The importance of the permanent thermocline to the cold water coral carbonate mound distribution in the NE Atlantic

Abstract

A prominent feature of the NW European continental slope is the presence of numerous cold water coral carbonate mounds that are clustered in a number of provinces. These provinces occupy a relatively narrow depth range along the continental slope: 95% of all coral carbonate mounds identified on the Irish seabed have their mound bases between 500 and 1000 m water depths, with a peak in distribution at ∼ 650 m water depth. The distribution in mound base depths is skewed with a tail extending from the maximum at 650 m to deeper depths. This distribution brackets the depth of the permanent thermocline in the NE Atlantic (600–1000 m) formed below the base of the winter mixed layer. It is shown that the permanent thermocline is associated with the strongest residual near seabed current flow, with typical residual current speeds up to 2–3 times larger at the thermocline depth compared to other depths. The strong vertical density gradient associated with the permanent thermocline, together with the steep continental slope at those depths, also enhances the energy of certain periodic motions such as internal waves and baroclinic tidal currents. These dynamic conditions favour mound growth through the promotion of significant along-slope sediment transport and also provide large across-slope sediment movement and organic matter fluxes. The stability of the thermocline structure is likely the key in providing favourable conditions over long time scales that allow mound growth through sediment baffling processes.

Introduction

A prominent feature of the continental margin of the NE Atlantic is the occurrence of numerous cold water coral carbonate mounds, located at the continental slopes of the Rockall and Porcupine Banks and the Porcupine Seabight (e.g. de Mol et al., 2002, Kenyon et al., 2003, van Weering et al., 2003, Roberts et al., 2006). These biogenic seabed structures are composed of open frameworks of scleractinian corals (mainly Lophelia pertusa or Madrepora oculata) filled with hemipelagic sediments and dead coral fragments. They can reach heights in excess of 250 m and may have a base of up to 3 km in diameter (Freiwald, 2002, Kenyon et al., 2003, Roberts et al., 2003, van Weering et al., 2003, Wheeler et al., 2005) but mostly only elevate tens of meters above the surrounding seafloor. These cold water coral carbonate mounds occur clustered in so called mound provinces and provide a significant proportion of the cold water coral occurrences in the NE Atlantic (Roberts et al., 2003). A characteristic of the mound distribution is that the majority of these mounds at the continental margin fall within a relatively narrow depth range between 600–1000 m (de Mol et al., 2002, Kenyon et al., 2003, Roberts et al., 2003, van Weering et al., 2003; Fig. 1a), thus suggesting that oceanography and/or local hydrology may likely be responsible for the mound distribution characteristics.

At the ocean basin scale, the scleractinian corals L. pertusa and M. oculata are generally found over a large depth range and wide range of hydrographic conditions, including temperature and salinity. As an azoothanthellate coral, requiring an external energy source, there is a natural relationship of their occurrence in regions with large overlying surface productivity (Freiwald, 2002, Roberts et al., 2006). In the NE Atlantic, several carbonate mound clusters are present along the flanks of the Rockall and Porcupine Bank. It has been suggested that elevated surface productivity over these banks, driven by increased nutrient levels there, may be a significant contributory factor to the presence of the mounds at these locations (White et al., 2005).

At intermediate spatial scales in the order of 10–100 km, there has been much speculation on the environmental control of carbonate mounds in terms of their setting, depth distribution and growth (Freiwald, 2002, Roberts et al., 2003). Generally the initiation of mound growth at the continental margin has occurred at hard erosional surfaces associated with dynamic boundary currents (van Weering et al., 2003, Mienis et al., 2007, Mienis et al., 2009, van Rooij et al., 2007, Dorschel et al., 2009). The interplay of coral growth and sediment input result in the formation of these mounds (Wheeler et al., 2007, Dorschel et al., 2009). As a sessile filter feeder, cold water corals require a hard substrate on which to settle, a dynamic environment with little sedimentation and large organic fluxes (Freiwald, 2002, White et al., 2005, Roberts et al., 2006). For example the depth distribution of corals has been related to regions where internal waves cause both enhanced productivity in the upper layers through vertical mixing and nutrient fluxes, and by generating enhanced organic fluxes in the benthic boundary layer due to wave-generated, large near seabed shear stresses (Frederiksen et al., 1992). Other processes that have been proposed are enhanced sub-inertial tidal period currents near the seabed (White, 2007, White et al., 2007), or local flow acceleration around smaller scale topographic features such as seamounts (Genin et al., 1986).

The region associated with carbonate mounds of the NE Atlantic encompasses the region between the two main Atlantic gyres and the eastern portion of the sub polar gyre. The water masses associated with the depth range where carbonate mounds are found are mainly the northward spreading Eastern North Atlantic Water (ENAW) at intermediate depths (200–700 m), and below these depths, there is significant influence of Mediterranean Outflow Water (MOW,) most influential to the south. Also Sub Arctic Intermediate Water (SAIW) is present at depths (400–800 m) which has a source from the west, under the North Atlantic Current (NAC), and generally has diminishing influence as the continental slope is approached (van Aken and Becker, 1996).

