Carrying behavior in the deep-sea crab Paromola cuvieri (Northeast Atlantic)
- First Online:
- Cite this article as:
- Braga-Henriques, A., Carreiro-Silva, M., Tempera, F. et al. Mar Biodiv (2012) 42: 37. doi:10.1007/s12526-011-0090-3
- 161 Views
Observations of deep-sea homolids are becoming more common, but good-resolution imagery of these crabs in the natural environment is still scarce. Sixteen new in situ observations of Paromola cuvieri from various locations within the central and eastern groups of the Azores Archipelago (Northeast Atlantic) are described here based on video footage collected by two submersible vehicles. Crabs were found on coral gardens and deep-sea sponge aggregations, which are priority habitats of conservation importance under OSPARCOM. Diverse sessile megafauna were recorded (>59 taxa), including sponges, hydroids, corals, brachiopods, crinoids and oysters. Overall, 75% of the crabs were carrying live specimens of sessile invertebrates, mainly sponges and cold-water corals. Object selection shows to be a more complex process than previously thought, in which factors such as morphology, size and weight of objects and also palatability seem to be more important in the process of object selection than their availability.
The crab species Paromola cuvieri (Risso, 1816) belongs to the family Homolidae de Haan, 1839, which encompasses more than 14 genera and 62 species (Guinot and Richer de Forges 1995; Ng 1998; Ng and Chen 1999; Richer de Forges and Ng 2007, 2008). This species is known to occur in the Atlantic Ocean, from the Hebrides and southern Scandinavia south to Angola, including the Azores, Madeira and the Canary Islands, the Tripp Seamount, and the Mediterranean, at depths between 10 and more than −1,000 m (usually deeper than 150 m) (Manning and Holthuis 1981; Guinot and Richer de Forges 1995).
Guinot and Richer de Forges (1981) and Wicksten (1985, 1986) inferred carrying behavior in all species of the family Homolidae based on pereiopod morphology of the crabs. It consists of lifting or holding sessile animals or shells dorsally over the carapace, gripping these between the propodus and the dactyl of each subchelate fifth pereiopods, hereafter referred to as P5 (Guinot et al. 1995). Similar behavior has been documented or inferred for other crabs (e.g., Dorippidae MacLeay, 1838; Dromiidae de Haan, 1833; Latreilliidae Stimpson, 1858 and Cyclodorippidae Ortmann, 1892) with chelate or subchelate posterior pereiopods (Wicksten 1986). Carrying behavior differs from decorating or masking of majoid crabs. Homolids use their legs instead of the chelipeds to lift the object from behind and move it upward and anteriorly over the carapace (Wicksten 1986; Wicksten 1993).
In situ observations of this behavior in the genus Paromola are scarce and most of them lack good-resolution imagery (Wicksten 1985; Tyler and Zibrowius 1992). Moreover, the reasons for carrying behavior and the selected objects remain largely unknown. State-of-the-art observational technology currently used for deep-sea research allows the capture of high-resolution footage that can be used to closely examine carrying behavior of homolid crabs and identify the objects, and also to survey the distribution of carried species.
This paper reports 16 new in situ observations of P. cuvieri at six benthic habitats within the Azorean region (NE Atlantic). The main goal of this study is to provide a better description of the carrying behavior of this species, investigating the relationship between selected objects and surrounding habitat. We also aim to explore new insights on the function of this behavior. The biotopes where the crabs were found are of conservation importance under OSPARCOM (the Convention for the Protection of the Marine Environment of the North-East Atlantic) and are known as coral gardens and deep-sea sponge aggregations (OSPAR Commission 2010a, b). The knowledge about their ecology and associated fauna is still rudimentary when compared with other areas such as coral reefs (Reed 2002; Mortensen and Fossä 2006). Therefore, these new data will also improve the understanding of these communities and of their importance as hotspots of biodiversity.
