Gonionemus agilis Watson & Govindarajan 2017, sp. nov.

Gonionemus agilis sp. nov.

Figures 1A–C, 2A–G

Material examined. Holotype, NMV F222410, one specimen, sex undetermined, from brown alga Cystophora monilifera in tide pool, Shoreham, Victoria, collected by Audrey Falconer, 20 October 2013; Paratype, NMV F211740, one specimen, sex undetermined, on brown alga Cystophora monilifera in tide pool, Shoreham, Victoria, collected by Audrey Falconer, 18 May 2014.

Description (from holotype, paratype and other live material). Umbrella of resting medusa subhemispherical, without apical peduncle, mesoglea moderately thick, velum almost closed (Fig. 1C). Manubrium long, cylindrical to folded, lips deeply folded into four sections, four radial canals with a thin inner band, outer part of canal broad, widest at manubrium, decreasing in width to bell margin (Fig. 1C). Fifty to 60 long, straight, hollow cylindrical unbranched tentacles evenly distributed around bell margin, each with a basal marginal bulb. Proximal third of tentacle without nematocysts, distal two thirds with nematocysts scattered in clusters or in arcuate pads, end of tentacle with nematocyst pads and a terminal leaf shaped appendage armed with nematocysts. No ocelli present; small spherical statocysts present between tentacles.

Gonads on radial canals, developing in small transparent clusters at junction of radial canal with ring canal and at manubrium, but not completely extending along canals, each developing cluster with internal luminescent green spots, the spots becoming separated and fading as gonad matures. Measurements of Gonionemus agilis are given in Table 1.

total width across extended tentacles 13–14 width across bell 2.6–2.7 height of bell 3.6–4.0 length of extended tentacle 6.75 Cnidome: two categories of nematocysts; one category in three size classes.

microbasic euryteles, capsule elongate oval, 28– 30 x 11–12 µm, shaft 27–28 µm long, head with moderately long spines, thread long, ropy, with spirals of small spines; abundant in tentacular nematocyst pads; easily discharged.

microbasic euryteles, capsule broadly loaf-shaped, 11– 15 x 6–8 µm, shaft 6 µm long, sparingly scattered over umbrella, discharges sideways; difficult to discharge.

microbasic euryteles, capsule small, slender loaf-shaped, 10– 11 x 4–5 µm, shaft 10 µm long, head with a few spines, in nematocyst pads in tentacles, and sparingly scattered over medusa; difficult to discharge.

atrichous isorhizas, capsule elongate oval, 37– 48 x 11–12 µm, thread long, tubular, without spines, abundant in tentacular nematocyst pads; difficult to discharge.

Umbrella of living medusae pale transparent orange, tentacles transparent white, tentacular nematocyst pads white, radial canal and manubrium orange-brown, lips pale pink, developing gonads clear transparent with luminescent green spots, mature gonads pale fawn.

Etymology. The species name refers the very active grasping adhesive pad.

DNA sequences and phylogenetic analysis. All sequences were deposited in GenBank (accession numbers MF135184 - MF135189, Table 2). Gonionemus agilis COI differed by 98 out of 639 base pairs, or 15.3%, from Gonionemus sp. and 92 out of 639 base pairs, or 14.4%, from Gonionemus vertens. There were no gaps and 4 amino acid substitutions between G. agilis and G. vertens / G. sp.; one amino acid substitution between G. vertens and G. agilis / G. sp; and one amino acid substitution between G. sp. and G. agilis / G. vertens. G. agilis 16S differed by 60 and 56 out of 578 base pairs, or 10.4% and 9.7%, from Gonionemus sp. and G. vertens, respectively (including one gap). G. agilis 16S differed by 49 out of 427 base pairs, or 11.4% (including one gap), from Scolionema suvaense (Agassiz & Mayer, 1899).

The maximum likelihood and Bayesian phylogenetic analyses were generally consistent with each other and similar to Collins et al. (2008). Gonionemus agilis grouped in a clade that was highly supported in the Bayesian analysis but not in the ML analysis and that included Cubaia, Gonionemus and Scolionema (Fig. 3), which fell within another strongly supported clade that included Aglauropsis aeora Mills, Rees & Hand, 1976. The Olindias group formed another strongly supported clade, as did the geryoniids, and a group including Maeotias, Limnocnida, Craspedacusta, and Astrohydra. However, the arrangement of these clades relative to each other is not well-supported.

Remarks. T here are three accepted species of Gonionemus (see Schuchert 2013): G. vertens A. Agassiz, 1862, G. chekiangensis Ling, 1937 and G. hamatus Kramp, 1965. G, vertens is a variable species with a reputation for painful stings in certain populations and is widely distributed in the northern hemisphere (Arai & Brinckmann-Voss 1980, Govindarajan & Carman 2016). Based on mitochondrial COI sequences and morphological observations, the western Pacific and western Atlantic populations that are associated with stings may represent a distinct species referred to here as Gonionemus sp. (see Govindarajan et al., 2017). G. vertens has been recently recorded from Argentina in the southern hemisphere where it is assumed to have been anthropogenically introduced (Rodriguez et al. 2014), G. chekiangensis recorded from the northern Pacific is a much larger species than G. agilis (see Xu et al. 2014) and G. hamatus has been found only once in southern Australia where it is known from one specimen captured in 1961 in the plankton at Adelaide, South Australia. Attempted rehabilitation of the dehydrated holotype specimen (SAM H143, loaned by the South Australian Museum) was unsuccessful. The specimen is flattened and the contents of the bell degraded, however enough of one tentacle remained to show the nematocysts are arranged in circumferential bands as described by Kramp (1965).

