GORE and SCOTTO: LARVAL DEVELOPMENT OF CYCLOGRAPSUS INTEGER 



spinous (protopodal) process (i.e., type B) and 

 bearing at least 10 medial setae on the basis of 

 the first maxilliped. Wilson (1980) subsequently 

 demonstrated that Ewihirogra/psus larvae (Var- 

 uninae) have extremely shortened antennal exo- 

 pods (type C) plus only 8 basipodal setae, and 

 therefore are more allied to Grapsinae and Plag- 

 usiinae larvae than to those of the Sesarminae or 

 other Varuninae. As noted above, C. integer also 

 refute Rice's suggestion in regard to the Sesar- 

 minae, by having a type A antenna and by bear- 

 ing 9 (instead of 10) basipodal setae. Larvae of C. 

 cinereus also have 9 basipodal setae on maxilli- 

 ped 1, but these occur in a grouping different 

 from that seen in C. integer, C. lavauxi larvae 

 have 6 and C. insularum have 12 setae (Table 2). 



As to other features for distinguishing among 

 the larvae of Cyclograpsus, setation and arma- 

 ture of abdominal somites can be useful. Begin- 

 ning with the second (C. integer), third (C. ciner- 

 eus), or fourth zoeal stage (C. punctatus), 

 nonpaired, usually elongate or spinelike setae 

 are found on the posterodorsal margin of the first 

 abdominal somite. As development proceeds 

 these setae either increase in number (1, 3, 5, in 

 C. integer stages), or remain unchanged (3, C. 

 punctatus; 5, C. cinereus). Somite armature 

 shows similar diversity, with a hooklike spine or 

 knob on the second (C. cinereus), second and 

 third (C. punctatus, C. lavauxi), or second 

 through fourth somites (C. integer, C. insularum 

 first zoea). Regrettably, neither of these charac- 

 ters are specific for Cyclograpsus larvae because 

 they occur in other brachyuran zoeae and are 

 seen, for example, in the Goneplacidae(Carc*'»o- 

 plax, Lee and Hong 1970; Tritodynamia, Bouc- 

 quet 1965), as well as several other families less 

 closely related to the Grapsidae (Lebour 1928, 

 fig. 5, p. 483). 



The telsons in Cyclograpsus zoeae all seem 

 referrable to Aikawa's (1929) type B (i.e., with- 

 out supernumerary lateral spines, and typically 

 brachyuran in shape). The telson formula of 1+3 

 (= furcal spine, plus movable spiny seta; Gore 

 1979) changes in stage III to 1+4 in C. integer, C. 

 cinereus, and C. punctatus; the latter species, 

 however, adds an additional medial pair of setae 

 in stage V, becoming 1+5. Table 2 provides a 

 summary of all of these features. 



Megalopal Stage 



The megalopae of the three Cyclograpsus spe- 

 cies in which complete development is known 



differ substantially from one another and should 

 prove more easily separable than their respec- 

 tive larvae. The frontal region bears a strongly 

 deflexed rostral spine in C. punctatus (Fagetti 

 and Campodonico 1971), has a ventrally deflexed. 

 bluntly rounded rostrum with a median cleft in 

 C. integer, and is only slightly produced and 

 without a rostral spine in C. cinereus (Costlow 

 and Fagetti 1967). Other easily observed char- 

 acters not requiring dissection include terminal 

 setation on the telson, aesthetascs on the anten- 

 nule, exopodal setae of the third maxillipeds, 

 pleopods, and uropods. These, plus characters 

 requiring some dissection to observe, are sum- 

 marized in Table 3. 



None of the megalopal stages in any of the 

 three species considered resembles the juvenile 

 or adult crabs. Moreover, they do not exhibit 

 easily noticeable differences from many other 

 brachyuran megalopae, let alone grapsid mega- 

 lopae. In general, lack of rostral spines, or with 

 the rostrum only poorly developed, usually de- 

 flexed, and unarmed, is seen in many grapsid 

 postlarvae. Many of the species have the lower 

 ramus of the antennule appearing as a 1-seg- 

 mented, simple, palplike process (as in Chasmag- 

 nathus, Helice, Cyrtograpsus, and others, Cost- 

 low and Fagetti 1967), or even reduced to a 

 simple seta (Sesarma; Costlow and Bookhout 

 1962). But because of the paucity of descriptions 

 there is little use in attempting further classifi- 

 cation at this time. 



In the discussion above we have demonstrated 

 that several suggestions proposed by Rice (1980) 

 for classifying grapsid larvae can no longer be 

 considered useful. Although the distinctions 

 among the larvae of the subfamilies Grapsinae 

 and Varuninae, and Varuninae and Sesarminae 

 have become blurred, we nonetheless reiterate 

 the value of Rice's classification attempt, and 

 draw special attention to his key to the brachy- 

 uran families based on zoeal characters. By 

 using the characters he proposed, one may still 

 arrive within the Grapsidae using the key, pro- 

 vided that the subfamilial headings are disre- 

 garded. Rice's couplet 26 may then be modified 

 to read as follows: 



26. Carapace without lateral spines in all 



zoeal stages • • 27 



Carapace usually with lateral spines in 

 all zoeal stages or without only in first 

 zoeal stage 28 



519 



