Oxydromus humesi, (PETTIBONE, 1961), 2015

Martin, Daniel, Nygren, Arne, Hjelmstedt, Per, Drake, Pilar & Gil, João, 2015, On the enigmatic symbiotic polychaete ‘ Parasyllidea’ humesi Pettibone, 1961 (Hesionidae): taxonomy, phylogeny and behaviour, Zoological Journal of the Linnean Society (Zool. J. Linn. Soc.) 174 (3), pp. 429-446 : 434-441

OXYDROMUS HUMESI ( PETTIBONE, 1961) COMB. NOV.

FIGURES 3–7

Parasyllidea humesi . Pettibone (1961: 173–175, fig. 4A– H); Pleijel (1998: 162); Martin et al. (2012: 147–148, figs 2–3).

Material examined: Paratypes. USNM 30012 , c. 50 specimens fixed in formalin and preserved in 70% ethanol, plus one prepared for SEM, Luango , Republic of Congo (West Africa); intertidal, mangrove muddy bottoms . Additional material. MNCN 16.01 /16090, 20 specimens, and CEAB.A.P. 854 A, 20 specimens, fixed in a 4% formalin–seawater solution and preserved in 70% ethanol, Rio San Pedro ( Cádiz Bay , southern Atlantic coast of the Iberian Peninsula), intertidal muddy bottoms. GNM Polychaeta 86193: voucher specimen from molecular analysis (Table 1), fixed and preserved in 95% ethanol. Specimens for SEM. SMNH 8677–8679 , three specimens, preserved in osmium tetroxide solution, conserved in 70% ethanol, dehydrated and gold coated; CEAB.A.P. 854 B, one specimen, fixed in a 4% formalin– seawater solution, conserved in 70% ethanol, dehydrated and gold coated; all collected at Rio San Pedro ( Cádiz Bay , southern Atlantic coast of the Iberian Peninsula), intertidal muddy bottoms .

Description: Luango population: length up to 23 mm, width up to 3 mm (with parapodia), chaetigers up to 79. Cádiz population: length up to 37 mm, width up to 3 mm (with parapodia), chaetigers up to 75.

Live animals rusty orange to tan, with reddish brown eyes, becoming light pink to yellow when preserved, with dark brown eyes ( Fig. 3 ).

Body outline in dorsal view long, rounded anteriorly, tapering gradually posteriorly, with median parapodia much longer than anterior and posterior ones ( Fig. 3A ). Body flattened dorso-ventrally, elliptical in transverse view, with dorsal side more vaulted than ventral side. All cirri smooth, very deciduous in fixed material.

Prostomium rounded subrectangular, wider than long, with a median ridge, less evident or absent when pharynx everted ( Figs 3B , 4 , 7A ). Two pairs of small reddish dark brown eyes very close to each other; anterior pair slightly larger, positioned slightly further apart, forming a narrow trapezium with posterior pair ( Fig. 7A ). Antennae inserted frontally on anterior prostomial margin. Median antenna very small, button like, inserted centrally on anterior edge of prostomial median ridge ( Figs 4F , 5B ), probably eversible, visible under LM on c. 70% (Cádiz) and 45% (Luango) of speci- mens. Lateral antennae well developed, conical, with tapering tips, inserted laterally to furrows of median prostomial ridge ( Figs 3 , 4 , 5A, B , 7A ). Palps biarticulated, inserted lateral and slightly ventral to lateral antennae ( Figs 3C , 4 , 5A, B ). Palps and paired antennae subequal in size. Palpophores (cylindrical) and palpostyles (flask-shaped) equal in length. Nuchal organs on lateroposterior prostomial margin, just behind ocular region, mid-dorsally separated ( Fig. 4D ). Prostomium without posterior incisions. Facial tubercle present. Lip pads absent.

Proboscis solid, muscular ( Fig. 7A ), with wider basal ring and terminal narrower ring, which is shorter middorsally and mid-ventrally. Both rings with smooth surface. Terminal ring ciliated, without papillae. Jaws absent. Non-everted proboscis extending internally from prostomium to about segments 8–9.

