FISHERY BULLETIN; VOL. 87, NO. 3. 1989 



endopod of the 2nd antenna has been used for 

 generic identification of protozoea (Cook 1966a; 

 Hassan 1974; Haq and Hassan 1975). While this 

 character is often referred to as the setal formula 

 of the 2nd antennal endopod, this is not correct 

 as the first seta referenced arises from the proto- 

 pod. 



We compared the 2nd antennal formulae of 

 protozoeae in our reference collection with pub- 

 lished formulae. All Penaeus larvae in our collec- 

 tion have the formula 1-t- 1-1-2, in agreement with 

 other published descriptions. The three species 

 of Trachypenaeus in our collection (7". an- 

 choralis, T.fulvus, and T. granulosus) have the 

 formula 0-1-2-1-2, which is in agreement with the 

 only existing pubhshed descriptions by Kirkega- 

 ard (1969) and Pearson (1939). Parapenaeopsis 

 is represented in our reference collection by the 

 protozoeae of P. conuda, which also have the 

 formula 0-1-2-1-2. This confirms the formula found 

 by Hassan (1984) for P. stijlifera. Rao (1973) also 

 described larvae of P. stylifera, and his figures 

 seem to indicate the same formula of 0-1-2-1-2, 

 although no detailed drawings of appendages are 

 provided. 



However, the 2nd antennal formula is not con- 

 sistent in all genera. There are five species of 

 Metapenaeus in our refei'ence collection: M. ben- 

 nettae, M. eborocensis, M. endeavouri, M. 

 ensis, and M. insoUtus. These larvae normally 

 have the formula 1-1-2-1-3, but on some occasions 

 the formula is 1-1-2-1-2. This difference occurs at 

 random (although rarely) among most species. 

 Menon (1951), Moms and Bennett (1951), Raje 

 and Ranade (1972), Kurata and Vanitchkul 

 (1974), Courties (1976), and Hassan (1980) 

 agreed that Metapenaeus have H-2-1-2, while 

 Vanitchkul (1970), Hassan (1974), and Haq and 

 Hassan (1975) found both 1-^2-^2 and 1-^2-^3. In 

 our larvae, the third seta in the most distal 

 group was often small and hard to distinguish 

 when present, and could easily have been over- 

 looked. Similarly, the 2 species of Metapenae- 

 opsis we have reared are not consistent. Meta- 

 penaeopsis palmensis is always 1-1-2-1-2, but M. 

 7iovaeguinae, although normally exhibiting the 

 same formula, occasionally has H-2-1-3. 



In the mysis stage, the presence of a serrate 

 anteroventral carapace margin (Fig. llbi) is a 

 character that has not previously been described 

 for laboratory-reared penaeid larvae. At present 

 the only genus in which it is definitely present is 

 Metapenaeopsis — both M. pabnensis and M. 

 novaeguinae from our reference collection. 



The number and placement of thoracic sternal 



spines have not been mentioned in previous 

 studies. In cleared and stained specimens it is an 

 easy character to assess and makes postlarval 

 identification straightforward, and it is an im- 

 portant generic character in our key for post- 

 larvae. Unfortunately these spines have not 

 been described by other workers and so we have 

 no information about the sternal spine formulae 

 for genera that are not represented in our refer- 

 ence collection. Partly owing to this we have 

 used other characters to help identify some gen- 

 era; however, when both the size and position of 

 the sternal spines are taken into account, they 

 are sufficient to identify five genera: Atypo- 

 penaeus, Metapenaeopsis, Metapenaeus, 

 Penaeiis, and Trachypenaeus (Fig. 15). Within 

 the genus Penaeus, the sternal spine formula 

 has taxonomic value at the subgenus level. We 

 have examined postlarvae from 3 species within 

 the subgenus Litopenaeus {P. setiferus, P. 

 stylirostris, and P. vannamei), all of which have 

 only a single, small sternal spine on the 4th seg- 

 ment (formula O-l-O-l-O-l- 1-1-0). All other Penaeus 

 examined have a large spine on the 4th segment 

 and a small spine on the 5th (formula 0-1-0+ 

 0-1-1 -1-1): P. aztecus from the subgenus Farfan- 

 tepenaeus; P. indicus and P. merguieyisis from 

 the subgenus Fenneropenaeus; P. japonicus 

 from the subgenus Marsupenaeus; P. lati- 

 sulcatus, P. longistylus, and P. plebejus from 

 the subgenus Melicertus; and P. esculentus, P. 

 monodon, and P. semisulcatus from the sub- 

 genus Penaeus. Species of Litopenaeus do not 

 occur in the Indo-west Pacific region and so will 

 not be wrongly identified by our key. 



We have not used the distribution of ab- 

 dominal spines in protozoea III or mysis sub- 

 stages (e.g.. Cook 1966a; Muthu et al. 1978), 

 since their presence can be variable, at least for 

 M. pabnensis. Paulinose (1977) suggested that 

 the dentition and asymmetry of the mandibles 

 might be useful for both generic and specific 

 identification of penaeid larvae. We have not 

 investigated this possibility since the operational 

 use of this character would not be practical due 

 to the time-consuming dissections necessary. 



Key to Penaeid Genera 



Our key was constructed from both our own 

 reference collection and published descriptions. 

 Many species were referred to in characterizing 

 the genera Penaeus, Metapenaeus, and Trachy- 

 penaeus. However, there is much less reference 

 material available for other genera. Although 



724 