Materials and methods
Detail of the dives from the submersible ‘LULA’ and the remotely operated vehicle ‘LUSO’ at the sites surveyed where Paromola cuvieri was video-documented
Time at bottom (hh:mm)
Time of observation (hh:mm:ss)
For characterization of the biotope where P. cuvieri was found and description of its behavior, a video sequence from each dive was considered (1 min before and after crab observation while the vehicle was moving, plus the time of crab observation; see Table 1). Each analyzed video sequence covered an approximate distance of 5–25 m. None of the vehicles was equipped with laser beams, except for ‘LUSO’ in dive 020-CO3A, therefore scale had to be estimated by reference to fauna or objects of known size (e.g., robotic arms). The same methodology was used to estimate other dimensions such as crab carapace length.
Benthic observations consisted of sessile megafauna that were distinctly identifiable in the images (>5 cm, but mainly >15 cm). Organisms were identified to the lowest practicable taxonomic level and counted with the aid of image analysis software (Ocean Floor Observation Platform, OFOP 3.3.0 and ImageJ 1.43). Material deposited in the collection of the Department of Oceanography and Fisheries of the University of the Azores, as well as reports of classic deep-water expeditions (e.g., cruise reports of the Prince of Monaco expeditions 1886–1915) were used to validate the identification of the benthic species. Moreover, fresh specimens were also collected during many of the dives. Identification of sponges to the species level was often not achieved, since this is very difficult to accomplish through video analysis. In those cases, a code number has been attributed to each sponge according to the coloration and gross morphology presented (see Boury-Esnault and Rutzler 1997; Bell and Barnes 2001).
Sex, missing appendages, presence or absence of a carried object and type of object were recorded for each crab observation. P. cuvieri shows a marked sexual dimorphism, which enabled sexing of each crab (e.g., size of the first pair of pereiopods, P1). In males, the P1 are much stouter than the other legs and longer than the carapace. Each specimen carried was identified to the lowest possible taxon, and its size estimated. Notes on the health condition of the carried specimen were taken (i.e., alive or dead, entire specimen or fragment). Presence of tissue injuries or loss of the basal part of the stem was also examined for carried corals.
Summary of Paromola cuvieri observations at the sites surveyed
Position (Lat, Lon)
Right P3, P4
Auletta cf. sycinularia
Both P5, right P3
Chrysogorgia cf. agassizii
Southern slope of Faial-Pico Channel
A total of nine crabs were observed at site 1 and carrying behavior was detected in all of them. Five females and two males carried diverse sponges identified as axinellid sp.1 (‘LULA’ dive 070, Fig. 2a), Porifera sp.1 (‘LULA’ dive 105, Fig. 2b), demosponge sp.15 (‘LULA’ dive 110, Fig. 2d), demosponge sp.5 (‘LULA’ dives 136, 147-record 2 and 152) and Auletta cf. sycinularia from the family Axinellidae (‘LULA’ dive 139). All of them possessed foliaceous/flabellate or lamellate shape with exception of Auletta cf. sycinularia and were frequently larger than the size of the carapace. This specimen had an arborescent morphology (tube-shaped and stalked) and measured ca. 7 cm high. Axinellid sp.1 resembled a sponge of the genus Axinella or Phakellia and had a creamy white color. The remaining observations consisted of a female holding an unidentified item, which looks like dead organic material (‘LULA’ dive 109, Fig. 2c), and a male carrying a gorgonian Acanthogorgia hirsuta (‘LULA’ dive 147-record 2, Fig. 2e). This colony was oriented upwards and still possessed the basal part of the stem, showing no evidence of damage or areas with missing polyps. Moreover, of all dives this was the largest crab found with a remarkable span of approximately 1.20 m across the P1.
