Methane Derived Authigenic Carbonate (MDAC) Aragonite Cemented Quaternary Hardground from a Methane Cold Seep, Rathlin Basin, Northern Ireland: δ 13 C and δ 18 O Isotopes, Environment, Porosity and Permeability

: A block of sandstone retrieved by divers from near Rathlin Island, Co. Antrim, Northern Ireland, represents an aragonite cemented sand formed during the Quaternary. Strongly negative δ 13 C of the aragonite cement ( − 50 to − 60‰ δ 13 C) indicates that the hardground was formed by the anaerobic oxidation of methane (AOM), resulting in the formation of a methane-derived authigenic carbonate (MDAC) hardground. Such hardgrounds have previously been recorded as forming extensive pavements in deeper waters in the mid Irish Sea (e.g., Croker Carbonate Slabs), although the latter also contains high-magnesium calcite. Sand was initially deposited as part of a storm lag deposit, with a reworked bivalve and gastropod fauna. This sand was then colonised by a probable crustacean fauna, producing horizontal open dwelling burrows ( Thalassinoides ). After aragonite cementation, the hardground was colonised by boring bivalves, with slightly negatively elevated levels of δ 13 C. Finally, the hardground was colonised by an encrusting fauna (bryozoans, calcareous algae and serpulids), by then in warmer seas. Continued depleted levels of δ 13 C present within the encrusting fauna ( − 1 to − 5‰ δ 13 C) indicate continued methane generation and seepage, which may still be active to the present day, and to the possibility of shallow gas reserves. The δ 18 O values change between macro-infauna vs. encrusters, indicating a warming in water temperature, reﬂecting glacial and post-glacial environments. The aragonite cemented sandstone has a highly variable porosity, with large vugs (open burrows and borings), smaller mouldic porosity within gastropods and bivalves and complex micro-porosity associated with acicular aragonite cements. Overall permeability was recorded at the 2.5 to 23 Darcies level, reﬂecting the highly variable vuggy porosity, although matrix permeability was around 100 mD and controlled by the MDAC fabric. Actual permeability will likely be controlled by the extent to which larger pores are interconnected. The sea around the Rathlin Island area contains a diverse fauna, which is worthy of future study in the context of cold seep and MDAC pavement formation.


Introduction
Methane-derived authigenic carbonate (MDAC) forms within sediments near to the seafloor, within the sulphate-methane transition zone (SMTZ), through the mixing of methane derived by thermocatalytic cracking of kerogen, or the microbial decomposition of organic material in shallower muddy sediments, interacting with SO 4 2¯f rom seawater; with anaerobic oxidation of methane (AOM) and sulphate reduction by sulphate reducing bacteria (SRB's) and methane utilising archaea ( Figure 2) [1,2]. Methane seep carbonates are known from around the world [3], for instance offshore Costa Rica, Norwegian Sea, Black Sea, the Mediterranean, and are diverse niche ecological refugia as well as indicators of deep, leaking hydrocarbon reserves [3]. The complex methane-AOM-sulphate reaction promotes the growth of carbonate cemented hardground pavements, which have been reported from a number of locations within the Irish Sea [2,[4][5][6][7][8][9][10][11][12], the best three documented examples being Holden's Reefs (Tremadog Bay, offshore Wales), the Codling Fault Zone (Irish Sea, Ireland) and the Croker Carbonate Slabs (Irish Sea, Ireland) [2]. The last occurs at a water depth of 70 to 100 m, covers an area of some 500,000 m 2 , and may have been a continuously active methane seep for 17,000 years [2]. MDAC phases include acicular aragonite and high-magnesium calcite, with associated pyrite [2]. Other authors have also reported high-magnesium calcite (up to 48 mol% MgCO 3 ) and dolomite from the Croker site [4,5], while the Codling Fault Zone has been described as aragonite dominated [11]. Pavements are typically relatively thin, and have commonly been exposed by local erosion and colonised by modern encrusting organisms [2]. The Croker site records MDAC with a depleted δ 13 C of between −34 to −54‰, which indicates a thermogenic methane origin from the cracking of kerogen, with a possible Carboniferous source [1,2]. Material collected from offshore, Rathlin Island, Northern Ireland, indicates an extension of the known range of MDAC cementation within the Irish Sea area, and comes from shallower water depths than those previously described. The occurrence is herein described in terms of its composition, δ 13 C and δ 18 O isotopes, porosity and permeability and discussed in terms of its potential as an indicator of environmental information, as well as hydrocarbon exploration in the area and potential significance as an MDAC associated marine protected area.