Segment 1 dorsally reduced, visible only in lateral view; segments 2–3 distinct in dorsal view. Parapodial lobes and chaetae absent on segments 1–3. All cirrophores of anterior cirri clearly distinct, not fused. Dorsal cirri of segments 1–3 not clearly distinguishable from following ones, with stronger cylindrical cirrophores and tapering cirrostyles; first cirrostyle shortest, second longest; first pair of dorsal cirri lateral to prostomium, extending to about segment 4; second and third pairs posterior to prostomium, both extending to about segment 5. Ventral cirri of segments 1–3 progressively long, clearly longer, thicker and stouter than following ones, with well-delineated cylindrical cirrophores.

Parapodia prominently subbiramous from segment 4 onwards ( Figs 3 , 6 , 7A ), about as long as body width; notopodial lobes achetous, with dorsal cirri of slightly different lengths; neuropodial lobes chaetigerous, with ventral cirri similar to following ones. Notopodia with about five slender, curved acicula reaching cirrophores basally; parapodial lobe long, half as long as neuropodial one, with cirrophore on tip. Notopodia slightly elevat- ed on segments 5, 8, 10, 12, 15 and 17, thereafter on every second segment. Longer dorsal cirri not always on elevated notopodia, usually on segments 4, 5, 7, 8, 10, 12, 15 and 17, thereafter on every second segment. Short dorsal cirri as long as second dorsal cirri; long dorsal cirri about one-third longer. Cirrostyles of dorsal cirri reaching slightly further than chaetal tips. Neuropodia elongated, subconical, with blunt tips, long digitiform prechaetal lobe and rounded postchaetal lip. Ventral cirri slender, fusiform, with blunt tips, about as long as prechaetal lobes, extending beyond acicular lobe tips, usually reaching only following segment, ventrally inserted at one-third from neuropodial tips. Cirrophores absent or indistinct. About three ambercoloured, thick acicula, with tips inserted in basal region of prechaetal digitiform lobe.

Notochaetae absent. Neurochaetae 40–60, ambercoloured, compound, with heterogomph shafts ( Fig. 5C ); blades unidentate with curved, blunt tips, finely denticulate cutting edges ( Fig. 5E, G ), and a long subdistal spine reaching tip of curved end ( Fig. 5D, F ); blades 45–250 μ m long, being longer in median region of bundle, slightly shorter dorsally, and much shorter ventrally.

Pygidium with one pair of pygidial cirri, similar in shape and length to dorsal cirri ( Figs 3A , 5H ). Pygidial papilla absent. Anus terminal.

Remarks: From a morphological point of view, the Iberian and Congolese specimens are identical. The differences in size (the former being slightly larger than the latter) can be attributed either to growth (i.e. specimens collected at different phases of their life cycle) or to the preservation (i.e. directly placed into the fixative solution or being previously relaxed as in the Congolese and Iberian specimens, respectively).

In the cephalic region, the main differences concern the central antenna and the median ridge. The central antenna was more frequently seen in the Iberian than in the Congolese worms. However, although this could be a relevant character when identifying the species, it is certainly not relevant from a systematic point of view, as the fact that the antenna can be retracted or detached (i.e. non-visible) does not necessarily mean that it is absent in the species. Accordingly, both populations must be described as having central antennae and so, morphologically, both belong to Oxydromus . Concerning the median ridge, there is high variability in the traces of its existence. While it is clearly marked in all specimens from both populations with a non-everted pharynx, the degree to which this structure is patent depends strongly on the level of ever- sion of the pharynx, to the point that it almost disappears if the pharynx is fully protruded. However, even in these cases it is always possible to distinguish the traces of the presence of the central ridge in all specimens we have observed. Again, these traces tend to be more evident in the Iberian than in the Congolese specimens, but we strongly suggest that this depends on the differences in the fixation procedure.

As for the morphology of the chaetae, this tends to be a very conservative character among the species of Oxydromus . There is considerable intraspecific variation in the length of chaetal blades ( Pleijel, 1998) and only slight differences in the shape of distal tips (F. Pleijel, pers. comm.). Therefore, both characters proved difficult to delineate, and this is further complicated by the high variation we observed within parapodia. Despite this variability, the blades of some chaetae from the Iberian population appear to be slightly wider and have slightly more hooked tips and slightly longer marginal spines than the Congolese ones ( Fig. 5F , left), while others have marginal spines similar in length to the Congolese ones ( Fig 5F , right). Moreover, parts of some chaetae of the Iberian population appear to be markedly rough ( Fig. 5F ), while others are clearly smooth ( Fig. 5G ), a character that we attribute to a fixation/ preservation artefact. Therefore, we consider that the observed differences based on chaetal morphology are not conclusive.

