Methane-derived authigenic carbonate is a rocky biogenic substrate, formed just beneath the seabed. This substrate precipitates as a by-product of anaerobic methane oxidation by microbes, during the flow of methane from an underlying geological source. Methane-derived authigenic carbonate structures and their associated microbiota provide a range of important ecosystem functions, and occur globally from the intertidal to the deep ocean, at many thousands of ‘cold seeps’. When exposed at the seabed, the carbonate concretions are colonized by fauna that favour hard substrate, resulting in ‘oases of life’ in comparison with the surrounding, generally featureless, soft sediments. Where gas release is active, assemblages of seep-specialist macrofauna may occur, especially in waters beyond the continental shelf, and chemoautotrophic primary production by Bacteria and Archaea can influence the entire food web. Microbial assemblages also perform an important role in climate regulation. Precipitated carbonates serve as major carbon sinks, with anaerobic oxidation by microbes estimated to convert between 10 and 80% of mobilised methane to benthic biomass and carbonate. Significant active microbial biomass can be harboured within precipitated carbonates, capable of ongoing methane oxidation, even after exposure at the seabed. In light of the unique environmental services provided by methane-derived authigenic carbonate habitats, the need to further understand, assess, and mitigate anthropogenic disturbance to these systems is recognized, particularly in areas where petroleum exploration, mining, and demersal trawling are prevalent, or are likely to become so in the future. Global protection and management of methane-derived authigenic carbonate habitats is, however, currently limited in comparison with other biogenic habitats, such as Coral reefs.
In a
paper published in the journal
Aquatic Conservation on 21 February 2020, Tamsyn Noble-James of the
Joint Nature Conservation Committee, and the
Centre for Environment, Fisheries and Aquaculture Science, Alan Judd of the Alan Judd Partnership,
Markus Diesing, also of the Centre for Environment, Fisheries and Aquaculture Science, and of the
Norges Geologiske Undersøkelse, David Clare, again of the Centre for Environment, Fisheries and Aquaculture Science, Andrew Eggett, also of the Joint Nature Conservation Committee, Briony Silburn, once again of the Centre for Environment, Fisheries and Aquaculture Science, and Graeme Duncan, again of the Joint Nature Conservation Committee, present the first UK effort to develop a multidisciplinary approach to monitoring shallow methane-derived authigenic carbonate feature attributes, using the
Croker Carbonate Slabs Special Area of Conservation as a case study.
The Croker Carbonate Slabs Special Area of Conservation is located in the UK sector of the Irish Sea, and covers an area of approximately 116 km², with a water depth ranging from 65 to 109 m below Chart Datum (the water level that depths displayed on a nautical chart, derived from the lowest astronomical tide). This area is influenced by high natural water turbulence, with strong tidal currents and with sediments subject to frequent resuspension and redistribution. The influence of these hydrodynamic forces is evidenced by the presence of a deep-water channel to the east of the marine protected area, and sedimentary bedforms such as sand waves. Mobile demersal, static demersal, and pelagic fishing activities are known to have taken place within the marine protected area, and three telecommunications cables intersect it; however, evidence of anthropogenic pressures within its boundary is generally limited, and at the time of the study the marine protected area was considered to be in ‘Favourable Conservation Status’.
The location of the Croker Carbonate Slabs Special Area of Conservation, within the UK Exclusive Economic Zone. Noble-James et al. (2020).
A multidisciplinary survey was conducted between 24 October and 6 November 2015, onboard the
RV Cefas Endeavour. The broad aim of the survey was to investigate key methane-derived authigenic carbonate feature attributes, with the following objectives: (i) investigate the extent and distribution and characterize the physical structure of the methane-derived authigenic carbonate across the marine protected area; (ii) investigate whether methane release is ongoing, estimate the age of the surficial methane-derived authigenic carbonate, and explore the possibility of its continued precipitation; (iii) investigate whether chemoautotrophic taxa occur on the methane-derived authigenic carbonate or in the surrounding sediments; (iv) describe the epifaunal assemblages associated with ‘high relief’ and ‘low-relief’ forms of methane-derived authigenic carbonate; and (v) test the efficacy of combined multidisciplinary survey methods to optimise methane-derived authigenic carbonate monitoring strategies.
An
EK60 single-beam echosounder and an
EM2040 multibeam echosounder were deployed continuously and simultaneously throughout the acoustic survey operations. The single-beam echosounder data were used to identify potential gas seepage plumes, whereas the multibeam echosounder provided data on the depth, morphology, and backscatter strength of the seabed. Sub-bottom data were acquired using a deep-towed boomer, operated alternately at 135 and 240 J, to investigate the sub-seabed geology and to examine potential fluid transport pathways. Side scan sonar data were acquired using an
EdgeTech 4800MP 300/600 kHz dual frequency digital system to identify areas of suspected gas seepage in the water column.
Video and still-image data were collected at 128 locations using a
SeaSpyder ‘Telemetry’ drop camera system. Sediment samples were acquired at 56 stations using a 0.1 m² mini-Hamon grab. The sediment samples were sieved over a 1 mm mesh and processed for benthic macrofauna. Additional grab samples were acquired at three locations where suspected gas bubbles were observed leaking from the seabed in acoustic data, using a 0.1 m² Day grab. These samples were subsampled for benthic meiofauna and sieved over a 38 μm mesh. Full macrofauna samples were processed and identified according to
National Marine Biological Association Quality Control scheme protocols, and meiofaunal samples were processed and identified. All three faunal datasets were checked for chemoautotrophic taxa.
