Onisimus turgidus (Sars, 1879) (Amphipoda, Uristidae), an overlooked amphipod from sea anemones in Northern Norway

Two Norwegian uristid amphipods, obligate associates of sea anemones, have for a long time been confused sub nomine Onisimus normani Sars, 1890. In reality this species only occurs in south Norway, while the north-Norwegian material belongs to O. turgidus (Sars, 1879), described from the Barents Sea and for a long time forgotten. This paper fully illustrates both species, gives a key, and provides data on their distribution and ecology.


Introduction
In his great monograph on Norwegian amphipods G.O. Sars (1891) described the uristid species "Onesimus normani, Schneider, ms" on the basis of a single specimen, found in a sample of O. edwardsi (Krøyer, 1846), "probably from some place in Finnmark". He noted (op. cit.: 106): "The present new species was first detected by Mr Schneider in the neighbourhood of Tromsø, Finnmarken (…) I have had an opportunity of seeing his detail drawings of the species." Unfortunately, neither Schneider's drawings nor his specimens seem to be extant. Vader (1967Vader ( , 1970 discovered that Onisimus normani Sars, 1891 lives in the gastrovascular cavity of the sea anemone Bolocera tuediae (Johnston, 1832) and studied this association in the vicinity of Bergen, W Norway; the amphipods enter the host as 2-3 mm long juveniles in autumn, and leave the sea anemone as 8-9 mm long adults 18 months later. Strangely, ovigerous females have as yet never been found anywhere. (The specimen from Oslofjord described by Sars (1895) as having "reddish ova in the marsupial pouch" in reality carries the isopod parasite Parapodascon Hansen, 1916 (Vader, unpubl. obs.: Parapodascon sp.), which also commonly infested O. normani in the Bergen population (Vader 1967)). Onisimus Boeck, 1871 thus spends a large part of its life-cycle inside the sea anemone and it is almost immune to the proteolytic enzymes in the gastrovascular cavity that quickly kill and digest other amphipods (Vader & Lønning 1973). A later study showed that these amphipods do not, as originally assumed, feed primarily on the semi-digested remains of the prey of the sea anemone, but instead on the mesenterial filaments of the host itself (Moore et al. 1994).
When the first author moved from Bergen to Tromsø and again found reddish Onisimus specimens inside sea anemones there, it was assumed at first that they too must be O. normani, a species originally described from this area (Vader 1975). A few points of difference between the west and north Norwegian animals were nevertheless noted. Although both amphipods were reddish, there was a clear difference in hue between the two, and in addition the animals in the population of Ullsfjorden N of Tromsø were primarily found in the sea anemone Actinostola callosa (Verrill, 1882) and only to a lesser degree in Bolocera tuediae. The question was tackled by Johnsen (2001) in an unpublished MSc thesis, and he came to the unexpected conclusion that the N Norwegian material does not belong to O. normani, which is largely confined to southern Norway, with the northernmost record in Andfjorden in Nordland, but instead represents Onisimus turgidus (Sars, 1879), a species further described by Sars in his report on the crustaceans of the Northern Norwegian Sea Expedition (Sars 1885), on the basis of a few specimens collected in the Barents Sea, in fact also with sea anemones. This species was for a long time almost forgotten; it was not mentioned at all in Sars' monograph on the Norwegian amphipods (Sars 1890(Sars -1895, and although it figures in Stephensen's "The Amphipoda of Northern Norway and Spitsbergen" (Stephensen 1935(Stephensen -1942, it is misrepresented in his key to Onisimus species and is impossible to identify correctly with this key. Johnsen (2001) studied presumed O. normani samples from different parts of the world (W Atlantic Ocean, E Siberia) and came to the conclusion that they all are closer to O. turgidus than to O. normani, but that both the W Atlantic and the E Siberian material show some peculiarities, that may lead to these entities needing to be described as separate species. At least the W Atlantic material is also associated with sea anemones (Vader 1967). Johnsen's drawings of O. turgidus (Johnsen 2001) are here reproduced in Figs 4-10. They are both re-illustrations of specimens from the type material (from the Barents Sea) (Figs 4-6) and material from Ullsfjord (Figs 7-10). This species has recently also been reported once more from the Barents Sea (Bryazgin 1997;Lyubina et al. 2014, again as O. normani;Zimina et al. 2019); Johnsen has seen some of Bryazgin's material.

Material and methods
This study is based on a combination of museum material (Arctic University Museum of Tromsø (TSZCrnumbers) and the Natural History museum Oslo (F-numbers) and material collected in recent field studies (deposited at the Arctic University Museum of Tromsø). All historic material was initially preserved in formaldehyde before being transferred to 70% ethanol for storing at the museums. The material sampled for the Johnsen thesis work was directly fixed in seawater-diluted ethanol (75%), before being transferred to freshwater-diluted 70% ethanol for longterm storage at the Tromsø Museum (now known as the Arctic University Museum of Tromsø). All material was initially studied and sorted using stereo microscope, dissected and mounted in rose Bengal-stained polyvinyl-lactophenol on microscope slides for morphological examination. All drawings were made using a camera lucida mounted on a Leica compound microscope fit with a camera lucida for pencil drawings. Initial pencil drawings were subsequently inked using the methods described by Coleman (2003Coleman ( , 2009

Distribution
Johnsen (2001) studied samples of presumed Onisimus normani from different parts of the world, and concluded that they all (except the ones from S Norway) belonged in the turgidus species complex; unfortunately there was not sufficient material available to be able to revise this entire complex. The type material was collected from the Barents Sea and it has been collected from this area several times since (Bryazgin 1997;Lyubina et al. 2014, sub nomine O. normani;Zimina et al. 2019). Material from N Norway (Ullsfjorden) is quite similar to the type material, but the animals are smaller, 10 mm against 14 mm for the type. No fully adult or ovigerous specimens of O. turgidus have been found as yet, similar to the situation in O. normani. Onisimus turgidus also spends most of its life inside a sea anemone and like O. normani, likely leaves the host when almost adult.

Biology
Onisimus turgidus appears to be an obligate associate of sea anemones for most of its life-cycle. The preferred host of Onisimus turgidus in N Norway is Actinostola callosa. The numbers vary quite a lot from year to year (Moore et al. 1994;Vader & Krapp-Schickel 1996:  There is a significant positive correlation between the size of the Actinostola hosts and the number of Onisimus found inside: in the large sample of January 1974 sea anemones of 10-50 g wet weight contained 0.03 amphipods per host, those of 60-100 g 0.54 and the few of 100-150 g 3.00. Actinostola callosa in Ullsfjorden also is the host of another amphipod species, Stenothoe brevicornis Sars, 1883, that likewise feeds on host tissue, but this amphipod lives on the outside of the host, among the tentacles (Moore et al. 1994;Vader & Krapp-Schickel 1996).
Onisimus turgidus in Ullsfjorden is also sometimes found inside Bolocera tuediae, the almost exclusive host of O. normani in S Norway (Vader 1973), but here it is even less common than in A. callosa: in 297 specimens of Bolocera Gosse, 1860 from Ullsfjorden only 14 specimens of Onisimus were found (0.05 amphipods per host). The parasitic isopod Parapodascon sp. that is quite common on O. normani in S Norway, has also been found a few times on O. turgidus from the Ullsfjord.

Discussion
The discovery that all Onisimus material collected in N Norway and identified as O. normani in reality belongs to O. turgidus, causes a nomenclatural problem, since Sars (1891)