Arthropoda; Crustacea; Decapoda of deep-sea volcanic habitats of the Galapagos Marine Reserve, Tropical Eastern Pacific

Abstract Background The deep-sea biome (> 200 m depth) is the world’s last great wilderness, covering more than 65% of the earth’s surface. Due to rapid technological advances, deep-sea environments are becoming more accessible to scientific research and ocean exploration around the world and, in recent years, this is also true for the Galapagos Islands. Deep-sea habitats cover the largest proportion of Galapagos Marine Reserve (GMR), yet to date, no comprehensive baseline exists on the biodiversity of the benthic fauna associated with volcanic seafloor formations within this region. Closing this knowledge gap is essential to provide information for decision-making for the management of marine resources within the GMR and assessing any potential changes in biodiversity resulting from climate-driven alterations that deep-sea environments are expected to experience. In 2015, the Charles Darwin Foundation’s Seamounts of the GMR Research Project, together with the Galapagos National Park Directorate (GNPD) and Ocean Exploration Trust (OET), conducted a joint expedition on board the EV Nautilus. Using Remotely operated vehicles (ROVs), the aim of the expedition was to characterise the geological formations and biological communities present on seamounts, lava flows and other deep-sea habitats (> 200 m) within the GMR. New information We provide the first comprehensive image inventory for the phylum Arthropoda from 260 to 3400 m of depth within the GMR. Past studies on deep-sea macroinvertebrates in the GMR have been limited to voucher samples collected from dredging (restricted to soft bottom environments) or by submersibles (only allowing limited biological sampling). The image inventory, presented here, is based on high-definition video transects conducted by the Hercules ROV on board the EV Nautilus. Images of macroinvertebrate morphospecies were captured, catalogued and identified, thus providing the first known image inventory of in-situ macroinvertebrate species from the deep-sea region of the GMR. We present 32 distinct morphospecies occurrences within the class Malacostraca and order Decapoda. We also report 17 different families, three species that are new records to the GMR, in-situ images of two new species to science recently described and one possible new squat lobster, as well as interesting behavioural observations.


Introduction
The Galapagos archipelago is a volcanic island chain located in the Tropical Eastern Pacific, south of the Galapagos Spreading Center and forms part of the western end of the Carnegie Ridge (White et al. 1993). The archipelago originated approximately 10 Myr from the active volcanism of a melting anomaly (or "hotspot") in the Earth's upper mantle in the eastward-moving Nazca Plate (Glass et al. 2007). Resulting from this region's active volcanic history, both the terrestrial and marine environments of the archipelago are defined by complex and diverse lava formations (Sinton et al. 1996, Hoernle et al. 2000, with the seafloor on and around the Galapagos platform being a heterogeneous seascape shaped by seamounts, lava flows, slopes, rift zones and hydrothermal vents (Geist et al. 2008). Therefore, these environments are likely to host rich deep-sea invertebrate communities (McClain andBarry 2010, Souza Júnior et al. 2014). To date, however, relatively little is known about the life and physical environments of these deep ocean seascapes compared to their shallow water counterparts.
The Galapagos Marine Reserve (GMR) covers approximately 138,000 km and was established in 1998 ( Fig. 1) to protect the archipelago's marine biodiversity by banning large-scale commercial exploitation of marine resources (Henry and Roberts 2017). While deep-sea environments cover by far the largest proportion of the GMR, the technological challenges of studying deep waters mean that most baseline studies on biodiversity are limited to coastal and shallow pelagic ecosystems (Glynn andWellington 1983, James 1991). However, there are notable historical exceptions. In the late 1890s, the Albatross Expedition pioneered the collection and study of deep-sea fauna in the Tropical Eastern Pacific (Agassiz 1905). Employing dredging methods, benthic sampling was mostly unsuccessful owing to the volcanic nature of the seafloor in the region. Still, some samples were retrieved, but only for soft bottom environments (Bergh andAgassiz 1894, Grove andLavenberg 1997). Nearly one hundred years later, in 1986, the Harbor Branch Oceanographic Institutes submersible, Johnson SeaLink, collected the first targeted collections at a maximum depth of 915 m (Pomponi et al. 1988), with subsequent expeditions in the following years (Banks et al. 2014). Since these expeditions, research efforts have been made to list and describe the biodiversity of Galapagos' deep-sea environment (Agassiz 1905, Pomponi et al. 1988, Manning 2016, especially on cnidarians (Cairns 1991, Cairns 2018), chordates (McCosker and Smith 1997, Grove and Lavenberg 1997, Barnett et al. 2006, McCosker et al. 2012), echinoderms (Pawson and Ahearn 2000 and isopods (Faxon 1895, Wetzer 1990. Nonetheless, to date, there are no comprehensive image inventories of the deep-sea macrofauna within this region. In recent decades, the use of Remotely operated vehicles (ROVs), equipped with effective sampling gear and high-resolution recording technologies, has greatly accelerated exploration and surveying of deep-sea habitats (Danovaro et al. 2014) and the use of such deep-submergence technology has contributed to our knowledge of deep-water decapod crustacea in both the western (Komai and Tsuchida 2014) and southern (Poupin et al. 2012) Pacific. In the same way, the Nautilus expedition conducted in 2015 marks the future of deep-sea exploration in the GMR and we hope that this inventory, based on in-situ images of the Arthropoda, will facilitate the standardisation of the morphospecies and will be useful for targeted specimen collections during future deep-sea studies in the GMR and the Eastern Tropical Pacific.
Being the first organisms to colonise the islands, even before the appearance of macroscopic plants, the terrestrial arthropods from the Galapagos Islands have been a subject of interest for many years (Peck 1990, Peck 1994. By studying these terrestrial organisms, ecologists and geologists gain a better understanding on how the terrestrial environment of the islands came into being at its present state (Peck et al. 1998). There were intensive collections of arthropods in the intertidal and shallow subtidal regions of the Galapagos Islands during 1931-1938, performed by the ship Velero III of the Allan Hancock Foundation, University of Southern California (Meredith and Hancock 1939). These collections resulted in work on brachyurans by Garth (1946), Garth (1991) and records of carideans in various taxonomic publications. James (1991) published a series of accounts of various Galapagos invertebrates but these also were from shallow waters. However, after the initial descriptions in the 1890s of the collections of the Albatross Expediton (Faxon 1893, Faxon 1895), deep-sea marine arthropods of the Galapagos received very little attention, compared to their terrestrial and shallow-water counterparts.

