Discussion

Since the specimens of Paeonodes nemaformis described here bore a superficial resemblance to Wilson's description (1944, p. 550) of P. exiguus I obtained his type specimen from the United States National Museum (Cat. No. 79642). As in my material, the body in Wilson's specimen (fig. 14) consists of head, neck and trunk, not separated by joints, and a small genital segment and abdomen bearing caudal laminae; the head is expanded dorsolaterally into rounded processes; the first antenna (fig. 16) is five-segmented, some setae present but in damaged condition, others presumably lost; the second antenna (fig. 17) is three-segmented, subchelate, with spines on the first and second segments. All of this agrees with Wilson's description. I was unable to locate the mouth parts. The pereiopods were in poor condition, covered in mucus, dust, and crystalline material which could not safely be removed since this is the only recorded specimen of this species. However, with the aid of phase contrast microscopy, I was able to determine that the structure described by page 37

Paeonones exiguus Wilson, female—Fig. 14: whole view; fig. 15: posterior part of body, showing positions of pereiopods 2-4, broken lines indicate the position of structures seen from another aspect; fig. 16: first antenna; fig. 17: second antenna.

page 38 Wilson as the first two pairs of pereiopods are lacking. The first pereiopods are situated medially on the ventral surface of the trunk and the remaining three pairs on the ventral posterior margin of the trunk (fig. 15) as in P. nemaformis. The last two pairs were mistaken by Wilson for a dwarf male of the shyriid type. This mistake is not surprising considering the condition of the specimen. As in P. nemaformis the first three pairs of pereiopods are biramous, the rami three-segmented, the last pair biramous, with one-segmented rami. There is also some similarity in the relative lengths of the segments of the rami.

The similarities between P. exiguus and P. nemaformis are so great that they must certainly be assigned to the same genus.

However, the facts that P. nemaformis has its cephalic processes bilobate, not fully rounded, lacks the transverse striations found on the base of the neck of P. exiguus and has different body proportions are sufficient to make it a separate species.

Paeonodes exiguus badly needs redescribing from fresh material. Wilson states that it was taken from near the eye of an unidentified fish. However, the label in the tube reads "near eye, Verillida grayula, May, 1911, locality unknown". It is possible that further specimens may be obtained from members of the family Verillidae.

The occurrence of P. nemaformis on New Zealand salmonids is something of an inigma. From the records available to me New Zealand salmonids have been introduced into New Zealand as fertilised ova (Thomson, 1922, Chapter 6; Stokell, 1955; Scott, 1964) which could not have brought this parasite with them. In any case, if it occurred at all commonly on salmonids in Europe or North America one would expect that it would already have been recorded.

Salmonids have apparently become infected by other parasites from native fishes in the short time since they were introduced, e.g. the nematodes Hedruris spinigera Baylis and Eustongylides sp. (Stokell, 1937, p. 80). However, these nematodes have been recorded by Stokell from a variety of endemic freshwater and esturine fish species in addition to their salmonid host, illustrating the adaptability which one would expect on the part of a parasite which has infected an introduced species in 100 years or less. Dix (1968, pp. 371-372) reports that a number of helminth parasites have adapted to the brown trout in this way after it lost its natural parasites during importation. P. nemaformis has not, as yet, been recorded from any of the endemic fish species. Further, Dr. R. M. McDowell, Fisheries Laboratory, Marine Department, Wellington, has examined in detail thousands of specimens of the families Retropinnidae, Galaxiidae and Eleotridae without sighting parasitic copepods (pers. comm.). This suggests that the most likely endemic hosts are the shark bully (Cheimarrichthys forsteri Haast), anguillids or the tidal flounder (Rhombosolea retiaria Hutton). Even should this prove to be the case, it is hard to see why this parasite has been able to adapt so quickly to the introduced salmonids when it has apparently failed to parasitise endemic species in the same habitat.

It is possible that P. nemaformis is normally parasitic on a marine host, but this would require that the parasite has, in less than 100 years, adapted to a new host family and also to a freshwater or anadromous existence.

As described here, the genus Paeonodes has much in common with Therodomas Kroyer and Mugilicola Tripathi. In particular it shares the same body divisions, it has antennae and mouth parts that are very similar, and the pereiopods are similar in form and restricted to the trunk region. The genera are separated in a key given below.

Thomsen (1949, pp. 34-35) placed the genus Therodomas in the family Ergasilidae (Cyclopoida) because of the similarity between the mouth parts found in this genus page 39 and in other genera usually assigned to the Ergasilidae. Tripathi (1960, p. 545) described Mugilicola and pointed out its similarity to Therodomas. He agreed with Thomsen that the mouth parts indicated a relationship with the ergasilids but argued that the modification of the body to a parasitic habit was so profound as to necessitate their separation as a new family (Therodomasidae). This view is accepted in this paper and Wilson's genus Paeonodes becomes the third genus placed in this family.