Suspected methane-derived authigenic carbonate concretions from 30 grab samples were retained for further analysis, following carbonate testing in the field using 10% hydrochloric acid. These samples underwent petrological, mineralogical, and isotope analysis. Eight samples were selected for strontium isotope (Strontium⁸⁷/Strontium⁸⁶) analysis and nine samples were selected for Uranium–Thorium age dating. X-ray diffraction optical microscopy and backscatter scanning electron microscopy were undertaken, primarily to ascertain the mineralogical composition of carbonate cements. Carbon and Oxygen stable isotope analyses provided indications for the origin of Carbon in the cements and for seawater temperature at the time of carbonate precipitation, respectively. Strontium isotope analysis of aragonite extracted from five methane-derived authigenic carbonatesamples was conducted to establish the source of pore water fluids, and on three samples of recent calcareous encrusting biota to provide a reference value for the Strontium⁸⁷/Strontium⁸⁶ of the Irish Sea seawater signature. Uranium–Thorium age dating analysis was carried out on a subset of nine methane-derived authigenic carbonate samples previously analysed by petrographic and stable isotope methods.
A pump-driven molecular electronic transducer methane detector system was attached to the camera frame and operated continuously during 125 of the 128 deployments. Data were recorded at 5 second intervals and average methane concentrations were calculated per minute deployed. Water sampling was conducted at six locations, using a 10 L Niskin bottle (a plastic cylinder with stoppers at each end in order to seal the bottle, used to take water samples at a desired depth without the danger of mixing with water from other depths), to provide chemical validation of the methane values detected by the molecular electronic transducer sensor. Four samples were collected within the marine protected area at locations where the molecular electronic transducer sensor, single-beam echosounder, multibeam echosounder, or video data suggested that gas seepage was occurring. In addition, two control locations were sampled outside the marine protected area: one over the deepwater channel east of the marine protected area, and one to the south of the marine protected area. Three water samples were acquired at each location: one just above the seabed, one in mid-water, and one just below the sea surface. They were subsequently analysed by gas chromatography.
Statistical analyses were undertaken on the still-image dataset, which excluded images of insufficient quality based on clarity, visibility of fauna, and confidence in methane-derived authigenic carbonate identification. Each image was assigned to a habitat class based on the
European Nature Information System classification. The topographic character of the methane-derived authigenic carbonate varied on a continuum from small exposed patches to large outcropping cliffs. To further explore the variation in epifaunal assemblages associated with areas of the methane-derived authigenic carbonate that were more elevated from the seabed, the methane-derived authigenic carbonate images were assigned to one of two, broad, ‘highrelief’ and ‘low-relief’ methane-derived authigenic carbonate classes. High-relief methane-derived authigenic carbonate were defined as outcropping from the surrounding sediment and showing distinct elevation, whereas low-relief methane-derived authigenic carbonate formed low-lying slabs and isolated chunks that did not stand substantially above the surrounding seabed. The epifaunal abundance data were transformed according to the semi-quantitative Superabundant-Abundant-Common-Frequent-Occasional-Rare scale, were assigned numerical values; (6) Superabundant; (5) Abundant; (4) Common; (3) Frequent; (2) Occasional; and (1) Rare, and analysed in
PRIMER 6. The use of Superabundant-Abundant-Common-Frequent-Occasional-Rare data enabled the inclusion of the entire epifaunal community, which was recorded using different units (percentage cover for colonial and count data for solitary taxa), and reduced noise in the data caused by field-of-view variance. Mixed sediment and coarse sediment stillimage datasets were extremely large and were randomly subsampled to match the number of low-relief methane-derived authigenic carbonate class images (205 images), whereas all other classes contained fewer images. Analysis of similarity was conducted to test the null hypothesis of there being no difference in epifaunal assemblage structure between the habitat classes, and similarity percentage measures were generated for each class to identify which taxa contributed to similarity within and dissimilarity between classes.
Several epifaunal taxa were major contributors to similarity within the high-relief methane-derived authigenic carbonate class and were also observed in association with low-relief methane-derived authigenic carbonate, albeit at lower abundances. Where the occurrence frequency of these was deemed sufficient at least 20% of images in the high- and low-relief methane-derived authigenic carbonate classes), univariate analyses were conducted to evaluate whether the differences in abundance corresponded to a preference of taxa for high-relief methane-derived authigenic carbonate, or simply reflected the more fragmented nature and reduced continuous coverage of low-relief methane-derived authigenic carbonate. Abundance was compared between: (i) high-relief and low-relief forms; and (ii) different methane-derived authigenic carbonate percentage cover classes. The null hypothesis predicted no difference in abundance of each selected taxon between the high- and low-relief methane-derived authigenic carbonate forms. The test was repeated for three different data subsets to determine whether differences in abundance reflected a genuine preference for high- or low-relief methane-derived authigenic carbonate, or whether they were related to variance in methane-derived authigenic carbonate cover. The three datasets comprised: (A) all images including methane-derived authigenic carbonate; (B) images in which methane-derived authigenic carbonate covered at least 30% of the seabed; and (C) images in which methane-derived authigenic carbonate covered at least 70% of the seabed.
