INTRODUCTION
Marine shallow-water species currently face a multitude of ecological challenges. Coastal waters are increasingly subject to pollution, pathogens, hypoxia, overfishing, non-native species and other anthropogenic and natural disturbances. Additionally, shallow-water species are more exposed to increased UV radiation, warming temperatures and acidification than inhabitants of deeper water. We are only beginning to understand the long-term consequences of these environmental stressors, separately and in combination, on biological communities. Shallow-water coral reefs, in particular, are in decline worldwide, causing concern about the loss of biodiversity associated with hermatypic corals (e.g. Pandolfi et al., Reference Pandolfi, Connolly, Marshall and Cohen2011; De'ath et al., Reference De'ath, Fabricius, Sweatman and Puotinen2012; Descombes et al., Reference Descombes, Wisz, Leprieur, Parravicini, Heine, Olsen, Swingedouw, Kulbicki, Mouillot and Pellissier2015). Some shallow-water invertebrate species appear to be better equipped than others to cope with particular challenges. For example, calcifying species are especially vulnerable to acidification (Kroeker et al., Reference Kroeker, Kordas, Crim and Singh2010, Reference Kroeker, Kordas, Crim, Hendriks, Ramajo, Singh, Duarte and Gattuso2013).
In this contribution, we examine the potential of the bearded fireworm, Hermodice carunculata (Pallas, 1776) (Figure 1), to survive – and potentially thrive – in a changing ocean environment. We argue that its ability to withstand environmental extremes and fluctuations, its predator avoidance strategies and its non-selective diet will probably benefit its survival and possibly lead to its widespread emergence as a nuisance species although the effects of ocean acidification could interfere with its defence mechanisms in the long term.
Fig. 1. The bearded fireworm, Hermodice carunculata, observed off the South Florida Atlantic coast.
Hermodice carunculata is a common species of amphinomid polychaete with a distribution throughout the Atlantic Ocean, the Caribbean, Gulf of Mexico, Mediterranean and Red Seas (Yáñez-Rivera & Salazar-Vallejo, Reference Yáñez-Rivera and Salazar-Vallejo2011; Ahrens et al., Reference Ahrens, Borda, Barroso, Paiva, Campbell, Wolf, Nugues, Rouse and Schulze2013). The common name refers to the tufts of ‘harpoon chaetae’ which are flared when the worm is threatened, causing serious irritation at the site of contact.
Hermodice carunculata is primarily reported from warmer waters, but one questionable record exists from as far north as the Dogger Bank in the North Sea (Fauvel, Reference Fauvel1923). In the south, it ranges to Rio de Janeiro in the west and to the Gulf of Guinea in the east (Ahrens et al., Reference Ahrens, Borda, Barroso, Paiva, Campbell, Wolf, Nugues, Rouse and Schulze2013). It has also been reported from Ascension and St. Helena Islands in the South Central Atlantic (Yáñez-Rivera & Brown, Reference Yáñez-Rivera and Brown2015). Hermodice carunculata inhabits primarily shallow water, including the intertidal zone, but has been reported to a maximum of over 300 m depth (Ehlers, Reference Ehlers1887). The species is common in a variety of habitats, such as coral reefs and seagrass beds, as well as artificial structures like pilings, bridge spans and shipwrecks. Hermodice carunculata is primarily active from dusk to dawn and often hides in crevices, under overhangs or underneath rocks throughout the day.
Ahrens et al. (Reference Ahrens, Borda, Barroso, Paiva, Campbell, Wolf, Nugues, Rouse and Schulze2013) showed that H. carunculata is genetically homogeneous throughout its distribution range, suggesting high dispersal capabilities. Unfortunately, little is known about the larval development of the species. Based on chaetal characteristics, a planktotrophic larval type known as rostraria is perhaps associated with amphinomids (Bhaud, Reference Bhaud1972), but this association has never been confirmed by direct observation of metamorphosis into a juvenile or by DNA barcoding of the larva. Even accepting that the rostraria is an amphinomid larva, assignments to particular species cannot currently be made.
Apart from pelagic larvae, H. carunculata may also disperse by rafting. Some studies have reported amphinomids rafting on marine debris (Donlan & Nelson, Reference Donlan and Nelson2003; Thiel & Gutow, Reference Thiel and Gutow2005; Farrapeira, Reference Farrapeira2011; Borda et al., Reference Borda, Kudenov, Bienhold and Rouse2012), although none specifically mentions H. carunculata. McIntosh (Reference McIntosh1885) mentions a large specimen of H. carunculata swimming near the water surface.