Study sites
In June 2015, the EV Nautilus conducted a 10-day collaborative research expedition (NA064) between the Ocean Exploration Trust (OET), the Charles Darwin Foundation (CDF) and the Galapagos National Park Directorate (GNPD) to explore the deep-sea environments of the GMR. All methods were carried out in accordance with relevant guidelines and regulations by the GNPD under research permits PC-26-15 & PC-45-15. All experimental protocols were reviewed and approved by a GNPD's committee which evaluates animal care during research activities. We conducted a total of six exploratory dives to the far north, west and central part of the Galapagos archipelago ( Fig. 1, Table 1).
ROV dives began at the base of each feature and conducted a general upslope transect, following sonar and visual surveys along this transect. Dives H1435, H1436 and H1440 explored three seamounts around the most northern islands of the archipelago, which are part of the Wolf-Darwin volcanic lineament that extends to the Galapagos Spreading Center (Harpp and Geist 2002). All seamounts, located in this area, are conically shaped with small summit craters and relatively flat tops. These are also the youngest seamounts of the Galapagos platform estimated to be less than 1 million years old (Sinton et al. 1996, Harpp andGeist 2002). The deepest ROV transects were conducted during dives H1441 and H1442, which targeted the lava flows and abyssal plains to the west of Fernandina Island. These lava flows are part of the hotspot found beneath the Island where the largest and most active volcano of the Galapagos platform is located (Sinton et al. 1996 The final dive, H1443, explored two small conically-shaped shallow seamounts located in the central part of the Archipelago, between the islands of Santiago and Isabela. Seamounts from this part of the platform were once centred over the hotspot and are estimated to be between 5 to 6 million years old (Sinton et al. 1996, Harpp andGeist 2002).

ROV operations
Seafloor exploration was carried out using the two-body ROV system, Argus and Hercules, each rated to 4000 m water depth. Video and still images of the sites were taken using "Insite Pacific Zeus Plus" HD colour video cameras on both vehicles, each equipped with a 10× mechanical zoom lens. All in-situ images used for the inventory were obtained by Hercules' mounted camera system. Additionally, environmental parameters were also recorded using Hercules' telemetry sensors, which included oxygen concentration (Aanderaa Oxygen Optode 3830), temperature and salinity (Seabird FastCat 49Plus).
While the majority of the species analysed for this study were identified from image only, a few specimens were opportunistically collected using the ROV's hydraulic manipulators. After recovery of the ROV, the collected specimens were preserved following standardised protocols and this information is specified in the 'preparations' and 'notes' sections for each organism listed on the species checklist below.

