**4. Discussion**

This report presents, for the first time, extensive evidence for the association of feather duster worms with corals, other sessile invertebrates, and algae in the Caribbean. This discovery is remarkable because of the strikingly large wounds and deformities inflicted by them on their host corals. Photographs of the worms indicate that these associations have been present at least since 2014 on the coral reefs of Curaçao (Southern Caribbean) and since 2015 at St. Eustatius (Eastern Caribbean). Prior to that, they may have remained unnoticed because of the worm's withdrawal behavior, because it was perhaps less abundant in the past, or because scientists studying the worms did not pay much attention to the hosts.

**Figure 6.** Overview (**A**) and close-up images (**B**–**F**) of coral damage caused by split-crown feather dusters (*Anamobaea* sp.) on a large *Colpophyllia natans* colony at Curaçao (2021). The images show various developmental stages of coral injuries (dead skeleton covered by turf algae) forming coves at the coral margin (**A**,**B**) and circular patches over the colony surface (**C**–**F**). The maximum width of each tube is ca. 5 mm.

**Figure 7.** Close-up images of coral injuries around tubes made by split-crown feather dusters (*Anamobaea* sp.) shown in retracted condition at Curaçao (2021). The coral injuries are observed in various host species, such as (**A**) *Agaricia lamarcki*, (**B**) *Diploria labyrinthiformis*, (**C**) *Montastraea cavernosa*, (**D**) *Orbicella annularis*, (**E**) *Pseudodiploria strigosa*, and (**F**) *Stephanocoenia intersepta***.** In some species, the live tissue around the wound shows a discoloration (**C**,**D**,**F**). The maximum width of each tube is ca. 5 mm.

**Figure 8.** Split-crown feather dusters (*Anamobaea* sp.) hosted by noncoral invertebrates and algae that have overgrown corals: (**A**,**B**) The encrusting soft coral *Erythropodium caribaeorum* acting as a host on dead coral at Curaçao (2021), with tentacles extended (**A**) and retracted (**B**). (**C**,**D**) The encrusting colonial ascidian *Trididemnum solidum* at Curaçao overgrowing scleractinian host corals and worm tubes (except for the tube opening): on a scleractinian coral *Eusmilia fastigiata* in 2017 (**C**) and on dead coral in 2021 (**D**). (**E**,**F**) The phaeophyceaen alga *Lobophora* sp. at Curaçao (2021). Arrows indicate worm tubes. The maximum width of each tube is ca. 5 mm.

**Figure 9.** Split-crown feather dusters (*Anamobaea* sp.) at Curaçao (2021 and 2022) hosted by sponges that probably act as secondary hosts: (**A**) An unidentified black sponge partly overgrowing a worm tube and its host coral, *Siderastrea siderea*. (**B**) A zoantharian-infested sponge, *Niphates* sp., with an expanded worm. (**C**) A dark-red sponge, *Plakortis* sp., with a worm tube (arrow) next to the original host coral, *Orbicella franksi*. (**D**) An orange-red sponge, *Scopalina ruetzleri* (Wiedenmayer, 1977) with one worm tube (arrow). The maximum width of each tube is ca. 5 mm.

Coral deformities around sabellid worms embedded in the host's skeleton appear to be limited to a few scleractinian species of which *Pseudodiploria strigosa* appears to be the most common. Because coral-dwelling sabellids have been observed deep inside the coral skeleton, and the life span of sabellids may be over 10 years [55], these deformations have taken several years to develop. Morphological anomalies are not exceptional among corals inhabited by associated fauna. For example, the sabellid *Perkinsiana anodina* lives in short tube-shaped protuberances on the surface of an encrusting mushroom coral, which are part of the host's coral skeleton [32,51]. Coral gall crabs have received much attention because of the crescent-, canopy, slit-, and basket-shaped pits inside various coral species [44,56–59]. Coral cysts and pits made by other crabs and by shrimps in stony corals have also been described [60,61], which should not be confused with gall-shaped excavations made by shrimps [62,63], bivalves [64–67], and gastropods [34,68]. Copepods are also known to induce the forming of dwellings in corals, either as galls [69,70] or as tubular outgrowths [71,72]. Ascothoracidan crustaceans of the genus *Petrarca* Fowler, 1889 are known to form conspicuous galls in shallow-water and deep-sea corals [73–75].

