Dense cold‐water coral garden of Paragorgia johnsoni suggests the importance of the Mid‐Atlantic Ridge for deep‐sea biodiversity

Abstract Mid‐ocean ridges generate a myriad of physical oceanographic processes that favor the supply of food and nutrients to suspension‐ and filter‐feeding organisms, such as cold‐water corals and deep‐sea sponges. However, the pioneering work conducted along the Mid‐Atlantic Ridge failed to report the presence of large and dense living coral reefs, coral gardens, or sponge aggregations. Here, we describe the densest, near‐natural, and novel octocoral garden composed of large red and white colonies of Paragorgia johnsoni Gray, 1862 discovered at 545–595 m depth on the slopes of the Mid‐Atlantic Ridge, in the Azores region. This newly discovered octocoral garden is a good candidate for protection since it fits many of the FAO criteria that define what constitutes a Vulnerable Marine Ecosystem. The observations described here corroborate the existence of a close relationship between the octocoral structure and the ambient currents on ridge‐like topographies, providing new insights into the functioning of mid‐ocean ridges' ecosystems. The ubiquitous presence of biogenic and geological topographies associated with mid‐ocean ridges, which could act as climate refugia, suggests their global importance for deep‐sea biodiversity. A better understanding of the processes involved is, therefore, required. Our observations may inspire future deep‐sea research initiatives to narrow existing knowledge gaps of biophysical connections with benthic fauna at small spatial scales along mid‐ocean ridges.


| NATUR AL HIS TORY DISCOVERY
Mid-ocean ridges generate a myriad of physical oceanographic processes at various temporal and spatial scales, one of which corresponds to increases in the upthrust exchange between the deep and the upper ocean (St Laurent & Thurnherr, 2007). Vertical mixing and horizontal advection favor the supply of food and nutrients to suspension-and filter-feeding organisms, such as cold-water corals and deep-water sponges (Genin et al., 1986;Parrish & Oliver, 2020;van Haren et al., 2017). Pioneering work conducted along the Mid-Atlantic Ridge (MAR), however, failed to observe the presence of large and dense living coral reefs, coral gardens, or sponge aggregations (Mortensen et al., 2008); likely demonstrating the lack of scientific explorations in the area.
Here, we describe the densest, near-natural, and novel octo-

| RE SULTS AND PARTI CUL AR A S PEC TS OF THE OBS ERVATION
A total of 255 colonies were reported in the video images, belonging to two clearly distinguishable morphotypes, with an approximate 4:1 ratio of white to red colonies (Dominguez-Carrió, 2021 (Grasshoff, 1979a(Grasshoff, , 1979bSánchez, 2005). The specimens' terminal branches were smaller than 5 mm in diameter and had similar sclerome in both morphotypes. Surface cortex sclerites had a smooth ornamentation, dominated by six-radiate sclerites averaging 0.05 mm in length. Based on these observations, both color morphotypes were identified as the octocoral species Paragorgia johnsoni Gray, 1862. Together with P. arborea (Linnaeus, 1758), these are the only two Paragorgiidae species reported for the Azores so far (Sampaio et al., 2019). Both species are commonly known as bubblegum corals because of the bulbous tips of their branches (with clumps of polyps) and are characterized by the presence of dimorphic polyps, reproductive siphonozooids, and feeding autozooids, without axial skeletal structures other than a medulla formed by unfused sclerites (Sánchez, 2005). Even though the biology, ecology, and distribution of P. johnsoni are far from comprehended, it appears to be widespread in the northern Atlantic Ocean, found over a wide depth (~400-4000 m depth) and temperature range (~4-13°C) (Arantes et al., 2009;Grasshoff, 1979a;Lapointe et al., 2020). The projection of the parallel laser beams over the seabed allowed measuring the height and width of 178 and 92 colonies, respectively. Still images for each of the colonies observed were taken from the video footage and colony size was estimated using the software Macnification (Orbicule). Along the patch evaluated, P. johnsoni colonies measured between 6 and 107 cm in height, with an average of 45 ± 22 cm (mean ± SD). The most frequent sizes were between 40 and 60 cm, with at least 25% of the colonies larger than 60 cm (Figure 4a). Since small-sized colonies of the white morph were difficult to tell apart from specimens of the white coral Pleurocorallium johnsoni (Gray, 1860), also detected in the same area, only colonies for which of biophysical connections with benthic fauna at small spatial scales along mid-ocean ridges.

