1 Introduction

A steady population growth, increasing metal demand and lower grades of mineral deposits mined on land have fostered an interest in the economic potential of marine mineral resources. While mining of some commodities has already been carried out for many years in the shallow marine environment, others—especially in the deep sea—are still in the state of exploration.

1.1 Marine Mineral Deposits

The seabed is already an important source of commodities for humankind. Sand and gravel used for construction purposes or beach restoration as well as oil and gas have been mined from the sea for decades. Additionally, placer deposits of diamonds have been extracted off the coast of South Africa and Namibia for a long time, in addition to deposits of tin, titanium, and gold along the coasts of Africa, Asia, and Australia (Cronan 1992; Rona 2003, 2008). These placer deposits are accumulations of minerals with a high density and chemical resistance to weathering, which form along beaches by gravity settling due to wave or current movement. Marine phosphorite deposits may potentially be used to extract phosphate for fertilizers (Rona 2003). Such deposits can be found on the continental shelf in tropical upwelling regions, where cold deep water is rising to the surface (Föllmi 1996, Murton 2000). Dissolved phosphate in the low temperature waters precipitates along the upper slopes of continental margins and on the continental shelf. The continental shelf itself is underlain by bedrocks that are submerged and covered by sediments. Since these bedrocks host a variety of mineral deposit types that are currently being mined onshore, but close to the coast lines, the interest in shelf areas may increase even further if extensions of onshore mines can be traced into the shallow ocean (Hannington et al. 2017).

Raw materials are, however, not confined to shallow waters. Potential deep-sea mineral resources include manganese nodules, cobalt-rich manganese crusts and seafloor massive sulphides (Rona 2008; Hein et al. 2013; Petersen et al. 2016). Manganese nodules occur on the sediment-covered abyssal plains at depths of about 4000–6500 m. They are mineral concretions made up largely of manganese and iron oxides and oxyhydroxides. These form around a hard nucleus and incorporate economically interesting metals, such as copper, cobalt, and nickel, together with potentially valuable metals such as lithium, molybdenum, titanium, and rare earth elements (REE) from the sediment and seawater. Cobalt-rich ferromanganese crusts form on the sediment-free flanks of volcanic seamounts, ridges, and plateaus and are composed of manganese oxides and iron oxyhydroxides that precipitate directly from seawater. They are usually several cm thick, with the thickest and most metal-rich crusts forming in depths between 800 and 2500 m (Hein et al. 2013). The elements of interest are similar to those in manganese nodules.

Seafloor massive sulphides (SMS) are another deep sea mineral resource, associated with volcanically active areas along divergent plate boundaries. These deposits form at or below the seabed as a consequence of the interaction of seawater with a heat source in the sub-seafloor. Hot fluids leach economically interesting metals such as copper, zinc, gold and silver from the ambient rocks and transport them to the seafloor where they precipitate as metal sulphides forming SMS deposits. Their composition is highly variable as a consequence of variations in source rock composition, water depth, and input of magmatic volatiles (Hannington et al. 2005).

Marine mineral resources fall under two different legal regimes depending on their location: they either occur within the legal boundaries of national jurisdiction of a coastal state or occur in areas beyond national jurisdictions. In the latter case exploration and exploitation are managed by the International Seabed Authority (ISA) that was established in Kingston (Jamaica) in 1994 based on the United Nations Convention on the Law of the Sea (UNCLOS).