**Figure 10.** Magnificent feather dusters (*Sabellastarte magnifica*) at Curaçao (2022) in close proximity to corals. (**A**) An extended worm on a dead coral patch of *Orbicella annularis*. (**B**) Same individual retracted inside its tube. (**C**) A worm in a colony of *Madracis auretenra* surrounded by healthy branch tips but attached to their dead base. (**D**) A worm surrounded by colonies of *Millepora alcicornis* and *Pseudodiploria strigosa*. (**E**) A worm underneath a colony of *Diploria labyrinthiformis.* (**F**) A worm on a dead patch of *Orbicella annularis* surrounded by healthy coral tissue. The width of each worm tube is ca. 2 cm.

**Figure 11.** Social feather dusters (*Sabellastarte magnifa*) at Curaçao (2022) in close proximity to live coral. (**A**) A worm colony on a dead patch of *Orbicella annularis*. (**B**) Another worm colony on dead coral underneath a *Montastraea cavernosa*.

Tube-dwelling gammarid amphipods and chaetopterid polychaete worms have been reported to induce the forming of densely distributed finger-like structures in *Montipora* corals [76,77]. Coral barnacles usually become embedded in the coral skeleton and become partly overgrown by coral tissue [44,60]. Some alterations in the coral skeleton morphology are microscopic and hardly visible, such as those caused by coral-dwelling hydroids of the genus *Zanclea* [78–80]. In contrast, vermetid snails that live inside branching *Stylophora* and massive *Porites* corals are known to modify the host's morphology on a larger scale by flattening its surface relief, which is attributed to growth inhibition caused by the snail's toxic mucus webs [81–83]. In contrast, large growth alterations in massive, branching, and encrusting corals consisting of deep fissures can be formed by *Pedum* scallops embedded in corals [34,66,67]. Some aggressive coral-dwelling sponges are not considered long-term associated fauna because they usually tend to overgrow and kill their hosts, but in some foliose corals, they evoke a morphological response, which is visible as the growth of flap-like protrusions that overlap the approaching sponges [84]. A modified morphology is also seen in other foliose corals that overgrow sponges as if the coral shape is molded by that of the sponge [85]. All these examples indicate that some corals may adapt their shape to resist the presence of potentially harmful associated fauna or competitors for space.

Coral injuries caused by feather duster worms have not been reported before. These appear to be much larger than those caused by coral-dwelling *Spirobranchus* worms [18,19,86]. On the other hand, large densities of serpulid worms overgrowing live coral may eventually cause partial coral mortality [87]. Many feather duster worms in the present study were found on dead coral (Figure 3A,B), and some of them formed clusters (Figure 3C,D). In some cases, the worm-infested dead-coral area was next to live coral, suggesting that the worms contributed to partial coral mortality (Figures 3E,F and 5A). A few patches of dead coral were surrounded by discolored live-coral tissue (Figure 6C,D,F). This may represent a reaction to stress as seen in some massive *Porites* corals in which polyps in contact with algae or epifauna show pink or purple pigmentation [18,66,88–90]. The difference is that there may not be extra pigmentation in the examples of the present study.

Some split-crown feather dusters at Curaçao were not hosted by stony corals but by other invertebrates. These invertebrates may have either colonized dead coral or overgrown living corals and became secondary hosts when the worms were able to resist becoming overgrown as well. The last scenario has been shown by serpulid Christmas tree worms of the genus *Spirobranchus*. The encrusting octocoral *Erythropodium caribaeorum* is recognized as an aggressive competitor for space in the Caribbean [91], which is able to overgrow corals but apparently not their symbiotic *Spirobranchus* worms [92], similar to some feather

duster worms at Curaçao in the present study (Figure 8A,B). Similarly, the colonial tunicate *Trididemnum solidum* is notorious for overgrowing Caribbean corals [93,94], except for their associated *Spirobranchus* [95] and seemingly also individuals of symbiotic *Anamobaea* sp. (Figure 8C,D). Sponges are also able to overgrow corals with the exception of symbiotic *Spirobranchus* [96,97] and apparently also *Anamobaea* sp. (Figure 9). The feather duster worm was also observed in association with algae, in particular the brown algae *Lobophora* sp. (Figure 8E,F). *Lobophora* has increased in abundance over the last decades at Curaçao and is able to overgrow live coral [98,99]. It is likely that it is able to overgrow dead and live coral containing *Anamobaea* sp., but apparently the worm tubes protrude too far to become outcompeted.