Gonionemus agilis differs in the following morphological respects from known species of Gonionemus:

(i) Location of adhesive pads. In both G. vertens and G. chekiangensis the adhesive pads are not at the tips but further up the tentacles. Kramp (1965) described the pads of G. hamatus as being at the tip of the tentacles. (ii) Distribution of tentacular nematocysts. In G. agilis they are not arranged in circumferential bands as described for the other three species but are in scattered discrete clusters or arcuate pads.

(iii) Morphology of gonads. All other species of Gonionemus are reported to have gonads folded in a zig-zag pattern along their entire length. In various stages of maturity those of G. agilis showed no evidence of extending completely along the radial canals and only a few specimens showed incipient zig-zag folding of the clusters just above the ring canal.

Statocysts are very difficult to find in G. vertens (Thomas 1921, Brinckmann-Voss pers. comm.). Statocysts may also be an incipient structure in Gonionemus agilis as examination (x 1000) of several crushed specimens found only a few spherical bodies (c. 15 µm diameter) believed to be statocysts. Yet another specimen (see Fig. 1C) clearly showed statocysts between some of the tentacle bases.

Taxon Gene Genbank Reference

Accession number

Gonionemus agilis 16S MF135187 This study

COI MF135184 This study

Gonionemus vertens 16S MF135188 This study

COI MF135185 This study

Gonionemus sp. 16S MF135189 This study

COI MF135186 This study

Scolionema suvaense 16S AB720909 Gotoh,R.O., Ito,C., Mochizuki,S. & Hanzawa,N., unpublished

Olindias sambaquiensis 16S EU293977 Collins et al. 2008

Olindias phosphorica 16S EU293978 Collins et al. 2008

Olindias formosus 16S KF184031 Patry, W., Knowles, T., Howard, M. & Christianson, L., unpublished

Maeotias marginata 16S AY512508 Collins et al. 2005

Limnocnida tanganyicae 16S EU293972 Collins et al. 2008

Astrohydra japonica 16S EU293975 Collins et al. 2008

Aglauropsis aeora 16S EU293973 Collins et al. 2008

Craspedacusta sinensis 16S AY512507 Collins et al. 2005

Craspedacusta sowerbii 16S KY077294 1 Grange et al. 2016

16S EU293971 2 Collins et al. 2008

16S NC018537 3 Zou et al. 2012

Cubaia aphrodite 16S NC016467 Kayal et al. 2012

Geryonia proboscidalis 16S EU293979 Collins et al. 2008

Liriope tetraphylla 16S EU293980 1 Collins et al. 2008

16S KT266624 2 Maronna, M.M., Miranda, T.P., Pena Cantero, A.L., Barbeitos, M.S. & Marques, A.C., unpublished 16S EU999222 3 Kirby R.R., Licandro, P. & Conway, D.V.P., unpublished 16S JQ715945 4 Zheng et al. 2014

16S JQ715944 5 Zheng et al. 2014

16S KJ859220 6 Liu, M., unpublished 16S AY512510 7 Collins et al. 2005

Monobrachium parasitum 16S EU293970 Collins et al. 2008

Little attention has been given to the cnidome of Gonionemus. Arai and Brinckmann-Voss (1980) reported without description or figure, microbasic euryteles and basitrichous or atrichous isorhizas in G. vertens. The cnidome of living G. agilis supports their observations.

Green luminescent spots at the junction of the radial canal with the ring canal were mentioned by Ling (1937) in his description of G. chekiangensis. It has not been reported by other authors. In G. agilis the brilliant transparent green spots are in clusters in the earliest developing gonads at the ring canal and manubrium and are sometimes visible along the radial canals as the gonads mature.

The living medusa is very active, the tentacles reacting immediately to touch and the terminal adhesive pad is in constant waving motion. The manubrium is very extensile, changing shape and sweeping well beyond the umbrella margin when the bell is flattened. The lips are in constant motion, rapidly changing shape from quadrate to fluted and swastika shapes.

Medusae collected in May, 2014 were kept in a laboratory dish for 25 hours during which time they remained immobile with tentacles extended. When a small branch of Cystophora was placed in the dish the tentacles became very active, the adhesive pads grasping the alga, the medusa then scrambling onto the alga with which it closely blended in colour. The association of the crawling G. agilis with Cystophora monilifera is of interest since Schuchert (1996) recorded another crawling medusa, Staurocladia wellingtoni Schuchert, 1996 from Cystophora in New Zealand.

Gonionemus agilis may possibly be conspecific with Gonionemus hamatus as they are both from southern Australian localities separated by 1,200 km of coastline. The degraded condition of the type specimen and lack of detail in Kramp’s figures of G. hamatus preclude more detailed comparison. Thus for the present and until more material is found in the Adelaide region, G. agilis is considered a distinct species.

The molecular results also support the morphological observations that G. agilis is a distinct species. The position of G. agilis relative to G. vertens and Gonionemus sp. in the 16S phylogeny is consistent with the observed morphological uniqueness of G. agilis, at the same time demonstrating its position in the Gonionemus – Scolionema clade.

The relatively large genetic distances between G. agilis and other species of Gonionemus for 16S (~10%) and COI (~15%) indicate that both of these markers will be useful for DNA barcoding. In general, one of the limitations of DNA barcoding is that barcode sequences are still not available for most species (Bucklin et al. 2011). By providing DNA barcode sequences as part of a formal taxonomic description, accurate barcode-based taxonomic assignment is ensured (Dayrat 2005).