44 



R. HUYS AND S. CONROY-D ALTON 



Goniopsyllus tenuis (Lubbock, 1860) comb. nov. 



Clytemnestra tenuis Lubbock, 1 860 



Lubbock's (1860) description is very incomplete and based on a 

 single specimen. The antennule was figured as 7-segmented but 

 comparison with other clytemnestrid descriptions indicates that the 

 author had erroneously shown the second segment as subdivided 

 into two distinct segments. The segmentation of the distal half of the 

 antennule conforms with the Goniopsyllus pattern, justifying its 

 placement in this genus. Giesbrecht (1892) regarded C. tenuis as a 

 likely synonym of G. rostratus but in the light of the discovery of 

 several closely related species we regard this course of action 

 premature. Conversely, Marques (1973) listed C. tenuis in the 

 synonymy of C. scutellata. Although Lubbock doubted the sexual 

 maturity of the holotype female this is contradicted by his state- 

 ments that the specimen was ovigerous and that the second and third 

 abdominal somites had almost completely coalesced (this being in 

 conflict with his illustration of a 6-segmented urosome lacking any 

 trace of a genital double-somite). With the scanty information 

 available it is extremely unlikely that C. tenuis will ever be recog- 

 nized; it is ranked here as species inquirenda. 



Sapphir rostratus Car, 1 890 



Conspecificity between S. rostratus, described from Trieste (North 

 Adriatic), and G. clausi, recorded from the South Adriatic (this 

 paper), seems conceivable on zoogeographical grounds. The rela- 

 tive lengths of the distal antennulary segments in both sexes and the 

 length of caudal ramus seta II, however, do not agree with those of 

 G. clausi. It is questionable whether these discrepancies are real or 

 reflect observation bias since Car's (1890) illustrations contain 

 other, more significant errors such as the P5 which is shown with 

 only 3 setae and the P4 which allegedly lacks an outer spine on the 

 distal endopod segment. A final obstacle to conspecificity is the 

 small size of S. rostratus which, based on the dorsal view of the 

 male, measures only 0.58 mm. Rather than proposing a new replace- 

 ment name in anticipation of potential secondary homonymy with 

 the type species, we maintain this species as species inquirenda 

 under its current name. If S. rostratus and G. clausi are conspecific 

 then the former becomes a invalid senior synonym of the latter. 



Other records 



Monard's (1928) description of 'C. rostratd' from Banyuls-sur-Mer 

 contains several inconsistencies such as his illustration of the P5 

 exopod which shows only 4 setae and his statement that the P2-P4 

 enp-3 setal pattern is 6-5-5, indicating that he has confounded P2 

 and P3. The author also claims that the male P5 is modified and the 

 female antennule 7-segmented. The small size (0.65 mm) seems to 

 rule out conspecificity with G. clausi. 



Chen et a/.'s (1974) record of G. rostratus from the East China 

 Sea and Mori's (1937) from Japanese waters are indeterminable on 

 the basis of the few illustrations provided. The short female P5 

 suggests a species different from G. rostratus. Similarly, Marques 

 ( 1 958) did not give convincing evidence for her record from Angola 

 since only the habitus of the male and body length measurements ( 9 

 : 0.4- 0.94 mm; 3: 1 mm) were provided. 



DISCUSSION 



Generic concepts and species discrimination 



The generic concepts of Goniopsyllus and Clytemnestra (as 



Goniopelte) introduced by Claus (1891b), but dismissed by subse- 

 quent authors, are reinstated here. Claus based the distinction on 

 differences in antennule segmentation and setation of the antennary 

 exopod, and on the presence or absence of sexual dimorphism in the 

 caudal rami. Goniopsyllus is clearly more advanced than 

 Clytemnestra, being illustrated by several reductions in the cephalic 

 appendages, PI and male P6 which provide additional discrepancies 

 between both genera. In Goniopsyllus the number of distal setae on 

 the antennary endopod is reduced (the missing elements being 

 marked by rudiments; arrowed in Fig. 21B), the armature of the 

 maxillule is represented by a single apical element, the distal 

 syncoxal endite of the maxilla bears only 2 elements and the long 

 syncoxal seta representing the proximal endite is lost. The latter 

 character should be used with caution in generic discrimination 

 since convergent loss of the proximal endite has happened in at least 

 one representative of Clytemnestra (Fig. 16E). All species of 

 Goniopsyllus lack the outer basal seta of PI and have lost the inner 

 seta of its exopod. The male sixth legs are weakly developed bearing 

 only 1 seta in Goniopsyllus (Fig. 11C) but are produced into con- 

 spicuous, elongate, trisetose processes in Clytemnestra (Fig. 1 1 A-B), 

 resembling the condition found in the Aegisthidae and Cerviniidae. 



Although Clytemnestra is the more primitive genus, it can be 

 readily identified by the absence of the outer spine on P2 exp- 1 . As 

 far as we could ascertain this is a unique character in harpacticoids 

 with a 3-segmented P2 exopod. The caudal ramus sexual dimor- 

 phism displayed only by Clytemnestra requires further ontogenetic 

 study before it can be considered a potential autapomorphy for the 

 genus. The typical caudal ramus condition found in the majority of 

 the Harpacticoida shows normally developed terminal setae IV and 

 V. In the Clytemnestridae this condition is exhibited only by the 

 males of Clytemnestra (e.g. Fig. 5B), the atypical female state (Fig. 

 5A) showing reduced setae. In contrast to swimming leg sexual 

 dimorphism which is nearly always the result of deviations in male 

 ontogeny, secondary sexual characters in the caudal rami are exclu- 

 sively expressed by the female, and as a rule are not expressed until 

 the final moult. This timing of expression has been demonstrated in 

 various families displaying caudal ramus sexual dimorphism, in- 

 cluding the Canuellidae, Cylindropsyllidae and Canthocamptidae. 

 In these families it is intrinsically linked with precopulatory mate 

 guarding where female caudal ramus modification shows substan- 

 tial congruence with male antennule morphology. Since the atypical 

 condition in female Clytemnestra is also found in both sexes of 

 Goniopsyllus - and thus unlikely to be the result of transformation at 

 the final moult - a different ontogenetic explanation must apply. 

 This is further corroborated by examination of early copepodids 

 (including Cop V 6) of C. asetosa and G. clausi which revealed 

 similarly reduced caudal setae in both species. The male caudal 

 setae in Clytemnestra must therefore undergo transformation at the 

 final moult. Hence, it is assumed here that reduction of setae IV-V 

 represents the ancestral state in the family and that elongation 

 evolved only secondarily in male Clytemnestra, not being linked to 

 mate guarding but possibly enhancing its capacity during mate 

 location. 



Examination of the genital field has revealed significant differ- 

 ences between both genera. In Goniopsyllus the copulatory pore is 

 located halfway down the genital double-somite in a large circular 

 depression (Fig. 27A) and connects via a strongly chitinized duct 

 with the anteriorly positioned seminal receptacles (Fig. 23C-D). In 

 Clytemnestra the copulatory pore is represented by a posteriorly 

 directed minute slit (arrowed in Fig. 27B), located between the 

 genital apertures far anteriorly on the genital double-somite, and a 

 copulatory duct is hardly differentiated (Fig. 5A). The polarity of 

 copulatory pore displacement is difficult to assess, however, outgroup 