1.2 General Exploration Methods for Resource and Environmental Impact Assessment

Hydro-acoustic mapping of the seafloor with multibeam echosounder (MBES) and side scan sonar (SSS) systems are the most important tools for the exploration of submarine mineral resources. These methods provide a quick overview of the bathymetric and substrate properties of the area of interest. Both are very important for evaluating the possible minable terrain and to estimate resource potentials. After mapping with hull-mounted MBES systems, which provide information about major terrain structures, areas of interest will be mapped in detail using deep/close to the seafloor operating platforms. This is especially important for deep sea resources, where the resolution of hull-mounted systems provides only a resolution of several tens to a hundred of metres, not enough for a detailed assessment. In contrast hydro-acoustic mapping (MBES, SSS) with autonomous underwater vehicles (AUVs) operating at several tens to a few metres altitude (depending on mission and terrain variability) provides a detailed insight in small scale terrain structures with few metres or even less than a metre resolution. Backscatter (BS) data, photos or video data, recorded by either AUV’s, ROV’s or towed camera systems, as well as specific sampling platforms are used for ground truthing, substrate characterisation. Sub-bottom profiling reveals valuable information about sediment thickness and underlying structures, which can be important for resource evaluation of some deposits. Apart from resource assessments, the same methods are used to study mining-related environmental impacts; this needs to be part of any exploration action undertaken under the jurisdiction of the ISA [ISBA/19/LTC/8]. Habitat mapping, which forms the basis for environmental impact assessments (EIA), requires information about bathymetry and substrate characteristics on various scales. Video and photo data, as well as biological sampling, reveal the respective benthic and epi-benthic fauna and community structures, which are only poorly studied, particularly in the deep sea. As benthos and ecosystem services of the top sediment layers are essential to local and regional biogeochemical processes and ecosystems, the consequences of habitat destruction through seabed mining need to be properly evaluated.

In the following sub-chapters, the exploration, exploitation and monitoring methods for environmental impact assessments are described in more detail for sand and gravel, Mn-nodules and SMS deposits.

2 Resource Description

2.1 Sand and Gravel

Sand and gravel deposits are terms used to describe marine aggregate deposits based on grain sizes and performance, rather than the mineralogical components (Padan 1983). The size thresholds defining the nomenclature varies between countries, but scientifically, sand defines grain sizes between 0.63 and 2 mm in diameter, and larger grains are classified as gravel (Collins 2010). These deposits occur on the continental shelf at or close to the seabed surface and as relicts from river and coastal bank deposits. These were formed under low sea level conditions during the last glacial period, when rivers extended onto large parts of the continental shelf, filling their river-beds with eroded continental runoff material. Modern deposits are a product of present hydrodynamic and sedimentary particle reworking and are therefore strongly related to coastal sediment budgets and dynamics (Collins 2010). The composition can be very variable depending on their formation history. Relict deposits mainly consist of land derived particles, while modern sand and gravel bodies also contain a considerable amount of shell fragments, which influences the resource quality (Collins 2010).

2.2 Mn Nodules

Ferromanganese Nodules are widely abundant in global oceans and occur mostly at the sediment surface (Fig. 1) or in near-surface sediment layers. They are concretions of Mn-oxides and Fe-hydroxides that precipitated around some kind of a nucleus, such as rock fragments, shark teeth, shells or fragments of micro-nodules.

Fig. 1
figure 1

Mn-Nodules covering the sediment surface within the CCZ and organism associated with the nodules as hard substrate habitat (ROV Kiel 6000, GEOMAR)

Full size image

Besides their name-giving main components, the nodules also contain trace metals adsorbed within the oxy-hydroxides, which make them of interest to the mining industry. Two major formation types can be distinguished: hydrogenetic nodules that form by precipitation from the ambient seawater, and diagenetic nodules, which grow by precipitation from the sediment pore water. Most of the nodules are mixed-types rather than grown from only one of the end member formation types (Hein and Koschinsky 2013). Depending on the formation type and geographic occurrence, the growth rate varies, which influences the metal enrichment within the concretions. The formation itself is affected by the metal supply, the seafloor morphology that influences the depositional environment (bottom current strength and sediment composition), as well as by bioturbation (Hein and Koschinsky 2013). Only in some places of the world’s oceans the nodule grade, according to size and content of the most precious metals (Cu, Co, Ni, REE’s), is high enough for a possible mining. These places lie in sediment-covered abyssal plains at water depths between 4000 and 6500 m. Here the sedimentation rate is very low (<1 cm/ky−1), which allows slow nodule growth rates (only 1–20 mm/Ma) in a semi-liquid surface layer with sufficient pore water for the diagenetic growth. The bottom waters are also well-oxygenated enabling the formation of oxy-hydroxides (Petersen et al. 2016). Areas of young oceanic crust (<10 Ma), high sedimentation rates and a local morphology of >300 m relief are generally not suitable as a possible exploration area (Petersen et al. 2016). The largest high-grade Mn-nodule occurrences are known to be in the Central Indian Ocean Basin, the Peru Basin and in the Clarion-Clipperton Zone (CCZ) in the equatorial Pacific between Hawaii and Mexico. Here, a 4.2 million km2 area is managed by the ISA, with several countries having shown a commercial interest for decades and where mining could become reality in the near-future.