The methane-derived authigenic carbonate was mapped in several stages. First, the area of interest was spatially constrained to an area north and west of the channel and south of a sandwave field. Second, the methane-derived authigenic carbonate was mapped by extracting class descriptions from methane-derived authigenic carbonate observations (still images). Various features were trialled, and the following uncorrelated features were used: area (size of image objects), mean bathymetric roughness, mean backscatter, mean bathymetry, standard deviation of backscatter, and standard deviation of the bathymetric position index with a kernel size of 3 × 3. Third, a rule-based approach was subsequently used to map sand waves, i.e. areas where the methane-derived authigenic carbonate was not present or was buried beneath a substantial volume of sediment. All objects classified as sand waves or left unclassified were then assigned to the class ‘sediment’. Finally, the results were simplified by reclassifying ‘sediment’ objects fully or largely surrounded by methane-derived authigenic carbonate, and below a size of approximately 100 m², as methane-derived authigenic carbonate.
The extensive acoustic surveys provided a comprehensive physical description of the Croker Carbonate Slabs Special Area of Conservation, which was found to comprise an inclined plateau bordered to the south and east by deeper water. This plateau is covered by mobile surficial sandy sediments, in places forming sand waves. These are underlain by postglacial sediments that lie on coal-bearing Carboniferous rocks, believed to be a source of methane. The imagery data generally showed an extremely heterogeneous seabed. High- and low-relief forms of the methane-derived authigenic carbonate were visually identified across the majority of the marine protected area, whereas the remainder of the seabed comprised a range of sediment habitats (sand, mixed sediments, and coarse sediment) according to the European Nature Information System habitat classification.
There was an increase in the predicted extent of the methane-derived authigenic carbonate within the marine protected area, from approximately 10 km² detected by a 2008 survey, to 20 km² determined by the new model. This new interpretation largely confirmed the extent of the ‘hard substrate’ identified in a 2012/2013 study in areas where the two datasets overlapped; however, differences were observed in the south-east of the site, where the 2012/2013 study indicated a more widespread occurrence of the methane-derived authigenic carbonate that was not confirmed by ground-truthing results. Still images and grab samples provided evidence that surficial sediment was present in areas interpreted acoustically as methane-derived authigenic carbonate, and which the software interpretation delineated the extent of the methane-derived authigenic carbonate ‘at or just below the sea bed’. It was not possible to distinguish between high- and low-relief methane-derived authigenic carbonate in the model interpretation, as none of the predictor features yielded a suitable separation of the two classes (i.e. the image locations for high- and low-relief methane-derived authigenic carbonate did not consistently correspond to distinct ‘objects’ identified from the multibeam echosounder data).
The extent of the high- and low-relief methane-derived authigenic carbonate classes across the marine protected area could therefore not be predicted, only observed at discrete locations from imagery data. The imagery data indicated that low-relief methane-derived authigenic carbonate was present across the majority of the marine protected area, being most densely concentrated in the central and north-eastern areas, whereas the high-relief methane-derived authigenic carbonate was more prevalent in the south, and a discontinuous methane-derived authigenic carbonate ‘cliff’ line bounded the channel at the north-east edge of the marine protected area, with the cliff being visible from both multibeam echosounder and imagery data. The physical character of this cliff was comparable with a smaller cliff feature to the south of the site centre which was confirmed to be present in 2015. The north-eastern cliff line was also identified from multibeam echosounder in the south-east of the site, although no surficial methane-derived authigenic carbonate was observed here from the imagery.
Combined multidisciplinary evidence for (a) high-confidence and potential methane-derived authigenic carbonate extent and (b) active methane release within the Croker Carbonate Slabs Special Area of Conservation. Noble-James et al. (2020).
Sub-bottom profiler data revealed evidence of acoustic turbidity across the marine protected area: this is interpreted as gas within the seabed sediments. The lateral increase in the amplitude of some individual reflections (‘gas brightening’) also suggested an increased concentration of gas within sediments. ‘Gas chimneys’ were apparent in the form of vertical columns of acoustic turbidity. Boomer profiles were crossreferenced with acoustic, visual, and physical evidence, which validated the interpretation of shallow gas.
Low-relief methane-derived authigenic carbonate within the Croker Carbonate Slabs Special Area of Conservation. Noble-James et al. (2020).
Apparent gas bubble plumes were identified from single-beam echosounder (21 locations) and side scan sonar water column data (five locations), and bubbles were directly observed leaking from the seabed in video imagery at two locations.
Images of gas bubbles within the Croker Carbonate Slabs Special Area of Conservation. Noble-James et al. (2020).
Black, potentially sulphidic, sediment patches were recorded adjacent to methane-derived authigenic carbonate concretions at four locations, and at one location white mats thought to comprise thiotrophic (sulphur oxidising) Bacteria (e.g. Beggiatoa spp.) were recorded at their edges.
Twenty-nine of the 30 suspected methane-derived authigenic carbonate samples acquired in grab samples were determined to be carbonate cemented, and contained high-magnesium calcite, aragonite, or a combination of both cements. Carbon¹³ values generally ranged from −34 to −54‰ (parts per thousand) when compared with the Pee Dee Belemnite, a global standard; this is consistent with an authigenic origin in which carbonate precipitation is a consequence of anaerobic methane oxidation. Strontium⁸⁷/Strontium⁸⁶ analysis suggested the aragonite and calcareous encrusting biota were precipitated in sea water that was similar to present-day sea water (strontium isotopes vary a great deal with local geology and hydrology,
and are useful for determining the origin of biominerals).
Uranium-Thorium dating conducted on aragonite extracted from two samples indicated that methane-derived authigenic carbonate formation occurred during the period from approximately 17 000 to 4000 years before present (Uranium-thorium dating works because uranium decays to thorium at a
known rate, so that the ratio of the two elements in minerals that
naturally incorporate uranium but not thorium can be used to establish a
date for the minerals). It is likely that gas release has been ongoing for a considerable period of time (for tens of thousands of years, at least), and that methane-derived authigenic carbonate formation is likely to have occurred continuously since sea water returned to the site as the sea level rose following the last glaciation, approximately 17 000 years ago.
