GENERIC CONCEPTS IN CLYTEMNESTRIDAE 



45 



Table 1 Diagnostic characters of Clytemnestra species [Al = antennule; GDS = genital double-somite; AS = first adominal somite]. Length measure- 

 ments are based on material examined in this paper. 





scutellata 



gracilis 



farrani 



long i pes 



asetosa 



size 9 (in urn) 



1121 



1309-1562 



927-947 



9 



758-830 



size 6 (in urn) 



1064 



1420-1531 



939-945 



1211 



920 



cephalothoracic processes 



present 



present 



absent 



obsolete 



absent 



setal number segment 2 Al 9 



12 



12 



12 



? 



10 



proximal endite maxilla 



present 



present 



present 



present 



absent 



P2 exp-3 formula 



223 



223 



222 



222 



222 



P3 exp-3 formula 



323 



323 



323 



322 



322 



P4 exp-3 formula 



323 



323 



323 



322 



322 



setal number P5 exopod 91 S 



6 



6 



5 



5 



5 



P5 apex $ vs GDS posterior margin 



coinciding 



distad 



coinciding 



? 



proximad 



P5 apex $ vs AS posterior margin 



proximad 



coinciding 



proximad 



coinciding 



proximad 



spinules 2nd abdominal somite 9 



absent 



present 



present 



7 



absent 



spinules 1st abdominal somite 6 



absent 



absent 



absent 



present 



absent 



comparison with the Tegastidae, Peltidiidae and Tisbidae suggests 

 that migration happened anteriorly and the condition in Clytemnestra 

 is apomorphic. 



Species discrimination in Clytemnestra is most easily achieved 

 by comparing primarily cephalothorax shape, swimming leg spine 

 pattern, urosomal ornamentation and setation of the maxillae and 

 antennules (Table I). Conversely, identification of Goniopsyllus 

 species is strenuous and largely based on size, maxillipedal orna- 

 mentation and proportional lengths of caudal ramus setae. The 

 reported variability in body size and/or P5 setation for both C. 

 scutellata and G. rostratus (e.g. Boxshall, 1979; Huys et al.. 1996) 

 is based on erroneous identifications and observational errors. 



Relationships 



Prior to Claus' ( 1 89 la) study the relationships of the Clytemnestridae 

 were believed to lie with the planktonic poecilostomatoid families, 

 in particular the Sapphirinidae (Car, 1890). This concept was partly 

 based on the superficial similarity in dorsoventrally depressed body 

 shape, laterally displaced fifth legs and the failure to recognise the 

 geniculate antennules in the male (Car, 1890). More significantly, 

 this assignment was based also on the strongly reduced mouthparts 

 and the sexual dimorphism displayed by the maxillipeds, two 

 characters regarded as highly diagnostic for the Poecilostomatoida 

 (Huys & Boxshall, 1991). 



Sexual dimorphism in the maxillipeds is uncommon in the 

 Harpacticoida. Huys ( 1988) reviewed the topic, showing that there 

 is clear dimorphism only in the Aegisthidae (as a result of male 

 atrophy), some Tisbidae (e.g. Boxshall, 1979) and deepwater 

 Huntemanniidae (Metahuntemannia, Talpina). Dahms & Schminke 

 (1993) demonstrated that in Tisbe bulbisetosa the male maxilliped is 

 involved in precopulatory mate guarding by holding the female's 

 caudal setae IV and V prior to spermatophore transfer, the antennules 

 playing only an auxiliary role during this process. We speculate that 

 the modified male maxillipeds in clytemnestrids perform a similar 

 function, the elongate endopodal claw probably being involved in 

 holding the female's caudal rami or swimming legs. 



Boxshall & Huys (1998) pointed out that the antennulary chemo- 

 sensory system of C. rostratus (= G. clausi sp. nov.) is secondarily 

 enhanced in both sexes by transformation of three setae into 

 aesthetasc-like elements. The middle and distal of these elements 

 are fused basally to an aesthetasc. This study has revealed this 

 pattern to be diagnostic for all Clytemnestridae and can be consid- 

 ; ered an apomorphy for the family. Examination of copepodid stages 

 showed these transformed setae to be present from at least copepodid 



III onwards. Modification of antennulary elements into putative 

 chemosensors is rare in harpacticoid copepods and has thus far only 

 been recorded in some deep-sea species. Gee & Huys (1991) 

 described a densely opaque, bulbous element on the distal antennulary 

 segment in both sexes of the paranannopid Leptotachidia iberica 

 Becker, 1974. The only report of a similar structure is that by Por 

 (1969) who figured a modified bulbiform element on the antennule 

 of Cerviniopsis obtusirostris Brotskaya, 1963 (Cerviniidae) which 

 he called the 'Brodskaya organ'. 



The complete lack of swimming leg sexual dimorphism impedes 

 an assessment of the relationships of the Clytemnestridae. The 1- 

 segmented PI exopod is found in several interstitial Paramesochridae, 

 Leptastacidae and Laophontidae, yet it is diagnostic at the family 

 level only in the Rotundiclipeidae and Tegastidae. Lang (1948) 

 recognised a close relationship between the latter, the Peltididiidae 

 and the Clytemnestridae. He based this affinity solely on PI mor- 

 phology, including the non-prehensile nature of the endopod and the 

 presence of maximum 5 elements on the distal exopod segment. 

 Within this group of tisbidimorph families he placed the Peltidiidae 

 as the sistergroup of the Clytemnestridae on account of the dorso- 

 ventrally flattened body and the reduction of the P5 baseoendopod 

 in the female. The usefulness of Lang's (1948) characters is limited 

 due to their homoplastic nature, however, there are at least two other 

 features which appear to substantiate a close relationship between 

 these three families. First, the aesthetasc pattern on the male antennule 

 (with an additional aesthetasc on ancestral segment XI) is displayed 

 by all three families. Secondly, the modification of the distal palmar 

 element on the maxillipedal basis into a pad-like sensory element 

 (Fig. 10B) is a unique synapomorphy (see Huys et al. (1996) for 

 examples in Peltidiidae and Tegastidae). A detailed phylogenetic 

 analysis of the Peltidiidae is nevertheless required before its 

 sistergroup relationship with the Clytemnestridae can be substanti- 

 ated. Indeed, an alternative evolutionary scenario could be that the 

 latter represent only a specialized terminal branch of the former. 

 Most species of the peltidiid genus Alteutha Baird are common 

 members of the coastal plankton, performing pronounced diurnal 

 vertical migrations in the water column. This may well be viewed, 

 either ecologically or evolutionary, as a transitionary step towards 

 the holoplanktonic lifestyle exhibited by the Clytemnestridae. 



'Taxonomic Impediment' and Marine Plankton 



The present revision has quadrupled the number of species in the 

 family solely by examination of the relatively limited material 

 deposited in the NHM. There is no doubt that this number would 