Materials and Methods
A single block of Quaternary methane-derived authigenic carbonate (MDAC) was collected by divers, from the seafloor near Rathlin Island, County Antrim, Northern Ireland. The block was slabbed using a diamond thin sectioning saw and optically imaged ( Figure 1). Some of the material was resin impregnated and made into polished thin sections. The latter, along with smaller broken surfaces were additionally imaged using a Quanta 650 FEG scanning electron microscope (SEM) (Thermo-Fisher Scientific, Waltham, Massachusetts, USA). Samples were examined uncoated and imaged in low vacuum mode utilising backscattered electrons (BSE), with both single images and larger scale montages collected [13]. Biogenic components and aragonite cement samples were analysed for δ 13 C and δ 18 O isotopes (Scottish Universities Environmental Research Centre (SUERC), East Kilbride), using a VG SIRA 2 triple collector, dual inlet isotope ratio mass spectrometer (VG Instruments, Crawley, West Sussex, UK): for information on the technique see Mitchell et al. [14], Parkinson et al. [15]. X-ray diffraction (XRD) analysis was also carried out (University of the West of Scotland) for carbonate mineralogy, using a Seimens D5000 X-ray powder diffraction system (Bruker Corporation, Billerica, Massachusetts, USA). A 25 mm core plug was also scanned by X-ray tomography (XRT) using an RX Easy Tom ET150 (RX Solutions, Annecy, France), for porosity, and a portable TinyPerm permeameter (New England Research Inc., White River Junction, VT, USA) was used to collect permeability data. In addition, ImageJ (open source software, Wayne Rasband, National Institutes of Health, Bethesda, MD, USA) and pore analysis tools (PAT) (Heriot-Watt University, Edinburgh, Scotland) were used to assess porosity and permeability.

Results
Overall, the hardground has a chaotic rubbly appearance, which is only clarified through slabbing ( Figure 2) and thin sectioning of the material:

Matrix and Cement
The hardground is a mixed calcareous-siliciclastic sandstone, variably cemented by an aragonite cement. Grains are typically 100 to 300 µ m in size (very fine to fine, with some medium sand), often appearing to float within a matrix of carbonate cement. Some coarse silty horizons occasionally occur, but are restricted within bivalve shells and as fills to large horizontal burrows. Smaller bicarbonate particles include echinoderm spines (Figure 3e,f), and foraminifera (Figure 4b,c). Aragonite cement comprises of blade like to acicular shaped morphology (Figure 3b-d), forming rimming cements to grains, bivalves and burrows. Pyrite framboids (~5 to 20 µ m) are common (Figure 3g,h), as are also zircons (not illustrated). Porosity is highly variable (see Section 3.4).

Fauna
Owing to their occurrence within the cemented hardground, no material has been taxonomically identified to species or genus level. However, the fauna is dominated by bivalves, both burrowing and boring forms with some gastropods (Figure 2), and an encrusting fauna of bryozoans, calcareous algae and calcareous worm tubes (not illustrated). Complete (intact) closed bivalves are often observed within the hardground, and are variably orientated, with no sign of a preferred life position.