Multivariate analysis of still-image data showed that epifaunal
assemblages associated with the high-relief form of the methane-derived
authigenic carbonate were not significantly different from those of the
low-relief form, and the statistics indicated a high level of
similarity. Assemblages associated with both forms of the
methane-derived authigenic carbonate were highly significantly different
to all other habitat classes except for coarse sediment, in comparison
with high-relief methane-derived authigenic carbonate; however, it
should be noted that although statistically significant, the differences
between some group pairs were small (e.g. low-relief methane-derived
authigenic carbonate versus mixed sediment, and low-relief
methane-derived authigenic carbonate versus coarse sediment).
High-relief methane-derived authigenic carbonate with attached sessile epifauna. Noble-James et al. (2020).
High-relief and low-relief methane-derived authigenic carbonate faunas were both dominated by the clump-forming and solitary Hydrozoa, which occurred in 75 and 76% of high- and low-relief methane-derived authigenic carbonate images, respectively, contributing 30.9 and 42.9% to within-class similarity. A small group of taxa, including clump-forming and solitary Hydrozoa, Polychaetes, turf Hydrozoa, and Serpulids, were consistently dominant within all habitat classes, including sediment. Beyond these dominant taxa, differences in epifaunal assemblage composition between the two methane-derived authigenic carbonate forms and the remaining habitat classes appeared mainly attributable to a small number of taxa. Five taxa (identified to species or genus level) were substantially more abundant in association with the high-relief and/or low-relief methane-derived authigenic carbonate classes in comparison with other habitat classes. These were the soft coral Alcyonium digitatum, the Bryozoan genus Cellaria, the Hydroid genera Nemertesia (including Nemertesia ramosa and Nemertesia antennina) and Tubularia (including Tubularia indivisa), and the Polychaete family Sabellidae (including Sabella pavonina). In combination, these taxa contributed 32.8% to the within-group similarity of high-relief methane-derived authigenic carbonate and 13.0% to the within-group similarity of low-relief methane-derived authigenic carbonate. These five taxa were consistently amongst the top contributors to dissimilarity between the two methane-derived authigenic carbonate classes and other habitat classes (except for Alcyonium digitatum and Tubularia on low-relief methane-derived authigenic carbonate, and mixed and coarse sediments). The frequency of occurrence in still images for the five taxa was higher for both forms of the methane-derived authigenic carbonate than for other habitat classes, particularly for high-relief methane-derived authigenic carbonate.
The occurrence frequency of Alcyonium digitatum, Cellaria, and Sabellidae in still images was deemed sufficient to compare abundances between high- and low-relief forms of methane-derived authigenic carbonate (i.e. they occurred in over 20% of images for each methane-derived authigenic carbonate class). The abundance of Alcyonium digitatum was significantly higher in association with high-relief methane-derived authigenic carbonate than with low-relief methane-derived authigenic carbonate for all three data subsets, however, there was no significant difference in the abundance of Sabellidae between the two methane-derived authigenic carbonate forms in the data. The abundance of Cellaria was significantly higher on high-relief compared with low-relief methane-derived authigenic carbonate for dataset (A) (1–100% methane-derived authigenic carbonate cover), but this difference was not apparent when the test was repeated for subsets (B) (over 30% methane-derived authigenic carbonate cover) and (C) (over 70% methane-derived authigenic carbonate cover). These results indicate that only Alcyonium digitatum was consistently more abundant on high-relief than low-relief methane-derived authigenic carbonate, regardless of the percentage cover of methane-derived authigenic carbonate in the images.
No conclusive evidence of chemoautotrophic fauna associated with gas seepage was found in the epifaunal, macrofaunal, and meiofaunal taxa identified. Two Desmodorid Nematode genera were recorded within the marine protected area: Leptonemella and Catanema. These taxa exhibit relationships with thiotrophic ectosymbiotic Gammaproteobacteria. Although the presence of sulphidic materials may enhance the sediment conditions for these taxa, they cannot be categorically linked to gas seepage, as bubbles were not directly observed at the sample locations (i.e. only observed from acoustic data).
Petrological, mineralogical, and isotope analysis of suspected methane-derived authigenic carbonate samples confirmed that the mineral assemblage and Carbon isotope values are characteristic of methane-derived authigenic carbonate. The sample characteristics are comparable with those described previously from more limited surveys of the Croker Carbonate Slabs and methane-derived authigenic carbonate occurrences elsewhere, including the UK sector of the North Sea, and the Kattegat, where ‘Submarine structures made by leaking gases’ were first described in 1992.
Methane-derived authigenic carbonate sample with faunal excavations. Noble-James et al. (2020).
Noble-James
et al.'s interpretation of the 2015 acoustic and groundtruthing data increased the extent of the methane-derived authigenic carbonate mapped with high confidence within the marine protected area (either exposed at or subcropping close to the seabed), from approximately 10 km² to approximately 20 km². This makes the Croker Carbonate Slabs the largest known shallow methane-derived authigenic carbonate feature within the
Natura 2000 marine protected area network, mapped using ‘good’ quality data. The analysis did not confirm the presence of methane-derived authigenic carbonate in the south-east of the site, an area where the possible presence of methane-derived authigenic carbonate was inferred previously based on the presence of ‘hard substrate’ detected from acoustic backscatter data. The large methane-derived authigenic carbonate cliff feature observed from imagery in the north-east of the site was thought to continue southwards to this area below the seabed surface, however. It is possible that the hard substrate interpreted in the previous study may be methane-derived authigenic carbonate buried under a substantial veneer of mobile sediment and/or clay (which was observed as exposed here between sand waves). To reflect the uncertainty around the character of the highly reflective substrates observed from backscatter, additional areas totalling approximately 37 km² were classified as ‘potential methane-derived authigenic carbonate’ within the marine protected area.