HERMODICE CARUNCULATA IS TOUGH
Many amphinomid species occur in habitats that are commonly described as ‘extreme’. A few examples are Archinome species which inhabit hydrothermal vents and seeps (Borda et al., Reference Borda, Kudenov, Chevaldonne, Blake, Desbruyeres, Fabri, Hourdez, Pleijel, Shank, Wilson, Schulze and Rouse2013), Benthoscolex cubanus, a commensal or parasite in the body cavity of sea urchins (Emson et al., Reference Emson, Young and Paterson1993), and Linopherus canariensis, a potentially invasive species in a hypersaline lagoon in Sicily (Cosentino & Giacobbe, Reference Cosentino and Giacobbe2011).
Hermodice carunculata is common in Caribbean coral reefs, including reef crests exposed at low tide, with significant short-term fluctuations in temperature, salinity and dissolved oxygen. During these fluctuations, the metabolic rate, as measured by oxygen consumption, only changes marginally (Sander, Reference Sander1973; Ferraris, Reference Ferraris1981). The species can be abundant in organically enriched areas where microbial activity can lead to oxygen depletion, such as the benthos underneath fish farms (Heilskov et al., Reference Heilskov, Alperin and Holmer2006; Riera et al., Reference Riera, Pérez, Rodríguez, Ramos and Monterroso2014) or coral algae interfaces (Smith et al., Reference Smith, Shaw, Edwards, Obura, Pantos, Sala, Sandin, Smriga, Hatay and Rohwer2006). In the Azores, H. carunculata has been reported from the shallow-water hydrothermal vents at D. João de Castro Seamount (Cardigos et al., Reference Cardigos, Colaço, Dando, Ávila, Sarradin, Tempera, Conceição, Pascoal and Serrão Santos2005). Remarkably, the worms occur in very close proximity (<1.5 m) to the vents, where fluids with elevated temperatures of up to 63.3°C, low pH and high sulphide and heavy metal concentrations are released (Cardigos et al., Reference Cardigos, Colaço, Dando, Ávila, Sarradin, Tempera, Conceição, Pascoal and Serrão Santos2005). Shiber (Reference Shiber1981) reports that in the heavily polluted Ras Beirut, on the Mediterranean coast of Lebanon, H. carunculata appears to be unaffected by blasts from dynamite fishing and will feed on dead or paralysed fish sinking to the seafloor. Among benthic invertebrates in Ras Beirut, H. carunculata was the species with the highest concentrations of lead, cadmium, nickel, iron and zinc. Hermodice carunculata is also frequently reported from marine and anchialine caves in the Caribbean (Frontana-Uribe & Solís-Weiss, Reference Frontana-Uribe and Solís-Weiss2011), the Mediterranean (Gerovasileiou et al., Reference Gerovasileiou, Chintiroglou, Vafidis, Koutsoubas, Sini, Dailianis, Issaris, Akritopoulou, Dimarchopoulou and Voutsiadou2015; Knittweis et al., Reference Knittweis, Chevaldonné, Ereskovsky, Schembri and Borg2015) and the Azores (Micael et al., Reference Micael, Azevedo and Costa2006), ranging from the cave entrance to the dark zone.
Like many annelids, amphinomids have the ability to regenerate missing body sections after injury. Eurythoe complanata even routinely goes through cycles of asexual reproduction during which the worms fragment into two or more parts and can regenerate both anterior and posterior body sections (Kudenov, Reference Kudenov1974). To date, only posterior regeneration has been demonstrated in H. carunculata (Ahrens et al., Reference Ahrens, Kudenov, Marshall and Schulze2014). Posterior fragments without a head can survive and remain active for several weeks in an aquarium setting but no new head formation has been observed (pers. obs.).
HERMODICE CARUNCULATA IS ARMED
Annelid bristles, or chaetae, are generally chitinous structures. Amphinomid chaetae are unique in that they contain calcium carbonate in addition to chitin (Gustafson, Reference Gustafson1930). Each parapodium carries tufts of dorsal notochaetae and ventral neurochaetae. In Hermodice carunculata, the notochaetae may be smooth and hair-like or distally serrated ‘harpoon chaetae’ (Gustafson, Reference Gustafson1930; Yáñez-Rivera & Salazar-Vallejo, Reference Yáñez-Rivera and Salazar-Vallejo2011). Harpoon chaetae may be erected, or even ejected, for defence (Penner, Reference Penner1970; Halstead, Reference Halstead, Bücherl and Buckley1971). When touched, they will penetrate human skin and, thanks to the serration, remain stuck in it. The neuropodial tuft probably only contains a single type of chaetae (Gustafson, Reference Gustafson1930). The texture of the chaetae may be erodible and may depend on the status of regeneration after they have been shed. Therefore they are not used as diagnostic characters. However, as they play an important role in defence and possible prey capture, chaetal structure and arrangement should be further investigated.
