61
Marine Bird Impact Assessment Guidance Workshop Report Notes from the workshop held on 20 February 2020 May 2020

Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

  • Upload
    others

  • View
    3

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine Bird Impact Assessment Guidance Workshop Report

Notes from the workshop held on 20 February 2020

May 2020

Page 2: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

2 | P a g e

Contents Summary ................................................................................................................................. 5

Introduction ............................................................................................................................. 5

Data gathering – first steps of the EIA process ........................................................................... 7

1. Colony counts and population estimates (Julie Black – JNCC and Kate Thompson SNH) ....... 7

2. Bird biological parameters (Glen Tyler – SNH) ........................................................................ 7

3. Foraging ranges (Alex Robbins – SNH) .................................................................................... 8

4. Connectivity (Alex Robbins – SNH) .......................................................................................... 8

5. Baseline site characterisation (Gayle Holland – Marine Scotland Licensing Operations Team) .............................................................................................................................................. 9

Question & answer session ............................................................................................................ 9

Actions from data gathering session ........................................................................................... 10

Collision risk modelling ........................................................................................................... 11

1. Stochastic CRM and worked example (Alex Robbins – SNH) ................................................ 11

2. Research update (Tom Evans – MS Science) ......................................................................... 11

3. Issues and knowledge gaps (Tom Evans - MS Science & Alex Robbins - SNH) ...................... 11

Feedback from the break-out session on collision risk modelling .............................................. 12

Actions for CRM subgroup 1 ........................................................................................................ 14

Action points for CRM subgroup 2 ............................................................................................... 16

Action points for CRM subgroup 3 ............................................................................................... 19

Action points for CRM subgroup 4 ............................................................................................... 21

Action points for CRM subgroup 5 ............................................................................................... 24

Question & answer session .......................................................................................................... 24

Displacement assessment ....................................................................................................... 25

1. Displacement assessment during the breeding season (Tom Evans – MS Science) .............. 25

2. Displacement assessment during the breeding season (Francis Daunt – Centre for Ecology and Hydrology) .............................................................................................................................. 25

3. Displacement assessment during the non-breeding-season (Francis Daunt – Centre for Ecology and Hydrology) ................................................................................................................ 26

Feedback from the break out session on displacement .............................................................. 26

Action points for displacement subgroup 1 ................................................................................. 28

Action points for displacement subgroup 2 ................................................................................. 31

Action points for displacement subgroup 3 ................................................................................. 32

Action points for displacement subgroup 4 ................................................................................. 34

Apportioning predicted impacts.............................................................................................. 35

1. Apportioning predicted impacts during the breeding season (Tom Evans – MS Science) .... 35

2. Apportioning predicted impacts during the non-breeding season (Glen Tyler – SNH) ......... 35

Page 3: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

3 | P a g e

Question & answer session .......................................................................................................... 35

Action points from apportioning question and answer session ................................................. 36

Panel session ......................................................................................................................... 37

Action points from panel session ................................................................................................. 38

Conclusions ............................................................................................................................ 39

Marine bird impact assessment guidance workshop - Action plan ............................................ 39

Appendix 1 – questions from breakout sessions ...................................................................... 48

Appendix 2 – feedback received post workshop ...................................................................... 52

Appendix 3 – list of delegates ................................................................................................. 60

Page 4: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

4 | P a g e

ACKNOWLEDGEMENTS We would like to thank all those who attended the workshop and who had input into this workshop report. We would especially like to thank Marine Scotland, and in particular Tom Evans, for help with the workshop preparation and on the day. Big up to the presenters (Alex Robbins, Glen Tyler, Kate Thompson, Julie Black, Francis Daunt, Tom Evans & Gayle Holland) who provided a varied and interesting programme for all. Thanks also to the breakout session facilitators (Alex, Tom, Glen, Gayle, Karen Taylor), scribes (Erica Knott, Emma Philip, George Lees, Jen Graham, Zoe Crutchfield, Kate and Karen), panellists (Alex, Glen, Tom, Ross McGregor, Francis Daunt, Aonghais Cook, and Carl Donovan), to Erica and Chris Eastham for helping to coordinate the workshop and write this report, and to Bob Furness, SNH Board, for chairing the workshop. Finally, we’d also like to thank the SNH Service Delivery Team at Battleby for helping make the workshop run smoothly. Everyone who attended the workshop made valuable contributions towards the discussions throughout the day, and the workshop has undoubtedly helped us focus on the work ahead.

Page 5: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

5 | P a g e

Summary

The purpose of the workshop was to discuss current bird impact assessment guidance and tools for offshore wind farms, with a view to identifying and agreeing what issues and gaps may need to be addressed prior to the next offshore wind leasing round in Scotland - ScotWind.

Feedback from the workshop will inform our priorities for finalising guidance planned for completion prior to scoping for ScotWind. It is likely other actions will be identified that will require other partners to become involved in guidance or revisions to existing tools or the development of new tools. The workshop covered the key topics relating to bird impact assessment for offshore wind and we thank the many participants that helped make the workshop a very useful and productive day. As a result of the feedback received both on the day and after the workshop, we have developed an action plan, which can be found in the Conclusions.