Results
Overall, the hardground has a chaotic rubbly appearance, which is only clarified through slabbing ( Figure 1) and thin sectioning of the material:

Matrix and Cement
The hardground is a mixed calcareous-siliciclastic sandstone, variably cemented by an aragonite cement. Grains are typically 100 to 300 µm in size (very fine to fine, with some medium sand), often appearing to float within a matrix of carbonate cement. Some coarse silty horizons occasionally occur, but are restricted within bivalve shells and as fills to large horizontal burrows. Smaller bicarbonate particles include echinoderm spines (Figure 3e

Fauna
Owing to their occurrence within the cemented hardground, no material has been taxonomically identified to species or genus level. However, the fauna is dominated by bivalves, both burrowing and boring forms with some gastropods (Figure 1), and an encrusting fauna of bryozoans, calcareous algae and calcareous worm tubes (not illustrated). Complete (intact) closed bivalves are often observed within the hardground, and are variably orientated, with no sign of a preferred life position. Foraminifera are also commonly observed within the sediment, both in SEM and XRT ( Figure 4).
In addition, finger sized, horizontal, open burrows (Thalassinoides) have been observed (Figure 4a), although their three-dimensional morphology is not clear. The latter have their walls cemented by aragonite, and are partially infilled by repeated pulses of fine sand to coarse silty sediment which is variably carbonate cemented.

Isotopes
δ 18 O values can be roughly subdivided into two groups, with slight overlap, with bivalves and gastropods (δ 18 O = 1.64 to 3.33‰) and surface encrusting organisms (δ 18 O = −0.47 to 1.74‰). It is also notable that there is no clear separation between macro-fauna pre-or post-cementation (δ 18 O = 1.64 to 3.33‰ versus 2.14 to 2.91‰ respectively). Authigenic MDAC aragonite has the consistently highest δ 18 O values (3 to 3.29‰), which is only matched by one of the large bivalves (Table 1).

Description
Comp (XRD) In terms of δ 13 C the aragonite MDAC cement is clearly separated from all biogenic calcites (regardless of phase -aragonite, low magnesium calcite (LMC), high magnesium calcite (HMC)), with strongly negative values of −48.99 and −60.09‰, compared to values of between −8.1 to 1.52 (Table 1). The large bivalves have δ 13 C values centered around zero, with −1.75 to 1.52‰ (Table 1). All other faunas have essentially negative δ 13 C values, of which the boring bivalve fauna has the most depleted values (up to −8.1‰), but also records one positive value of 0.39‰ (Table 1). Nevertheless, none of the macrofauna nor encrusters can be clearly distinguished based on their δ 13 C ranges (Figure 5c).

Porosity and Permeability
The material examined from Rathlin is highly variable in porosity [13], with (1)  Porosity measured for 1214 2D slices (each ~230 mm 2 ) obtained from a reconstructed X-ray tomographic volume (XRT), indicates an average porosity of 14%, with a range between 6 and 28% ( Figure 8). Permeability extracted from the XRT reconstruction using the PAT software gives values between 2515 and 23,307 mD, depending on the direction in which the permeability is calculated (Table 2). However, direct measurements taken for sampled areas of the cemented matrix records permeability with an average of 100 mD ( Figure 9, Table 2).

Discussion
Material referred to as boring bivalves is so labelled on the basis that they occurred loose and 'imprisoned' within borings. Another possible interpretation is that such material represents a postdepositional fauna living within, or washed/trapped within, open vertical burrows formed just prior to cementation. Other bivalves either occurred as pairs of closed valves, or as large singular valves, both enclosed within sediment. The co-associated closed valves occur in a variety of orientations, suggesting that they are a death-assemblage deposited after a storm event. The orientation of the block is unknown, however, the large single valves are all the same way up, and occur on just one side of the block. Given the size of the valves and the storm deposited postulated origin, these are likely to occur on the base of the bed, and are therefore concave-up.

