100 



R. HUYS AND W. LEE 



Characters used in the analysis are listed in Table 3. Apomorphic 

 character states are explained inside square brackets using the multistate 

 system. The scores for each character and taxon are compiled in 

 matrix format in Table 4. A question mark indicates missing data, 

 either because the appendage or structure is unknown in that species 

 (certain sexually dimorphic characters could not be scored because 

 only one sex is known) or because it was impossible to score the 

 character accurately due to incompleteness or the lack of detail in the 

 original descriptions. Esola typhlops pontoica, E. longicauda var. 

 sensu Vervoort (1964) and the unnamed forms of E. longicauda 

 identified by Noodt (1955) and Wells & Rao (1987) were excluded 

 from the analysis because of their questionable status. 



Huys & Boxshall's (1991) study of ordinal copepod phylogeny 

 demonstrated that oligomerization was the dominant trend of evolu- 

 tionary transformation within the Copepoda. Armature counts used 

 in this analysis were scored according to this overall polarisation 

 mode. Most characters in Table 3 are self-explanatory but additional 

 notes are provided for the following: 



Integumental pores (characters 1-4) 



The conspicuous cup-shaped integumental pores on the 



cephalothorax and genital (double-)somite have remained unnoticed 



Table 3. Characters used in phylogenetic analysis. Apomorphic character 

 states are referred to in square brackets. 



1 Paired anterodorsal cup-shaped pores on cephalothorax absent 

 [present] 



2 Paired anteroventral cup-shaped pores on cephalothorax absent 

 [present] 



3 Paired cup-shaped pores on genital double-somite of 9 and genital 

 somite of cT absent [present] 



4 Caudal rami without large pore medially or ventrally [present] 



5 Cephalothorax without transverse spinular row dorsally [present] 



6 Caudal rami not sexually dimorphic [modified in 2] 



7 Antennule 2 7-segmented [6-segmented; failure in separation of 

 segments 6 and 7] 



8 Antennule $ with 3 segments distal to geniculation [with 2 

 segments: segments 7 and 8 fused] 



9 Aesthetasc of segment 4 in 2 (and segment 5 in c?) fused basally to 

 seta [fused to two setae forming trifid compound element] 



10 Antennule segment 1 without processes in 2/c? [with 3 spinous 

 processes along posterior margin] 



1 1 Antennule segment 2 with large spinous process arising from 

 posterior margin in 9/6 [absent] 



12 Antennule segment 5 of cT without anterior cylindrical process 

 (bearing large spine) [present] 



13 Labrum without conspicuous ornamentation on anterior surface 

 [with overlapping scales distally and dense pattern of fine spinules 

 proximally] 



14 Maxillulary endopod represented by 3 setae [2 setae, outermost seta 

 lost] 



1 5 Maxillipedal syncoxa with 3 setae [state 1 : 2 setae, proximal seta 

 lost; state 2: 1 seta] 



16 PI exopod 3-segmented [2-segmented; exp-2 and -3 fused] 



17 PI exopod 2-segmented, exp-2 with 3 outer spines and 2 apical 

 geniculate setae [exp-2 with 2 outer spines and 2 apical geniculate 

 setae] 



18 PI enp-1 with inner seta [absent] 



19 P2 enp-2 with outer spine/seta [absent] 



20 P3 endopod o* 3-segmented [2-segmented; neotenic development] 



21 P3 enp-2 6 with inner seta [absent] 



22 P5 baseoendopod 5 with 5 setae [state 1 : with 4 setae, middle inner 

 seta lost; state 2: with 3 setae] 



23 P5 baseoendopod 6 with 2 setae [setae absent] 



24 P5 basoendopod 916 without distinct setophore for outer basal seta 

 [basal seta positioned on long cylindrical setophore] 



25 P5 exopod 2 with all outer setae arranged around margin [proximal 

 2 outer setae displaced with overlapping insertion sites] 



in previous descriptions except for Vervoort (1962, 1964) who 

 briefly described the anterodorsal pores in C. bulligera and E. 

 vervoorti and suspected them to be eyes. Various authors (e.g. 

 Jakobi, 1953; Hamond, 1969; Mielke, 1981, 1997) have uninten- 

 tionally figured the modified pores on the caudal rami, however, 

 incorrect interpretation of the internal chitinized walls of the ducts as 

 external ridges ('Chitinleiste') made them fail to recognize these 

 structures as true pores. Huys (1990&) pointed out that the trans- 

 formed cup-shaped pores in Esola are not serially homologous with 

 the pleural glands of the Adenopleurellidae and consequently can- 

 not serve as a basis for phylogenetic affinity. With the exception of 

 Archilaophonte and the typhlops-group of Esola all other esolinids 

 appear to exhibit a propensity for developing modified secretory 

 pores. The functional correlation between pores of different body- 

 regions is unknown and in view of their positional disparity and 

 structural differences it is unlikely that their expression is controlled 

 by a single gene. We postulate that the cup-shaped pore type evolved 

 from a surface precursor pore by major integumental invagination 

 and secondary development of setular extensions. These marginal 

 extensions either protect the depression or (more likely) maintain 

 the secrete bolus in close contact to the body wall. The degree of 

 invagination is obviously morphologically constrained and this is 

 particularly the case in swimming leg segments which are typically 

 depressed along the antero-posterior body axis. Although the 'trap- 

 ping basket' seta on the P4 endopod of C. bulligera represents a 

 radically divergent modification, it can be viewed as an external 

 analogue of the internal cup-shaped pore which developed in response 

 to this constraint. The tube-pore, which is also found in most other 

 esolinids, is enclosed by the long setules arising from the proximally 

 dilated distal inner seta (Fig. 25E-F) which hold the secrete bolus in 

 position. Since there are no differences in pore pattern between the 

 sexes a possible role in mate recognition is considered unlikely. 

 Huys (1992) demonstrated that in the interstitial Leptastacidae the 

 mucopolysaccharid strands produced by the caudal ramus glands 

 are intimately involved in mucus-trap feeding. We suggest that in 

 esolinids the secretory products discharged by the cup-shaped pores 

 perform a similar role in trophic gardening. It should be noted that 

 the caudal ramus pores located near the insertion sites of setae I-HI 

 in E. typhlops (Fig. 3 1C-D) are not homologous to the large slit-like 

 pores found in Esola and Mourephonte. 



Caudal ramus sexual dimorphism (character 6) 

 Females of Esola typically have bulbous caudal rami, displaying a 

 variety of swelling medially, ventrally and/or dorsally. Although the 

 secondary expansion appears to be correlated with the size of the 

 transformed pores, it is decoupled here from character 4 (presence of 

 caudal ramus pores) and scored separately. This is justified by the 

 absence of caudal ramus sexual dimorphism in A. hirsuta despite the 

 presence of modified pores in both sexes. 



Setal fusion on antennules (character 9) 



In most esolinids (except Archilaophonte) the proximal aesthetasc 



(on segment 4 in 9, segment 5 in 8) is fused at the base to 2 setae. 



This trifid compound element is a unique character in the 



Harpacticoida. 



Antennulary processes (characters 10-12) 



Within the esolinid grouping a spinous process along the posterior 

 margin of the second antennulary segment (character 1 1 ) is present 

 only in Archilaophonte. This is not an autapomorphy for the genus 

 but considered a retention of the ancestral state, based on outgroup 

 comparison with the remaining families of the Laophontoidea (Huys, 

 1990a; Huys & Lee, 1999). The presence of auxiliary processes 



