LIFE CYCLE OF PARACYCLOPS 



67 



row of long spinules and row of minute spinules. Segment 2 with 

 4 setae. Segment 3 with 2 setae. Segments 4 and 5 partly fused, 

 with segment 5 (arrowed in Fig. 22A) defined only on dorsal side, 

 not ventrally. Segment 4 with 2 setae; segment 5 with 2 setae. 

 Segment 6 with 2 setae. Segments 7, 8 and 9 separated from each 

 other by extensive arthrodial membrane: segment 7 with 2 setae, 

 segment 8 with 2 setae, and segment 9 with 2 setae plus 

 aesthetasc. Segment 10 (= ancestral segment XV) produced on 

 one side into extensive sheath (arrowed in Fig. 22B) enclosing 

 segment 1 1 ventrally: armed with 2 setae, 1 ornamented with 

 long spinules unilaterally. Segment 1 1 bearing curved seta 

 ornamented with double row of strong denticles, plus 1 naked 

 seta. Segment 12 partly fused to segment 13; armed with short 

 seta ornamented with 2 rows of fine spinules, plus short naked 

 seta. Segment 13 partly subdivided by partial suture: armed with 

 short spinulate seta proximally, 2 short naked setae, plus 1 

 modified element attached to segment by short stalk, main part 

 of element lying along surface of segment and ornamented with 

 longitudinal ridges and small central pore (arrowed in Fig. 22D). 

 Geniculation located between segments 13 and 14. Segments 14 

 and 15 partly fused, forming curved subchela-like section: 

 segment 14 armed with 1 seta, 1 aesthetasc and 2 modified 

 elements each ornamented with longitudinal ridges and a central 

 pore, as distal element on segment 13. Apical segment tapering 

 distally; armed with 1 1 setae and 1 aesthetasc, mostly 

 originating on outer (= posterior) surface. 



All other appendages as in female except for fifth (Fig. 14E) 

 and sixth legs (Fig. 14F). Outer spinulose seta of leg 5 

 ornamented with some long setules distally. Sixth legs forming 

 opercular plates bearing row of large spinules along ventral 

 surface; armed with 1 inner spine, 1 well developed spinulose 

 seta and 1 inner naked seta. 



DISCUSSION 



The number of naupliar instars in the Cyclopoida has been the 

 subject of some controversy but it is clear, as Elgmork & 

 Langeland (1970) strongly indicated, that there are normally 6 

 naupliar instars. This is supported by recent works on free-living 

 freshwater Cyclopoida by Dahms & Fernando (1992, 1993, 

 1994) and by our data. The most difficult distinction is between 

 nauplius IV and V and these stages have often been confused. 



The complete naupliar sequence of P. fimbriatus was 

 previously described by Ewers (1930) and Dukina (1956). 

 Nauplius I and II were also described by Gurney (1933) but none 

 of these provides setation counts of sufficient accuracy. Ewers 

 (1930) described 6 stages but our descriptions differ as follows: 

 antennule is 3-segmented not 4-segmented; antennary exopod is 

 4-segmented at N I and becomes 6-segmented, rather than 

 remaining 4-segmented; caudal rami of N IV are represented by 

 2 pairs of setae and 1 pair of minute seta, rather than just 2 pairs 

 of setae; ventral body surface is ornamented with spinular rows 

 throughout the nauplius phase. In general, however, Ewers' 

 drawings are so small that it is not worthwhile making detailed 

 comparisons of appendage setation patterns. 



Six naupliar stages were also described by Dukina (1956). Our 

 descriptions differ from Dukina's as follows: antennule is 

 3-segmented; antennary exopod is 4-segmented at N I (given as 

 5-segmented by Dukina). Dukina's descriptions of appendage 

 setation also lack sufficient detail for meaningful comparisons. 



Gurney (1933) described the first 2 naupliar stages, but our 



findings indicate that the caudal rami of N II are represented by 

 a pair of setae not by 2 pairs as illustrated by Gurney. Apart from 

 this discrepancy our results differ only in that the exopod of 

 antenna is described as 3-segmented rather than 4-segmented as 

 in our material. 



The copepodid stages of P. fimbriatus were also partly 

 described by Gurney (1933). Although our results are in 

 substantial agreement, for example, with antennulary 

 segmentation throughout the copepodid phase, Gurney's 

 drawings are not sufficiently accurate to permit comparisons of 

 segmental setation. 



Analysis of the antennulary setation patterns of adult male 

 Paracyclops permits the identification of the pattern of segmental 

 homologies. The basic armature of each antennulary segment in 

 copepods is 2 setae plus one aesthetasc (Giesbrecht, 1 892), with a 

 few exceptions as identified by Huys & Boxshall ( 1 99 1 ). Using this 

 basic pattern, the 9 setae on the first segment of male P.fimbriatus 

 indicate that it can be identified as representing 5 ancestral 

 segments (segments I-V). The second segment, with 4 setae, can 

 similarly be identified as derived from 2 ancestral segments 

 (VI-VII). The third to twelfth segments all represent single 

 ancestral segments (VIII to XVII), as indicated by the presence of 

 a maximum of 2 setae on each. The fourth and fifth are 

 incompletely separated (Fig. 22A) but we have treated them as 

 distinct. The thirteenth segment has only 4 setae but is here 

 identified as representing 3 ancestral segments (XVIII-XX). This 

 decision is based on the presence of the neocopepodan 

 geniculation between the thirteenth and fourteenth segments 

 which unequivocally identifies the segmental boundary involved as 

 XX to XXI, and on comparison with other cyclopids such as 

 Etayte robusta Giesbrecht, 1900. In E. robusta males segment 

 XVIII is separate and carries a long naked seta and a short 

 spinulose seta, segment XIX-XX carries a short spinulose seta and 

 a modified spine proximally and a slender seta distally (Huys & 

 Boxshall, 1991). In P. fimbriatus the proximal part of the triple 

 segment is defined by a partial suture marking the original plane 

 between segments XVIII and XIX-XX. This part carries only a 

 single spinulose seta and lacks the long seta; the distal part 

 representing XIX-XX carries the same setation as in E. robusta. 

 This confirms our interpretation of the thirteenth segment as a 

 triple segment (XVIII-XX). The fourteenth and fifteenth 

 segments, lying distal to the geniculation represent ancestral 

 segments XXI-XXIII and XXIV-XXVIII, exactly as Huys & 

 Boxshall ( 1 99 1 ) found for E. robusta. 



Compound antennulary segments, such as the first and 

 second segments of P. fimbriatus, were simply referred to as 

 'fused' by Huys and Boxshall (1991) in their comparative 

 analysis of antennulary segmentation patterns in all copepod 

 orders, although such compound segments could be the result of 

 two different developmental processes: 



1) secondary fusion of segments that were separated earlier 

 during ontogeny 



2) failure of separation during development. 



The compound first and second segments of the male 

 antennule of P. fimbriatus result from the second process, the 

 failure to separate. In contrast, the compound apical segment 

 results from the secondary fusion of the sixth, seventh and 

 eighth segments of the copepodid V stage. (The eighth segment 

 of the copepodid V was already a compound segment, 

 representing three ancestral segments XXVI-XXVIII which are 

 not separately expressed by any known member of the order 

 Cyclopoida). 



Vertical tracking of the segmental boundaries as identified by 

 their setation elements allows us to identify the homologies of 