Porosity and Permeability
The material examined from Rathlin is highly variable in porosity [13], with (1) variably infilled finger sized pores within open burrow systems of Thalassinoides (Figures 4a and 6a); (2) millimetre to centimetre scale pores of variable inclination from boring organisms into the cemented hardground (Figures 1 and 6b); (3) variably filled centimetre scale cavities within bivalves (Figures 1 and 6c); (4) normal inter-particle porosity associated with sediment infill of burrows, borings and bivalves (Figure 4d,e and Figure 6d); and (5) irregularly shaped inter-crystalline porosity between the acicular or bladed cements, which surround the sedimentary and biogenic particles (Figures 3d, 6e and 7). Porosity measured for 1214 2D slices (each~230 mm 2 ) obtained from a reconstructed X-ray tomographic volume (XRT), indicates an average porosity of 14%, with a range between 6 and 28% ( Figure 8). Permeability extracted from the XRT reconstruction using the PAT software gives values between 2515 and 23,307 mD, depending on the direction in which the permeability is calculated (Table 2). However, direct measurements taken for sampled areas of the cemented matrix records permeability with an average of 100 mD ( Figure 9, Table 2).         Table 2. Permeability values recorded using X-ray tomography (XRT) in the X, Y and Z directions, and TinyPerm from points 1-6 in Figure 9.  Individual permeability values given in Table 2. Average permeability 100.35 mD. Table 2. Permeability values recorded using X-ray tomography (XRT) in the X, Y and Z directions, and TinyPerm from points 1-6 in Figure 9.

Comparison with Other MDAC Sites
Similar hardground pavements are also seen from the Quaternary of the North Irish Sea Basin (but generally in deeper water), Holden's Reefs, the Codling Fault Zone and the Croker Carbonate Slabs. These pavements also have acicular aragonite needles with depleted δ 13 C, and many also typically contain dolomite, non-ideal dolomite, or very high magnesium calcite [2,4,5,11]. The Codling Fault Zone is the most similar to the Rathlin material in terms of its petrography, being well cemented with acicular aragonite of a similar nature to that observed and illustrated herein [8], also containing pyrite framboids, but no Mg enriched forms of calcite. The Codling Fault Zone MDAC deposits are also closely associated with a major fault complex, as is the Rathlin MDAC (Tow Valley Fault), which may have acted as pathways for methane migration. The other Irish Sea MDAC pavements also have similar ∂ 13 C depleted values of −33‰ (Holden's Reefs), −41 to −46‰ (Codling Fault Zone) and −34 to −54‰ (Croker Carbonate Slabs) [2,8], but are in general consistently lower than that of the Rathlin samples. The other Irish Sea areas clearly have ∂ 13 C values that fall within the range associated with thermogenically generated methane [1, 16], with which they have commonly been cited [2,8]. Acicular aragonite crystals have also previously been recorded filling burrows, with  Table 2. Average permeability 100.35 mD.

Discussion
Material referred to as boring bivalves is so labelled on the basis that they occurred loose and 'imprisoned' within borings. Another possible interpretation is that such material represents a post-depositional fauna living within, or washed/trapped within, open vertical burrows formed just prior to cementation. Other bivalves either occurred as pairs of closed valves, or as large singular valves, both enclosed within sediment. The co-associated closed valves occur in a variety of orientations, suggesting that they are a death-assemblage deposited after a storm event. The orientation of the block is unknown, however, the large single valves are all the same way up, and occur on just one side of the block. Given the size of the valves and the storm deposited postulated origin, these are likely to occur on the base of the bed, and are therefore concave-up.
Finger sized open horizontal burrows, partially infilled by repeated ingress of silty sediment may represent Thalassinoides, or similar horizontal networks of open living burrows. Lack of sample material prevents more detailed analysis, although Figure 4a appears to have a potential vertical entrance burrow connected to a horizontal section of burrow.