The cause of the injurious effect of the feather duster worms is unclear. The size of the wounds suggests that the worms produce toxins, but there is limited information on toxicity as a defense mechanism in Sabellidae [100]. The use of toxins can perhaps prevent worms from becoming overgrown by their hosts, as seen in *Pedum* scallops [66]. The mucus secreted by some sabellid species proves to have antibacterial properties [101–103]. According to a recent review paper on polychaete toxins, no relevant information appears to be available on the negative effect of sabellid mucus on other organisms [104]. In contrast, coral-dwelling worm snails, which occupy the same ecological niche as the feather duster worms of the present study [20], are well known for their venomous mucus and the damage this may inflict on the host corals [105,106].

Unlike *Anamobaea* sp., the relation of *Sabellastarte magnifica* and *Bispira brunnea* to corals is unclear because they were never found in living coral tissue (Figures 10 and 11). A close proximity to live corals shown by these two species may be unusual since they were also commonly found at a distance from live corals. Therefore, it may be more appropriate to use the term "pseudo-association" for this kind of unclear relation. On the other hand, it is also possible that these worms cause damage to corals and are responsible for coral mortality in their proximity.

The present study shows that the Caribbean feather duster worm *Anamobaea* sp. is more common and harmful to corals than previously known. The species has a symbiotic relation with a large range of corals and other invertebrates, which was also unknown before. It is unclear if the species has increased in abundance recently. Because this worm has the potential to become a pest species, future research should focus on its population dynamics, its settling behavior on live corals (as done with larvae of symbiotic barnacles [107,108]), and the cause, growth, and extension of coral wounds around its tubes. The larval settlement behavior of *S. magnifica* and *B. brunnea* also needs to be investigated in order to find out whether these species prefer to live in close proximity to corals or not.

**Author Contributions:** Conceptualization, B.W.H. and R.J.v.d.S.; methodology, B.W.H., C.E.H., S.J.L.-D., R.J.v.d.S., A.S.-M., R.S., and R.F.T.; validation, B.W.H.; formal analysis, B.W.H.; investigation, B.W.H., C.E.H., S.J.L.-D., R.J.v.d.S., A.S.-M., R.S., and R.F.T.; resources, B.W.H., C.E.H., S.J.L.-D., R.J.v.d.S., A.S.-M., R.S., and R.F.T.; data curation, B.W.H., C.E.H., S.J.L.-D., R.J.v.d.S., A.S.-M., R.S., and R.F.T.; writing—original draft preparation, B.W.H.; writing—review and editing, B.W.H., C.E.H., S.J.L.-D., R.J.v.d.S., A.S.-M., R.S., and R.F.T.; visualization, B.W.H., C.E.H., S.J.L.-D., R.J.v.d.S., A.S.-M., R.S., and R.F.T.; supervision, B.W.H.; project administration, B.W.H.; funding acquisition, B.W.H., C.E.H., S.J.L.-D., R.J.v.d.S., A.S.-M., R.S., and R.F.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** The field research at Curaçao was funded by the Alida M. Buitendijk Fund, the Jan-Joost ter Pelkwijk Fund, the Holthuis Fund, the Groningen University Fund, and the Dutch Research Council (NWO) Doctoral Grant for Teachers Programme (nr. 023.015.036). Fieldwork at Bonaire was supported by the World Wildlife Fund (WWF) Netherlands. The Treub Maatschappij (Society for the Advancement of Research in the Tropics) funded research at Bonaire and Curaçao.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data sharing not applicable. **Acknowledgments:** We are grateful to the funding agencies mentioned above. We thank María Ana Tovar-Hernández (Universidad Autónoma de Nuevo León, Mexico) for confirming the identity of the sabellid worms and Jaaziel E. García-Hernández for confirming the identity of the host sponges. We thank the staff of CARMABI (Curaçao) and the Dive Shop for their hospitality and assistance during the fieldwork. BWH is also grateful to Stichting Nationale Parken Bonaire (STINAPA), the Dutch Caribbean Nature Alliance (DCNA) and Dive Friends (Bonaire) for logistic support at Bonaire, the Caribbean Netherlands Science Institute (CNSI), St. Eustatius National Parks Foundation (STENAPA), and Scubaqua Dive Center for facilitating research at St. Eustatius. We want to thank three anonymous reviewers for their constructive comments, which helped us to improve the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.
