Introduction
The genus Trematomus Boulenger comprises 15 species inhabiting the Antarctic continental shelf and associated islands. Most of these species have a circum-Antarctic distribution; one species, Trematomus hansoni Boulenger Reference Boulenger and Lancester1902, extends as far north as South Georgia Island; Trematomus vicarious Lönnberg Reference Lönnberg1905 has also been recorded from that locality, as well as the South Orkney Islands and Peter I Island (DeWitt et al. Reference DeWitt, Heemstra, Gon, Gon and Heemstra1990, Matallanas Reference Matallanas1997, Matallanas & Olaso Reference Matallanas and Olaso2007, Eastman & Eakin Reference Eastman and Eakin2021). Most trematomids are benthic species characterized by a robust body and high buoyancy index values (Eastman Reference Eastman2020). Trematomus loennbergii Regan Reference Regan1913 is exceptional in having a more streamlined adult body and a significantly lower buoyancy index value, both implying a semipelagic or epibenthic lifestyle (Eastman & Hubold Reference Eastman and Hubold1999, Eastman Reference Eastman2020). Moreover, this species has been collected from shallow to very deep water, having a depth range of 65–1243 m, which is exceeded only by that of Trematomus lepidorhinus, its closest relative, and the two Dissostichus species in the Nototheniidae (Eastman Reference Eastman2017).
Regan (Reference Regan1913) described T. loennbergii from three specimens collected in three different localities. These specimens included one adult and two juveniles, but the whereabouts of the two juveniles had been unknown (Fricke et al. Reference Fricke, Eschmeyer and van Der Laan2025). This may have led DeWitt et al. (Reference DeWitt, Heemstra, Gon, Gon and Heemstra1990) to consider the only apparently extant specimen as the holotype of T. loennbergii. In this paper, we report the finding of the two missing juveniles and we designate a lectotype for this species.
Materials and methods
Meristic and morphometric data for the three specimens used by Regan (Reference Regan1913) are presented in Table I. Counts and measurements follow Hubbs & Lagler (Reference Hubbs and Lagler1958). Measurements were made point-to-point on the specimens. Identification was made using the key and data for Trematomus spp. of DeWitt et al. (Reference DeWitt, Heemstra, Gon, Gon and Heemstra1990). The following collection and institution acronyms are used below: SL = standard length; TL = total length; BMNH and NHM = Natural History Museum, London; NRM = Swedish Museum of Natural History, Stockholm.
Table I. Meristic and morphometric data for the type series of Trematomus loennbergii from the Southern Ocean.
HL = head length.
Results and discussion
Regan (Reference Regan1913) described T. loennbergii from three specimens that had syntype status. The largest and only specimen recorded as a type (BMNH 1906.5.8.69) - an adult - was collected by the British National Antarctic Discovery Expedition (1901–1904) south-west of the Balleny Islands, at 465 m depth. The first missing specimen - a juvenile - was collected by the same expedition at Tent Island, a small, little-known island in the vicinity of the well-known Ross Island in the south-west Ross Sea. The second missing specimen - also a juvenile - was collected by the Swedish Antarctic Expedition (1901–1903) in a tide pool at Seymour Island, northern Antarctic Peninsula (Lönnberg Reference Lönnberg1905). Another interpretation of the Swedish-language description of the locality, ‘Vattenpölar på ebbstranden’, on the NRM register card (Fig. 1) could also be ‘puddles on the low tide shore’, indicating that the specimens may have washed up on shore after a storm. What prompted Regan to request the loan of the Swedish specimen is unknown, but it is not surprising, as his request was made in the midst of a period - termed ‘the Heroic Era’ - of high expeditionary activity in the Southern Ocean (Headland Reference Headland, Gon and Heemstra1990), and he must have been preparing ‘The Antarctic Fishes of the Scottish National Antarctic Expedition’ (Regan Reference Regan1913) at that time. In a letter dated 17 October 1912 (NHM Archives 1912) responding to Regan’s loan request, Dr Einar Lönnberg stated that he was sending six lots of nototheniids: five he had described earlier (Lönnberg Reference Lönnberg1905) as new species or subspecies, and also a specimen of Notothenia nicolai Boulenger Reference Boulenger and Lancester1902; he gave Regan the option to keep four specimens, but the N. nicolai specimen was not one of them. Possibly, Lönnberg may have included this specimen in the loan because he noticed some differences between the juvenile and the larger N. nicolai, and he was not sure whether they were the same species. Be that as it may, when Regan (Reference Regan1913) decided that this juvenile was an unknown species, he thanked Lönnberg by naming it after him.
Figure 1. The original register card of NRM 2980 and 2981 written by Dr E. Lönnberg.
Figure 2. The type series of Trematomus loennbergii: a. BMNH 1906.5.8.67, paralectotype, 63.7 mm standard length (SL), Tent Island, south-west Ross Sea; b. NRM 2981, 58.6 mm SL, paralectotype, Seymour Island, northern Antarctic Peninsula; and c. BMNH 1906.5.8.69, lectotype, 112.6 mm SL, Balleny Islands, Southern Ocean.
Figure 3. Dorsal view of the heads of a. Trematomus nicolai NRM 2980, 91.6 mm SL and b. Trematomus pennellii NRM 2981, 58.6 mm standard length (SL), paralectotype and c. BMNH 1906.5.8.67, 63.7 mm SL, paralectotype.