Comparison with Other MDAC Sites
Similar hardground pavements are also seen from the Quaternary of the North Irish Sea Basin (but generally in deeper water), Holden's Reefs, the Codling Fault Zone and the Croker Carbonate Slabs. These pavements also have acicular aragonite needles with depleted δ 13 C, and many also typically contain dolomite, non-ideal dolomite, or very high magnesium calcite [2,4,5,11]. The Codling Fault Zone is the most similar to the Rathlin material in terms of its petrography, being well cemented with acicular aragonite of a similar nature to that observed and illustrated herein [8], also containing pyrite framboids, but no Mg enriched forms of calcite. The Codling Fault Zone MDAC deposits are also closely associated with a major fault complex, as is the Rathlin MDAC (Tow Valley Fault), which may have acted as pathways for methane migration. The other Irish Sea MDAC pavements also have similar ∂ 13 C depleted values of −33‰ (Holden's Reefs), −41 to −46‰ (Codling Fault Zone) and −34 to −54‰ (Croker Carbonate Slabs) [2,8], but are in general consistently lower than that of the Rathlin samples. The other Irish Sea areas clearly have ∂ 13 C values that fall within the range associated with thermogenically generated methane [1, 16], with which they have commonly been cited [2,8]. Acicular aragonite crystals have also previously been recorded filling burrows, with depleted δ 13 C (−43.59 to −54.54‰) from localised cold seep mounds from the Pliocene of central Japan [18]. Finally, Kinnaman et al. [19] investigated shallow water (10 m deep) carbonate cemented sands from the Brian Seep, Santa Barbara, from the AOM of a thermogenically sourced methane (∂ 13 C−43‰). Although similar in many respects to the Rathlin material, the carbonate from the Brian Seep comprises of high magnesium calcite and forms carbonate concretions [19], rather than more widespread acicular aragonite cement. Many biominerals have 18 O values that precipitate in equilibrium with the isotopic composition of the host marine water, which is generally the case for calcite within bivalves [20], bryozoa [21,22], encrusting calcareous algae [23] and serpulid tubes [24]. As noted earlier, a separation in δ 18 O values between the macrofauna (bivalves and gastropod) and encrusting fauna can be noted with a small degree of overlap between ∂ 18 O 1.64 and 1.74‰. This equates to a temperature of 9 • C, suggesting a relative increase in water temperature between the two groups, with the macrofauna reflecting temperatures between 1 to 9 • C, while the encrusters record between 9 and 18 • C (Table 1, Figure 5b). This fits with an increase in water temperature since the formation of the hard ground in the Quaternary. As the sea warmed, additional encrusting organisms colonised the substrate. Encrustation may range from recent to modern, with the latter supported by comparison with monthly average sea water temperatures of 8-14 • C recorded at nearby Ballycastle [25]. This would agree with the findings of Judd et al. [2] for encrusters on the Croker Carbonate Slabs.
In addition, many of the bivalves interpreted as borers have a more positive δ 18 O signature, in comparison to the burrowing bivalves and bulk of the isolated large single bivalve valves (Table 1), suggesting the possibility of slightly cooler palaeo-water temperatures after formation of the hardground. Unlike other boring fauna such as from the Croker Carbonate Slabs, the similarity in temperature between all forms of bivalves, and the separation in terms of δ 18 O (temperature) from the encrusting fauna is indicative that at least some proportion of observed borings occurred soon after hardground formation. The wide range of temperature values interpreted from the large isolated single bivalve valves (minimum temperature 1 • C, maximum 7 • C) may reflect a longer range of depositional periods, some commensurate with cooling during the glacial maximum.
Oxygen isotopes in bivalves can also be influenced by salinity, becoming increasingly lower in δ 18 O [26], and can vary with age and shell layer [26,27]. The relationship with shell ultrastructure was not taken into consideration from the current samples, and may, in part, explain some of the variability recorded within the different bivalve groups. No evidence was noted to suggest that the bivalves were non-marine, or subject to lower salinities, with no negative δ 18 O excursions, and calcareous microfossils within the matrix material of the hardground are suggestive of normal marine conditions. The δ 18 O values for the aragonite cementing phase records amongst the most enriched values (3 and 3.29‰), which equate to a temperature of between 1 and 2 degrees. Assuming that these values are in equilibrium with seawater, this is likely to indicate that hardground formation was at its zenith during the height of glacial development. However, a number of previous authors [2,11,28] have pointed out that enrichment in δ 18 O from waters rising from gas sources, or from the dissociation of gas hydrates, mixing with seawater within pores can skew the results (resulting in cooler values than otherwise expected). This is highly likely, with similar but more δ 18 O enriched values of + 3.9 to + 4.8‰ recorded from aragonite crusts associated with methane seeps from deep-sea fans of the Congo, related to altered pore water chemistry caused by gas hydrate dissociation [29].