The still and video imagery, supported by side scan sonar and multibeam echosounder data, provide clear evidence of a mobile sediment veneer across a significant portion of the marine protected area, forming sand waves in places. These bedforms are likely to impede the detection of methane-derived authigenic carbonate in the affected areas, with the methane-derived authigenic carbonate being uncovered and reburied on various temporal and spatial scales. The exposed extent of the methane-derived authigenic carbonate is thus in a continual state of flux. Exposed methane-derived authigenic carbonate may be colonised by fauna but are also subject to erosion. It has previously been postulated that there may be an evolution from intact sheets of carbonate to sand- and gravel-sized fragments, as a result of erosion from strong currents and mobile sediments and the activity of boring organisms. methane-derived authigenic carbonate is a relatively hard substance, akin to limestone, however, and the persistence of the cliff feature observed in 2004 and 2008, and the high- and low-relief features that were delineated from 2008 data, indicate that this process is slow (in human terms).
The wide range of multidisciplinary data sources used in the study (molecular electronic transducer sensor, water samples, sub-bottom profiler, single-beam echosounder, side scan sonar, and imagery) provide strong evidence with which to conclude that methane seepage is ongoing across much of the site. Most methane concentrations detected by the molecular electronic transducer sensor were supersaturated when compared with the overlying atmosphere. The supersaturated sea surface waters indicate that the site is exporting methane to the atmosphere. It has been suggested that most of the methane rising from sub-seabed sources is used by anaerobic oxidation (the ‘benthic methane filter’) close to the seabed. The methane-derived authigenic carbonate and hydrogen sulphide are by-products of anaerobic methane oxidation beneath the seabed, so it can be concluded that methane-derived authigenic carbonate formation is likely to continue within the Croker Carbonate Slabs Special Area of Conservation. The Uranium-Thorium isotope dating of aragonite crystals in the methane-derived authigenic carbonate samples suggests that it has been forming since the Last Glacial Maximum. Evidence of ongoing gas seepage across the marine protected area supports the theory that methane-derived authigenic carbonate is present beneath the seabed within, and possibly beyond, the areas of ‘potential methane-derived authigenic carbonate’. The rate at which methane-derived authigenic carbonate is expected to be precipitated is unknown, although the Uranium-Thorium dating analysis suggests that the formation of the Croker Carbonate Slabs has taken place over an extended period of time, which is consistent with findings from the North Sea and the Barents Sea. As methane-derived authigenic carbonate is still thought to be forming at the Croker Carbonate Slabs, there may be potential for the generation, regeneration, and improvement of feature condition in terms of methane-derived authigenic carbonate extent and structure, which could counteract the natural erosion of the exposed methane-derived authigenic carbonate. Regeneration is, however, highly unlikely to mitigate any damage caused to the methane-derived authigenic carbonate by non-natural disturbances within the marine protected area (and presumably elsewhere), such as abrasion or physical damage by demersal towed or static fishing gear.
Methane concentrations (nmol/L) from water samples acquired at surface, mid-water and bottom-water. Noble-James et al. (2020).
Six of the seven habitat classes, including high- and low-relief methane-derived authigenic carbonate, were dominated by one or more of the same four high-level (and predominantly attached sessile) taxa, suggesting a hard substrate beneath a sediment veneer or stable lithic material not subject to transport by tidal currents. This apparent lack of distinction between the methane-derived authigenic carbonate forms and the surrounding sediment habitats is likely to reflect the adaptability of the taxa, with many of the same taxa that have colonised the methane-derived authigenic carbonate also attaching to the larger fractions of coarse sediment. The seabed was extremely heterogeneous, meaning that many images contained small areas of secondary substrate in addition to the primary classified substrate (e.g. a small number of pebbles in an image otherwise dominated by sand). As suggested by the results of the analysis, it is also likely that areas of non-visible methane-derived authigenic carbonate buried by a thin sediment veneer (and colonized by typical sessile taxa) have been classified as sediment habitats. Additionally, it should be noted that some areas of methane-derived authigenic carbonate may have been obscured by extensive low-lying aggregations of Sabellaria spinulosa (and therefore images classified as such), although the presence of methane-derived authigenic carbonate could not be verified because of the encrusting nature of this tubiculous Polychaete.
A reef formed by the tubiculous Polychaete, Sabellaria spinulosa, in the the Dutch Brown Bank area of the North Sea. Oceana.
Although five taxa were found to be associated with both highand low-relief methane-derived authigenic carbonate (being seldom recorded in the other habitat classes), four of them; Alcyonium digitatum, Cellaria, Nemertesia, and Tubularia, occurred in a notably higher number of high-relief methane-derived authigenic carbonate images. For Alcyonium digitatum, this relationship was not related to disparities in percentage cover between low- and high-relief methane-derived authigenic carbonate. The high-relief form of methane-derived authigenic carbonate clearly provides enhanced conditions for this species (and possibly others that could not be detected from imagery data), potentially through the greater elevation above the seabed and the increased structural complexity. An elevated position above the seabed is likely to afford protection to sessile organisms from intense smothering or scouring by mobile sediments. This seems unlikely to be the sole driver of the increased abundance, however, as Alcyonium digitatum does not appear to be particularly sensitive to these pressures. Tidal currents may become accelerated and water velocity increased with obstruction from high-relief methane-derived authigenic carbonate, potentially resulting in a greater volume of plankton and organic material available to suspension-feeding taxa. Particle uptake by Alcyonium digitatum is limited by low-velocity water flow (in addition to very high velocities at the opposite end of the spectrum). The increased structural complexity of high-relief methane-derived authigenic carbonate could also provide greater potential for the retention of, and colonisation by, planktonic invertebrate stages.