Taking into account that the specimens of both populations had no evident morphological differences, we tried to check them from a morphometric perspective using some parapodial characters (dorsal cirrostyle, dorsal cirrophore, neurochaetal lobe and ventral cirri). However, the measurements of these appendages, as well as their relative proportions, were highly variable ( Table 2). The morphometric results were thus nonconclusive and would have a real significance only if a large number of individuals preserved under the same conditions could be analysed (far beyond our current ability).

Locality: Known from mangrove swamps at Luango, north to Pointe Noire, Middle Congo ( Republic of Congo, West Africa) and from salt marsh creeks at Río San Pedro and Caño Sancti Petri (Cádiz Bay, Iberian Peninsula).

Reproduction: Some observed specimens fom Cádiz Bay were females, with intracoelomic oocytes being visible through the body wall, measuring about 140 μ m in diameter ( Fig. 3C ).

Ecology: The species is a strict bivalve symbiont, probably a parasite according to Martin et al. (2012), living in association with Tellina nymphalis Lamarck, 1818 in Luango (mangrove swamps) and Scrobicularia plana

DSC, dorsal cirrostyle; DCP, dorsal cirrophore; NCL, neurochaetal lobe; VC, ventral cirri.

(intertidal) and Psammotreta cumana (shallow subtidal) in Cádiz Bay saltmarshes. The original description by Pettibone (1961) reported the species as living in association with ‘ T. nymphalis Lamarck’. This species, however, apparently inhabited shallow sandy beaches and is currently considered as a taxon inquirendum ( Sartori, 2014). In turn, there was another species, Tellina nymphalis sensu Hanley, 1846 , currently accepted as Tellina innominata ( Bertin, 1878) which is a typical inhabitant of mangrove swamps ( R. Von Kossel, Museum National d’Histoire Naturelle, Paris, pers. comm.). Thus, we strongly suggest that the host species infested by the Congolese population of O. humesi should be T. innominata rather than T. nymphalis . However, this can only be demonstrated after examination of the original material, which is beyond the scope of the present paper.

The population of O. humesi associated with S. plana (hundreds of bivalves examined) has a regular distribution, with infestation intensity of one adult per host (except for a single finding of one male and one female together) and a prevalence of less than 5% ( Martin et al., 2012). The population associated with P. cumana was collected by accident and 56 specimens of this host were studied. However, the distribution was also regular, with one adult per host (except two occurrences of one male and one female together; and a single occurrence of three worms corresponding to one male, one female and one juvenile). The prevalence was very high, with 33 bivalves infested over a total of 56 (59%). However, note that 79% of the larger hosts (42 specimens with shell length> 20 mm) were infested, while only one within the 14 smaller hosts (<20 mm in shell length) harboured the symbiont. In parallel with this finding, there was also a double record (probably ac- cidental, see discussion below) of O. humesi inside the specimens of C. glaucum transported from the field to laboratory inside the same container as P. cumana .

All observed worms initially show the same behaviour when deposited in the respective Petri dish containing one specimen of S. plana . They swim or crawl until contacting the bivalve and then quickly move to lie between the shell and the glass bottom ( Fig. 8A 1– A 3 , Film S2). Soon after this, the worms start to crawl over and below the shell until reaching either the foot or siphon regions ( Fig. 8B 1–B 3 , Film S2). Three of the worms entered the shell so quickly that it was not possible to observe how they did it.

Some worms first reached the siphon region, triggering a characteristic shell-entering behaviour. Before attempting to enter, the worms gently touched the edge of the shell until the host protruded its siphons. The worms then started to touch the siphons with their anterior appendages, crawling gently around them. Attempts to use the exhalant siphon failed ( Fig. 9A 1– A 3 , Film S3), and also failed to enter through the inhalant siphon of a host that was relatively too small in comparison with the worm. This particular worm immediately tried to enter into the small host through the basal area of the inhalant siphon. However, this position did not allow the worm to enter into the branchial cavity and so its head emerged through the opposite side at the basis of the inhalant siphon. After several movements, the bivalve attempted to use the foot to expel the worm, which was finally accomplished by a sudden compression of the valves ( Fig. 9B 1– B 3 , Film S4).