Trends ∂ 13 C Isotopes
The δ 13 C isotopes recorded from the acicular aragonite cement are strongly depleted. Depletion in δ 13 C can have a number of sources, from the production of methane through microbial breakdown of organic matter in sediment, to thermogenic release of methane during kerogen 'cracking' [1,16].
Isotopically heavier values (−20 to −60‰) were recognised by Judd [1] as being of a more thermogenic methane origin, while the more δ 13 C depleted (−60 to −80‰) being indicative of biogenic (microbially) sourced methane. Vigneron et al. [16], although generally in agreement with this placed the boundary between thermogenic and microbially derived methane at −50‰, with an upper limit for biogenic methane at −110‰. For the majority of Irish Sea MDAC pavements, this does not pose a major problem, as, in either case, most would fall into the thermogenically derived window (see Section 4.1). However, for the Croker Carbonate Slabs (−34 to −54‰) and the Rathlin MDAC (−49 to −60‰), it is likely that for the Croker Carbonate Slabs the cement may have a mixed methane source, while that of Rathlin may be substantially more microbially derived. Stolper et al. [30] indicated that isotopic and compositional parameters may not be enough to clearly differentiate methane source within cold seeps, advocating the use of clumped isotope studies, which benefit also by providing data on temperature during methane generation. Even so, for high rates of methane production, not in equilibrium with the temperature at the time of methane production, some ambiguity in diagnosing the methane source still exists [30]. Nevertheless, whatever the source of methane, there is general agreement that MDAC formation would have occurred near the sediment surface, within the sulphate-methane transition zone (SMTZ), mediated through the biological anaerobic oxidation of methane (see Figure 2) [1,2]. Work on lipids from those microbiota responsible for anaerobic oxidation of methane (AOM), contained within MDAC, characteristically shows extreme depletion in δ 13 C, with −137 to −93‰ δ 13 C [31], as low as −127‰ δ 13 C [32] and at least −100‰ from MDAC associated lipids of the North Sea [33]. In addition, Himmler et al. [32] also recorded δ 13 C values of as low as −101‰ from biomarkers of sulphate reducing bacteria. Microbial activity during AOM is well known to have a fractionation effect for δ 13 C [31,34], consequentially offsetting δ 13 C methane by approximately 20 to 60‰ [31]. This effect was clearly illustrated by Himmler et al. [32], with δ 13 C carbonate −54.6‰, with actual recorded δ 13 C methane values of −70.3 to −66.7‰. Therefore, given the recorded δ 13 C aragonite of −49 to −60‰ from the Rathlin aragonite MDAC cement, it appears more plausible that, in this case, the source for methane is of a shallow microbially derived form. The latter may at least in part also be the case for the Croker Carbonate Slabs MDAC.
The majority of the fauna present within the sample are depleted in δ 13 C (Table 1). The δ 13 C of marine bivalves is commonly neutral to slightly positive, with Jurassic examples recorded with values of 0.3 to 5.1 [35]. However, it has been shown that fluviatile bivalves can have δ 13 C values in the range of −13.7 to −6‰ [36], Leng and Lewis [26] reporting similar values of between −12 and −8 for freshwater Unio bivalves, and similar recorded for Mytilus edulis from tidal zones [37]. Such negative values are likely a result of the influence of depletion within the filtered food source [36,37]. It is therefore plausible that the recorded negative values within bivalve shells from the present study reflect the nature of their food source, in itself influenced by the continuous presence of the methane cold seep. Fluctuations in δ 13 C values representing differences in food source, or possibly seasonal variations recorded within the shell.
Bryozoans can display a range of δ 13 C values, with Key et al. [38], recording values of between −2.2 and 2.5‰ from the Miocene, in which depletion in δ 13 C was attributed to the upwelling of nutrient-rich waters. Although this may be the case at Rathlin, which does possess a rich and diverse marine fauna [39], this may, in part, reflect the continued presence of the cold seep around the Rathlin area.
Williams et al. [40], illustrated that coralline algae hold records of depletion in δ 13 C within surface seawater, brought about by changes in the dissolved inorganic carbon content. Therefore, it is possible that the depletion in δ 13 C (−5.43, −4.8‰), recorded herein, may be due to upwelling, or possibly anthropogenic influences, but, given the relatively large surface area to volume ratio of the white coralline algae, and their encrustation on the MDAC surface, perhaps more reasonably bears witness to the continued emission of methane through the cold seep.
Doss [41] illustrated an example where δ 13 C depletion in serpulid tubes occurred under hyposaline (abnormally low salinity) conditions compared to fully marine, and other authors have also shown that freshwater annelid tubes can be highly depleted in δ 13 C [42]. The depletion herein is not high enough to indicate freshwater dilution. The δ 13 C and δ 18 O values for the annelid tubes are not that dissimilar to other normal marine annelid tubes [43], although the slight depletion in δ 13 C is likely to also be related to continued activity of the cold seep.
Judd et al. [2] reported δ 13 C values from the Croker Carbonate Slabs, for serpulids, bryzoans and shells of between −1.46 to +1.39‰, consistent with normal marine carbonates, although noting that some biogenic material was depleted with δ 13 C values of approximately −12‰. This was interpreted as a mixture of MDAC and normal seawater values, and hence reflects the influence of the active methane seep. Therefore, it is likely that as indicated, many of the boring bivalves (δ 13 C −3 to −6‰), some encrusting bryozoans (around −3‰), the encrusting calcareous algae (−5‰) and possibly the recorded annelid tubes (around −1.30‰) from the Rathlin MDAC location are also witnessing growth during active methane seepage and growth of the MDAC material.

