al. 1986; Anger 1996; Gebauer et possibly similar to those of true crabs. al. 2002; Charmantier et al. 2002). This phenom- If is carcinization perceived as the development enon would appear to represent a more physiolo- by an organism of a brachyuran crab-like body form, gical than morphological metamorphosis. Harvey only three major anomuran taxa need be investi- (1996) however, reported two metamorphic molts gated from the developmental view point, i.e., the in three species of hermit crabs, the first occurring monotypic Lomisoidea, the galatheid family Por- between the last zoea and the megalopa, and the cellanidae, and the superfamily Paguroidea. Like second occurring between the megalopa and first brachyurans, anomuran larvae most commonly hatch crab stage. Anger (2001) similarly defined the tran- from the parental egg as a zoea, a swimming larval sition from zoea to megalopa and from megalopa form with several For the of provided carapace, rostrum, pairs to juvenile as metamorphic. purposes of cephalothoracic appendages, antennules, anten- the present discussion, we consider the primary nae, mandibles, maxillules, maxillae, and first and metamorphosis occurring between the last zoea and second maxillipeds, elongate pleon of five pleomeres the megalopa as the focal point, and in our discus- of refer and a telson (Figs, la, b, e, f). These larvae usu- sion semaphoronts, we to stages subsequent zoeal this molt Addition- ally pass through one or more stages, during to metamorphic as “postlarval”. which time a sixth pleomere ordinarily is deline- ally, we have elected to use the currently more ated, uropods most frequently develop, and pleopod commonly recognized term megalopa, rather than buds appear, before a metamorphic molt to the ben- glaucothoe of numerous earlier authors (Gurney thonically adapted megalopa (decapodid of Kaest- 1942; MacDonald et al. 1957; Kurata 1956; Pro- ner 1970; Felder et al. 1985; Anger 2001) (Figs. venzano 1971, 1978; Bookhout 1972; Gore& Scotto Downloaded from Brill.com10/08/2021 10:07:18PM via free access Contributions to Zoology, 73 (3) - 2004 167 Fig. I. Larval, and megalopal early juvenile stages of Paguridae (a-d) and Lithodidae (e-h). a-d, Pagurus kennerlyi (Stimpson, 1864): a, zoea 1, dorsal dorsal view; b, zoea IV, view; c, megalopa, dorsal view; d, first crab stage, dorsal view, e-h, Lopholithodes mandtii 1848: Brandt, zoea dorsal e, I, lateral view; b, zoea IV, lateral view; c, megalopa, view; d, first crab stage, dorsal view, (a-d, modified from al. & McLaughlin et 1989; e, f, modified from Crain McLaughlin 2000; g, h, specimens used by Crain & McLaughlin 2000, (USNM 1006124). Not to scale. 1983; Forest 1987; Nakasone 1988; Gould 1992), tensive; although the majority of these descriptions decapodid of Kaestncr (1970), Felder, et al. (1985), have been limited to zoeal and megalopal develop- and