Long-term monitoring reveals unprecedented stability of a vent mussel assemblage on the Mid-Atlantic Ridge

Highlights

• This paper presents the longest daily time-series of plurisdisciplinary data within a vent assemblage.

• The diffuse-flow habitat was highly stable over 7 years  with small-scale variability mainly related to tidal periodicity and localised temperature anomalies.

• The Bathymodiolus azoricus mussel assemblage and zoanthid patches were remarkably stable.

• Microbial mats displayed heterogeneous spatio-temporal dynamics.

• Our results question the assumption that vent assemblages are ephemeral and highly dynamics.

Abstract

Understanding scales and drivers of ecological variability is essential to a full understanding of ecosystem functioning. At remote deep-sea hydrothermal vents, infra-annual dynamics remain poorly described. This study aims to characterise the factors that drive the dynamics of a vent faunal assemblage dominated by Bathymodiolus azoricus mussels from infra-daily to monthly time steps. We analysed a 7-year time series of images and environmental data collected at a depth of 1695 m at the base of the active Eiffel Tower edifice in the Lucky Strike vent field (Mid-Atlantic Ridge). Using images acquired by the TEMPO ecological module connected to the EMSO-Azores observatory, we assessed the dynamics of key species inhabiting the faunal assemblage in relation to changes in environmental conditions monitored daily.

Our results show that habitat conditions were generally stable over the 7-year period, with small-scale variability related to tidal periodicity and local temperature anomalies. Likewise, the mussel and zoanthid assemblages exhibited remarkable stability. Changes in fluid exposure and substratum instability induced decimetre-scale movements of the mussel assemblage. Microbial mats displayed infra-annual changes characterised by aperiodic growth and decline. Their development patterns could not be entirely attributed to environmental conditions, because other factors, including biotic interactions, appeared to be involved. The crab population preferentially occupied the mussel habitat, but no predation was observed. Scales of variation and driving factors were compared with those governing intertidal zones. The outcomes question the assumption that vent fauna experience extreme and highly variable conditions. On the MAR, mussel assemblages appear to experience relatively stable and mild environmental conditions compared with their coastal counterparts.

Introduction

Hydrothermal vents result from the emission of superheated fluids that are released on the seafloor through the advection of cold seawater in the oceanic crust where a variety of mixing and reactive processes occur. These fluids are enriched with reduced chemicals that are used by chemoautotrophic organisms to sustain exceptionally dense faunal communities in a generally food-deprived deep sea. A variety of microhabitats are spread along a dilution gradient between hot hydrothermal fluids and cold oxygenated seawater (Jannasch 1985). Despite the presence of environmentally stressful conditions, vent ecosystems sustain luxuriant communities of endemic species, often dominated by large endosymbiotic invertebrates (Tunnicliffe, 1991, Childress and Fisher, 1992, Léveillé et al., 2005). Vent species are distributed according to their nutritional needs as well as their physiological tolerance to environmental conditions (Vismann 1991). Their habitats are characterised by steep centimetre- to metre-scale gradients of physico-chemical conditions that can vary through time due to tidal and hydrodynamic forcing (Johnson et al., 1988a, Chevaldonné et al., 1991, Le Bris et al., 2006, Podowski et al., 2009, Lee et al., 2015). Biotic interactions also influence the spatial distribution of the vent fauna (Micheli et al., 2002, Mullineaux et al., 2003, Sancho et al., 2005, Lenihan et al., 2008). At longer time scales, succession mechanisms are also controlled by changes in venting activity, habitat modifications and stochastic events (Fustec et al., 1987, Tunnicliffe et al., 1990, Tunnicliffe et al., 1997, Sarrazin et al., 1997, Sarrazin et al., 2002, Shank et al., 1998, Marcus et al., 2009). Although we are beginning to understand the spatial distribution of vent assemblages, resolving the scales of ecological variability and underlying mechanisms is paramount to reaching a fuller understanding of vent ecosystem functioning (Levin, 1992, Wiens et al., 1993).