This search for the missing syntypes started when the first author of this paper noticed an entry for a specimen (BMNH 1913.12.4.118) of T. loennbergii from Inaccessible Island, Saint Helena District, in the online database of the Natural History Museum, London. The Saint Helena District is a British territory in the south Atlantic, including the islands of Saint Helena, Ascension and Tristan da Cunha; the small Inaccessible Island, located a short distance south-west of Tristan da Cunha, is included in this British territory. As a species endemic to Antarctica (Eastman & Eakin Reference Eastman and Eakin2021) and adapted to living in the subzero temperatures of Antarctic waters (Jin & DeVries Reference Jin and DeVries2006), it is highly improbable that T. loennbergii would be present at this island. The entry was found to be erroneous and should have referred to another Inaccessible Island located next to Tent Island in the south-west Ross Sea. Suspecting that this could be the missing syntype from Tent Island, we searched the NHM holdings of Trematomus for specimens of ~65 mm TL from Tent Island. Because Regan (Reference Regan1913: 230, 263) stated that the third specimen he had used was a loan from Lönnberg, we also searched the NRM holdings for specimens of similar size from Seymour Island.
As it turned out, the Inaccessible Island specimen (BMNH 1913.12.4.118, 78.3 mm SL, 94.2 mm TL) was collected by the ill-fated British Antarctic (Terra Nova) Expedition, 1910–1913, so it is not a missing syntype. However, we did find a more probable specimen (BMNH 1906.5.8.67) from Tent Island, collected by the British National Antarctic Discovery Expedition, although at 74.8 mm TL it is somewhat larger than the 65 mm TL measured by Regan (Reference Regan1913). As this is the only Trematomus specimen from Tent Island (other than two much larger specimens of Trematomus bernacchii Boulenger Reference Boulenger and Lancester1902 collected by the same expedition that captured the large syntype) and therefore matches the information given in Regan (Reference Regan1913), we believe it is the first missing syntype (Fig. 2a). This specimen was examined by H.H. DeWitt in 1965, who re-identified it as Trematomus pennellii Regan Reference Regan1914. At first, this juvenile specimen seemed to have no occipital and interorbital scales, but close examination of the top of the head revealed very small scales without ctenii and with only few circuli in the interorbital area. The small size and the few circuli of these scales suggest that they appeared late in the ontogeny of this juvenile.
Table II. Summary of morphological differences between Trematomus loennbergii, Trematomus nicolai and Trematomus pennellii from the Southern Ocean.
Data from DeWitt et al. (Reference DeWitt, Heemstra, Gon, Gon and Heemstra1990).
Lönnberg (Reference Lönnberg1905) reported four specimens that he identified as N. nicolai that were collected by the first Swedish Antarctic Expedition (1901–1903) at Seymour Island during 8–11 February 1903. The search at NRM yielded only two specimens from Seymour Island - both are currently registered as T. loennbergii. The change of identification was apparently made by Lönnberg himself (Fig. 1) after he had received the re-identified loaned specimen back from Regan (we compared the handwriting on the label with a letter he sent to Regan in 1912). One fish (NRM 2980), measuring 108.8 mm TL (91.6 mm SL), is larger than the size given by Regan (Reference Regan1913). At 69 mm TL (58.6 mm SL), the other specimen (NRM 2981; Fig. 2b) is only slightly longer than the size given by Regan (Reference Regan1913) and is probably the second missing syntype he examined. Our re-examination of both specimens (Fig. 3 & Table I) confirmed that the larger one is indeed Trematomus nicolai (NRM 2980). The smaller fish (NRM 2981) has scales on the snout reaching as far forward as the anterior nostrils (Fig. 3b), as in T. loennbergii and similar to the first missing syntype (Fig. 3c), but its pectoral and anal fin ray counts are lower than in the latter species (Table I). Moreover, the middle lateral line of this fish seems to have only pored scales, although some scales are missing. These characters are in better agreement with T. pennellii than with T. loennbergii (Table II), and we therefore consider the second missing syntype a juvenile T. pennellii.
The mismatch between the specimen length given by Regan (Reference Regan1913; 65 mm TL) for both missing syntypes and our length measurements of 74.8 mm TL (BMNH 1906.5.8.67) and 69 mm TL (NRM 2981) is difficult to explain given the usual shrinkage expected after many years in preservative; it may not have been a coincidence since it happened in two of the three specimens that Regan (Reference Regan1913) had used in his description of T. loennbergii. One possibility is that Regan (Reference Regan1913) actually measured the standard length, known at the time as body length without the caudal fin, but did not remember it when he was writing his paper. The identification of the two juvenile syntypes as T. pennellii as well as the uncertainties regarding their lengths and regarding the whereabouts of the remaining two of the four N. nicolai specimens from Seymour Island lead us to designate the largest syntype (BMNH 1906.5.8.69, 112.6 mm SL; Fig. 2c), from the Balleny Islands, as the lectotype of T. loennbergii Regan Reference Regan1913. The two remaining specimens we identified as T. pennellii (i.e. BMNH 1906.5.8.67, 63.7 mm SL and NRM 2981, 58.6 mm SL) become paralectotypes of T. loennbergii.
Acknowledgements
We thank Lucie Goodayle (NHM Photo Unit) and Andrea Hennyey (NRM) for producing the high-quality images presented in this paper. Thanks also go to Kathryn Rooke (NHM Archives) for providing the Lönnberg correspondence. We are indebted to the Trustees of the Natural History Museum, London, and of the Swedish Museum of Natural History, Stockholm, for permission to use images under their copyrights. We acknowledge that the opinions, findings and conclusions or recommendations expressed in this publication, generated by the National Research Foundation of South Africa (NRF)-supported research, are those of the authors, and that the NRF accepts no liability whatsoever in this regard.
Author contributions
OG initiated this study, prepared the tables and figures and wrote the first draft of the manuscript. JM and BD retrieved and examined specimens and had them photographed at their institutions. All three authors wrote, reviewed and agreed to all subsequent versions of this article.
Competing interests
The authors declare none.
Open access