Video transects image analysis
Each ROV dive ranged in duration from 11 to 18 hours. For the subsequent review of morphospecies, each dive was spilt into 2-hour segments. In-situ images of organisms were captured and extracted from video transects analyses using VLC software (Version 3.0.4) by "non-expert analysts". To avoid reviewer bias in capturing unique morphospecies, Table 1.
EV Nautilus Sampling sites details.
five "non-expert analysts" were assigned random video segments from all six ROV dives. Only organisms that appear to be larger than 3 cm were captured and considered for further identification. All images were then tentatively classified under their common names (i.e. squat lobsters, crabs, shrimps, etc.) and only images that appeared in sufficient detail to be determined beyond phyla were sent to taxonomic experts for further identification.
Taxonomists identified all images to the lowest taxonomic level possible. To identify the species, taxonomists consulted published literature (e.g. Wicksten 1989) to find out which species were previously reported in the vicinity of the Galapagos Islands. Images were compared to specimens (where available) or photographs with existing illustrations and images from monographs, species descriptions or expedition reports. Many identifications remain tentative because the characteristic features of a species are on the hidden ventral surface or are too small to see in the photograph.
The open nomenclature identification qualifiers presented here are modified from Sigovini et al. (2016). These qualifiers are being developed as part of an initiative by the National Oceanography Centre, Southampton, UK (Dr Tammy Horton, pers. Comm.), to standardise taxonomic nomenclature for image-based faunal analyses.
Below is a brief overview of each qualifier assigned to the different Arthropoda morphospecies. We assigned the qualifier, based on the original comments provided by each taxonomist.

indet. (indeterminabilis)
The sign 'indet.' is to be used as an abbreviation of indeterminabilis and to indicate that the specimen is indeterminable beyond a certain taxonomic level due to the lack of diagnostic characters visible in the image. This qualifier can also be used at higher taxonomic ranks and in conjunction with inc. (below) to indicate a difference between the uncertainties of the IDs at higher taxonomic ranks. We also used this term for some of the very poor-quality images.
inc. (incerta) the usage of this qualifier is to be restricted to the meaning of 'uncertain identification' and to be equated to the question mark. Since the latter may be considered as a 'wildcard' by some software, in data stored in digital form, it may be substituted by 'sp. inc.', 'gen. inc.' etc.

stet. (stetit)
Use the term stetit after the taxon name to explicitly express the identifier choice of not proceeding further.
A total of 32 distinct morphospecies of arthropods were identified, belonging to 17 families, 19 genera and 13 confirmed species of the class Malacostraca, order Decapoda.   Notes: In-situ images of Eumunida subsolanus described in Baba and Wicksten (2019), which is a new species discovered as a result of the NA064 expedition. Fig. 6 Sternostylus defensus (Benedict, 1902)

Material
Notes: This crustacean looks much like J. californiensis (Baba and Wicksten 1997). If so, it constitutes a major range extension, because previous records of the species are no further south than the Gulf of California, Mexico. To date, there is only one species known of Janetogalathea, but the animal in the photograph differs from those previously described, because of the slender fingers of the chelae, without spines. This might be a variant of J. californiensis or something undescribed. Fig. 9 Genus Munida Leach, 1820

Material
Notes: This species can be distinguished as different from the one shown in Fig. 9, because that one has hirsute chelae, not seen in Fig. 11.

Munidopsis mina
Notes: Munidopsis sp. indet. 1 (Fig. 15), Munidopsis sp. indet. 2 (Fig. 16), Munidopsis sp. indet. 3 (Fig. 17) and Munidopsis sp. indet. 4 (Fig. 18). These images were taken at too great a distance to see fine details of the antennae, spines of the anterior carapace or other distinguishing features. However, all can be determined to belong to different species by the shape of the chelae, the colour and the length of the pereopods. Fig. 15 Genus Munidopsis Whiteaves, 1874 Material Figure 15.

Material
Notes: This observation is a new record for Galapagos. Notice that a sea anemone has overgrown the abdomen of the crab, a common association found in this family of hermit crabs.  Notes: Appears to be a goneplacoid crab. The overall habitus, aspects of the shape and colouration of the claws and shape of the carapace all fit well with Mathildellidae, based on the author's (S.Ahyong) extensive examination of many species of this group.   Notes: Reported from Peru (Sample NA064-022-01-01-A). Genetics could not provide idenfication beyond genus. The images were taken too far away to see features of the teeth on the rostrum or the relative length of the rostrum to the carapace and so neither

Genus Nematocarcinus A. Milne-Edwards, 1881
Material Notes: The elongate, thread-like legs of this shrimp are easy to see in this photograph. However, as previously stated, the images were taken too far away to see features of the teeth on the rostrum or the relative length of the rostrum to the carapace and so neither Fig. 27 nor Fig. 28 can be identified beyond Nematocarcinus. Furthermore, it cannot be determined whether they are the same species or not. Fig. 29.