Discovered in 1993, Lucky Strike (LS) is a basalt-hosted vent field located in the Azores Triple Junction on the slow-spreading Mid-Atlantic Ridge, at a depth of ∼1700 m (Langmuir et al. 1993). This large hydrothermal field (∼1 km²) lies at the summit of a seamount that harbours a central fossilised lava lake surrounded by three volcanic cones (Fouquet et al., 1994, Langmuir et al., 1997, Cannat et al., 1999; Fig. 1A). More than 20 active hydrothermal sites have been discovered (Von Damm et al., 1998, Charlou et al., 2000, Ondréas et al., 2009, Barreyre et al., 2012), all fed by a unique source (Pester et al., 2012, Chavagnac et al., 2018) fuelled by an axial magmatic chamber (Singh et al. 2006). Differences in hydrothermal fluid composition occur among vent sites due to varying geological settings and permeability of the upflow zone (Charlou et al., 2000, Leleu et al., 2015, Chavagnac et al., 2018). Eiffel Tower (ET), located east of the lava lake, is the most studied hydrothermal edifice of the vent field, and its ecology has been thoroughly investigated for over 20 years (e.g. Sarradin et al., 1999, Desbruyères et al., 2001, Cuvelier et al., 2009, Cuvelier et al., 2011a, Cuvelier et al., 2011b, De Busserolles et al., 2009, Crépeau et al., 2011, Martins et al., 2011, Sarrazin et al., 2015, Sarrazin et al., 2020, Husson et al., 2017, Girard et al., 2020; Fig. 1B). This 11 m edifice consists of a massive sulphide deposit of ∼452 m² (Girard et al. 2020) surrounded by a peripheral zone extending more than 20 m from the summit (Cuvelier et al. 2009). Hydrothermal activity occurs on the main sulphide tower and at the periphery through focused releases, flanges and diffuse outflows (Cuvelier et al., 2009, Mittelstaedt et al., 2012).

Similar to several edifices of LS, ET diffusion zones are dominated by the symbiont-bearing mussel Bathymodiolus azoricus Cosel & Comtet, 1999 and shrimp Mirocaris fortunata Martin & Christiansen, 1995 forming two main assemblages: those found in warmer and more variable habitats (5.2–9.5 °C) and visually dominated by M. fortunata and those visually dominated by B. azoricus in colder habitats (4.4–6.1 °C; Cuvelier et al., 2011a, Sarrazin et al., 2015, Sarrazin et al., 2020; Husson et al. 2017). The biomass of ET is largely dominated by B. azoricus mussels (∼90%, Husson et al. 2017), which can thrive in a wide range of trophic niches. For their nutrition, they mostly rely on sulphur and methane-oxidising Gammaproteobacteria endosymbionts hosted in their gills (Fiala-Médioni et al., 2002, Duperron et al., 2006). However, they are also able to filter-feed (Riou et al. 2010); this nutritional mode is more common in smaller individuals (Martins et al., 2008, De Busserolles et al., 2009). B. azoricus is considered an engineer species, because the 3D structure of their aggregations provides shelter, feeding grounds and various microhabitats. At ET, over 79 species of macro- and meiofauna composed of grazers, predators and detritivores have been identified in these assemblages (review by Husson et al. 2017). Mussels can be further subdivided into distinct assemblages, corresponding to different microhabitats and various shell sizes (Cuvelier et al., 2009, Sarrazin et al., 2015, Husson et al., 2017). Faunal diversity varies along the mixing gradient between vent fluids and ambient seawater (∼4.4 °C), with higher densities and richness in lower temperature habitats (Sarrazin et al. 2015). Dense colonies of unidentified zoanthid colonise the bare substratum in the periphery of the ET sulphide edifice (Husson et al., 2017, Girard et al., 2020). The most mobile taxa, such as M. fortunata shrimp and the crab Segonzacia mesatlantica Williams, 1998, occupy a wide range of temperature niches (Husson et al. 2017). Both species occupy the highest level of the trophic network (De Busserolles et al. 2009), either as predators or scavengers. Moreover, S. mesatlantica shows territorial behaviour and is occasionally observed feeding on mussels (Matabos et al. 2015). The ichthyofauna consists of a few visiting species (Cuvelier et al. 2009, 2017). To complete the picture, microbial communities form visible mats that cover all kinds of hard substrata including mussel shells (Cuvelier et al., 2009, Crépeau et al., 2011). These microbial mats are dominated by Gammaproteobacteria sulphur-oxidisers such as Beggiatoa spp. which give them a white filamentous aspect (Crépeau et al. 2011). They are found in low-temperature areas (<6°C, Cuvelier et al. 2011a) that benefit from hydrothermal particles conveyed by bottom currents (Girard et al. 2020). Although the factors explaining the spatial distribution of these assemblages have been identified and niches of dominant species characterised, much less is known about their infra-annual temporal dynamics.