2.3 Seafloor Massive Sulphides (SMS)

Metal-rich hydrothermal deposits form as a consequence of seawater circulating through the oceanic crust in volcanically active areas. Cold ocean water penetrates through cracks into the sub-surface where it is heated by a heat source, commonly magma. The hot fluids leach metals and sulphur from the surrounding rocks within the crust and finally ascend back to the seafloor, where a part of the dissolved metals precipitate as a reaction with the cold seawater, forming the characteristic chimneys (black smokers) or mounds on the seafloor (Fig. 2; Hannington et al. 2005). Some portion of the fluid is expelled into the overlying water column, forming hydrothermal plumes that are commonly used to search for active deposits (Petersen et al. 2016, Baker et al. 2016). Active black smoker systems have been discovered at all plate boundaries as in spreading centres of Mid Ocean Ridges (MOR), along active submarine volcanic arcs and in back-arc basins. Only very few deposits are known to occur in association with intraplate volcanoes. The greatest economic potential of the deposits can be expected in water depths between 1000 and 5000 m (Petersen et al. 2016).

Fig. 2
figure 2

Active black smoker chimney at the Mid-Atlantic Ridge. Photo ROV Kiel 6000, GEOMAR

Full size image

3 Exploration Methods

3.1 Sand and Gravel

For exploring the horizontal and vertical extent and surficial character of sand and gravel deposits, MBES and SSS systems are employed and BS data play an important role. Based on these data, deposits of the required grain size are identified based on sampling, local knowledge about the sediment geology, as well as Angular Response Analysis (ARA). ARA uses the characteristic BS strength variations with incidence angle for different substrate types and frequencies. By comparison with model results, the substrate type can be evaluated. Highly detailed bathymetric information in correlation with sub-bottom profile data is used to identify deposits in the sub-seafloor; structures such as paleo-channels or paleo-river beds, which are most probably filled with the material of interest, are targeted. Detailed sediment sampling, coring and drilling finally provide validation of the acoustic mapping results and reveal the resource quality. Detailed MBES data from the extraction area before and after the mining (Fig. 3) are an important basis to evaluate the extraction process efficiency, amount of sediment removed and the related potential impact on the local hydrography/current regime and ecology.

Fig. 3
figure 3

a Detailed bathymetry before and b after mining; the dredge tracks are clearly visible. c Difference between before and after dredging, furrows are up to 1.2 m deep. Source DEME-Dredging environmental and marine engineering

Full size image

3.2 Mn-Nodules

Besides bio-geochemical factors and oceanographic parameters, the seafloor morphology is a key parameter in evaluating Mn-nodule resources in the deep sea. Satellite-derived bathymetric data even with a low resolution of 1–3.7 km (ETOPO2 or GEBCO30 data) allow the identification of abyssal plains in relation to seamounts, ridges, or troughs located in the admired water depths. One of these abyssal plains lies in the Peru Basin where the abundance and grade of occurring Mn-nodules were considered as economically valuable and which has been studied in detail since the 70s (Von Stackelberg 2000). In 1989, the seafloor of an area called “DISCOL Experimental Area” was disturbed by a plough harrow and has been revisited since then several times in order to study the ecological recovery from this disturbance and thus gain knowledge with respect to deep sea mining EIAs (Thiel 2001; www.discol.de). The following example is from the DISCOL area (Greinert 2015), located approximately 500 miles west off the Peruvian coast. MBES mapping with the hull-mounted system of RV SONNE (EM122; 0.5° by 1° beam angle, 18 kHz) reveals the bathymetry of the area with a resolution of 50 m (Fig. 4a, b); the relief of this ‘abyssal plain’ shows depth variations of 300 m.