The presence of biological excavations on the methane-derived authigenic carbonate samples and in imagery data indicates that boring fauna are likely to be a functionally significant component of the fauna within the marine protected area. Such fauna increase the habitat complexity and therefore niche availability to other epifauna, whilst also contributing to the erosion and eventual breakdown of the carbonate. The boring Sponge Cliona celata was recorded on methane-derived authigenic carbonate, although only from a small number of images. It is likely that boring fauna have been underrepresented in Noble-James et al.'s study, as a result of the dense cover of attached fleshy, branching, and foliose Hydroids, Bryozoans, and Soft Corals, and the generally low taxonomic resolution achievable from the imagery data. The characterising and monitoring of such taxa would involve acquiring replicate samples of intact methane-derived authigenic carbonate, a method that is unlikely to be justified in the context of a long-term monitoring programme, given the slow formation rate of the feature.
White patches of suspected thiotrophic Bacteria were observed close to some methane-derived authigenic carbonate concretions; these Bacterial mats are recognized as ubiquitous where there is active anaerobic oxidation of methane. Despite the extensive visual and physical sampling regime, and the analysis of several meiofaunal samples from areas of suspected gas seepage no other chemoautotrophic organisms thought to be directly associated with methane release were recorded. A previous study found no seep specialist macrofauna at the Codling Fault Zone, approximately 30 km south-west of the Croker Carbonate Slabs. The authors of that study noted that seep assemblages in shallow water (0–200 m) are less likely to contain seep-specialised taxa than at deep water sites, as strong competition from fauna that use photosynthetic carbon is expected at shallow depths. In the Kattegat there were found to be ‘major differences’ in the abundance and biomass of Nematode populations close to seeps (5–20 cm) and at control sites 30 m away, which may support this assertion. Another study did, however, record the seep-specialist Bivalve, Thyasira sarsi, and Nematodes of the genus Astomonema at the Scanner Pockmark in the North Sea (roughly 150 m below Chart Datum), indicating that chemoautotrophic infauna could also exist within the marine protected area (beyond the observed Leptonemella and Catanema), but may not have been detected. Another study recorded high densities of Astomonema southwardarum in the Scanner pockmarks, but not at the Braemar pockmarks.
The results of the multidisciplinary survey have highlighted various implications for monitoring shallow water methane-derived authigenic carbonate habitats and faunal assemblages under the feature attribute framework, which can be used to assess condition and to inform management measures.
Geological, mineralogical, and petrographical analyses provided conclusive evidence that previously suspected areas of methane-derived authigenic carbonate did in fact comprise this habitat. These specialist analyses are, however, resource and cost intensive. It is expected that an initial application of these methods will be sufficient to provide evidence on the ‘Natural gas seepage and methane-derived authigenic carbonate formation’ and ‘Climate regulation’ feature attributes, after which a simplified scope can be used to monitor whether gas seepage continues (and probable continued precipitation can be inferred). The use of the molecular electronic transducer sensor (with validating water samples) was extremely cost-effective in comparison with the acoustic methods used to detect gas seepage (sub-bottom profiler, side scan sonar, and single-beam echosounder), as it was deployed on the camera frame (with camera transects taking 23 minutes on average, across the 128 deployments), and readings were instantly available.
The mineralogical and petrographical analyses supported the postulation that methane-derived authigenic accretion is likely to continue at present, indicating that the habitat and its microbial consortia could perform an ongoing role in climate regulation. Methane-derived authigenic carbonates, including sediment-hosted nodules and well-lithified carbonate rocks and pavements at the seabed, host substantial active microbial biomass capable of methane oxidation under both oxic and anoxic conditions. Further sampling and analysis of intact carbonates (at and below the seabed) in areas of active methane release could improve our understanding of endolithic microbial assemblages, and their role (if any) in climate regulation. The value of this (and other) detailed ‘one-off’ study (or studies) would need to be carefully evaluated in the context of the available resources and the competing time-series evidence requirements of the other feature attributes.
At a number of ground-truthing locations methane-derived authigenic carbonate was predicted from the acoustic backscatter interpretation but were not visually observed at the seabed. Multibeam echosounder pulses penetrate the seabed to some extent, depending on the acoustic frequency and the sediment type. Here, it appears that exposed methane-derived authigenic carbonate and methane-derived authigenic carbonate covered by a thin veneer of sediment have essentially the same acoustic signature. Furthermore, the analysis might be confounded by spatial heterogeneity at short length scales on the same order or less than the positioning accuracy of the drop camera system. Therefore, it should not be assumed that perceived methane-derived authigenic carbonate signatures consistently indicate exposed methane-derived authigenic carbonate. Equally, in shallow habitats dynamic regimes can result in the fluctuating resuspension and deposition of sediments and the presence of mobile bedforms, such as sand waves. An apparent reduction of methane-derived authigenic carbonate extent from imagery data alone will therefore not necessarily indicate a loss of habitat and should not automatically be interpreted as such. The acquisition of photo-mosaic data by remotely operated vehicle or autonomous underwater vehicle would allow direct comparison of changes in the extent of methane-derived authigenic carbonate exposed at the seabed, improving our understanding of natural fluctuations in seabed composition over time. Subsurface cores (e.g. using a vibrocore) would confirm the depth and likelihood of exposure of any methane-derived authigenic carbonate underlying the sediment veneers.