Introduction

(Chair – Professor Bob Furness, SNH Board) Last year a climate emergency and a biodiversity crisis were declared. In 2020, we will see the COP26 in Glasgow1 convene and scrutiny from all over the world on what Scotland is doing to address both the climate emergency and the biodiversity crisis. Scotland’s nature is special. All the work we do as an organisation aims to protect and enhance our natural environments, and to connect people to nature. However, the impacts on Scotland’s nature since the start of the industrial revolution have been severe and the rate of global change in nature during the past 50 years is unprecedented in human history. The workshop held on the 20 February 2020 focused on the impact on marine bird species, but we need to remember the context that renewables energy has a key role in reducing the effects of climate change. How then do we travel to a net zero carbon economy as well as addressing the biodiversity crisis? One area we can assist is by being clear in our guidance on our expectations to assist in the project impact assessment (EIA and HRA) process for new offshore wind applications. ScotWind - the new round for offshore wind farms - will be finalised later in 2020, and is likely to result in new project applications. The aim of the workshop (and this report) is to have reviewed, updated and provide new guidance, that is fit for purpose, recognises the importance of the EIA / HRA process and is enabling and proportionate in our requests for information with regard to bird impact assessment. We have reflected on the process that started back in 2009 / 2010 for the current suite of consented

offshore wind projects - we have more knowledge than we did then. We are also in the fortunate position that there have been a large number of research projects commissioned over the last decade, particularly by Marine Scotland Science, to help improve our understanding of the interactions of offshore wind farms and marine birds. We are preparing for the next round of project applications. We are keen to provide more certainty and detailed guidance to help us all in taking forward applications. The first stage will be advice on

1 COP26 is postponed until 2021 due to Covid-19 pandemic.

Page 6: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

6 | P a g e

baseline surveys and scoping, right through to the tools to assess impacts whether for collision, displacement or identifying which designated sites to consider. The workshop was about us largely advising on what guidance is being prepared, where we are seeking your feedback, and how this will be taken forward in finalising guidance. Some aspects of our guidance will be completed quickly; other aspects may take longer as we determine how to resolve some of the issues both ourselves and workshop participants may have encountered. This report provides information and key messages from the presentations, details from the question and answer sessions, and feedback from the break-out sessions on collision risk modelling and displacement assessment. Workshop attendees were provided with questions to promote discussion within the breakout sessions (see Appendix 1 – questions from breakout sessions) prior to the workshop. Facilitators used these questions to help steer each breakout session. This report summarises the notes taken by the scribes from the breakout sessions, which have largely followed the questions provided. Where questions were amalgamated, not answered, or the discussions followed a different (but important) thread, this report clearly states this. The notes have only been edited to either provide anonymity to the workshop attendees and/or provide context and clarity for those that were not in attendance. Importantly, this report also provides actions and recommendations for finalising guidance. Feedback received after the workshop (see Appendix 2 – feedback received post workshop) has also been used in the actions and recommendations mentioned in this report. We encourage all parties to continue discussions regarding how guidance and impact assessment tools may be improved, and provide feedback as we proceed with the next round of project applications.

Page 7: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

7 | P a g e

Data gathering – first steps of the EIA process

1. Colony counts and population estimates (Julie Black – JNCC and Kate Thompson SNH)

This presentation set the scene for current colony count status. Julie provided an update on the Seabirds Count census on behalf of Daisy Burnell (Seabirds Count Project Coordinator – JNCC), and an update on Seabird Monitoring Programme (SMP) database. Kate presented information on applying colony counts for impact assessments. Key messages:

Seabirds Count - the majority of seabird colonies have been counted, with plans to survey the remaining colonies in 20202.

Counts available as soon as input into SMP database, but may not have been checked validated.

Official launch of Seabirds Count is planned for December 2021.

There are plans to replace the current SMP database with an improved database providing greater functionality, and a Spatial Portal which will provide an online mapping and visualisation tool3.

Seabirds Count should be used when apportioning impacts from colonies within foraging range of the proposed development.

Developers should identify key colonies / species not counted in Seabirds Count, and discuss potential options, including carrying out new surveys, with Marine Scotland Science (MSS) and SNH.

2. Bird biological parameters (Glen Tyler – SNH)

Glen presented a summary of the approaches and resources used in modelling. The key references for bird biological parameters were highlighted, including any changes or updates from previous offshore wind rounds. Key messages:

Biological parameters need to be agreed with SNH and MS.

Bird sizes (range of wing-length and body-length) need some additional data, but current estimates are useable.

Flight speed and flight height data should be collected from ‘live’ situations where possible.

Collation of data from tagged birds should be used to determine nocturnal flight activity.

There are no absolute rules over models for population viability analysis (PVA) other than they must be capable of delivering required metrics. The Natural England tool provides a standard option. If semi integrated population models (SIPM) can be supported by colony data then that is preferred.

2 Since the workshop the outbreak of Covid-19 has meant that the Seabirds Count surveys planned for 2020 are cancelled, but the period of the census is extended to 2021 to complete the survey work. 3 SMP recently launched online at https://app.bto.org/seabirds/public/index.jsp

Page 8: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

8 | P a g e

3. Foraging ranges (Alex Robbins – SNH)

This presentation by Alex provided an update on key changes in evidence on seabird foraging ranges since the previous offshore wind leasing round. Key messages:

Woodward et al. (2019)4 improves the robustness of foraging range estimates, and includes a wider range of species.

The estimates of foraging ranges of some species have increased (e.g. kittiwake and razorbill)

SNH recommends the use of Woodward et al. (2019) for impact assessments.

4. Connectivity (Alex Robbins – SNH)

Following on from the presentation on foraging ranges, this presentation provided worked examples using foraging ranges to establish connectivity between an offshore wind site and SPAs. Foraging ranges, together with information on impact pathways, may be used to determine the second Habitat Regulations Appraisal (HRA) test of likely significant effect. Key messages:

Previously recommended using mean maximum foraging range +1 SD. However, further work is required to determine which metric from Woodward et al. (2019) should be used to establish connectivity between an offshore wind site and a colony SPA.

Recommend using site specific colony foraging ranges from Woodward et al. (2019) for apportioning.