Fig. 4
figure 4

a Ship-based bathymetric map revealing the morphology with a horizontal resolution of 50 m. Black square marks the AUV-mapped central DISCOL area. b Extract from the ship-based bathymetry showing the AUV-mapped area with the 50 m cell size resolution. c AUV-obtained bathymetry with a horizontal resolution of 2 m revealing small-scaled structures that were not resolved in the ship-based bathymetric map

Full size image

The different bathymetric settings (seamounts, ridges, basins, plains) influenced sediment depositional settings that also cause variability in Mn-nodule abundances and compositions. Backscatter data recorded with ship-based MBES systems (Fig. 5a, b) allow the identification of higher/lower Mn-nodule abundances (Rühlemann et al. 2011). Areas of higher Mn-nodule abundance are typically associated with higher BS values because the existence of ‘many’ hard and irregularly shaped bodies on the seafloor increase the scattering potential in comparison to no-nodule areas (all this needs to consider the slope/incidence angle of the respective hydro acoustic beam). The results need validation by direct sampling (e.g. box coring) but in general enable the evaluation and segmentation of potentially low and high Mn-nodule abundance on large scale in a relatively short time.

Fig. 5
figure 5

a BS data from hull-mounted MBES system; resolution 50 m. Bright to dark grey scales indicate high to low BS signal strength. White square marks the AUV-mapped central DISCOL area. b Extract from the ship-obtained BS data showing the AUV-mapped area in a low resolution. c AUV-obtained BS data with a horizontal resolution of 2 m revealing small-scaled structures and nodule-free areas (very low BS signal strength) that were not resolved in the ship-based BS map

Full size image

Recent studies show that a more detailed knowledge of the local terrain is important for a realistic resource assessment that also needs to consider the minable terrain (Madureira et al. 2016; <10° slope, Fig. 6) determined by the technical limitations of the crawler capabilities. For planning mining operations, ship-based bathymetry is not sufficient. Over the last decade MBES systems have been reduced in size so that they can now routinely be used on remotely operated vehicles (ROVs) and on autonomous underwater vehicles (AUVs) providing bathymetric data with a resolution down to a few decimetres, depending on the altitude at which these instruments are deployed (Caress et al. 2012; Yoshikawa et al. 2012; Clague et al. 2014). The AUV-based bathymetry presented was typically acquired at 80–50 m above the seafloor, delivering data of about 2 m horizontal and decimetre vertical resolution, from which smaller-scaled in-homogeneities, also in habitat distribution, can be derived (Fig. 4c). Terrain analysis clearly shows flat areas suitable for mining (Fig. 6) and correlating AUV- or ROV-obtained photo data reveal variabilities in Mn-nodule coverage within several metres to tens of metres distance, clearly linked to bathymetric changes.

Fig. 6
figure 6

a Slope calculated from the ship-obtained bathymetric data. b Extract of the AUV-mapped area with the ship-based bathymetry resolution. c Slope map of the central DISCOL area derived from AUV-obtained bathymetric data. Green and orange areas are suitable for mining (color figure online)

Full size image

AUV-obtained BS and SSS data provide an equally detailed insight into the local substrate properties with a resolution of up to 1 m (SSS). With this mapping method, nodule-free patches or substrate types that could be challenging for the mining gear (e.g. tallus block fields) can be identified. These small-scale changes in Mn-nodule abundance can be visualised in the BS data of Fig. 5c where nodule-free patches stand out as very low BS areas. In a seafloor classification effort different statistical values of BS measurements falling in a larger grid cell (here 6 m) can be used in an un-supervised approach to improve knowledge about small-scale seafloor/habitat changes (Fig. 7). At the same time these results lead to a better understanding of driving processes in nodule formation, in particular in combination with additional sub-bottom profile data, specific direct sampling and optical investigations.