The seabed surface at this site consisted of a mixed sedimentary seafloor with outcropping basaltic rock. Sponges, hydroids and corals comprised the majority of sessile epifauna (92.1%) with specimens from 35 different species. Sponges were quite diverse, showing a large range of sizes (diameter: 5–60 cm) and morphological types. In a total of 19 species recorded, demosponge sp.5 (23.1%; growth form: lamellate, erect, vertical orientation; color: white), Auletta cf. sycinularia (13.4%), demosponge sp.2 (8.54%, growth form: erect; color: creamy white), demosponge sp.1 (6.21%; growth form: repent and also erect; color: yellow) and Phakelia ventilabrum (2.72%) were dominant. Among hydroids, Polyplumularia flabellata (Plumulariidae) and Errina dabneyi (Stylasteridae) were the most common (9.13%). These were fan-shaped colonies, ranging from 10 to 45 cm for P. flabellata and from 15 to 40 cm for E. dabneyi and, most of the time were growing on the more current-exposed ridges and overhang tops of the bedrock outcrops. Twelve coral species from the orders Alcyonacea (families Acanthogorgiidae, Coralliidae, Ellisellidae, Plexauridae and Primnoidea), Antipatharia (Aphanipathidae and Leiopathidae) and Scleractinia (Dendrophylliidae) were also found. The fan-shaped gorgonians of the genus Acanthogorgia had the highest percentage of occurrence (4.27%).
South slope of Terceira Island
Only one female P. cuvieri was observed at site 2 (‘LULA’ dive 096, Fig. 2f). The crab was moving over a sedimentary bottom carrying a demosponge sp.5. This sponge resembled the same species carried by three of the crabs observed at sites 1e (‘LULA’ dive136), 1 g (‘LULA’ dive 147, record 2) and 1 h (‘LULA’ dive 152). Two sponges A. cf. sycinularia and one whip-coral Viminella flagellum were the only species observed in the video sequence analyzed from this site (Fig. 3 site2). There was no match between carried object and dominant species.
São Jorge-Pico Channel
One male P. cuvieri was found still and facing downwards on a vertical wall of basalt interrupted by small plateaus covered with sediment and small areas of dead reefs of Lophelia pertusa mixed with cup corals Desmophyllum dianthus (‘LUSO’ dive 019, Fig. 2g). No object was being carried by the crab which had its P5 erect in a postero-dorsal position over the carapace.
Sessile fauna observed in the video sequence included a variety of species from different taxonomic groups (Fig. 3 site3), with the small living crinoid cf. C. foresti, measuring ca. 5 cm height being the most abundant one (26.7%). The crinoids co-occurred with the deep-sea oyster cf. Neopycnodonte (22.7%), which measured ca. 12 cm height and 7 cm width. Other taxa comprised the hexactinellid Farrea cf. occa (17.3%), cup corals of the family Dendrophylliidae (cf. Leptosammia sp., 13.8%), brachiopods (8.9%), and even the irregularly arborescent antipatharian Leiopathes grimaldi (2.2%).
Furnas de Fora (SW of São Miguel Island)
Two female P. cuvieri were observed moving very slowly along a sediment veneer plateau. One of the crabs was missing its right third pereiopod as well as the P5 appendages and therefore was not carrying any objects. The other crab, observed 6 min later, was holding a living young colony of the gorgonian C. verticillata (‘LUSO’ dive 020R-record 2, Fig. 2h). The gorgonian was much wider than the crab’s carapace and was oriented upwards.
The two video sequences from this site showed that sponges were less abundant than in the previous 11 analyzed sequences, representing 5.4% of total observations. Seabed was mainly colonized by alcyonacean corals Acanthogorgia (61.6%), C. verticillata (15.9%), Plexauridae Indet. (11.7%) and Dentomuricea sp. (3.6%). The highest number of specimens recorded in both sequences belonged to the genus Acanthogorgia (Fig. 2 sites4-1, -2). These fan-shaped gorgonians made-up 61.9% at ‘LUSO’ dive 020R (record 1) and 60.6% at ‘LUSO’ dive 020R (record 2) of the observations. The sizes observed ranged from 10 to 45 cm height. In record 2, the carried specimen did not belong to the most abundant taxa at that site.