For marine proposed SPAs, the conservation objectives should be used to establish: o which seabird colonies are within foraging range of the marine pSPA; o will disturbance / displacement result in a redistribution of birds within the

marine pSPA; o will the offshore wind site provide a barrier in terms of access to birds flying

to the marine pSPA; o are there any direct impacts on prey or supporting habitat within the marine

pSPA; and o are there any indirect impacts that may affect the food resource?

Additional methods of establishing connectivity may be used, such as tagging studies and flight directions from colony surveys and site-based surveys.

4 Woodward, I., Thaxter, C.B., Owen, E., and Cook, A.S.C.P. 2019. Desk-based revision of seabird foraging ranges used for HRA screening. BTO research report number 724.

Page 9: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

9 | P a g e

5. Baseline site characterisation (Gayle Holland – Marine Scotland Licensing Operations Team)

A summary of baseline site characterisation survey requirements for proposed offshore wind farm developments was provided by Gayle in this presentation. Key messages:

Early engagement with Marine Scotland and SNH to discuss requirements and design of baseline site characterisation surveys.

Two years of baseline surveys are considered a minimum, with further surveys required if, for example, there is high variation between years.

Digital aerial surveys are preferred over other survey platforms.

Encourage collaboration between developers when lease areas are within the same Draft Plan Options (DPOs)5, or in close proximity. This will help to decrease costs, improve the quality of information, and provide regional context.

Additional survey techniques, such as tagging, may be required, and colony counts may be required if there are gaps in the Seabirds Count census.

Interim survey reports, covering the breeding and non-breeding season and initial data analysis, are very useful and will be requested.

Encourage early discussion with Marine Scotland and SNH on data analysis and outputs, and impact assessment methodologies.

Question & answer session

Question 1. Why use mean maximum foraging range +1 standard deviation for determining connectivity? Answer: This provides a long list of colonies and species with potential connectivity, many of which can be screened out with information from site surveys. We are still considering Woodward et al. (2019) and how we use it to provide advice. We will ensure that advice provided makes biological sense and considers the size and number of close proximity colonies as well as the species and its rarity. (Additional comments received after the workshop may be found in Appendix 2 – feedback received post workshop) Question 2. Will guidance include information on seasonality and how it varies depending on location and / or individuals? Answer: SNH has guidance on seasonality for Scottish birds. We acknowledge there is still likely to be variation and will consider this further in our guidance at scoping. Question 3. What parameters have the biggest influence on the assessment? Is a sensitivity analysis required?

5 Draft Plan Options (DPOs) will become ‘Plan Options’ once the draft Sectoral Marine Plan for Offshore Wind is adopted.

Page 10: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

10 | P a g e

Answer: Bird density, flight height and avoidance / displacement rates have the biggest influence on the assessment. We do err on the precautionary side because we need to consider whether a proposal is likely or not to have no adverse effect on site integrity. Question 4. Does the response from SNH / NatureScot to the HRA for the Sectoral Plan include information from Woodward et al. (2019)? Answer: Our response does not include specific information from Woodward et al. (2019) we consider the approach taken by Marine Scotland has captured the likely key species and sites. Question 5. Does the guidance include information on connectivity during the non-breeding season? Answer: This will be discussed in the break-out session. Question 6. Foraging range data for breeding gannets shows that the foraging areas of neighbouring colonies do not tend to overlap, so it is reasonable to apportion impacts to one colony. What about other seabird species? Answer: Woodward et al. (2019) includes tracking data and shows that species behave differently. We need to consider this and base our advice on the evidence that is available to us. Question 7. What bird biological parameters do we need further information on? Answer: It is useful to gather site specific information for some biological parameters, but some will be generic (see key messages from the bird biological parameters presentation above). Question 8. Different areas within the DPOs will have different levels of bird activity (e.g. using foraging ranges). Is this taken into consideration when awarding leases? Answer: Applicants need to demonstrate they have taken account of other environmental information – these are part of the requirements of the ScotWind leasing process. Foraging range data can be used to predict areas where there is less risk (i.e. areas further away from breeding seabird colonies).

Actions from data gathering session

Actions Who (suggestions)

1 Develop guidance on the use of Woodward et al. (2019) SNH

2 Review SNH guidance on seasonality / seasonal definitions.

SNH

Page 11: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

11 | P a g e

Collision risk modelling

Presentations

1. Stochastic CRM and worked example (Alex Robbins – SNH)

This presentation provided an introduction to the stochastic CRM and the user-friendly shiny app using a worked example.

Key messages:

Preference is to use the stochastic CRM (sCRM) rather than offshore Band Model (2012)6.

The shiny app provides a clear user friendly front to the stochastic CRM R-code, with helpful input parameter graphics.

Some input parameters have changed from the deterministic excel spreadsheet.

There are still questions over how to approach stochasticity for certain parameters e.g. calculation of SD for monthly mean bird densities.

2. Research update (Tom Evans – MS Science)

This presentation provided a summary of past and present research projects that MS Science is managing in relation to collision risk assessment.

Key messages:

sCRM developed by Masden (2015)7 to incorporate variation and uncertainty in input parameters.

Refinements made sCRM including user-friendly interface.

GitHub used to address issues raised here and here on sCRM tool.

Project currently at procurement looking at how existing GPS tracking data from seabirds can be used to improve estimation of key parameters in CRM.

Project underway to develop conceptual framework for the simultaneous assessment of seabird collision risk, displacement and barrier effects.

3. Issues and knowledge gaps (Tom Evans - MS Science & Alex Robbins - SNH)

Key messages:

Potentially new species for CRM from the ScotWind draft plan options.

Update required to the migratory CRM.

Provision of seasonal outputs including breeding, non-breeding and passage season.