It is still unclear whether the irritation the chaetae cause is merely mechanical or whether they are actually venomous. Although no toxins specifically associated with the chaetae have been identified to date, there are indications that venoms are utilized. Localized reactions in the affected area include an acute, intense stinging pain, itchiness, numbness and swelling (Smith, Reference Smith2002). These symptoms can last up to several weeks. More notably, however, in rare cases, systemic reactions such as nausea, cardiac and respiratory problems can occur (Ottuso, Reference Ottuso2013). The recommended treatment is to remove the bristles with tape, to treat the area with vinegar and to apply hot water (Smith, Reference Smith2002). The vinegar may dissolve the calcium carbonate in the chaetae. The heat treatment implies that a toxin is involved which can be denatured by heat.
It has long been assumed that toxins are released through a hollow core of amphinomid chaetae (e.g. Nakamura et al., Reference Nakamura, Tachikawa, Kitamura, Ohno, Suganuma and Uemura2008; von Reumont et al., Reference von Reumont, Campbell, Richter, Hering, Sykes, Hetmank, Jenner and Bleidorn2014), but some studies have shed doubt on this interpretation. Under light microscopy, the clear core does appear hollow and sometimes a small amount of fluid seems to be released from the tip of the chaeta (Figure 2A). However, histological sections do not reveal any glands near the bases of the chaetae in H. carunculata (pers. obs.) or Eurythoe complanata (Eckert, Reference Eckert1985). Gustafson (Reference Gustafson1930) found that the core is actually filled with a clear gelatinous substance consisting of individual fibrils with a hexagonal cross-section. He attributes the toxic nature of the chaetae to this substance. He described that only the outer sheath of the chaetae, including the recurved hooks, when present, are calcareous. In contrast, Tilic et al. (Reference Tilic, Pauli and Bartolomaeus2016), based on ultrastructural observations on Eurythoe complanata, postulate that the central core is also filled with calcium carbonate, contributing to the brittleness of the chaetae. According to their study, the calcium carbonate is deposited after the large central microvilli of the chaetoblast retract and their canals fuse together. When exposed to acidic conditions (e.g. many fixatives), the calcium carbonate may dissolve and leave a central cavity. In some cases, we have observed pieces of tissue adhering to the base of the chaetae (Figure 2C, D) which we interpret to be the chaetoblasts. Scanning electron micrographs reveal that chaetae may also be grooved, adding another potential conduit for toxins (Figure 2E).
Fig. 2. Light micrographs (LM) or scanning electron micrographs (SEM) of chaetal structure in Hermodice carunculata. Images were taken of chaetae released after the worms were mechanically irritated with a stream of water from a pipette. (A) LM of chaetal tip, showing the serration, a clear core and the release of a drop at the tip (arrow). (B) SEM of a chaetal tip, showing the serration. (C) LM of base of a chaeta, showing the putative chaetoblast adhering to the insertion point. (D) SEM of base of chaeta with tissue at insertion point. (E) basal portion of a chaeta with a groove (arrow). All scale bars: 10 µm.
While it is uncertain which, if any, toxins are associated with amphinomid chaetae, several studies have documented the presence of toxins in whole body extracts. Nakamura et al. (Reference Nakamura, Tachikawa, Kitamura, Ohno, Suganuma and Uemura2008) isolated complanine, an inflammatory compound, from Eurythoe complanata. Hermodice carunculata sequesters palytoxin (PTX) from its zoanthid prey, Palythoa spp. (Gleibs et al., Reference Gleibs, Mebs and Werding1995). PTX maintains its haemolytic activity on human blood when isolated from the worm tissues. Researchers observed H. carunculata preying on Cassiopea spp., the upside-down jellyfish, in the Bahamas. Cassiopea contains numerous toxins, indicating that H. carunculata may sequester their toxins from various benthic cnidarians (Radwan et al., Reference Radwan, Román, Baksi and Burnett2005; Stoner & Layman, Reference Stoner and Layman2015).