Dom João de Castro bank
Two crabs were observed at site 5 (‘LUSO’ dive 022). The first one, a male, was moving downwards along a rocky wall colonized by diverse sponges, corals, brachiopods, crinoids and oysters. The crab was carrying an upward-oriented Acanthogorgia armata — again, not the most common species — which was approximately the size of the crab’s carapace. Dominant species were Farrea cf. occa with 28.6% of total observations in the video sequence examined from this site (Fig. 3 site5-1). Other fauna such as cf. Neopycnodonte sp. or Cyathidium foresti comprised 31.9% of all the sessile megafauna.
The other record from this site consisted of a female with no P5 exposed dorsally over the carapace and transporting no object. The crab was observed on a rocky plunge densely colonized by V. flagellum (80.8%), Anthomastus cf. agaricus (10.1%, usually ca. 5 cm high—when polyps are current-exposed — and 2.4 cm diameter), gorgonians Acanthogorgia (5.6%) and plexaurid sp.1 (1.5%) (Fig. 3 site5-2).
Of all the records this was the deepest one (‘LUSO’ dive 020-CO3A, 1002 m). A female P. cuvieri was observed moving along a plateau of a sedimentary rock carrying a bottlebrush-shaped gorgonian resembling a Chrysogorgia agassizii. This coral was oriented upwards and exhibited no clear damage. Epibenthic megafauna was dominated by a species that resembles Leptosammia sp., with 45.2% of the records (Fig. 3 site6). This species, previously found at site 3 (‘LUSO’ dive 019), is usually small with 5 cm height and 1.8 cm diameter. Small-sized gorgonians Candidella imbricata, with heights ranging between 20 and 30 cm, were also very abundant (44.6%). A few gorgonians Chrysogorgia cf. agassizii were also documented (4.5%). There was no match between carried object and dominant species.
Response behavior to the submersible
All P. cuvieri were moving slowly over the seabed with their walking legs exposed. In close proximity to the submersible ‘LULA’, three of the crabs that were carrying objects stopped moving, lowered the carried object to cover the carapace, thus producing a camouflage effect, and remained still for at least 1 min (‘LULA’ dives 070, 096 and 147-record 1; see supplementary material). One of those crabs, immediately after the immobility period, moved very quickly away from the submersible's floodlights (‘LULA’ dive 096). This crab also exhibited what appeared to be an agonistic behavior towards the submersible, raising the chelipeds to a medium ventral plane while standing in front of the submersible and moving its P5 backward with the carried object. Video footage of the crab behavior in response to ‘LULA’ can be found in the electronic supplementary material accompanying this paper.
The data presented here indicate that P. cuvieri is not only common in deep-sea areas of the Northeast Atlantic (0.005 individuals per minute of video survey), but it is also usually found on coral and sponge habitats. ROV-based assessments of benthic communities that are currently being conducted across European waters corroborate these results. Websites of on-going projects as well as imagery archives of Research Institutes show still-images of P. cuvieri living on coral habitats (e.g., a male P. cuvieri carrying a gorgonian over a Lophelia pertusa reef; image captured using ROV ‘QUEST’ MARUM-University of Bremen; http://www.marum.de/en/Cold_Water_Corals.html).
Carrying was documented in 81% of the P. cuvieri observations, which indicates that this is also a frequent behavior among crabs of this species in these deep-sea habitats. This activity is likely even more prevalent, considering that two of the three crabs without objects were missing both P5, and thus could not carry. Therefore, we would expect that many historical records contained observations of associated objects, which is often not the case (Gordon 1956; Figueira 1964: 69, pl.1 and 2; Crosnier 1969; O' Riordan 1983). One possible reason is that most specimens have been collected by traditional sampling methods (e.g., trawling nets), and carried objects were probably lost.