A prominent feature of the margin circulation is the presence of the poleward flowing slope or shelf edge current (SEC, e.g. Huthnance, 1981, Pingree and LeCann, 1990, Pingree et al., 1999). This flow is driven principally by the north–south density and associated zonal and cross-slope sea level differences. Residual flow speeds of between 5 and 20 cm s^− 1 have been reported and flows are concentrated as relatively narrow cores located over the mid slope (e.g. Pingree and LeCann, 1990, Pingree et al., 1999, White, 2007). The current plays a significant role in driving the benthic boundary layer (BBL) at the continental margin — the turbulent near seabed frictional layer of direct importance to the resident benthic fauna, including the cold water corals. In addition to the SEC, tidal currents and various wave motions also determine the spatial and temporal variation in the BBL turbulent structure on the slope (e.g. Thorpe et al., 1990, White, 1994, White, 2007).

Another characteristic of the dynamics at the NE Atlantic margin is the relatively large amount of energy that is transferred from the barotropic tide to periodic, baroclinic, motions (Baines, 1974). These baroclinic motions take two forms; i) bottom trapped baroclinic waves or ii) freely propagating internal waves. Oscillations with a frequency ranging between the buoyancy frequency (N, Eq. (2)) and the local inertial frequency (f, the Coriolis parameter) will be freely propagating and generate internal waves, principally at the forcing period (e.g. Pingree et al., 1986, New and Pingree, 1990, Thorpe, 1992). An example would be the deformation of the seasonal thermocline by tidal excursions across rapidly varying topography such as the continental slope (e.g. Sharples et al., 2009).

Of particular relevance for internal waves is the characteristic angle (to the horizontal, β) of energy propagation (group velocity) along the internal wave beams which, according to, for example Cacchione et al. (2002), depends on the wave frequency (σ), the vertical stratification (N) and the Coriolis parameter (f):

$$c=tan(\beta )={[({\sigma }^2–f^2)/(N^2–{\sigma }^2)]}^{\mathrm{½}}$$

and the vertical stratification is given by the expression;

$$N^2=-(g/{\rho }_o)⁎d\rho /dz$$

Here g is the acceleration due to gravity, ρ is the density, ρ_o a mean density and z the vertical coordinate, such that dρ/dz represents the vertical variation (gradient) in density.

Furthermore, this angle is maintained upon wave reflection, so that the angle relative to the seabed will remain the same for a flat seabed but changes for reflection off a sloping seabed (of angle α), due to the invariance of β. Reflection from a sloping seabed, therefore, changes the wave number of the internal wave and hence the energy density within the wave, with the likelihood of increased currents and vertical mixing (Thorpe, 1987). Of particular importance is the condition when β = α which corresponds to “critical” conditions at locations where internal waves can be generated and where impinging internal waves will be reflected along the seabed.

If the period of oscillation is longer than the local inertial period (2π/f), the wave motion may be trapped in the vicinity of the topographic feature that they have been generated from, e.g. the continental slope (e.g. diurnal period waves for latitudes greater than 30° (e.g. Rhines, 1970)). Two characteristics of this type of wave are i) that the tidal motion may become rectified, i.e. a residual current will be generated directed along the slope and ii) that under certain conditions of vertical stratification both the residual and tidal period motion may be amplified (Huthnance, 1981). The extent to which any periodic tidal current, and associated residual rectified flow, will be amplified will depend on a coupling of the forcing frequency (tidal period) and the natural oscillation period within the water column. This is determined by the vertical density stratification (N) and the bottom seabed slope (α). According to Huthnance (1981), the maximum amplification, if any exists, will be expected at a bottom depth where

$$N⁎sin(\alpha )=\text{maximum}\text{.}$$

The degree of amplification will be dependent on the degree of resonance between tidal forcing and the response of the water column stratification to the forcing. This resonant period (T_res) of the density oscillation can be found from the expression

$$T_{\text{res}}=N⁎sin(\alpha )⁎sin(\gamma )$$

Here γ is the angle of the baroclinic wave to the orientation of the isobaths defining the continental slope, where γ = 0° is directed along the isobaths. If the waves are directed across the slope then sin(γ) = 1 and the maximum resonance frequency for the response to the periodic forcing will be found, or correspondingly, the minimum wave period (Rhines, 1970).

In this paper we investigate the local importance of the thermocline structure for the generation of dynamic hydrographic environments near the seabed and correlate our findings with the vertical and spatial distribution of cold water coral carbonate mound provinces in the Rockall Trough and Porcupine Sea Bight, NE Atlantic. We highlight the correlation of this distribution to the depth of the permanent thermocline, the depth range where the vertical temperature (and density) gradients are largest (with the exception of the shallower seasonal thermocline). This is achieved through an analysis of historical current meter data from the region along the NE Atlantic margin in comparison to simple theoretical considerations of baroclinic wave dynamics generated by the presence of the thermocline.