Eumida cf. longicirrata Hartmann-Schröder, 1975

Fig. 21

Material examined. NHMUKANEA 2022.406, NHMUK ANEA 2022.407 – 408, NHMUKANEA 2022.439, IN2017 _ V03 _100; 9 June 2017; off Byron Bay, NSW, Australia, beam trawl, start: 28.05°S 154.08°E, 999 m, end: 28.10°S 154.08°E, 1013 m. DNA vouchers: NHMUKANEA 2022.404 (COI, 16 S, 18 S), AMW.52192 (COI, 16 S), AMW.52193 (COI, 16 S), AMW.52194 (COI, 16 S), same locality .

Description. Complete specimens measuring 2.8–7.6 mm long with 20–42 segments, appearingbiannulate (Fig. 21A). Segments widest anteriorly (0.53–0.67 mm) and tapering posteriorly.

Prostomium pentagonal with rounded corners, wider than long (Fig. 21B). Anterior end of prostomium with pair of antennae slightly longer than prostomium dorsally and similar pair of palps ventrally. Amedian antenna, shorter and thinner than frontal antennae, inserted near middle of prostomium. Eyes absent.

Proboscis fully everted in one specimen (NHMUK ANEA 2022.404), funnel-shaped, with ~40 terminal cirri and low papilla in concentric rows on the column (Fig. 21C). Partially-everted proboscis appears columnarin other specimens (NHMUKANEA 2022.439).

First and second segments appear fused, with the first pair of tentacular cirri inserted ventral to the prostomium. Four tentacular cirri on segments 1–3 cylindrical, tapering distally (1 + C1/1 + C1/ N). In preserved specimens, the cirrus on segment 1 reaches segments 5–6, the dorsal cirrus on segment 2 reaches segments 6–7, the ventral cirrus on segment 2 reaches segments 5–7, and the cirrus on segment 3 reaches segments 7–8.

Dorsal cirri present from segment 4, ventral cirrus from the segment 3. Dorsal cirrus lanceolate, 1.5–2× longer than parapodial lobe, around 2.5–3× longer thanwide (Fig. 21D). Ventralcirrus lanceolate, slightlylonger thanparapodial lobe, around 2.5–3× longer than wide.

Parapodia uniramouswith a single aciculaand numerous heterogomph spinigers (Fig. 21E). Shaft of chaetae with apical teeth and blade with fine serration.

Pygidiumbroad and blunt, with two terminal tear-shaped cirri easily detached (Fig. 21F).

Distribution. IN2017_V03, Station 100. Pilot whale carcass, off Byron Bay, New South Wales, Australia in 999–1013 m.

Remarks. Recognized Eumida species lacking eyes include Eumida alvini Eibye-Jacobsen, 1991, Eumida angolensis Böggemann, 2009, Eumida (Eumida) longicirrata Hartmann-Schröder, 1975, and Eumida nuchala (Uschakov, 1972) . Among these, the IN2017_V03 Eumida specimens most closely resembled E. longicirrata, which has the median antenna inserted slightly anterior to the middle of the prostomium unlike the other species where it is inserted closer tothe posterior end. The other species are also distinct from the IN2017_V03 specimens in the following ways: E. alvini have the median antenna longer than the frontal pair as wellas very long tentacular and dorsal cirri; E. angolensis have oval prostomium and “bottle-shaped” tentacular cirri (Böggemann, 2009); E. nuchala haveenlarged nuchalorgans and ventral cirri much longer than the parapodial lobe.

Raised semicircular structures on theposterior end of the prostomium, described for E. longicirrata, were difficult to observe in the IN2017_V03 specimens. This was also the case for other specimens identified as E. longicirrata (Ravara et al., 2017) . The micropapillae observed in the proboscis of the IN2017_V03 specimens is contrary to the smooth proboscis described for E. longicirrata . The presence of>40 oral papillae (NHMUK ANEA 2022.404) is unusual since most Eumida species often have 17, except for E. alvini, which can range from 22–50 (Eibye-Jacobsen, 1991). Estimates of tentacular cirri lengths, measured by the extent of reach along body segments, are similar to E. longicirrata . It should be noted that live specimens were observed to have more contracted segments compared to preserved specimens (Fig. 21), so this measurement differs between the two. Based on the illustrations of the holotype of E. longicirrata, estimates of cirri length were measured fromlivespecimens while the IN2017_V03 specimens were measured from preserved specimens.

Incorporating the COI sequences of IN2017_V03 specimens and all available sequences of the closely allied genus Sige in GenBank with those used by Teixeira et al. (2020) places these samples within a clade of Sige spp. and sister group to Sige fusigera Malmgren, 1865 (Fig. 22). The Australian specimens, however, lack the characteristic pointed superior parapodial lobe of other Sige species (Eklöf et al., 2007). San Martín et al. (2021) found Sige, Eumida and other closely related genera to be paraphyletic and polyphyletic in their analyses. Further investigation is warranted to explore the relationship among these genera. This isthe first report of eyeless Eumida species occurring at 1000 m. Previousrecordswerecollectedatdepths> 3000 m.