Fig. 7
figure 7

Unsupervised classified map based on statistics of AUV-obtained BS data (mean, mode, min, max; ISODATA algorithm in SAGA-system for automated geoscientific analyses) for identifying areas of different Mn-nodule coverage

Full size image

3.3 Seafloor Massive Sulphides

For the prospection for SMS occurrence, global satellite-derived topographic data can be used to define regions of interest, e.g. the location of the spreading axis in areas where no other data are available. The small size of the resource targets—tens to a few hundreds of metres for individual deposits—implies the need to obtain higher resolution data. Even ship-based bathymetric data (Fig. 8) will not identify individual occurrences, but will allow defining target areas more precisely due to detectable indications of e.g. faulting and recent volcanic activity. The capability of acquiring high resolution AUV-based data has profound impacts, not only on our ability to detect SMS, but also on assessing the resource potential. In a recent two-week long survey at the Endeavour Segment in the Northeast Pacific Ocean, AUV-based bathymetry quadrupled the number of identified hydrothermal chimneys and mounds (Jamieson et al. 2014), although this vent site has seen well over one hundred submersible and ROV dives over the past 30 years and is considered one of the best studied vent fields in the oceans.

Fig. 8
figure 8

a Ship-obtained bathymetry of a section of the Mid-Atlantic Ridge with the TAG vent field located on the hanging wall of an active detachment fault. b Zoomed-in view of TAG field (ship-based bathymetry; Resolution: 30 m)

Full size image

Such high-resolution hydro-acoustic surveys overcome another shortfall of SMS exploration. So far prospecting technologies have mainly been developed for the search of actively venting sites that can easily be traced through physical and chemical anomalies in the water column as temperature, concentration variations in Mn or Fe, redox potential (Eh), and particle concentration. These plume surveys have been a primary tool for the exploration of SMS systems, but they only identify active, and therefore mostly young and thus small mineral occurrences. Economically more interesting deposits are inactive and have gone through their entire life cycle (Hannington et al. 2005). It has been recognised that these deposits are not only much larger than active deposits, but that they and can often be identified based on their shape, aspect ratio, or slope (Fig. 9) derived from high-resolution bathymetric surveys. The combination with other geophysical parameters such as magnetic or self-potential-sensor data recorded during the same survey, may be used to further distinguish volcanic mounds from hydrothermal occurrences.

Fig. 9
figure 9

a AUV-obtained bathymetry of the TAG field (resolution: 2 m) revealing the active TAG mound and inactive mounds. b Zoomed-in view showing the active TAG mound and the surroundings in more detail

Full size image

A first resource assessment can be made by estimating the thickness of such mound-style occurrence by comparison with the surrounding seafloor and estimates of bulk sulphide density derived from previous studies (Jamieson et al. 2014). However, former sampling has revealed that such deposits can be very variable in terms of internal composition, which makes sampling and drilling a necessity for a proper resource assessment. Detailed AUV-based MBES mapping (Fig. 9) allows the identification of SMS deposits and distinction of mounds, based on their slopes and shape (Jamieson et al. 2014). For the identification of old and possibly larger deposits that are unfortunately sediment covered, new exploration methods still need to be developed (Petersen et al. 2016).

Current exploration models assume that large extinct SMS occurrences can be found in a strip of up to few tens of kilometres away from the mid-ocean ridge. Further away from the ridge axis, the coverage by sediments or lava may become too thick for exploitation and/or the sulphuric minerals are ‘destroyed’ by dissolution from oxidized seawater. Still, the potential to find extinct sulphide deposits far from the ridge axis adds an additional vast seafloor area for future exploration.

