60 



R. HUYS AND J.M. GEE 



Wilson (1966). There are slight discrepancies between these 2 

 descriptions which should be noted because of their possible 

 phylogenetic significance. The rostrum does articulate with 

 the cephalothorax and there are 2 accessory setae at the base 

 of the maxillipedal claw. In the female there is no aesthetasc 

 on the terminal segment of the antennule and the outer distal 

 corner of enp-2 in P2 and P3 is significantly attenuated but 

 that of enp-1 is normal (especially compared to the condition 

 in Psammis). In the male the antennule is distinctly 

 9-segmented with segment 6 being swollen and bearing an 

 aesthetasc; the distal outer element of P2 enp-3 is not 

 completely fused to the segment but articulates along the 

 anterior surface and partially articulates on the posterior 

 surface; and there is no sexual dimorphism in P4 enp-2 (Fig. 

 8B). 



We have been unable to locate the type, or any other, 

 material of D. sibirica and therefore, like Wilson (1966), have 

 had to rely on Sars' (1898) original description and figures. A 

 study of these has led us to conclude that there are no real 

 differences between D. stefanssoni and D. sibirica. We agree 

 with Wilson's interpretation of the swimming leg setation. 

 The original copy of Sars' paper in our possession also shows 

 faint lines where the seta should be on P2-P4 exp-1 and in no 

 other species in any paranannopid genus is this seta absent in 

 the female when it is present in the male. Although Sars 

 (1898) states in his text that the female antennule is 

 5-segmented, close scrutiny of his drawing (Plate X, Fig. 4) 

 reveals that the terminal segment is at least partially divided 

 into two segments and that the arrangement of the pectinate 

 spines (3 on segment 5 and 1 at the apex of segment 6) is 

 exactly the same as in D. stefanssoni. Further Sars has clearly 

 misinterpreted the segmentation of the male antennule in 

 that he has combined segments 1 and 2 as shown by the fact 

 that his segment 1 bears 2 setae, a condition found in no other 

 Paranannopidae. Wilson (1966) concluded that the only real 

 difference between D. stefanssoni and D. sibirica was the 

 absence in the latter of the distinctive outer distal spine on P2 

 enp-3 in the male. However, Sars' drawing of this limb (Plate 

 X, Fig. 16) shows only 4 elements on enp-3 instead of 5, as in 

 the female, and it is possible that the crucial one (the outer 

 spine) is concealed behind the large apophysis on enp-2. This 

 interpretation is supported by the fact that, irrespective of the 

 degree of modification in males, the number of elements on 

 P2 enp-3 is always the same in both sexes, except in the 

 genera bearing claviform aesthetascs on the mouthparts (Gee 

 & Huys, 1991) where the distal outer spine is further modi- 

 fied into a non-articulating apical apophysis. 



These observations, coupled with such factors as similarity 

 of size (they are the largest known members of the Paranan- 

 nopidae except for Psammis borealis Klie, 1939 whose taxo- 

 nomic position is unclear) and the peculiar identical 

 distribution of setae on the distal margin of the female P5, 

 lead us to the conclusion that the two species are synonymous 

 and have a circum-polar distribution. Fig. 13 shows that the 

 most easterly record of D. sibirica at Wrangell Island (Yash- 

 nov, 1935) is very close to the most westerly record of D. 

 stefanssoni on the Chukchi Sea coast of Alaska (Wilson, 

 1966). The only record of the species outside the Arctic Circle 

 is that of Wells (1965) from Loch Nevis on the west coast of 

 Scotland and this must be regarded as doubtful (original 

 specimens no longer available for re-examination). Further, 

 both D. stefanssoni and D. sibirica have been recorded from 

 estuaries and in brackish water, a most unusual habitat for 

 members of this family. All other species are found only in 



sublittoral marine habitats although Danielssenia typica 

 Boeck, 1872, the other species with a known global distribu- 

 tion, has been recorded from the Baltic Sea (Veldre & 

 Maemets, 1956; Arlt, 1983), a region of lowered salinity. 



(ii) Autapomorphies of Archisenia 



Turning now to consider the taxonomic status of D. sibirica, 

 its distinction from other members of the genus Danielssenia 

 was suggested by Lang (1944) who divided Danielssenia into 2 

 groups, the typica group and the sibirica group. The latter he 

 characterized by: (i) antenna exp-1 with 2 setae; (ii) P4 enp-3 

 with 2 inner setae; (iii) the male P2 enp-1 with the inner seta 

 transformed into a non-articulating process; (iv) the male P3 

 enp-2 with an outer hooked apophysis. This last character is 

 now known to occur in all species of Paranannopidae and 

 might even be a diagnostic character for a wider group of 

 families. The first two characters, although diagnostic of D. 

 sibirica in combination, are the plesiomorphic condition in 

 the family and are found in a number of other genera. 

 Paranannopus, Psammis, Micropsammis Mielke, and 

 Bathypsammis gen. nov. also retain 2 setae on exp-1 of the 

 antenna (all other genera bear only 1 seta on this segment) 

 and Sentirenia Huys & Gee (= Jonesiella Brady, see below) 

 and the male of Fladenia retain 2 inner setae on P4 enp-3. 

 However, the transformation of the inner seta into a non- 

 articulating spine on P2 enp-1 in the male is unique to this 

 genus, as are the following autapomorphies: (i) the outer 

 extension on P3 enp-2 in both sexes; (ii) the sigmoid, heavily 

 sclerotized female copulatory duct; (iii) the sexual dimor- 

 phism of the inner basal spine of the male PI. Another 

 character of phylogenetic significance is the loss of the inner 

 seta on the male P2 enp-2. This character is also found in 

 Afrosenia spinipes (Wells, 1967) and is regarded here as a 

 product of convergence. A further character which is unique 

 but difficult to quantify is the arrangement of the setation on 

 the exopod of P5, with the inner seta well separated from the 

 remaining setae. It is on the basis of all these characters that 

 we have removed D. sibirica to a new genus leaving the typica 

 group as the only species group in the genus Danielssenia. 



(iii) Intersexuality 



The male from East Greenland (Fig. 12), collected by Jes- 

 persen, proved upon examination to be an aberrant intersex- 

 ual specimen. It has the male body facies, including a 

 6-segmented urosome (Fig. 12A), a well developed testis and 

 vas deferens (however, a spermatophore has not been 

 observed), and the male form of the P5 and P6. The 

 endopods of P2 and P3 are also modified but differ from the 

 typical male condition by the retention of certain female 

 features. 



The antennules resemble the female condition in all 

 aspects: they are 6-segmented, lack any trace of a genicula- 

 tion mechanism and possess the female armature pattern. 

 The PI (Fig. 12B) basis and endopod show the same spinule 

 arrangement as in the female but size and shape of certain 

 spinule rows approach the male condition. The P2 endopods 

 are not identical on both sides (Figs. 12C-D) and show a 

 combination of male and female characteristics. The proximal 

 segment and its basal pedestal are moderately enlarged but 

 the spinule at the base resembles the female condition and 

 the inner seta is — though being shorter than in the female — 

 not transformed into a spinous process. The outer apophysis 



