42 



R. HUYS AND S. CONROY-DALTON 



pinnate setae and 3 pores; inner margin with long plumose seta; apex 

 with 1 pinnate and 1 plumose seta. 



MALE. Unknown. 



REMARKS. Upon re-examination Boxshall (1979) concluded that 

 the holotype, identified by Brady (1883) as a male, was in reality 

 female. The true sexual identity however, had already been noted by 

 both Poppe (1891) and Claus (1891 a-b) who based their conclusion 

 on the 5-segmented urosome and the female facies of the antennule 

 and maxilliped. This opinion was also confirmed by Giesbrecht 

 (1892) but not by Car (1891b) who continued regarding it as a male 

 on the basis of the internal spermatophore drawn by Brady. The most 

 plausible explanation is that Brady (1883) had misinterpreted the 

 strongly chitinized copulatory duct, a suspicion reinforced by in- 

 spection of the holotype. 



Giesbrecht (1892: 573) pointed out the discrepancy between the 

 size mentioned in Brady's text and that inferred from his habitus 

 figure reproduced at x80 magnification. According to Brady the 

 holotype is only 0.65 mm long (' l-40th of an inch') but Giesbrecht 

 considered 1.16 mm a more realistic figure. Re-examination of the 

 slides strongly suggests that Brady must have made a morphometric 

 error of at least a factor 2. The urosome (excl. P5-bearing somite) 

 which is mounted intact measures 0.43 mm. Extrapolation by using 

 the urosome/body length ratio found in its congeners G. clausi and 

 G. brasiliensis (about 0.3) gives an estimated total body length of 

 1.43 mm. This large size rules out possible conspecificity with G. 

 brasiliensis (x = 0.96 mm). 



Brady (1883) assumed all four swimming legs to be similar, 

 having 3-segmented rami and resembling the leg illustrated in his 

 Fig. 15 (i.e. the P2). His lateral habitus view suggests that the PI 

 possesses 3-segmented exopods and endopods, however Poppe 

 ( 1 89 1 ) suspected that Brady had overlooked the exopod and instead 

 had superimposed both left and right endopods. For some unknown 

 reason he assumed the PI exopod to be 2-segmented but failed to 

 confirm this against the holotype due to the absence of the PI on 

 Brady's slide. 



G. rostratus can be readily identified from the other South- 

 American species G. brasiliensis by the large body size (compare 

 urosomes in Fig. 28A-B drawn at the same scale), the elongate 

 caudal ramus setae IV- V, the long seta I clearly extending beyond 

 the distal margin of the ramus, and additional differences in the 

 ornamentation of the maxilliped (spinule pattern on palmar margin). 

 Brady (1883) also illustrated well developed posterolateral exten- 

 sions on the cephalothorax which are completely absent in G. 

 brasiliensis. 



Goniopsyllus brasiliensis sp. nov. 



? Clytemnestra rostrata (Brady, 1883) sensu Ramirez (1966): 291; 

 Lam. II, figs 12-15. 



TYPE LOCALITY. Rio Grande do Sul (Brazil); outside opening of 

 Lagoa dos Patos to ocean; 32°11'S 52°7'W. 



Etymology. The species name refers to the type locality. 



Type material. Holotype 9 dissected on 8 slides (BMNH 

 1999.1056). Paratypes are 8 9? in alcohol (BMNH 1999.1057- 

 1064). Collected by G.A. Boxshall, February 1996, plankton haul. 



Description. 



female. Total body length from tip of rostrum to posterior margin 

 of caudal rami: 892-1057 urn (x = 958 urn; n = 8). Maximum width 

 (265 urn) measured at posterior margin of cephalic shield. Postero- 



lateral angles of cephalothorax rounded, virtually not expanded 

 laterally (Fig. 29A). Rostrum (Fig. 29A) rounded and less pro- 

 nounced than in G. clausi. Backwardly produced alate processes of 

 somites bearing P2-P4 distinctly shorter and less pointed than in C. 

 clausi. Integument generally less chitinized than in G. clausi. 