Where methane-derived authigenic carbonate is exposed above the seabed, the physical structure is likely to change over time, through erosion by sediment scour and bioturbation. The challenge will lie in determining whether changes have occurred as a result of these natural processes or through human impacts. Given that the cliff structure observed in 2005 and 2008 was physically comparable in 2015, Noble-James et al. postulate that the physical structure should not change dramatically between monitoring events without physical damage having taken place; however, we cannot state this with certainty unless further studies on erosional processes and the rate of methane-derived authigenic carbonate formation are undertaken. The results of the methane-derived authigenic carbonate dating analysis suggest that methane-derived authigenic carbonate formation is extremely slow (in human terms) and that the methane-derived authigenic carbonate has been assessed as highly sensitive to physical loss, physical change, removal of substratum, and abrasion/disturbance of the surface of the substratum. Fishing equipment, such as bottom trawling nets, has been known to tear pieces off the carbonate structures. Methane-derived authigenic carbonate surveys could therefore acquire targeted lines of multibeam echosounder data across high-relief methane-derived authigenic carbonate and cliff features in areas where these pressures are known to occur (e.g. mobile or static demersal fishing grounds, or cable routes) and areas of no impact. Comparing georeferenced bathymetric profiles of these features over time would provide a cost-effective means of monitoring the physical structure in the areas most likely to exhibit detectable change. As the hydrodynamic and sedimentary regimes are thought to exert strong influences on the distribution and character of methane-derived authigenic carbonate and associated communities, an improved understanding of these processes, the natural erosive forces exerted on the methane-derived authigenic carbonate across the entire marine protected area, and the extent to which it is uncovered and reburied, could be achieved through hydrodynamic modelling, sediment capture, seabed coring, and/or the investigation of sediment veneer depth using high-frequency sub-bottom profilers (e.g. a parametric echosounder). It should also be noted that the joint effects of ocean acidification and the production of hydrogen sulphide by anaerobic oxidation of methane could create conditions accelerating the erosion of methane-derived authigenic carbonate, and therefore pH measurements at the seabed would be a valuable addition to future monitoring surveys.
Key and influential species and characteristic communities supported by the methane-derived authigenic carbonates and surrounding methane-rich sediments (and in turn, associated primary and secondary production) are important components of habitat condition assessments. Given the active methane seepage observed within the marine protected area, chemoautotrophic community structure was initially considered for future monitoring; however, Noble-James et al.'s study supports the findings of a number of previous publications indicating that chemoautotrophic assemblages, if present, are highly spatially constrained at shallow water methane seeps through competitive pressure from background fauna. It therefore appears that ‘blind’ infaunal sampling (i.e. using samplers that cannot be guided to specific targets) is of limited value for detecting chemoautotrophic communities in shallow areas, where substrate heterogeneity creates a large number of microhabitats, and when steep geochemical gradients are present. Chemoautotrophic communities, if present, are likely to be more effectively sampled using a remote operated vehicle. This would, however, require direct sampling over visually identified seeping gas bubbles (which were only encountered twice during the survey of Croker Carbonate Slabs) or dark patches of sulphidic sediment. This approach is likely to prove costly and is likely to yield relatively few data with which to evaluate any changes in chemoautotrophic community structure.
The results of Noble-James et al.'s study suggest that future biological monitoring at this marine protected area should focus on epifaunal ‘Characteristic communities’; however, the inherent limitations of epifaunal identification from imagery data have been highlighted at this marine protected area, as a result of both the turbidity of the water column and the problematic nature of remotely identifying taxa that do not have distinctive characteristics to aid the identification from images (including many of the dominant fauna at this marine protected area, e.g. Hydroids, Bryozoans, and Polychaetes). Given that future epifaunal monitoring will be centred on non-destructive imagery data, a morphotaxonomic approach to imagery analysis could be used, as these rely on morphological descriptors rather than the recognition of potentially cryptic taxonomic identification features to record epifauna from imagery. If adopted as the first step in a nested classification approach, morphotaxonomic classification could have advantages in improving the consistency and reliability of epifaunal identification where achievable taxonomic resolution is low. Such an approach would require testing before implementation, as well as the inclusion of a second step using traditional taxonomic identification, where species identification is possible. A fully quantitative approach to monitoring the ‘Characteristic communities’ is also required for robust and accurate assessments of change over time. Although the semiquantitative Superabundant-Abundant-Common-Frequent-Occasional-Rare scale was effective for characterizing both the sessile and motile fauna associated with methane-derived authigenic carbonates, the habitat provision function of methane-derived authigenic carbonates may be best assessed using the percentage coverage of epifauna (with post-hoc standardization of the image field of view). Given the highly variable and discontinuous nature of methane-derived authigenic carbonates exposure in some areas, epifaunal percentage cover should be recorded as the percentage of the methane-derived authigenic carbonates substrate when assessing the role of methane-derived authigenic carbonates in habitat provision, as opposed to the percentage within the image.