As an interesting ethnographic side note, Davis (Reference Davis1983) reports that H. carunculata is one of many ingredients in a potent concoction used in Haitian voodoo rituals during which victims are turned into ‘zombies’. During the preparation of the poison, the worms are combined with a toad in a closed container, stimulating the toad to increase its own toxin secretions. It is unclear, however, whether H. carunculata actually contributes any toxins to the final potion in which tetrodotoxin is probably a key ingredient (Davis, Reference Davis1983).
Even though H. carunculata is powerfully armed with chaetae and toxins, it does have some natural predators. Most notably, it provides a primary source of nutrition for at least three species of cone snails in the Caribbean (Kohn et al., Reference Kohn, Nybakken and Van Mol1972; Vink, Reference Vink1974; Vink & von Cosel, Reference Vink and von Cosel1985). Recently Ladd & Shantz (Reference Ladd and Shantz2016) published the first observations of two fish species, the white grunt (Haemulon plumierii) and the sand tilefish (Malacanthus plumieri) feeding on H. carunculata in Florida. Whitebone porgies (Calamus leucosteus) also seem to have an appetite for amphinomids, although the species of amphinomid prey has not been identified (Sedberry, Reference Sedberry1989). D. Meyer, pers. comm. in Sebens (Reference Sebens1982), noted the predatory anemone Phyllactis flosculifera consuming H. carunculata when the worms were trapped in eddies in sand depressions. Specimens of H. carunculata have also been fatally injured by snapping shrimp (Alpheus armatus) living as symbionts with the anemone Bartholomea annulata. The snapping shrimp thus successfully defend their host anemones from fireworm predation (McCammon & Brooks, Reference McCammon and Brooks2014). In aquarium settings, the coral-banded shrimp (Stenopus spp.), the six-lined wrasse (Pseudocheilinus hexataenia) and cleaner shrimp (Lysmata spp.) have been observed preying on bristle worm species. Whether this also occurs in natural settings remains to be determined.
HERMODICE CARUNCULATA IS OMNIVOROUS
Most reports of feeding activity of H. carunculata are on live cnidarians, such as hermatypic corals (Ott & Lewis, Reference Ott and Lewis1972; Miller & Williams, Reference Miller and Williams2007; Wolf & Nugues, Reference Wolf and Nugues2013; Miller et al., Reference Miller, Lohr, Cameron, Williams and Peters2014), gorgonians (Vreeland & Lasker, Reference Vreeland and Lasker1989), fire corals (Whitman, Reference Whitman1988; Lewis & Crooks, Reference Lewis and Crooks1996), zoanthids (Sebens, Reference Sebens1982; Francini-Filho & Moura, Reference Francini-Filho and Moura2010), anemones (Lizama & Blanquet, Reference Lizama and Blanquet1975) and upside-down jellyfish (Stoner & Layman, Reference Stoner and Layman2015). Barroso et al. (Reference Barroso, Almeida, Contins, Filgueiras and Dias2016) recently reported feeding on several species of sea stars. Due to its relatively slow movements, H. carunculata is limited in its feeding activity to slow moving, sedentary or sessile prey. When feeding on cnidarians, it apparently remains unaffected by their stings or toxins. It feeds by everting its buccal cavity over a portion of its prey and drawing soft tissue into its complex, muscularized pharynx. The digestive tract was described by Marsden (Reference Marsden1963) and consists of five regions: (1) the buccal cavity, (2) the pharynx, (3) a short oesophagus, (4) a long intestine and (5) a short rectum which terminates in the anus. Using micro-computed tomography, Faulwetter et al. (Reference Faulwetter, Vasileiadou, Kouratoras, Dailianis and Arvanitidis2013) demonstrated the presence of a rasping organ in the buccal cavity, which would explain how the worms remove soft tissues from the hard skeleton of corals or gorgonians.
Apart from feeding on a variety of live prey, H. carunculata is also an opportunistic scavenger which will feed on virtually any dead animal or animal parts on the seafloor (pers. obs., Wolf et al., Reference Wolf, Nugues and Wild2014). It actually seems to prefer decaying corals, corals overgrown with algae or dead fish to live cnidarians (Wolf et al., Reference Wolf, Nugues and Wild2014). In captivity, H. carunculata will even devour injured members of its own species (pers. obs.).