In general, AUV-based high-resolution surveys, as needed for deep sea resource exploration, will need to cover large areas of the seafloor. It is unlikely that this can be done with single AUV surveys, and fleets of AUVs working together will be operated in the future, as time is an additional factor in exploration and each contractor to the International Seabed Authority has to explore 10,000 km2 within the 15-year runtime of the contract, during which 50% of the area has to be returned to ISA after eight years. Deep sea exploration needs to be fast and cost-effective and therefore acoustic mapping of the seafloor will be the method of choice for decades to come.

4 Exploitation Methods

4.1 Sand and Gravel

Sand and gravel deposits commonly get extracted by static or trailer suction dredging, where the material is hydraulically sucked into the dredging vessel by pumps through a large pipe system (Fig. 10). During static suction, dredging the material is extracted punctually, creating up to 25 m deep and 200 m wide extraction craters (Collins 2010). This method allows the exploitation of deeper located and thick deposits as is typically done for diamonds (e.g. offshore Africa). For the trailer suction method, the vessel slowly moves, dragging the suction pipe over the seafloor, which creates extraction furrows up to a few metres deep (Fig. 3b, c). This method is used for the exploitation of thinner surficial deposits as sand and gravel. Although a wide area will be affected, the environmental impact is believed to be lower than the deep pit structures created by static suction dredging. Besides these two most common dredging methods, numerous other concepts are available for special circumstances (Cruickshank and Hess 1978).

Fig. 10
figure 10

Antigoon Trailing Suction Hopper Dredger from the DEME group, which can reach a maximum dredging depth of 45 m. The diameter of the suction pipe is 1200 mm. Image kindly provided by DEME—Dredging environmental and marine engineering

Full size image

Depending on the application, the material is screened at the vessel and immediately discharged onto a barge, to the shore through a pipeline system or back into the water column (Padan 1983) when not matching the required size or properties, to minimise the amount of unwanted material. The material is removed from the vessel onshore by different kinds of machinery. For beach nourishment, hydraulic discharge functions are also available on the vessel.

4.2 Mn Nodules

For several years, different methods for Mn-nodule extraction have been discussed. Early considerations in the 1970s included the use of an “air lift dredge” (a system consisting of a collector head, which is connected with a suction system pumping the material to the surface) and a “Continuous Line Bucket System” (consisting of a cable loop, where buckets are mounted in 25–50 m intervals; this cable loop lowers the buckets, skims them over the bottom where they get filled with Mn-nodules and finally rises the filled buckets back up to the sea surface) (Pearson 1975). The methods considered nowadays comprise a mining technique consisting of 3 sub-systems (Fig. 11). At the sea surface there is the mining vessel, from which the mining operation on the seafloor can be controlled and from which the lift pipe system (riser pipe) will be suspended. The actual mining will be done by crawler(s) which move over the seafloor and extract the nodules hydraulically or mechanically, depending on the mining concept (Oebius et al. 2001). According to a German concept developed by Aker Wirth GmbH, the crawler is approximately 17 m wide and 15 m long (Kuhn et al. 2011) and will move with approximately 1 m/s, removing the upper 10–20 cm of sediment. The collectors, which are still under development, need to extract 5–8 thousand tons of nodules per day for an economically valuable mining operation while keeping the environmental harm to a minimum (Madureira et al. 2016). After the separation of sediment and nodules within the crawler, the crushed nodules are transported through the lift pipe system to the surface platform, where the material might be further processed and transported to the shore.