The five epifaunal taxa found to associate with the methane-derived authigenic carbonate exhibit similar life-history traits (i.e. they are attached sessile suspension feeders), which are common among hard substrate-inhabiting invertebrates, suggesting that they could be considered characteristic of the methane-derived authigenic carbonates at this marine protected area, when in Favourable Conservation Status. All five taxa were noted in association with methane-derived authigenic carbonate following the 2008 survey, with Alcyonium digitatum, Nemertesia, and Tubularia also recorded at the nearby Codling Fault Zone marine protected area in Republic of Ireland waters. Alcyonium digitatum is widely distributed across the North-East Atlantic, being recorded on hard substrates from Norway to Portugal, and has been noted in association with methane-derived authigenic carbonates elsewhere in Europe. This species is of particular interest as a ‘characteristic’ species for future monitoring (at Croker Carbonate Slabs and elsewhere) because of its large body size, conspicuous shape, and colour range, which can be identified from sediment-obscured images and video footage where the camera is suspended high above the seabed. Although monitoring these easily identifiable and methane-derived authigenic carbonate-associated taxa would contribute to an assessment of whether the ‘status quo’ has been maintained (i.e. the methane-derived authigenic carbonate is in Favourable Conservation Status), these taxa are not thought to be highly sensitive to physical abrasion by fishing gear (to which some areas of the marine protected area have been exposed). Hard substrates such as methane-derived authigenic carbonate are likely to be rapidly colonised by opportunistic Hydroids, Bryozoans, and Ascidians following such a disturbance; three of the five ‘characteristic’ methane-derived authigenic carbonate-associated taxa were Hydroids and Bryozoans, whereas the ‘characteristic’ Soft Coral Alcyonium digitatum is also not thought to be highly sensitive to physical abrasion. The percentage cover of slow-growing sensitive taxa, such as Sponges and Anthozoans (both of which were observed on high- and low-relief methane-derived authigenic carbonate), may therefore be more effective as an indicator of physical damage to methane-derived authigenic carbonate, although the validity of this metric would require further investigation.
The results of the epifaunal analyses indicated that treating the low and high-relief methane-derived authigenic carbonate forms as separate sampling strata in future monitoring designs may provide greater insights into ‘characteristic’ assemblage composition between different forms, which could potentially be more or less vulnerable to anthropogenic impacts (e.g. demersal trawling). If consistently observed through time, the occurrence ratio of observed taxa between the two methane-derived authigenic carbonate forms could contribute to an understanding of habitat condition. The inability of the model to differentiate the two methane-derived authigenic carbonate forms is most likely a result of fine-scale heterogeneity that could not be resolved using ship-mounted multibeam echosounder data. Mapping the different morphological forms of the methane-derived authigenic carbonate would require higher resolution acoustic data such as that achieved when using an multibeam echosounder or synthetic aperture sonar mounted on an autonomous underwater vehicle. An autonomous underwater vehicle-mounted synthetic aperture sonar could deliver acoustic data with a resolution of a few centimetres, but this is a cost- and time-intensive survey method. Stratification between known areas of high and low methane-derived authigenic carbonate (verified as such by an initial collection of imagery data), is likely to prove sufficient to investigate communities associated with each form in the early stages of a monitoring time series. If a more targeted approach to monitoring specific areas was developed over time, detailed synthetic aperture sonar monitoring could form part of a resource-balanced monitoring programme.
In summary, the multidisciplinary techniques used in Noble-James et al.'s study enabled a comprehensive study of the methane-derived authigenic carbonate feature attributes; however, they do not offer the most cost-effective use of resources for monitoring beyond an initial characterization. For subsequent monitoring events, at the Croker Carbonate Slabs Special Area of Conservation and elsewhere, a simplified combination of methods could provide sufficient evidence for the assessment of whether the methane-derived authigenic carbonate is in Favourable Conservation Status, and whether human impacts have affected the integrity of the methane-derived authigenic carbonate in discrete areas. This simplified sampling scope could primarily consist of towed camera transects A molecular electronic transducer sensor attached to the camera (and validated by water samples) would allow for the assessment of natural gas seepage and methane-derived authigenic carbonate accretion, whereas targeted multibeam echosounder lines could provide further evidence for changes in the physical structure of the methane-derived authigenic carbonate.
Although this approach would deliver sufficient evidence for a condition assessment, a more holistic understanding of the ecosystem, involving research and development studies, would provide valuable context for these assessments and would improve the ability of scientists to make robust and defensible recommendations on future management. Such studies could include subsurface methane-derived authigenic carbonate investigations using acoustic and coring techniques, detailed characterizations of infaunal and endolithic chemoautotrophic assemblages, the improvement of habitat models using photo-mosaics and high-resolution acoustic data, and modelling of the hydrodynamic and sedimentary regimes.
Noble-James et al.'s study has demonstrated that the Croker Carbonate Slabs form an extensive, stable, and complex lithic substrate in an area of the Irish Sea that would otherwise be dominated by sandy sediments; it is one of the largest known examples of shallow methane-derived authigenic carbonate in European waters. This substrate supports dense aggregations of epifauna (being a rich source of secondary production), increases habitat complexity and niche availability, and presumably contributes to larval dispersal by taxa that attach to hard surfaces and are typically rare in the offshore region of the Irish Sea. The observed active methane seepage, and the presence of Bacterial mats, suggests that chemosynthetic primary production is ongoing, and the habitat is still thought to be precipitating subsurface, with chemoautotrophic assemblages performing important climate regulation functions.
Noble-James et al. have demonstrated multidisciplinary approaches to characterising and monitoring these functionally important habitats. They have developed recommendations to maximize the cost effectiveness of future monitoring, whilst optimizing the scientific robustness of data to inform effective decision making on methane-derived authigenic carbonate conservation and management. As mentioned previously, the conservation status of shallow methane-derived authigenic carbonates within the European marine protected area network is largely unknown.
See also...