Significance of Porosity and Permeability
Although areas with inter-particulate and inter-crystalline micro-porosity are common, permeability data suggest that flow through this matrix material may be limited to the order of 100 mD. X-ray tomographic reconstructions of the hardground (Figure 8), illustrates the complex relationships between the different styles of observed pores and a high likelihood of connectivity between the macro-pores (open burrows-'Thalassinoides'-borings and large open spaces between paired bivalve shells), strongly suggesting higher levels of permeability in certain situations; which will be further supplemented by inflow from micro-pores within the matrix (inter-particle and inter-crystalline pores). The latter pore type is too small to easily and accurately measure by XRT, and would require 3D digital rock modelling (DRM) methods, utilising higher resolution SEM images. Although such techniques are not uncommon [44][45][46], the unusual pore shape and complex three-dimensional structure of the MDAC aragonite cements and their associated pore spaces (Figure 7), would make it difficult to model permeability using traditional DRM techniques. Nevertheless, future analysis would greatly benefit the understanding of the detailed three-dimensional geometry and flow properties of these unusual cements. However, as indicated, permeability modelling of the whole rock XRT reconstruction, which takes into account the larger observed macro-pores and a range of the simpler inter-particle micro-pores, indicates permeability levels in the tens of Darcies ( Table 2). As would be expected, flow through the macro-pores is highly variable, depending on the orientation and connectivity of such structures. The hardground would be expected to act similar to microporous materials within a fractured rock system, having at least both dual porosity and permeability. Zwicker et al. [47] discussed the possibility that the burrows of callianassid decapods (Sponeliomorpha, Ophiomorpha and Thalassinoides) may act as conduits in the upper parts of methane-seep deposits, thus assisting in the distribution of mineralising fluids and gasses, and the continuation of AOM mediated cementation. Likewise, the observed 'Thalassinoides' burrows recorded from the Rathlin material could have provided a similar function. Haas et al. [29] illustrate similar material to 'Thalassinoides', which were also lined with aragonite, but represent non-biological gas flow conduits, associated with higher gas flow rates. Therefore, it is possible that the Rathlin 'Thalassinoides' have a similar non-biological origin, instead being formed due to high gas flow rates. In either case, such structures will greatly increase connectivity between areas of subsurface AOM carbonate formation and the seafloor [29,47]. The occurrence of aragonite within pipework/Thalassinoides, which are interpreted as conduits for the free flow of methane is of particular interest as Judd et al. [2] postulated that areas with Mg calcite within the Croker Carbonate Slabs represented parts of a free flowing system, while aragonite formed when pores were occluded, resulting in slower diffusion of methane through the sediment. Clearly this is at odds with observations from other MDAC systems, indicating that the form of calcite precipitation (Mg calcite versus aragonite) must depend on more than a single parameter such as flow rate.