Among the successful entries observed, one worm first contacted with the foot and quickly entered into the host by the apex-orientated side of the foot ( Fig. 10A 1–A 3 , Film S5). The remaining worms spent some time gently touching the inhalant siphon opening, which opened widely. This allowed the symbionts to introduce their whole body inside, starting from the anterior end. Once completely introduced, the host usually excreted pseudofaeces ( Fig. 10B 1–B 3 , Film S6).

All worms that entered their hosts remained inside until the end of the experimental observations, with some occasional external excursions, re-entering the host soon after.

The three specimens placed with C. edulis , R. decussatus and M. galloprovincialis initially showed the same behaviour, lying first below the bivalve shells and then crawling on their upper surfaces. No attempts were made to enter the first two species. The worm in contact with the mussel entered the bivalve through its opened valves, then left and entered several times before remaining finally below the shell. When

PARASYLLIDEA SYNONYMIZED WITH OXYDROMUS 441

placed again into Petri dishes with S. plana , all three worms quickly repeated the typical host-entering behaviour and remained inside their hosts until the end of the experimental observations.

Some of the worms kept outside their hosts were observed to attack their conspecifics ( Fig. 11 , Film S7). In some attacks, one worm bit the other on its anterior end, holding it until the shaking movements of the victim forced the aggressor to release it. Other worms attacked the lateral regions of their conspecifics, causing the loss of some parapodia in the victim. After several hours outside the hosts, only a few, highly aggressive worms were able to survive.

Figures

Figure 1. Location of sites where hosts were collected at Cádiz Bay. 1, Scrobicularia plana; 2, Cerastoderma glaucum and Psammotreta cumana. Images obtained from Google Earth, version 7.1.2.2041, (c) Google 2015.

Figure 3. Oxydromus humesi. Digital photographs of a living specimen from the Cádiz Bay. A, entire specimen, dorsal view. B, anterior end, dorsal view. C, anterior end, ventral view; intracoelomic oocites visible through body wall (white arrow). Scale bars are in centimetres.

Figure 4. Oxydromus humesi. SEM photographs of the specimens from Congo (A, C, E) and Cádiz Bay (B, D, F). A, B, anterior end, dorsal view. C, D, detail of the anterior end, dorsal view; white arrow indicates the position of the nuchal organ. E, F, detail of the prostomium, dorsal view; the arrows indicate the prostomial median ridge and the expected position of the minute central antennae (visible only in F). Scale bars are micrometres.

Figure 5. Oxydromus humesi. SEM photographs of the specimens from Cádiz Bay (A, B, F, G) and Congo (C, D, E, H). A, detail of the anterior end, ventral view. B, detail of the prostomium, ventral view; arrow pointing to central antennae. C, mid-body compound chaetae. D, F, details of the blade tip. E, G, details of the basal cutting edge of the blade. H, posterior end showing the pygidial cirri, lateral view. Scale bars are micrometres.

Figure 6. Oxydromus humesi. SEM photographs of the mid-body segments of specimens from Congo (A, C) and Cádiz Bay (B, D). A, B, dorsal view. C, D, ventral view. Scale bars are 300 μm (A), 500 μm (B–D).

Figure 7. Oxydromus humesi. Stereomicrographs of a preserved specimen from Cádiz Bay. A, anterior region, dorsal view. B, chaetiger 5 (without chaetae). C, chaetiger 20 (without chaetae).

Figure 8. Oxydromus humesi and Scrobicularia plana. First contacts with the host. A1–A3, contact with the bivalve and displacement to lie between shell and glass bottom. B1–B3, crawling over the host shell until reaching the siphons.

Figure 9. Oxydromus humesi and Scrobicularia plana. Failed attempts to enter the host. A1–A3, through the exhalant siphon. B1–B3, through the basis of the siphon; B1, anterior and mid-posterior ends of the symbiont outside the host at the lateral of the siphonal region; B2, foot of the host trying to discourage the symbiont from entering the host; B3, the symbiont fully outside the host after being expelled. Arrows indicate the head of the symbiont.

Figure 10. Oxydromus humesi and Scrobicularia plana. Successful attempts to enter the host. A1–A3, through the apex- orientated side of the foot. B1–B3, through the inhalant siphon.

Figure 11. Oxydromus humesi. A, B, sequences of two specimens fighting each other. Arrows indicate the heads of both worms.