Marine renewable developments in Scottish waters: review of benthic ecological surveying

9 Study discussion and conclusions

This study has undertaken a structured process of context building; scoping; evaluating survey techniques available and strategies required; and then formulating a framework of recommendations about what surveys can be implemented, most appropriately, at various stages of development and for various development types. Building from those survey framework recommendations, the following sections provide additional reflections, recommendations and conclusions in the key topic areas defined for this study.

9.1 Review of existing benthic ecology survey guidance, survey techniques, and data flows currently employed at offshore renewables sites in the UK and internationally

A significant volume of guidance and recommendations have been published about practices for surveying seabed ecology, but to date they have been prepared for geographies other than Scotland and to meet different sets of objectives, purposes and policies. Consequently, although most of the materials are generally coherent, and provide useful insights and findings, they are not necessarily consistent nor particularly relevant to the permitting and monitoring of MREDs in Scottish waters. For example, many documents have been developed for reference and theoretical validation rather than for practical use.

Therefore, the conclusion reached was that although certain principles and approaches could be harvested from existing guidance and study results, the application of this previous work to Scotland’s and MRED’s needs would require some adaptation.

The key areas of differentiation were:

• Recognising that there is a need to achieve a balance between the detail/volume of data gathered and its functional use or purpose. More data does not necessarily provide greater understanding or insight into an issue.
• Considering the likely extent of any impact mechanisms and the levels of feature sensitivity should feed into the survey design at an early stage to enable survey activity to focus on priority impact pathways.
• Adopting the MHC/EUNIS levels of marine habitat and community classification as a core descriptor of needed survey intensity and delivered survey capacity.
• Recognising that sea and seabed conditions are different to those encountered in other parts of the UK and Europe and contribute as a major factor to planning and undertaking seabed surveys in Scotland. Typical conditions include greater water depths, wider prevalence of rock, more exposed areas, different daylight patterns, greater transit times, etc.
• MREDs, and particularly offshore wind farms in Scotland, will be installed at a larger scale than in many other places. This could mean that they interact with many different habitats, leading to a greater range of survey techniques being needed on any given project.
• The close proximity of a number of likely development sites to each other may heighten the need for cumulative effects to be considered.
• Acknowledging that a different pathway to sectoral development has taken place in Scotland, in comparison to the rest of UK, with earlier implementation of wave and tidal testing and demonstration, but later expansion of offshore wind – enabling more established learning to be applied.
• If a regional-scale strategic benthic sampling approach was adopted for Scotland, a suitable division of regions would be required to ensure sampling is locally relevant. At present the Scottish marine area is divided into eleven regions for regional marine planning purposes. These may provide a suitable basis for regional survey management.
• Recognising that under a standard level of community sensitivity and project complexity, the evaluation and assessment of seabed impacts takes place at a descriptive level equivalent to Levels 3 and 4 on the MHC/EUNIS scale.
• Appreciating that there is existing data and established understanding about Scottish waters in various locations and organisations that to date have not been uploaded to present data sharing platforms. The data sources include survey contractor records, academic research records, ecological advisors, fishing organisation, fishers themselves, scuba divers etc.
• Acknowledging that surveying is not always necessary for reaching a clear and robust conclusion for certain objectives or questions being asked. There are times where modelling, detailed analysis and sometimes conceptual analysis of a situation can lead to the required clarity and help to determine the best next steps.

9.2 Review and assessment of new, emerging survey technologies and analytical techniques that could be applied to MREDs

Within the range of seabed ecology surveying tools and techniques reviewed for this project, there were a number that were considered to be new and/or emerging. The new and emerging techniques reviewed were primarily relevant to the analysis stage of a sampling approach. As such, these are described in section 3.5.3 which reviews the literature relevant to these techniques as well as further information in the dedicated section on Novel Analytical Techniques (3.6.1). Some of the methods (e.g., eDNA) were assessed in the Evaluation Matrix and associated results in section 6.4. Of these tools and techniques, a few were considered to be potentially relevant to supporting MREDs consenting processes. However, during the assessment it became clear that at present, these should be viewed as being optional additions to the existing suite of tools and techniques rather than being replacements.