6 Band, B. (2012). Using a Collision Risk Model to Assess Bird Collision Risks for Offshore Wind Farms. Report by British Trust for Ornithology (BTO). Report for The Crown Estate. 7 Masden, E. (2015) Developing an avian collision risk model to incorporate variability and uncertainty. Scottish Marine and Freshwater Science Vol 6 No 14. Edinburgh: Scottish Government, 43pp. DOI: 10.7489/1659-1

Page 12: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

12 | P a g e

Improved estimates for input parameters (see Bird biological parameters (Glen Tyler – SNH).

Commercial roll out will require multiple inputs for different scenarios – this is currently labour intensive and may require batch modelling.

The key bugs in the sCRM have been identified and rectified. Any new bugs should be reported here.

Feedback from the break-out session on collision risk modelling

The break-out session consisted of 5 sub-groups discussing different topics related to collision risk modelling. Each subgroup was asked to discuss a number of questions sent in advance of the workshop, and provide feedback. Subgroup 1 – Turbine parameters The group decided to answer the first two questions together: Q1) Pitch - how much does this vary? Variation is also seasonal, should monthly pitch mean and SD be used in sCRM? Q2) Do we want to include seasonal options or a monthly variation for rotor speed?

It was agreed that pitch varies with rotor speed as the blade is moved to keep the rotor at operational speed; it’s not a continual variation rather the pitch will be appropriate for a specific rotor speed. Rotor speed is influenced by wind speed, hence the seasonal variation. Much of which is determined by the turbine cut in, cut out and operational speeds. While it is not the most sensitive parameter (compared with avoidance rate, bird density, flight height and flight speed) pitch can still have an impact on collision risk outputs. It was highlighted that generic wind speed relationship information on this is from published sources and may not be correct for specific projects. Blade shape information is old and is based on onshore turbine information. There was agreement that better data is needed, however, it was acknowledged that non-disclosure agreements and commercial in confidence issues may hamper this. One possible solution discussed by the group was to include confidential annexes within EIA. It may also be desirable to update this collision predictions when turbine specific information becomes available post-consent. To allow for comparison and cumulative assessment a fixed (single value) pitch parameter may be more useful rather than adding levels of variations. Q3) Turbine densities – how important are these and does it vary with turbine size? Within project assessment collision risk is calculated for a single turbine. Discussion therefore arose around how best to calculate this for the DPOs, which are not only large but could include multiple windfarms. Potential solutions discussed included the use of heat maps to consider areas of exposure to different numbers of turbines. However, the group acknowledged that the collision risk model would need further development to accommodate this type of approach. Theoretical available headroom was also discussed and whether the DSLP layout (i.e. as built) could be used for cumulative collision risk calculations – MS identified that they are thinking about this and

Page 13: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

13 | P a g e

have commissioned a research project titled ‘Production of cumulative effects framework for key ecological receptors’. Q4) Tidal offset. How should this be incorporated? Mean air gap or some other measure? (except for floating wind when airgap is constant so sd = 0). For fixed bottom turbines, hub height is based on highest astronomical height (HAT) which is precautionary, noting that this is different from floating wind. The group felt that the tidal range within the North Sea was such that this is unlikely to influence collision risk significantly. Areas where this may have more influence might include Isle of Man and possibly the Solway Firth but in general it was not considered a big issue for the DPOs. Q5) Blade tip speed – is this more useful than rotation speed? Based on fact that a larger diameter turbine would have a faster tip speed for the same rotation speed. Is this reflected in modelling collision risk or just swept area? Any data on this?

The collision risk model takes account of this, the group noted that speed is highest at the tip of the turbine, but collision risk is highest in the middle. Interestingly, it was pointed out that blade shape parameter within the CRMs is based on a 4MW onshore blade. The group agreed that there was a need to look to update this to make it more applicable to large offshore turbines. A suggestion was also made to approach ORE Catapult to see if further modelling could better understand how blade-shape influences collision outputs. The group agreed that access to RPM information would be difficult but that blade tip speed information may be more readily accessible. However, blade tip speed isn’t relevant as the model needs to integrate speed across the whole rotor face. It does this from the rotor radius and rpm.

Q6) Related to above, where do strikes occur? As most flights at rotor height are in lowest band of heights then would expect most strikes to be near tip of turbines. Is this the case?

While it was agreed that strikes will vary across the rotor swept area, little evidence is available from post consent monitoring to narrow this down and it’s hoped that the Forth and Tay developer bird monitoring proposals8, Kincardine DT Bird and EOWDC camera and radar monitoring will help with this. We also need to understand whether strikes vary with different weather and to what influence rising hub height will have.

Overall conclusions from subgroup 1:

1. There will be variation in collision risk densities across the development site – the collision risk probability won’t be the same for every turbine therefore we need to understand whether removing certain turbines decreases the CRM output? Could this be taken forward through cold spot analysis?

2. The relationship between wind speed and rotor speed, and rotor speed and pitch needs to be teased out further.

3. The relationship between blade chord and length as blade shape input parameters are based on old onshore information needs to be reviewed and this will need to account for the large variation in blade-shape expected with new larger turbines.

8 This project is under development and the specific project details are yet to be confirmed.

Page 14: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

14 | P a g e

Actions for CRM subgroup 1

Action Who (suggestions)

1 It was highlighted that generic information on turbine pitch and rotor speed is old and is generally based on onshore turbine information. There was agreement that better data is needed, however, it was acknowledged that non-disclosure agreements and commercial in confidence issues may hamper this. Marine Scotland to consider if confidential annexes within EIA would overcome these issues.