Hydrocarbon Potential
The area around Rathlin Island (Rathlin Basin), and the closely associated North Channel, have received interest in terms of hydrocarbon exploration [48,49]. In terms of the Rathlin area, good reservoir potential exists within Carboniferous age sandstones and the Triassic Sherwood Sandstone Group, with hydrocarbon sourced from Carboniferous coals and organic-rich sediments [48]. However, no clear seals have been identified, with the absence of the Mercia Mudstone Group halite deposits which act as seals in the nearby Larne Basin, leaving only potential localised seals within the Carboniferous aged sequence [48]. Nevertheless, drilling within the Rathlin Basin area, has indicated a number of oil shows within cored materials, and the presence of seeps [48]. The major Tow Valley Fault, which passes to the east of Rathlin Island, which juxtaposes Pre-Cambrian meta sediments against post-Triassic strata [49], may act as a conduit for the migration of shallow methane and the formation of cold methane seeps. Further investigation on the origin of methane (microbial versus kerogen cracking), confirmation on the presence and extent of MDAC pavements (in the Rathlin area), and evidence of continued methane cold seeps, may help to further constrain hydrocarbon potential in the Rathlin Basin and surrounding area. Although the recorded δ 13 C values of −49 to −60‰ for the Rathlin MDAC are most likely indicative of microbial, rather than thermogenic, methane (see Section 4.2.2), these only represent two samples. Further investigation and analysis may indicate a wider range of values similar to many of the other Irish Sea MDAC pavements (see Section 4.1), and the possibility of a more mixed methane source.

Conservation Habitat
Twenty-seven cold seep MDAC marine protected areas (MPA's) from shallow water (less than 200 m) have been designated in European waters, which are important habitats for conservation [50]. The Croker Carbonate Slabs area was chosen by Noble-James et al. [50] to characterise such shallow water MDAC pavements, and its relatively unimpacted associated biological community. The waters surrounding Rathlin Island are already recognised as an important biological conservation area, with a rich and diverse fauna [39]. The shallower water depth of the Rathlin MDAC make this location a potentially important site for additional research of such important habitats. This would allow for the assessment of the current state of cold seep activity in the Rathlin area, the abundance and relative distribution of MDAC, as well as the relationship with the diverse associated fauna.

Conclusions
Our analysis of this methane-derived authigenic carbonate (MDAC) cemented sandstone off the coast of Rathlin Island extends the known range of this process within the Irish Sea. Furthermore, we demonstrated the complex interaction of methane expulsion, sulphur-reduction and a palimpsest of faunal inhabitation that will add to the understanding of MDAC [51]. Further work would ideally confirm whether this is an extensive pavement, or patchy in extent. In terms of δ 13 C isotopes, the Rathlin MDAC appears to differ from other known areas in the Irish Sea, having a signature that suggests a shallower microbially derived source for methane. Nevertheless, further investigation would be helpful in confirming this difference, and an assessment for potential gas exploration in the Rathlin Basin area.
The Rathlin MDAC hardground is similar to others recorded from the Irish Sea, although apparently lacking high-magnesium calcite or dolomite cements, with only aragonite being developed-therefore being more similar to the Codling Fault Zone than the Croker Carbonate Slabs. Although much of the encrusting fauna may be modern, as is the likely case for those observed at the Croker Carbonate Slabs location [2], at least some of the boring at Rathlin appears to be related to Quaternary activity, shortly after hardground formation. These differences may be due to the shallower depth of the Rathlin material.
MDAC has a complex structure with consequentially variable porosity, with macro-porosity related to burrows, borings, and cavities within bivalves, and inter-particle and inter-crystalline micro-porosity variably developed dependant on the level of cementation. The matrix has a permeability of around 100 mD, although flow through the larger pores has been modelled in the range of 2.5 to 23 Darcies. Actual permeability in the field will be controlled by the connectivity and orientation of the larger more vuggy pores, and the extent of MDAC cementation.
Interest in MDAC pavements and their associated fauna as important MPA's, suggests that the Rathlin MDAC could play an important role in such studies, in conjunction with on-going investigations of other areas such as the Croker Carbonate Slabs.