Fig. 11
figure 11

Simplified scheme for the Mn-nodule mining system and the associated environmental impacts of the different system sections. The nodules will be collected and crushed by crawlers and the material will be transported through the Riser pipe to the production support vessel. The waist material from on-board processing will be discharged through a return pipe. Source GRID Arendal: http://www.grida.no/graphicslib/detail/example-of-a-sea-floor-manganese-nodule-mining-system-and-related-sources-of-potential-environmental-impact_f38d)

Full size image

4.3 Seafloor Massive Sulphides

Since SMS deposits can be several tens of metres thick and the terrain they occur in is very rugged, the mining technology needs to be able to move not only in a difficult terrain but also needs to include cutting devices for fragmenting the SMS deposit rocks (Liu et al. 2016). The crushed material may be collected and transported through the riser system to the surface support vessel to be pre-processed and transported to the shore. The leading technology is orientated on mining technology deployed in land mines. Seafloor mining machines will be equipped with some kind of cutter head, similar to gear used for cutting through rock while coal mining (Ishiguro et al. 2013; Liu et al. 2016). In most scenarios a collector machine is added to collect the fragmented rocks, similar to a Mn-nodule ‘harvester’. The fragments will be pumped into the crawler, possibly transferred into a buffer system, and then through the riser to the support vessel. According to the concept of Nautilus Minerals three different kinds of seafloor production tools will be deployed (Fig. 12): The Auxiliary Cutter (AC) will flatten the rough terrain and create benches to move on for the following tools. The Bulk Cutter (BC) will cut material from the deposits moving on the flat terraces created by the AC. Finally, the extracted material will be collected and transferred to the Riser and Lifting System by the collecting machine (CM, www.nautilusminerals.com). Other mining systems try to include the fragmentation and collection within one vehicle (Ishiguro et al. 2013). Alternatively large-scale grab systems may be used but further developments of mining systems based on currently ongoing tests will determine the most efficient exploitation technology for SMS deposits.

Fig. 12
figure 12

Mining concept of nautilus minerals for SMS exploitation. Three kinds of seafloor production tools will be deployed: The Auxiliary Cutter will be used to flatten the rough seafloor. The bulk cutter will then be moving along the created flat benches extracting material from the deposits. The extracted material will then be collected by a collecting machine, which also transfers the slurry to the Riser and Lifting System. Source Nautilus Minerals

Full size image

5 Monitoring Exploitation and Environmental Impact

5.1 Sand and Gravel

The environmental impact from sand and gravel extraction includes physical and biological effects, most prominently the complete destruction and removal of the seafloor surface, the creation of sediment plumes with unnatural turbidity as well as more or less pronounced impacts on the local current regimes and wave patterns. Due to the coastal proximity of the deposits, this impacts will eventually affect the coastal environment (Cruickshank and Hess 1978). The recovery of the excavation site is likely to take several decades in some areas. To avoid increased coastal erosion, prediction models are run prior to any exploitation activity—based on water depth, distance to the shore, depth and distribution of the resource, mobility of the seabed as well as oceanographic parameters—to secure coastal integrity (Collins 2010). Consequently, the evaluation of these data needs to be included in the exploration phase.

Unavoidable ecological impacts during sand and gravel mining include changes in fauna/flora diversity, density and community structures, particularly of benthic organisms due to the physical effects (Collins 2010). Substrate removal and changes in the local current regime considerably alter the habitat properties of the benthic communities. Increased turbidity can affect filter feeding sessile fauna by either clogging/hampering their filter capacity or burial by the re-settling sediment plume even at larger distance from the actual mining site. The final impact assessment of dredging on organisms needs to be based on a reliable knowledge of the local community and stress factors the organisms are naturally exposed to, such as tide-storm-caused enhanced sediment load. Ecosystems and fauna communities accustomed to such natural variabilities will likely be less affected by the dredging activities (Collins 2010). Studies of benthic community structures with direct sampling, visual observations and hydro-acoustic mapping (MBES, SSS) are needed for an accurate habitat mapping which is of great importance for the sand and gravel mining-related EIAs.