Marine Scotland

2 Update the blade shape parameter within the CRM model to make it more applicable to large offshore turbines.

Turbine manufacturer & Marine Renewables Ornithology Group (MROG)

3 Approach ORE Catapult to see if further modelling could better understand how blade-shape influences collision outputs. The group agreed that access to RPM information would be difficult but that blade tip speed information may be more readily accessible.

ORJIP / ScotMER

Subgroup 2 – Bird parameters The subgroup considered the following parameters: 1) Means of assessing bird flight speed

Some of the group were not convinced by radar as the resolution is not fine enough and cannot distinguish between flocks or individuals. It’s also likely that state of the art analyses could improve on what is currently applied, e.g. https://doi.org/10.1111/2041-210X.12699. It was suggested that a combination of radar and camera could be a good option - primarily for species ID rather than distinguishing between individuals/flocks.

A laser rangefinder could be useful but need to be careful about the representativeness of this sort of data, and there are still issues around where samples could and would be taken from.

The group suggested that it is likely that tags will improve and this could be a better means of assessing flight speed.

2) True vs straight line [speed]

The group highlighted that in terms of true vs straight line metrics, the most important metric is the speed at which a bird approaches a turbine – this is high value information.

One contributor suggested that the more detailed information (True) is more valuable than straight line data as true metrics can be converted to straight line but not vice versa.

A contributor raised a question about whether predictions are realistic, and that passage rates at the resolution of a wind farm would be very valuable.

Page 15: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

15 | P a g e

3) Wind speed

Currently ground speed is used. There is an R package, which can help to understand the [influence of] headwind, etc., but it is noted by BTO that wind direction is more important than wind speed9.

4) How to measure nocturnal activity

Tags and thermal cameras were suggested.

The Aberdeen [EWODC] study has used thermal cameras to assess nocturnal activity, and micro-behaviour changes.

A BTO report was published on this last year, which showed increased resting behaviour [at night], however this research did not survey reactive behaviour.

A contributor asked if it is necessary to look at data at this granular level, when it can’t be ground-truthed at sea.

5) Age Classification

Recognition that all methods of age classification have limitations and that misclassification is common. It was highlighted that age classification is heavily biased towards flying birds due to digital aerial images being restricted in picking up the concealment of features of birds on water.

Sample size was highlighted as a significant issue to this problem.

It was suggested that boat-based surveys would improve age classifications due to disturbance rates being higher, therefore higher proportion of birds in flight. Combinations of sampling techniques could be used to improve sample sizes.

It would be useful to look at the age structure of birds on sea vs in flight in archive digital imagery versus boat based survey data.

6) Sabbatical bird data

It was highlighted that it is important to include ‘pre-breeding’ birds as well as sabbatical birds.

It was discussed that marked individuals seem to be the only practical way of getting this type of data and that trapping birds at sea could be an option.

It was highlighted that one of the OWSMRF kittiwake projects might provide data of this type.

7) New species (DPOs)

Skuas were identified as species likely to feature in new DPO areas.

Storm petrels were discussed, including the possibility of using Motus tracking or nano-gps tags.

Issues were raised about shearwater attraction to light, specific question asked - what proportion of the population is attracted to light?

A key issue raised included shearwater displacement and disturbance from boats.

9 See Monitoring and forecasting avian collision risk at an operational offshore wind farm. BTO. NERC funded project. Unpublished.

Page 16: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

16 | P a g e

8) Age of survey data – is 5 years appropriate?

It was suggested that 5 years is reasonable, although industry would like it to be longer and consultants would like it to be shorter.

It was highlighted that older data tends to be limited by much smaller survey areas.

Power analysis could be useful but is quite hard to work out from historic data.

The seabird sensitivity index had a cut off of 20 years.

The age of colony density data and offshore populations should be matched. 9) Bird length / wingspan

Standard deviations have been derived from sources such as BWP10. There are currently limited studies providing species specific measurements with standard deviations.

A contributor asked whether the accuracy of the standard deviations would make much difference or not, and how wrong it could realistically be. Could it be adjusted to maximum length?

General comments from subgroup

Foraging range – Mean max is not as important as the amount of time they might spend in the site. Also that mean max might result in a dilution of local effects in favour of distant effects.

Action points for CRM subgroup 2

Action Who (suggestions)

4 Nocturnal activity: 1. Investigate nocturnal activity including reactive

behaviour through tags and thermal cameras. 2. Review Aberdeen [EWODC] study that has used

thermal cameras to assess nocturnal activity, and micro-behaviour changes.

ScotMER

5 Age classification: it was suggested that boat-based surveys would improve age classifications due to disturbance rates being higher therefore higher proportion of birds in flight. Combinations of platforms could be used to improve sample sizes.

Develop study to explore the age structure of birds on sea vs in flight in archive digital imagery versus boat based survey data.

ScotMER / ORJIP / TCE Enabling Actions

6 Sabbatical birds: Consider whether OWSMRF kittiwake and/or additional feasibility study (including wider range of species) on trapping birds at sea could be developed to gather data on sabbatical birds as well as pre-breeding birds.

ScotMER

7 New species: MSS/SNH

10 Cramp S. and Simmons, K.E.L. 1977, 1980, 1983, 1985 - Handbook of the Birds of Europe, the Middle East, and North Africa: The Birds of the Western Palearctic Volumes I-IV. Oxford, Oxford University Press.

Page 17: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

17 | P a g e

More information likely to be needed on skua, petrel, and shearwater species, with a focus on:

Consider the use of Motus tracking or nano-gps tags of storm petrels.

Identify what proportion of the shearwater population is attracted to light.

A key issue raised included shearwater displacement and disturbance from boats.

8 Age of survey data: Update SNH guidance to include:

5 years is reasonable.

It was highlighted that older data tends to be limited by much smaller survey areas.