In the coral conservation community, H. carunculata has a bad reputation, not only because it feeds on live corals, especially new recruits (Miller & Williams, Reference Miller and Williams2007; Miller et al., Reference Miller, Lohr, Cameron, Williams and Peters2014), but also because it can act as a vector and reservoir for coral pathogens. This has been demonstrated so far only for the Oculina patagonica/Vibrio shiloi system in the Mediterranean Sea (Sussman et al., Reference Sussman, Loya, Fine and Rosenberg2003) but there is concern that the phenomenon is more widespread.
CONCLUSIONS
We have reviewed the ability of amphinomids in general, and Hermodice carunculata in particular, to withstand environmental extremes, including a wide range of and fluctuations in temperatures, salinities, oxygen levels, heavy metals and other disturbances. Thanks to its arsenal of chaetae and toxins, whether produced endogenously or sequestered from prey, H. carunculata has few natural predators and its own diet is highly flexible.
Hermodice carunculata is clearly an opportunistic species with broad environmental tolerances. One factor that could potentially affect it negatively is ocean acidification, as a diminished pH could interfere with the formation or structural integrity of the calcified chaetae, a key feature for its survival. On the other hand, the occurrence of H. carunculata in very close vicinity to acidic vent sites (Cardigos et al., Reference Cardigos, Colaço, Dando, Ávila, Sarradin, Tempera, Conceição, Pascoal and Serrão Santos2005) suggests that a minor decrease in pH leaves adult H. carunculata relatively unaffected. Larvae generally tend to be more strongly impacted by acidification than adults (Kurihara, Reference Kurihara2008; Dupont & Thorndike, Reference Dupont and Thorndike2009; Byrne & Przeslawski, Reference Byrne and Przeslawski2013), presenting another reason to investigate the complete life cycle of H. carunculata. In the short term, H. carunculata will probably increasingly become a nuisance species. In particular, it may interfere with coral reef restoration efforts due to its feeding behaviour (Bruckner & Bruckner, Reference Bruckner and Bruckner2001; Wolf & Nugues, Reference Wolf and Nugues2013; Miller et al., Reference Miller, Lohr, Cameron, Williams and Peters2014). This would be even more troubling if new evidence emerges that it is involved in transmission of other coral pathogens, in addition to the reported Oculina patagonica/Vibrio shiloi system (Sussman et al., Reference Sussman, Loya, Fine and Rosenberg2003).
It is also noteworthy that H. carunculata probably has significant invasive potential, as is the case with other amphinomids (Cosentino & Giacobbe, Reference Cosentino and Giacobbe2011; Arias et al., Reference Arias, Barroso, Anadón and Paiva2013). Its genetic homogeneity throughout the Atlantic and its adjacent basins (Ahrens et al., Reference Ahrens, Borda, Barroso, Paiva, Campbell, Wolf, Nugues, Rouse and Schulze2013) suggests that it has remarkable capabilities for long-distance dispersal. The existence of a long-lived planktotrophic larva is likely and its potential to colonize new habitats may be increased by anthropogenic vectors such as ships’ ballast water. Additionally, juveniles and adults may be transported on ship hulls, natural and anthropogenic marine debris, or ‘live rock’ in the aquarium trade. ‘Live rock’ is a common hiding place for amphinomids which can become aquarium pests (Calado et al., Reference Calado, Vitorino, Dionísio and Dinis2007). To date, there are no reports of H. carunculata in the Pacific or Indian Oceans, except for the Red Sea. Oddly, it has been referred to as a Lessepsian species which invaded the Eastern Mediterranean through the Suez canal from the Red Sea, not vice versa (Fishelson, Reference Fishelson2001). Considering that the Red Sea is the only location not originally connected to the Atlantic Ocean, it does appear that it was introduced there at some point, but whether this happened through the Suez Canal or by other means cannot be confirmed.
Hermodice carunculata is widespread, common and easy to maintain in captivity. It therefore lends itself to experimental studies of physiology, toxicology and behaviour. In the future, it will be important to fill some gaping holes in our understanding of its biology. The most important of these are its reproduction and development and the origin and nature of its toxins. Microbiome studies could additionally shed some light on toxin synthesis as well as their potential to transmit coral and other pathogens. Future studies should also consider the effects of ocean acidification on this calcifying annelid.
ACKNOWLEDGEMENTS
We wish to thank Andy Mackie and his team for hosting the 12th International Polychaete Conference. Dr Ekin Tilic (University of Bonn) and two anonymous reviewers provided helpful feedback on the manuscript.
FINANCIAL SUPPORT
TAMU-CAPES Collaborative grant program (Grant 2015-16) has provided support for TER.