A short commentary of each of the options considered and their possible pathway to greater use is provided below:

• Reconfiguring BRUV timelapse photo/video surveillance to record movement and behaviour across an area of seabed surface - this might illuminate the movement of macrofauna (crustaceans, echinoderms, molluscs, small fish) through the area, for example. The data could also show the behaviour of tentacles, siphons, burrows and other feeding/foraging mechanisms of infauna and the behaviour of any epifauna present. Such outputs could be applied to assessments on potential organic enrichment effects from biofouling deposits, for example, or used for establishing new ecosystem function measures of community health linked to defining recovery from impacts such as excavation or sediment deposition.
• Enhanced imaging techniques clearly add new dimensions to imaging and expand what can be seen and understood, but they are unlikely to replace ROV/sledge video or high-resolution drop-down photography since these techniques provide alternative or additional real colour visual capacity at relatively low costs. A possible new development could be to integrate this technology with autonomous survey drones that could operate untethered, but which could still send back real-time imagery to an offshore or onshore control centre.
• Two clear uses for eDNA are to check for any INNS (where DNA has already been catalogued) and/or to confirm the presence of species that are too sporadic, cryptic, or hidden to be effectively sampled by existing video, photo, grab and coring techniques. For both purposes, improving the DNA genetic sequence databases to include sequences for more seabed species will be key for the development of this technique. Further research is also required to calibrate the method to be confident that a positive DNA result is representative of a particular locality within a given range. Linked to that is the significance of an absence result. Present techniques are accepted as useful despite rogue absence of species in some samples due to the species being too thinly spread or located too deep in the sediment for example. There is a question therefore about how an absence of eDNA result should be interpreted given the longevity of eDNA, how much it might get re-distributed and at what levels it becomes detectable. However, these limitations should be considered in the context of limitations of existing methods where species may be too difficult to detect or identify.
• The application of AI is already taking place at some level within the seabed surveys to help surveyors, taxonomists and seabed survey experts. The key question is perhaps whether AI will open-up or majorly advance certain approaches which traditionally use non or limited AI processes. Some possible examples of AI applications include:
• Shape/pattern recognition: developing suitable recognition protocols to stimulate a camera or sensor to take video, stills image or a sample when something specific has happened.
• Shape/pattern recognition: analysis of captured video clips and stills to recognise specimens, species, behavioural traits in sequences and shots.
• Rapid quantification/counting: Real-time tallying of identifiable features in acoustic, laser or visual spectrum data sets.
• 3D scanning: recognition of taxonomic specimens or features on surface of seabed.
• Seabed tractor technology is already used to assist in the laying of cables and pipelines as well in checking their status and condition. A key next step is to enable such vehicles to manoeuvre untethered over the seabed. This could be especially useful in areas where tethered craft are difficult to control such as in strong currents or heavy seas. They could also be used to cover transects repeatedly over time, perhaps servicing a cluster of sampling stations with one vehicle.

9.3 An assessment of the most effective methods and sampling designs for different monitoring requirements and objectives

This study has shown that there are a large array of potential tools and approaches that could be applied to surveying seabed ecology in Scotland around MRE projects. However, when the context for this surveying activity is considered, including the purpose; the type of habitats present; the communities and species found; the ambient operating conditions; and the quality and utility of the data gathered, there are nine core tools/techniques (summarised below).

This suite of selected tools creates a gradient of increasingly detailed insight into ecological conditions and status. This gradient can be broadly structured into three categories as follows:

• The geophysical remote sensing techniques give insights about the ecology present.
• The visual tools start to provide evidence of the key indicator species present.
• The physical biological material samples recovered by divers, grabs or box cores provide tangible evidence of the species present and increasing intensity of sampling gives increasingly comprehensive insights into the details of seabed ecosystem character.

The recommended tools/techniques are listed below:

• Swath bathymetry which reveals the depth and shape of the seabed, and from that some key seabed forms and textures. These can infer certain seabed types, objects and/or bedforms (e.g., relict rocks, faults, sandbanks, sand waves, debris, wrecks, cables, pipelines, trawl scars).
• Geophysical surveying techniques which reveal the near surface structure of the seabed confirming certain seabed types (e.g., bedrock, broken rock, gravel, sand, mud) and depths of these features.
• Video transects along pre-defined seabed routes or exploring new territory (e.g., smaller scale surface features such as ripples, sediment veneers on rock, larger epifaunal and surface living fauna and flora). These videos obtained by cameras deployed by towed sledge, tethered ROV, diver and possibly untethered AUV.
• Still images of seabed communities, taken vertically or obliquely, giving high resolution and wide depth of field images. These enable a greater range of smaller and quicker moving species to be more easily identified as well as finer scale detail of underlying habitat type to be described. The choice of vertical or oblique views will link to the specific purpose and target features of interest, with horizontal detection and quantitative use needing vertical orientation and more vertical target features and background context suiting a more oblique orientation.
• Diver observations of seabed conditions along with photography or specimen collection – this is particularly useful in shallow, and tide swept areas where other forms of sampling may be more difficult to deploy.
• ROV recovery of seabed rocks, sediment and sea life specimens – again, this is particularly useful in shallow, and tide swept areas where other forms of sampling may be more difficult to deploy.
• Grab sampling of seabed sediments and biota – undertaken across a wide range of sediment habitat types but becomes limited in coarser gravels and areas with many stones present. Both conditions stop the grab system closing, leading to sample wash out.
• Box core sampling of seabed sediments and biota – particularly for medium to finer sediments where a larger sample is beneficial.
• Sediment coring – used where relatively undisturbed samples need to be taken.

These tools are also not necessarily applicable ‘en masse’ to every surveying task. There are key differences over the data needed for projects of different scales, the MRE technology types deployed and the location and therefore ambient conditions encountered. The right tool needs to be applied in the right circumstance.

This study has not specifically defined the relationship between condition and technology since there are so many variables to consider. The principles and options underlying that choice have been set out in the previous sections and particularly in the accompanying Appendix 1 Evaluation Matrix for tools and technologies, which reviews the suitability of all tools and techniques. However, the following sections discuss how different surveying approaches may be applied to address different objectives, including to assess feature extent, distribution and condition and monitoring changes over time.

9.4 How can the proposed scheme be used to describe the extent, distribution and condition of benthic species and habitats?

The marine environment comprises a mosaic of species and communities that transition continually from one to another. The extent and distribution of habitats can be determined through physical surveying techniques such as bathymetry and geophysical acoustic surveys. Almost all of the physical features and many of the indicator species for particular habitats are visible in video and photographic images. However, the spatial scale of a baseline or monitoring approach is dependent on the type and diversity species and/or habitats in question and also the heterogeneity of the conditions.

For seabed communities, their extent, distribution and condition can be assessed at different levels using different techniques. The extent of a reef habitat might best be confirmed through ground-truthed geophysical data, the abundance and distribution of indicator species by video or still image records, and the presence or absence of a wider range of macrofaunal species may require samples to be taken and analysed. While direct biological samples (e.g., via grab sampling) can provide a high level of detail on the species composition and diversity which may indicate condition, the small surface area of such samples means that the habitat between samples is unknown and extrapolation or predictive mapping is required to assess other metrics such as extent.

9.5 How can the proposed scheme be used to monitor changes in seabed diversity and community composition over time at different spatial scales?

The capacity for the scheme to meet this objective depends critically upon the scale of spatial and temporal change that is anticipated. At key visual indicator species level and 10s to 100s m resolution the task can be relatively simple; at a full species list level and individual metre by metre level or at an even more localised range it will be much harder, time consuming and expensive.

To address a monitoring objective to detect change, it is necessary to understand what levels and types of changes are typical or happening elsewhere outside the influence of any project (e.g., natural variation), as well as ascertaining what factors may be contributing to any change. This is where control site monitoring may be very helpful or indeed critical (see section 3.6.3 and 3.6.4 for a comparison of different sampling designs including BACI – Before After Control Impact and BAG – Before After Gradient designs).

The length of time and intensity of seabed sampling needed to build an understanding of seabed community changes could also be impractical in the context of the pre-operational consenting process but may be more feasible over the lifetime of a project.

9.6 How can the proposed scheme be used to measure potential habitat recovery and enhancement at different stages of MREDs?

The applicability of a particular tool or technique to assess habitat recovery or enhancement has been included in the Evaluation Matrix in the Evaluation of Performance tab in Appendix 1, where it is considered under the secondary use category.

All the shortlisted and recommended sampling techniques suggested in this study and listed in Section 8.1.3 are able to support a seabed habitat enhancement programme linked to an MRED if desired. The specific tool(s) and approach(es) used will be dictated by the level of community definition required to detect the anticipated level of change. It is expected that a higher level of survey intensity will be required to draw conclusions on habitat enhancement or positive effects. Hence the inclusion of nature positive effects in the “comprehensive” survey option in section 6.2. For example, underwater video can be used to confirm general community habitat type, 2-dimensional imagery can help identify surface living species and grab sampling can give before and after details on actual seabed community type.