Power analysis could be useful but is quite hard to work out from historic data.

The seabird sensitivity index had a cut off of 20 years.

The age of colony density data and offshore populations should be matched.

SNH guidance

Subgroup 3 – CRM outputs and presentation Q1) Which sCRM outputs do we require presentation for? (e.g. annual, monthly – which options, all scenarios?)

The group noted that there are breeding season definition inconsistencies between projects and what is used to feed into sCRM. It was agreed that it would be helpful if there was consistency here.

A question was raised on how would/could/should we incorporate climate change shifts. There were no answers here but noted that it is something that we should be considering.

It was agreed that monthly outputs are still good/advisable (those joining the group later also agreed with this).

It was also agreed sCRM outputs need to be repeatable – this is key especially for advisers and regulators to improve transparency of calculations.

Which options?: The group agreed that moving towards an App based approach is helpful and will improve transparency allowing advisers and regulators to see all the information that has been included to run the model.

Which scenarios? It was agreed that the ‘worst-case’ scenario should always be included. Additionally advisers and regulators would look for ‘most likely’ scenario as well as ‘best-case’ scenario.

It was noted that parameters change during the course of a proposal. These would need to built-in/taken account of as the proposal develops.

Q2) What are your views on presenting the outputs from sCRM to aid SNH & MS in determining an application?

The group agreed it’s important for SNH/MS to be able to reproduce what’s been done by the applicant.

Page 18: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

18 | P a g e

The group raised that there is commercial sensitivity around providing some of the outputs – but it was felt that this could be addressed through confidential annexes to EIA Reports.

The group also discussed parameters and the Shiny App. Overall it was agreed that the outputs looked good and comprehensive (nothing seemed to be missing). There was a question about how many simulations would be required. The general view that 1000 would be a minimum however, advice would be required on this. Q3) Do we require presentation of a model log? (e.g. pdf stating parameters used) The group unanimously agreed that a model log should be presented. It would provide a quick audit and better understanding of what is going into the model. The group then outlined that the log should record:

Which version of the model is being used? This prompted a question about whether there would be a requirement to rerun models as and when new versions are released?

There needs synchronisation of different versions on different servers (i.e. local/App/other tools such as the Cumulative Effects Assessment Tool)

R-package versions also need to be recorded – there is the potential for impacts on functionality if different versions of R are being used for the sCRM (i.e. when it is being used locally)

Q4) To help manage future cumulative issues, can we use outputs such as standard report templates from sCRM to help us? What should these standard templates include?

Group felt this was important because when you are doing cumulative assessment you may have to rerun the model if there have been changes to project designs.

Yes – agreed that standard templates would be useful.

Group also suggested having a data library for storing and easy access.

Question: How do we take account of changes in parameters as the development is going forward? MS advised that the Cumulative Effects Assessment Tool hopes to start to address this.

General comment that the more prescriptive MS/SNH can be the better. Q5) Do we want assessment to give impacts after periods other than 25 years? (Early consents were for 25yrs, more recent consents for 50 years)

Only comment here was that this is most relevant to PVA/density distribution changes. Because of large changes in bird numbers it does not make sense to model over 50 years.

Q6) Is it possible to separate out displacement/barrier effects and collision impacts or are we still needing to assume all (flying) birds of relevant species are subject to both?

All agreed that it would be useful to do this for as many species as possible.

There are currently inconsistencies in the way this done in Scotland and England (NIRAS & King Consultancy): - Scotland accesses collision and displacement for kittiwake but only collision for gannet. - England assesses collision and displacement for gannet but only collision for kittiwake.

Page 19: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

19 | P a g e

- CEH suggested an individual base model could be used. However – it can’t be parameterised for every species. There is an MS project nearly completion titled ‘Study to examine how seabird collision risk, displacement and barrier effects could be integrated for assessment of offshore wind developments’.

Question over whether it would reduce double counting.

Action points for CRM subgroup 3

Action Who (suggestions)

9 CRM outputs Ensure updated guidance continues to recommend monthly outputs. Ensure sCRM outputs are repeatable – this is key especially for advisers and regulators to improve transparency of calculations. Overall it was agreed that the outputs looked good and comprehensive (nothing seemed to be missing). Seek advice on how many simulations would be required. The general view that 1000 would be a minimum but statistical advice would be beneficial.

SNH guidance

10 CRM presentation Update sCRM to produce a model log which should be presented in any EIA. It would provide a quick audit and better understanding of what is going into the model. The log should record which version of the model is being used and R-package versions also need to be recorded

MSS & SNH

11 Explore the possibility of standard report templates for sCRM (especially useful for cumulative assessments). Develop a data library and storing for easy access. MS/SNH guidance to be as prescriptive as possible.

MSS & SNH

12 Consideration of consistency across the UK for assessment of potential impact pathways of collision and displacement.

SNCBs / Regulators

Subgroup 4 – Non-breeding season Q1) Should there be a seasonal calculation and output? E.g., inclusion of a slider with recommended dates from guidance.

Yes to both, however, this needs to be flexible e.g. to account for different seasonal definitions applied in different countries / regions.

Any in-built flexibility needs to be transparent, i.e. it needs to be obvious if and when any departure from a standard agreed approach (below) is undertaken.

A standard, consistent approach to determining non-breeding season collisions needs to be applied, with standard values / and parameters set by MS and SNH

Page 20: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

20 | P a g e

Some discussion was also held around whether defining the non-breeding season at the monthly level was appropriate or whether, instead, the start and end should be on specific dates (not necessarily the first or last day of the month). There was no agreement reached as there are benefits and dis-benefits either way.

Q2) In some of our previous advice we advised on the need for two non-breeding calculations e.g. for gannet and kittiwake to account for spring and autumn passage movements - what new information is there, what advice do we need to consider for the next round of offshore wind applications?