Our results show that sessile invertebrates such as sponges and alcyonacean corals are the predominant carried objects (see Table 2, Fig. 2), which agrees with the few in situ observations of carrying behavior reported in this genus. Guinot et al. (1995) mention that four P. cuvieri collected off Port-Vendres in south France were observed by D. Guinot and B. Richer de Forges in the Banyuls-sur-Mer Aquarium in 1981 carrying large sponges over the body. Several Paromola japonica (Parisi 1915) were observed lifting pieces of unidentified sponges, antipatharians and gorgonians in Hawaii (Wicksten 1985: 476, fig. 1). Large crabs that appeared to be P. cuvieri were recorded holding a gorgonian branch in the continental slope, southwest of Ireland at 720 m and 534 m (Tyler and Zibrowius 1992: 215, fig. 3c, submersible Cyana, dive 36). These authors also reported a crab of the same species seated on a stone with a gorgonian branch resembling C. verticillata at 911 m (submersible Cyana, dive 40). Biscoito (1997) also reported one crab P. cuvieri over an arborescent coral off the active site of the Menez Gwen hydrothermal vent field (see photo in Biscoito 1997: 205). Additionally, a recent publication that mainly describes results from the European HERMES project shows an image of a large long-legged crab carrying an antipatharian coral photographed on a coral garden habitat on a seamount in the Northeast Atlantic (Weaver et al. 2009: 177, fig. 5). This specimen appears to be P. cuvieri.
Studies with other homolid species also reveal the use of sessile animals in carrying behavior. Specimens of Hypsophrys inflata Guinot and Richer de Forges 1981 were found carrying a sea anemone (Chintiroglou et al. 1996: 21, figs. 1–3) later described as Isanthus homolophilus (Chintiroglou and Doumenc 1998). Moreover, sponges with distinct morphologies (e.g., spherical, conical or massive and hard) and belonging to different species and families were found associated with Homola orientalis Henderson 1888 and Homola vigil A. Milne-Edwards 1880 (Guinot et al. 1995).
In a preliminary analysis, the diversity of carried items could suggest that homolids do not choose particular objects or species. However, when we combine this information with the type of habitat where the crab has been found, a clear indication is given that this is a selective process. Wicksten (1993) suggests that decoration in spider crabs can be influenced by a number of factors such as habitat, depth and available materials. Our results demonstrate that most invertebrates selected were not the most commonly found in the surveyed seabed (see Fig. 3), suggesting that, although object availability could influence the crab’s choice, it does not seem to be the main driver of the object selection. In a total of 16 video sequences analyzed, 13 revealed crabs carrying objects. Of these, only two (‘LULA’ dives 147-record 2 and 152) showed the recorded crab carrying the predominant species (see Fig. 3 sites 1 g-2, 1 h). Furthermore, video sequences of ‘LUSO’ dives 020R (record 2) and D20-CO3A showed crabs carrying common species at those habitats, but which were not the most abundant (Fig. 3 sites 4–2, 6). Thus, factors such as morphology and size of the objects seemed to assume particular importance. Every carried species of coral or sponge was soft and 83.33% were fan-shaped (corals) or foliaceous and lamellate (sponges). Objects possessing those characters are not only light, and therefore easier to carry, but also due to their morphology they can cover a greater fraction of the crab’s carapace. P. cuvieri females and males show similar carrying patterns and object preferences, unlike some decorator crabs. For instance, only adult females of the majoid Loxorhynchus crispatus decorate their carapace whereas males do not (Wicksten 1979).