K E Y W O R D S
biological conservation, cold-water corals, deep sea, Mid-Atlantic Ridge, oceanographic processes, vulnerable marine ecosystem there was a high degree of confidence in their taxonomic identification were annotated and used to determine their size. This could have produced an underrepresentation of small-sized colonies (<10 cm) in this study. The width of P. johnsoni ranged between 5 and 118 cm, with an average of 35 ± 21 cm (mean ± SD) ( Figure 4b). These measurements are comparable to those reported for other large octocoral species (e.g., Paragorgia arborea and Paramuricea placomus; see Buhl- Mortensen et al., 2010). It should be noted that although the ROV cruised very close to the seabed, the angle of the camera with respect to some of the measured colonies might have generated some underestimation of their real size. Additionally, underestimation of their natural size may have also resulted from a certain degree of structural damage. Although we acknowledge all these caveats, measurements of colony height and width are useful to understand the size structure of the population, determine the height-width relationship for this species (Figure 4c), which might be a relevant information to put the structural damage into a multispecies context, and have field data for future height-age relationships that could help infer the longevity of the population.
The P. johnsoni colonies were assigned to different degrees of structural damage following an adaptation of the categories defined by Pham et al. (2014). Noteworthy, most colonies observed along the video transect were found in good condition (Figure 5a), with 75% being intact or with very minor structural damage. About 14% of the colonies were found with major or massive structural damage likely caused by physical contact with bottom longlines, the most common fishing gear in the region (Pham et al., 2014;Sampaio et al., 2012). In fact, 20 portions of lost mono-and multifilament longlines were observed on the video transect, some of which were in close proximity of the damaged colonies.
Although the observed damage may also have been caused by other unknown natural reasons (e.g., local environmental conditions, excess turbulence, low colony fitness, or large predators), the past and present fishing footprint generally overlaps with the distribution of cold-water corals causing severe physical disturbances (Clark et al., 2016). Even on this ridge, the observed structural damage increased along the path of the transect (Figure 5b), with more affected colonies toward shallower and more fished areas closer to the summit. However, the overall good status of this P. johnsoni garden suggests reduced human-induced disturbance and a near-pristine status of this site. F I G U R E 1 A novel octocoral garden composed of large red and white Paragorgia johnsoni colonies was discovered on the western ridge of the Gigante seamount complex on the Mid-Atlantic Ridge, in the Azores region, during the Blue Azores 2018 Expedition, with the ROV Luso, onboard the NRP Almirante Gago Coutinho. The colonies were found on the slopes of both sides of the ridge, facing down the slope. The bathymetry data were collected by the Portuguese Hydrographic Institute and the Portuguese Navy. Numbers in screen captures refer to their position along the path (white line) of the ROV shown in the 3D map

| D ISCUSS I ON AND IMPLIC ATI ON S OF THE OBS ERVATI ON
These observations corroborate a close relationship between octocoral structure and ambient currents on ridge-like topographies, as The coral garden described here was found at much shallow depths (500-600 m) than many of previously reported occurrences, raising the hypothesis that shallow regions of the MAR may provide refugia from acidification impacts on P. johnsoni and on other similar octocoral species, as suggested for seamount summits (Tittensor et al., 2010). However, the upthrust exchange between the deep and the upper ocean may, at the same time, expose cold-water corals to deep corrosive waters (Feely et al., 2008). The increased food supply provided by upwelling currents in ridge-like topographies may pro- have been mostly recorded in waters supersaturated in carbonate that enable the bio-calcification of their skeletons (Bostock et al., 2015), although several records exist in slightly carbonate undersaturated waters in the Pacific Ocean (Bostock et al., 2015;Thresher et al., 2011). The axial skeleton of Paragorgids is made of high magnesium calcite sclerites, which are particularly vulnerable to dissolution under carbonate undersaturated waters, although the octocoral tissue (coenenchyme) intertwined among unfused sclerites may provide protection from corrosive conditions (Bostock et al., 2015;Gabay et al., 2014). Not surprisingly, their suitable habitat is forecasted to be significantly reduced under future conditions of ocean acidification (Morato et al., 2020

| CON CLUS IONS
Our observation provides new information for understanding the functioning of mid-ocean ridges, questioning previous observations documenting the lack of enhanced biological productivity over the MAR (Priede et al., 2013), and providing some evidence for a local-

CO N FLI C T O F I NTE R E S T S
All authors declare that they have no competing or conflicts of interest. Formal analysis (supporting); investigation (equal); methodology (supporting); writing-original draft (supporting); writing-review F I G U R E 5 Structural damage observed on the Paragorgia johnsoni colonies (a) and along the video transect path (b) following Pham et al. (2014). Damages were categorized as Category 1: intact, no evidence of physical damage; Category 2: minor damage, bent and/or 1-25% physical damage (e.g., broken/missing branches); Category 3: mild damage, 26-50% physical damage; Category 4: major structural damage, 51-75% physical damage; Category 5: dead/massive structural damage, 76-100% physical damage, displaced, and/or dead. The structural damage was not assigned to four colonies because the quality of the image was not appropriate & editing (supporting). Emanuel J. Gonçalves: Formal analysis (supporting); funding acquisition (equal); investigation (equal); methodology (supporting); project administration (equal); writingoriginal draft (supporting); writing-review & editing (supporting).

O PE N R E S E A RCH BA D G E S
This article has earned an Open Data, for making publicly available the digitally-shareable data necessary to reproduce the reported results. The data is available at www.doi.org/10.5281/ zenodo.4727163.