Proportions of non-breeding birds in any year

Proportions of birds from beyond Scotland

Quantification of immatures (flagged as especially important)

More tagging of birds at sea, i.e. in the DPOs, required to establish which, if any, SPAs they return to (and quantify the proportion of immatures staying at sea)

1. Recognised this is challenging and costly to deliver, especially to get a statistically significant number of records

2. Suggested that remotely operated gliders fitted with cameras could gather data more economically (but what about costs of processing?)

Need to pay specific attention this time round to migrating birds because some DPOs may affect migratory routes not previously affected by OWF development

1. Of specific concern here will be any OWFs built in the west coast as some species migrations may occupy relatively narrow corridors.

2. There is a need for a strategic study to identify and define migration routes of greatest significance to Sectoral Plan (viz birds flying at turbine height).

3. Some species migrate in large flocks just over a few days (as opposed to weeks or longer). These can be predicted and, in such circumstances, if passage through an OWF seems likely, a practical approach may be to shut down the OWF for the short period concerned.

Q3) Some aspects of assessment of non-breeding assessment was undertaken on a qualitative basis previously e.g. kittiwake and puffin, what should we do now in terms of these species? Do we still undertake a qualitative analysis or move to a quantitative analysis, if so how? Can we discount any species from a non-breeding season assessment?

Quantitative – virtually no debate on this.

No-one identified any birds that could be discounted; some thought though that a sensitivity index could be prepared indicating individual species sensitivity to collision on migration e.g. looking at flight heights and behaviour

Q4) Migratory CRM – do we need an updated model built into the sCRM app/stand-alone – or should this be undertaken as a strategic exercise?

Probably both; a strategic exercise would certainly be valuable for specific regions such as west coast of Scotland.

There was some debate about whether this would be too late now, as the DPOs have largely been defined; however strong view that it could inform later leases (this Plan is to last at least ten years) and individual projects; also some thought that, if done quickly, it could still influence leasing.

Birds of specific concern: 1. Barnacle geese 2. Corncrake

Page 21: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

21 | P a g e

3. Manx shearwaters 4. Geese and Swans 5. Whooper swan

Questions 5 and 6 were not covered in discussions.

Action points for CRM subgroup 4

Action Who (suggestions)

12 Develop a seasonal output within the model, which could be defined by a slider on the species parameter tab.

This needs to be flexible e.g. to account for different seasonal definitions applied in different countries / regions.

Any in-built flexibility needs to be transparent, i.e. it needs to obvious if and when any departure from a standard agreed approach (below) is undertaken.

A standard, consistent approach to determining non-breeding season collisions needs to be applied, with standard values / and parameters set by MS and SNH

MSS / SNH

13 Consider developing guidance on:

Proportions of non-breeding birds in any year

Proportions of birds from beyond Scotland

Quantification of immatures (flagged as especially important)

More tagging of birds at sea, i.e. in the DPOs, required to establish which, if any, SPAs they return to (and quantify the proportion of immatures staying at sea)

Recognised this is challenging to deliver, especially to get a statistically significant number of records

Suggested that remotely operated gliders fitted with cameras could gather data more economically (but what about costs of processing?)

SNH guidance

14 Need to pay specific attention this time round to migrating birds because some DPOs may affect routes not previously affected by OWF development:

Of specific concern here will be any OWFs built in the west coast as some species migrations may occupy relatively narrow corridors.

There is a need for a strategic study to identify and define migration routes of greatest significance to Sectoral Plan (viz birds flying at turbine height).

MSS & SNH

Page 22: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

22 | P a g e

Some species migrate in large flocks just over a few days (as opposed to weeks or longer). These can be predicted and, in such circumstances, if passage through an OWF seems likely, a practical approach may be to shut down the OWF for the short period concerned.

Develop a Migratory model that is within the sCRM app but also undertake a strategic project for use in this round.

Birds of specific concern:

Barnacle geese

Corncrake

Manx shearwaters

Geese and Swans

Whooper swan

Subgroup 5 – Overall model/code Q1) Ability to run batches – input to include multiple different windfarm scenarios, and select which scenarios and species to run.

The group members agreed that the ability to run batches of windfarm scenarios would be useful and considered the following:

a) Multiple windfarm scenarios – what are the scenarios - turbine parameters or collision rates? It was agreed that this would be relatively easy to achieve via a range of approaches. From use of template spreadsheets in folders (most user-friendly) to writing R code (greater functionality). It was highlighted that generally scenarios would be run in background and not presented.

b) Multiple species – the group noted this was already available within sCRM tool. Q2) Is there an alternative to the shiny app that we should be considering? Is it fit for mass roll out and use? The group noted that the ability to incorporate stochasticity was important. This can be achieved through:

using R code based on Masden (2015) - requires knowledge of R

using the Shiny App (not a model in itself)

trialling new approaches One approach would be R code that developers could use that generated the same form of outputs as the Shiny App This would provide more flexibility on the code, however, there are possible issues around how code would be audited and maintained. The group noted that transparency is important in the process to SNCBs and Regulators – this is advantage of applying a single model using the Shiny App.

Page 23: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

23 | P a g e

It was noted that there may be some appetite for a phased approach to modifying the underlying Band model (e.g. with respect to flux) It was also mentioned that there may be benefits in ability to add spatial aspects (i.e. moving beyond basic presumption of central place foraging): a post-doc in USA is exploring this by extending model to use GPS tracking data as an input rather than density/flux which may be helpful for modelling species on migration. Challenges in estimating spatially explicit densities. Q3) What issues do we anticipate with in combination/cumulative comparing sCRM and CRM? The group considered the key challenge is how to incorporate stochastic CRM estimates into cumulative impacts assessments (CIA). However, this may be explored at conceptual level by current MSS project for development of a cumulative effects framework. The following points were made in discussions:

This requires good understanding of population parameters.