5.2 Mn Nodules

The extraction of the Mn-nodules from the seafloor will come with a clear risk for the benthic ecosystem. On one hand the removal of substrate will cause habitat loss for sessile fauna depending on the nodules as hard substrate (Fig. 1; Vanreusel et al. 2016) and for organisms living in the uppermost fluffy sediment layer. Both nodules and the top 10–20 cm will be removed and either be transported to the surface support vessel (Mn-nodules with as little sediment as possible) or to a large extent re-sedimented on the seafloor. In addition, the movement of the crawler itself will create a sediment plume that will be distributed uncontrollably over larger areas with the consequence that the re-deposition of the suspended sediment buries sessile organisms and clogs filter feeders even outside the actual mining area resulting in an increased mortality (Markussen 1994; Sharma 2011). Large amounts of suspended and re-depositing sediment are very unusual for the deep sea low sedimentation environment and could also cause bio-geochemical disequilibria impacting the local environment (Shirayama and Fukushima 1997; Kotlinski and Stoyanova 1998; Sharma 2001). Estimating the size and distribution pattern of the re-settled sediment cloud is therefore of specific interest for correctly assessing the extent of the area which would be ecologically influenced/harmed by the mining activity.

The evaluation of the mining-related environmental impact needs a precise knowledge of the benthic habitat distribution on different scales. While ship-based MBES data points at large scale habitat changes on kilometre-scale, small-scaled habitats on metre to even sub-metre scale are important to identify. These can only be efficiently mapped by a combination of AUV/ROV-based MBES data, SSS and backscatter analyses providing information about substrate types, linked to visual observations and their expert annotation/evaluation.

Studies undertaken in 2015 in the CCZ and the DISCOL area show that the re-sedimentation of bottom near sediment plumes are affected by the local morphology even on sub-metre vertical scale changes. Such an approach could effectively monitor the impacted area after mining activities. With AUV-obtained SSS (resolution 1 m) and BS data (resolution 2 m) the tracks left behind can be monitored and used to better document the change of the seafloor ecosystem (Fig. 5c). A total recovery of the Mn-nodule ecosystem as it existed before the mining will not be possible as the Mn-nodules, being an important micro-habitat in the abyss, have been removed. Thus the primary goal is to mitigate the impact as much as possible and allow for similar or new ecosystem services fulfilling habitats to establish in the shortest time possible.

Besides the seafloor generated sediment plume, the waist material from the processing at the mining support vessel will be discharged back into the water column. Discussions are still ongoing at which depth these tailings should be finally released, considering technical, economic and ecologic aspects. Any chemical leaching from pre-processing steps should be avoided, as this might add an unnecessary toxicity component.

5.3 Seafloor Massive Sulphides

Hydrothermal vent fields are characterised by high biomass and low diversity, whereas inactive deposit sites are colonised by fauna also typically associated with hard substrate seamount communities (Boschen et al. 2013). The microbial communities of inactive occurrences will, however, be very specialised due to the presence of sulphides as a substrate. Like any other mining activity, SMS mining comes along with the removal of habitat substrate, reducing edifices and altering the texture of the substrate (Baker and Beaudoin 2013). Mining active vent sites would change the distribution of venting sites for tens to hundreds of metres, directly affecting sessile benthic fauna. However, due to the general small size of the deposit and the high temperature/low pH very close to the vent sites, targeted mining of active vent sites is rather unlikely.

The mining of the SMS deposits will expose fresh sulphide mineral surfaces, from crushed down material and the freshly exposed rock surfaces, to oxygen rich bottom water that cause the sulphides to be oxidised. Acid generation during this process is likely small (Bilenker et al. 2016), but a significant increase of toxic metals in the bottom water (e.g. Cu, Zn) will happen on short time scales (<day) possibly reaching lethal concentrations in the near field and cause accumulative effects further away on a longer run. This impact will vary from one deposit to another based on the differences in the geochemical inventory of the deposits (Hannington et al. 2005) but can be considered essential, as the spatial impact of SMS mining will be much smaller compared to Mn-nodule and sand and gravel extraction. As for these other two resources, hydro-acoustic mapping techniques paired with visual studies and direct sampling are the tools for evaluating the immediate and long-term impact of SMS mining.