In addition, novel techniques can play their role alongside older techniques. 3D Photogrammetry can be used to detect positive changes in biomass, eDNA can be used as a relative measure of diversity or species richness over time, while BRUVs can illustrate relative changes in abundance of mobile species or macrofauna. Therefore, a combination of metrics could be used depending on the species present and expected level of change, whilst considering natural variation and any influences from other anthropogenic sources or activities such as climate change and demersal fishing.

It would be of value to design surveys that consider the collection of data for reporting on the recovery and enhancement of species and habitats within MRED areas. If considered from the outset, the appropriate tools and technologies that score highly given the particular habitats and species of interest can be used, and the techniques must be quantifiable so that the output data can be compared over time. For example, macrofaunal data derived from repeated grab samples can be compared as they are obtained from a known volume, likewise for vertical still images with lasers where the seabed area surveyed can be quantified.

Survey planners could also consider the colonisation of new species onto hard substrates that were not present at the characterisation stage (artificial subsea infrastructure), as wells as those that are offered shelter and protection from physical pressures (e.g., fishing) and environmental influences (e.g., changes in sediment mobility caused by artificial obstructions).

One critical factor with regards to detecting and interpreting change is the need to have suitable control sites away from the influence of MREDs activities and restrictions on other sea users. The need for and resourcing of control sampling needs to be carefully considered in any survey plan targeting a 'change’ processes.

9.7 How can the proposed scheme be used to quantify potential habitat enrichment from enhanced biomass growing on hard structures?

As reported in a review of offshore wind reef effects by Degraer et al. (2020), there have been various studies undertaken around offshore wind turbines that have shown an increase in fine sediments and associated colonisation by species within 50 m or so of the turbine. There are also indications from these studies that jacket type foundation structures may lead to a greater effect than monopile structures.

Biofouling communities may ‘coalesce’ or ‘condense’ production into a smaller more compact area, and where there are no growth limitations (e.g., nutrients) then enhanced productivity in a localised area may occur. The various mechanisms for localised enrichment of sediments are explored by Degraer et al. (2020) with a key focus upon biofouling community production. However, it may not be possible to disentangle this cause-and-effect relationship with other drivers such as physical changes to current regimes, reduced fishing effort and the behaviour of aggregating fish which may all play a part alongside the production of the colonising biofouling itself.

The findings of Dannheim et al. (2020) agree that the reasons for any such changes are still uncertain. They also point out that the consequences of any such changes could be interpreted as ecologically beneficial, by adding to ecosystem diversity.

Whilst the investigation of nutrient pathways and sediment deposition processes are likely to be rather complex and challenging, designing a benthic survey to assess the consequential impact upon seabed species is a viable option. The sampling of seabed sediments to detect any signals of organic enrichment has been well practised around municipal outfalls, offshore drilling cuttings piles and fin-fish aquaculture sites. In these circumstances grab sampling has been used to gather samples for both biological and chemical analysis in the laboratory and therefore could be applied to an MRED scenario.

However, in these other sectors the enrichment effects have been linked to added direct organic inputs from the sewage, drilling mud or fish feed, and the zones of consequential effect have often extended to more than 100 m in radius. In the offshore wind case, where there are no artificially added nutrient inputs, the greatest challenge is likely to be getting the grab sampler close enough to the wind turbine, given the nominal 50 m range of effect detected so far (Degraer et al., 2020).

To help signal appropriate tools for these types of survey, a “biomass on hard structures” secondary use category is included in the tool Evaluation Matrix. The tools and technologies that score highest for quantifying seabed habitat enrichment are the grab samplers (such as Van Veen and Day grabs), with the tools for detecting enhanced biomass growing on hard structures include ROVs, high resolution imagery and laser profiling that can be used to produce 3D models (this is also discussed further in Section 6).

9.8 Strategic Sampling

Regional-scale strategic sampling approaches have recently been adopted across England and Wales to collect benthic baseline data and produce updated spatial models for key species (receptors), assemblages, and a suite of ecological metrics (diversity, functional traits). These models will be used to identify regions that are most vulnerable to potential offshore wind impact in the context of other activities and potential natural or global changes.

No such sampling strategy exists in Scottish waters, yet data gaps exist (e.g., outside MRED sites and MPAs). A recommended next step would be to improve data coverage to put site-specific assessments into context and improve understanding of broadscale environmental shifts.