FISHERY BULLETIN: VOL. 80, NO. 3 



carapace spines. No first stage grapsine zoeae, 

 no Sesarma (Sesarminae), nor Acmaeopleura (or 

 possible Gaetice) larvae in the Varuninae have 

 these spines (see summary in Wilson 1980). 

 Moreover, Pachygrapsus zoeae (Grapsinae), like 

 some larvae in Cyclograpsus, apparently lack 

 lateral spines in the first, but possess these in 

 subsequent, zoeal stages (Schlotterbeck 1976; 

 Bourdillon-Casanova 1960). On the other hand, 

 all known plagusiine larvae have lateral spines 

 from the first stage onward (Wilson 1980). 



Setation or spination on the ventrolateral cara- 

 pace margin is another widely shared feature 

 among grapsid genera. Examples include 

 Brachynotus (Bourdillon-Casanova 1960), Hemi- 

 grapsus (Kurata 1968), and Cyrtogr a ps : us (Scelzo 

 and Lichtschein de Bastida 1979) in the Varuni- 

 nae, Sesarma (Baba and Miyata 1971) in the Ses- 

 arminae, Leptograpsus (Wear 1970) in the Grap- 

 sinae, and Plagusia (Wilson and Gore 1980) in 

 the Plagusiinae. In many instances, however, 

 carapace and telson spine formulae differ sub- 

 stantially from that seen in larvae of Cyclograp- 

 sus, thereby allowing at least provisional separa- 

 tion among these zoeae. 



Regarding middorsal carapace setation, many 

 descriptions of brachyuran larvae either fail to 

 note its occurrence or do not allow judgment to be 

 made because of undetailed illustrations. A gen- 

 eral perusal of the literature available on grap- 

 sid larvae (Wilson 1980) shows that, in addition 

 to Cyclograpsus, only a few sesarmine genera 

 had this feature indicated, including Chasmag- 

 nathus (Boschi et al. 1967), Helice (Baba and 

 Moriyama 1972), and in the Varuninae Hemi- 

 grapsus (Hart 1935) and Cyrtograpsus (Scelzo 

 and Lichtschein de Bastida 1979). Several other 

 studies provide suitably detailed illustrations 

 which suggest that this character may be more 

 or less widespread among the zoeae of these sub- 

 families. However, these setae are apparently 

 absent in plagusiine and grapsine larvae, as far 

 as can be ascertained from the literature. Be- 

 cause these setae usually do not appear until 

 later zoeal stages (ZIII and beyond) their useful- 

 ness as an identifying character is somewhat 

 limited. 



One other carapace feature that seems note- 

 worthy, at least for the genus Cyclograpsus, is 

 the pterygostomian region of C. integer, this re- 

 gion is produced into a triangular, toothlike 

 prominence in the first stage, which becomes 

 more sharply pronounced as development pro- 

 ceeds. In C. cinereus, this prominence is always 



bluntly rounded until the last stage, when it be- 

 comes more acute. In C. punctatus the promi- 

 nence develops very slowly and apparently never 

 becomes acute. Only the first zoeal stages are 

 known in C. lavauxi and C. insularum (Wear 

 1970) and the prominence is not well developed 

 in either, being similar to that seen in C. puncta- 

 tus. Although the toothlike prominence is seen to 

 some extent in other grapsid zoeae, it does not 

 appear to be quite as prominent, based on the 

 illustrations provided in several studies. 



The type of antenna has always been consid- 

 ered an important classification feature in 

 brachyuran larvae (Aikawa 1929). Most brachy- 

 uran larvae have a type B antenna (i.e., exopod 

 about 0.5-0.75X the length of the protopodal 

 spine). This type is widely present throughout 

 the Grapsidae, being found predominantly in the 

 Sesarminae and Varuninae, but seen in only iso- 

 lated instances in either the Grapsinae or Plagu- 

 siinae. Nearly all Grapsinae have a type C an- 

 tenna (exopod substantially reduced in size to the 

 protopodal spine), an advanced character also 

 shared for the most part among the known larvae 

 of plagusiine genera (see summary in Wilson 

 1980). 



All Cyclograpsus larvae possess a type B an- 

 tenna, with the exception of C. integer, which has 

 a type A antenna (exopod and protopod about 

 equal). The type A antenna is considered to be 

 primitive (Aikawa 1929). The larvae of C. integer 

 are even more remarkable in having the anten- 

 nal protopodal spine and exopod both armed 

 along their respective lengths with rows of teeth, 

 in a manner similar to that seen in Eriocheir 

 zoeae (Varuninae; Aikawa 1929), and reminis- 

 cent of some antennae exhibited by larvae in sev- 

 eral xanthid genera (e.g., Scotto 1979). In other 

 Cyclograpsus zoeae, the exopod is entire, and 

 only the protopodal spine is so armed. Cyclograp- 

 sus integer is thus noteworthy for two exceptions: 

 1) an antenna of a form (i.e., doubly armed) 

 rarely noted within the Grapsidae, and 2) an an- 

 tenna type (A) found in no other zoeae of any 

 genus in the Grapsidae. 



Rice (1980) summarized the available knowl- 

 edge on the Grapsidae in a major paper dealing 

 with brachyuran zoeal classification. In attempt- 

 ing to delineate useful features among the four 

 subfamilies of grapsid crabs, he suggested that 

 the known zoeae of the Varuninae and Sesarmi- 

 nae might be distinguished from the Grapsinae 

 and Plagusiinae by always having a well-devel- 

 oped antennal exopod at least half as long as the 



518 