Genital double-somite (Fig. 28B) not constricted bilaterally and 

 relatively wider than in G. clausi; original segmentation marked by 

 minute, paired, chitinous patches ventrally. Genital field as in G. 

 clausi but with additional pores flanking copulatory pore (Fig. 28C). 



Urosomites with zone of small denticles around dorsal hind 

 margin (Fig. 29B); penultimate and anal somites also with larger 

 spinules around ventral hind margin (Fig. 28C). 



Caudal rami (Figs 28B; 29A-C) short, convergent. Setae I-II 

 bipinnate, spiniform and strongly developed; seta 1 1 .2 times as long 

 as seta II, not extending beyond apex of caudal ramus. Seta III 

 minutely bipinnate. Setae IV-V basally fused, multipinnate and 

 about as long as in G. clausi but seta IV more resilient (compare Fig. 

 23B); seta V about 1 .5 times ramus length. Seta VI extremely small; 

 seta VII Particulate at base, bare. 



Antennule (Fig. 28D) slender, 6-segmented; segment 2 shorter 

 than in G. clausi but armature pattern identical. 



Mandible and maxillule (Fig. 29D) somewhat more slender than 

 in G. clausi. 



Maxilliped (Fig. 29E-F) with similar armature as in G. clausi but 

 with different spinular ornamentation on palmar margin (Fig. 29F). 



P1-P4 with setal formula as for genus. 



P5 (Fig. 28B) markedly longer than in G. clausi, extending 

 beyond distal margin of genital double-somite. 



MALE. Unknown. 



REMARKS. Although many South- American authors have recorded 

 specimens that they attribute to C. rostrata, there is good reason to 

 believe that in fact often they have mistaken G. brasiliensis for this 

 species. In general, with the discovery of G. brasiliensis many of the 

 Brazilian records of G. rostratus are rendered doubtful (Bjornberg, 

 1963; Bjornberg et al, 1981; Campaner, 1985; Carvalho, 1944; 

 Gaudy, 1963; Montu, 1980; Montu & Gloeden, 1986; Montii & 

 Cordeiro, 1988; Santos, 1973; Vega-Perez, 1993). The same applies 

 to Legare's (1961, 1964) records of C. rostratus from Venezuelan 

 coastal waters. The species illustrated by Ramirez (1966) as C. 

 rostrata from Mar del Plata in Argentina differs from the one figured 

 in his later paper (Ramirez, 1970) by the complete absence of 

 posterolateral projections on the cephalothorax and is almost cer- 

 tainly conspecific with G. brasiliensis. The author described the 

 female antennule as 7-segmented but this clearly contradicts his 

 illustration which shows only 6 segments as in other species of 

 Goniopsyllus. The only anomaly remaining is the body size which 

 according to Ramirez (1966) is 1.8 mm for the female and 1.5 mm 

 for the male. Based on his illustrations and the accompanying scale 

 bars the female only measures 0.74 mm and the male 0.77 mm. 



It is not clear whether Carvalho's (1952) material of C. rostrata, 

 consisting of 5 males from the Bay of Santos (Sao Paulo State), also 

 belongs to C. brasiliensis. His size range (0.50-0.85 mm) precludes 

 possible identity with C. rostratus but the illustrations accompany- 

 ing the brief description are completely worthless and erroneous. 

 The caudal rami are exceptionally long for this genus, the P5 exopod 

 has only 4 elements, and the antennule is 8-segmented. The speci- 

 mens reported from Guaratuba (Parana State) in an earlier paper 

 (Carvalho, 1944) are also very small (0.5 mm) and their fragmentary 

 description is equally useless for identification purposes. 



Finally, there is no possibility of identifying any specimens from 

 Campos-Hernandez & Suarez-Morales' (1994) illustrations of C. 

 rostrata from the Gulf of Mexico. 