Thus, what is the main function of carrying behavior in P. cuvieri crabs? It is unlikely that they store on the carapace specimens for later consumption as has been reported for the kelp crab Pugettia producta by Mastro (1981) and been shown to be a function on other Majoidea species (Wicksten 1993; Woods and McLay 1994a, b; Woods 1995). All carried species were alive and did not seem injured. For instance, the presence of the basal part of the stem in both Acanthogorgia suggests that these colonies were detached from the substrate by the crabs without damage. The condition of the gorgonians should constitute an important factor if they are to serve as camouflage or defense, given that under stressful conditions some species lose their polyps and exhibit a naked skeleton (A. Braga-Henriques, M. Carreiro-Silva et al., unpublished data). Additionally, even though the diet of P. cuvieri is highly diverse, it is composed mainly of fish remains (e.g., teleost, sharks) and benthic decapods of the species Monodaeus couchii and Munida tenuimana (Cartes 1993). Wicksten (1985) discards the hypothesis of camouflage, which is also a well-known strategy of anti-predator defense that interacts deeply with habitat, as it requires an organism matching some part of their background using shape, color and/or patterning (Cott 1940; Ruxton et al. 2004). According to the same author, visual camouflage would be ineffective given the low levels of light at the depths inhabited by homolids. In a recent study, Wisshak et al. (2010: 2385, fig. 4B) indicated that the base of the dysphotic zone (0.01% surface irradiance) in the southern Faial Channel is at 150 m, which suggests that crabs recorded at site 1 were living in the aphotic zone. However, no data are available concerning the upper limit of this zone in the other study sites. Some of the crabs video-documented remained immobile for a period of time and covered their carapace with the carried object when disturbed by the submersible (‘LULA’ dives 070, 096 and 147). Thus, our observations provide evidence of a camouflage-type behavior. This suggests a strategy of defense and distraction display, which agrees with Wicksten (1985). Carrying is considered an old behavioral pattern and crabs that exhibit this behavior are primitive brachyurans (Wicksten 1986). Therefore, although carrying could be a relic behavior, it could also reflect an adaptive behavior of anti-predator defense, which benefits the species. Some studies suggest that animals that cover themselves with natural substrates can increase their protection by choosing objects that are eventually obnoxious for predators (Hay et al. 1990; Stachowicz and Hay 1999, 2000). In fact, with the exception of one object, all carried items were living poriferans and cnidarians. Thus, they could activate anti-predator defenses increasing the survival rate of crabs under a potential attack. For instance, alcyonacean corals could produce significant amounts of secondary metabolites such as diterpenoids, whose biological activity includes deterrence of predators (Coll et al. 1982; Barsby and Kubanek 2005). Sponges could increase spicule concentration as a mechanism of anti-predator defense (Hill and Hill 2002; Hill et al. 2005). Many tropical sponges possess spicules in their skeleton that seem to discourage predation by fish (Burns and Ilan 2003), just as sclerites from temperate and tropical gorgonians may significantly dissuade fish feeding (Harvell et al. 1988). Additionally, when crabs cover their carapace with objects this could distract predators, diverting the attention of the predator and so reducing the predation risk.
Dives with ‘LULA’ and ‘LUSO’ provided an opportunity to better understand how Paromola homolids interact with the surrounding environment and how this influences their carrying behavior. This interaction could also result in increased dispersal of some sponge and coral species, since many of these could use fragmentation to reproduce. Furthermore, our observations revealed that cold-water corals and deep-sea sponge aggregations provide important structural habitat for these large invertebrates.
This study was funded by the Portuguese Foundation for Science and Technology (FCT) and European Community's Seventh Framework Programme (FP7) under the projects CORAZON (FCT/PTDC/MAR/72169/2006), CoralFISH (FP7 ENV/2007/1/213144) and Hermione (FP7 ENV/2008/1/226354). EMEPC is acknowledged for sharing the data of ROV surveys in the Azores region. We also thank P. Ribeiro and I. Sampaio for notifying the existence of footage with homolid crabs obtained during the campaign Luso 09 from EMEPC. A. Braga-Henriques is funded by a FRCT Doctoral grant (ref. M3.1.2/F/016/2008) and M. Carreiro Silva by a Post-doctoral grant (SFRH/BPD/34634/2007). IMAR-DOP/UAz is Research and Development Unit #531 and ISR-Associated Laboratory #9 funded by the Portuguese Foundation for Science and Technology (FCT) through pluri-annual and programmatic funding schemes (OE, FEDER, POCI2001, FSE) and by the Azores Directorate for Science, Technology and Communications (DRCTC).