Compared with vent fields located on faster spreading ridges, catastrophic events at LS rarely occur (review in Glover et al. 2010). In fact, only one major seismic event — a dike intrusion in 2001 — has been recorded (Dziak et al. 2004). At ET, a temporal study based on imagery reported the stability of vent communities and environmental conditions over 14 years and suggested that faunal communities may have reached a climax state (Cuvelier et al. 2011b). Some authors have suggested that in conditions of low environmental stress and relative stability, biotic factors may play a crucial role in the structure of hydrothermal communities (Sarrazin et al. 1997; review in Glover et al. 2010). Negative interactions including predation, larviphagy, physical disturbance, grazing activities (Johnson et al., 1988b, Micheli et al., 2002, Sancho et al., 2005, Lenihan et al., 2008, Marticorena et al., 2021), as well as facilitation (Mullineaux et al., 2003, Sarrazin et al., 1997) influence faunal distribution. However, the absence of long-term, high-frequency observations has restricted our ability to determine the relative roles of biotic and abiotic factors in shaping vent communities (Tunnicliffe et al., 1990, Grelon et al., 2006, Matabos et al., 2015, Cuvelier et al., 2017). The long-term acquisition of high-resolution infra-annual time series of faunal and environmental changes is therefore essential to gain further knowledge on factors driving community dynamics in these ecosystems. The development of deep-sea observatories now offers this possibility (Matabos et al. 2016).

In 2010, after many years of scientific cruises at LS, a multidisciplinary observatory — EMSO-Azores — was installed to monitor the long-term dynamics of physical, chemical and geophysical factors and to evaluate their impact on faunal communities (Cannat et al., 2011, Cannat et al., 2016). Two Sea MOnitoring Nodes (SeaMON) are the energy suppliers and communication relays for a variety of sensors deployed on the seafloor (Fig. 1A). Data is acoustically transferred to a surface buoy (BOREL) that ensures the relay between the nodes and an Ifremer SISMER data centre on land via satellite (Blandin et al. 2010). The SeaMON East node is dedicated to ecological studies and includes, among others, the TEMPO ecological observation module (Sarrazin et al. 2007, Fig. 1C-D). Equipped with a camera and environmental sensors, TEMPO records high-resolution images as well as physico-chemical conditions (temperature, dissolved oxygen and iron concentrations) within the field of view of the camera (Fig. 2). The area chosen to study long-term vent faunal assemblage dynamics is located at the base of ET (Fig. 1B) and is colonised by a dense B. azoricus assemblage (Fig. 2A).

A pilot study using TEMPO imagery in this area provided the first insights into the day-to-day variations in the mussel assemblage for the 48 days during which the video camera was operational (Sarrazin et al. 2014). Daily observations showed that the assemblage was quite stable, reflecting the relative stability of environmental conditions during this period. B. azoricus mussels thrived in habitats with very limited hydrothermal fluid input and significantly influenced by ocean tidal signals (Sarrazin et al. 2014). Temporal variation in species abundance was observed, but — with the exception of M. fortunata shrimp — no link could be established with measured environmental factors (Sarrazin et al., 2014, Cuvelier et al., 2017). Although these imagery studies did not indicate a clear tidal influence on LS mussel assemblages, Mat et al. (2020) recently showed that the physiology and behaviour of B. azoricus were significantly influenced by these periodic variations. Nevertheless, questions about the processes influencing long-term variations remain. What are the underlying mechanisms acting on mussel assemblage dynamics over a period of several years? Can we observe biological processes such as interactions, settlement, mortality, reproduction? From infra-daily to pluriannual time scales, which environmental drivers explain habitat variability? Can any stochastic events be linked with biological responses? These questions will be addressed by analysing the spatio-temporal variability of biological processes and environmental conditions in the monitored diffuse-flow habitat. Imagery recorded between 2012 and 2019 by the TEMPO ecological observatory module at ET will be combined with in situ measurements to address the following hypotheses: (H1) local environmental conditions vary at scales of minutes to days, but generally remain stable over a long period of several years; (H2) mussel cover and (H3) microbial mat cover similarly remain stable over a period of several years; (H4) zoanthid abundance does not vary significantly over long time periods, (H5) the spatial distribution of fauna is linked to particular environmental conditions and/or substratum types and (H6) biotic interactions (e.g. facilitation, predation, competition) have a significant influence on faunal/microbial distribution.