Pentacheles laevis Bate, 1878
Material Notes: See discussion for more detailed comments on this observation.

Discussion
Here we provide the first and most complete image inventory of arthropods found in the deep waters of the GMR to date. Of particular interest was the presence of three species that are new records to the GMR; Sternostylus defensus, Tylaspis anomala and Paromola rathbunae sp. inc. (Figs 7,22,25), in-situ imagery of two new species recently described, Heteroptychus nautilus and Eumunida subsolanus (Baba and Wicksten 2019) (Figs 3, 6) and at least one species of squat lobster that is possibly new to science, Sternostylus sp. indet. (Fig. 8). Based on the occurrences presented here, these morphospecies could be targeted on future expeditions.
The species of Eumunida subsolanus ( Fig. 6) and the unidentified mathildellid crab (Fig.  24), to the best of our knowledge, represent the first records of their respective families for the eastern Pacific and the latter is also the first record for the family Eumunididae in the Tropical Eastern Pacific (Baba and Wicksten 2019). The mathildellid crab appears to represent either Mathildella or Neopilumnoplax, species of which have been previously recorded from seamounts and deep-sea ridges in the Indo-West Pacific and tropical Atlantic (Ahyong 2008, Ahyong and Ng 2016).
Overall, we observed many different types of arthropod behaviour and associations. For example, we found species from the genera Heteroptychus, Uroptychus, Eumunida and Sternostylus displaying a preference for gorgonians and black corals as hosts ( We also recorded two rarely-observed deep-water hermit crabs Probeebei mirabilis and Tylaspis anomala ( Figs 21,22), the latter displaying the behaviour of carrying a sea anemone as observed for specimens from New Caledonia (Lemaitre 1998). The hermit crab-anemone mutualistic relationship is very common in both shallow and deep-sea environments (Hazlett 1981, Jonsson et al. 2001, Mercier and Hamel 2008. Tylaspis anomala was recorded for the first time at the GMR (Fig. 22).
The brachyuran crab Paromola rathbunae sp. inc. was found carrying a sponge using its fifth pereiopods (Fig. 25a, b). This is commonly observed in several crustacean families, including homolids; however, most of these reports are from the Atlantic Ocean (Braga-Henriques et al. 2011, Capezzuto et al. 2012. Paromola is well documented from sponge and coral gardens (Capezzuto et al. 2012). In the Nautilus expedition, we found a large aggregation of homolid crabs all carrying different shaped-sponges ( Fig. 25c, d). To our knowledge, a large group of homolid crabs, such as the one presented here, has not been previously reported for the Tropical Eastern Pacific (Muñoz et al. 2011, Smith et al. 2011, Hendrickx and Wicksten 2016. Paromola rathbunae sp. inc. is also a new record for the GMR. The polychelid lobster, Pentacheles laevis ( Fig. 33), although already recorded from the Galapagos Islands (Galil 2000), was observed for the first time in-situ. Such observations of Polychelidae are rare and, to date, individuals have been observed only on the surface of, or more often, buried in the substrate (Ahyong 2009). Our observations of P. laevis actively swimming were the first such records for the family. The observed individual was not engaged in the typical 'caridoid' escape behaviour (i.e. rapid backward propulsion by pleonic flexion) of most decapods. Instead, the polychelid was swimming forwards using the pleopods and with a straightened pleon in typical 'natantian' fashion -a plesiomorphic trait not exhibited by other lobsters and crayfish or most other reptant decapods. Although the polychelids had long been placed amongst the achelate lobsters (Palinura), the 'natantian' swimming mode, observed here, is consistent with the basal position of polychelids amongst reptantian decapods as determined by many phylogenetic analyses (e.g. Scholtz 1995, Dixon et al. 2003, Ahyong and O'Meally 2004, Bracken-Grissom et al. 2014).
Since the 1950s, fisheries have been shifting towards deeper waters in most parts of the world, threatening deep-sea biodiversity and resulting in over-exploitation of seamount faunas (Morato Gomes and Pauly 2004, Morato et al. 2006, Clark et al. 2010. As a result, the last havens for commercial species are being exploited and, with only little over 4% of the world's deep-sea having been studied (Morato Gomes and Pauly 2004), we might be losing species that are still waiting to be discovered (Mora et al. 2011). This study highlights the importance of assessing deep-sea communities within protected areas and we strongly emphasise the need to publish taxonomic inventories, since these studies serve as a reference for understanding species ecology and for developing future conservation measures.