The current approach is to sum apportioned effects for each colony and run PVAs – however, how do you sum variation given auto-correlation?

The marine mammal P-COD example shows that a full model would be complex.

There is a need for consistency in description of stochasticity and post-hoc information for existing / consented developments; the group questioned whether we should continue to use deterministic estimates for CIA?

Q4) Is there a model log/id so that specific versions of the model can be run again – as per NE PVA? Each run will generate different outputs unless identical initialisation parameters are specified (Set Seed) – this could easily be done and these parameters are recorded in the output log. However the group considered that this shouldn’t be necessary as with sufficient runs (minimum is 1000) the means should converge (unless there are very high sensitivities around any input parameters); they highlighted that the only issue could be around the distribution tails. Overall conclusions from subgroup 5:

It would be helpful to use folder/templates system to enable users to set (multiple) input parameters externally – would still get full audit as the inputs are specified in output logs which is important to transparency.

An alternative to the sCRM was not identified, it was suggested that the sCRM would be best developed through a phased approach with different components gradually replaced or improved on (e.g. with respect to flux calculation).

Incorporating stochasticity into CIA is very challenging and without this being carefully thought through it may be better to keep with deterministic outputs for CIA.

It would be very feasible to include a Set Seed capability but in practice this shouldn’t be necessary given that the default is minimum of 1000 runs.

Page 24: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

24 | P a g e

Action points for CRM subgroup 5

Action Who (suggestions)

16 Upgrade sCRM app to include multiple windfarm scenarios. The project scope will need to consider which is the best approach: use of template spreadsheets in folders (most user-friendly) to writing R code (greater functionality). It was highlighted that generally scenarios would be run in background and not presented.

MSS/SNH

Question & answer session

Question 1. Can the turbine array configuration and spacing be incorporated into the model? Answer: It’s reasonable to assess each turbine and use variation in bird densities across the wind farm site. This may need to be considered for the life-time of the project. It would also be worth considering the variation in collision risk due to the size and spacing of turbines (i.e. larger and more widely spaced turbines may reduce impacts). Array configuration needs further research to look at what configuration is optimal for different species in terms of reducing impacts. Existing tracking data may be able to help with this. There may be multiple turbine configurations for different scenarios. Maritime navigational safety also requires consideration.

Actions from the CRM question and answer session

Actions Who (suggestions)

1 Research on how array configuration influences impacts to bird species.

ScotMER / ORJIP / ORE Catapult

Page 25: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

25 | P a g e

Displacement assessment

Presentations 1. Displacement assessment during the breeding season (Tom Evans – MS Science)

This presentation provided a summary of the Joint Statutory Nature Conservation Bodies (SNCB) Interim Displacement Advice Note (2017), based on the ‘matrix approach’, for displacement assessment. It also introduced the SeabORD displacement assessment tool, and the Scottish Waters East Region Regional Sectoral Marine Plan Strategic Ornithology Study (from the Strategic Environmental Assessment North Seas Energy initiative (SEANSE)). It also provided information on improving estimates of seabird body-mass survival rates. Key messages:

Displacement matrix uses an ‘assumed’ mortality rate.

Individual based models cover both displacement and barrier impacts and base mortality and productivity loss on condition rules. Condition is calculated from daily energy budget of the birds.

SeabORD software indicates mortality rates are likely to be higher than those used in the displacement matrix for Round 3 sites.

2. Displacement assessment during the breeding season (Francis Daunt – Centre for Ecology and Hydrology)

This presentation provided information on how to use SeabORD for displacement assessment during the breeding season. It provided a worked example and information on the limitations of the model. Key messages:

Displacement produces a sub-lethal effect.

Displacement matrix does not have biological underpinning (uses expert judgement to set mortality rate).

SeabORD can be used to produce a realistic (biologically defensible) mortality rate

Displacement rates are still uncertain – some species we know are strongly displaced and others are not, but the actual rates are not clear.

SeabORD only uses data from the chick rearing period, but could be extended to cope with other periods and eventually year round impacts.

MS project scoping the integration of collision and displacement effects is nearing completion. This will make recommendations on updates to enable integration.

Page 26: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

26 | P a g e

3. Displacement assessment during the non-breeding-season (Francis Daunt – Centre for Ecology and Hydrology)

The current approach to displacement assessment (i.e. the ‘matrix approach’) during the non-breeding season has limitations. This presentation provided information on these limitations and offered potential solutions to future assessments.

Key messages:

Current (matrix) approach lacks biological realism.

Most mortality occurs in the non-breeding season – it is a critical time.

Need empirical studies to fill data gaps and improve understanding of non-breeding season pressures and responses.

Effects on productivity are not included in the displacement matrix approach.

Need year-round models of behaviour, energetics and demography to estimate: o Interactions between birds and developments o Effects on survival o Effects on subsequent breeding

Feedback from the break out session on displacement

The break-out session consisted of 4 sub-groups discussing different topics related to displacement. Each subgroup was asked to discuss a number of questions sent in advance of the workshop, and provide feedback. Subgroup 1 - ScotWind species Q1) ScotWind has identified 17 DPOs. Will these areas present new species for consideration in displacement assessment? Species identified included:

Seabirds – skuas, petrels, Manx shearwater

Divers – great northern divers

Migratory species – swans and geese This list was not exhaustive, so there may be more species identified during baseline surveys. However, one issue raised was that digital aerial surveys are unlikely to be able to help identify age classes. Further consideration of this, and inclusion of a steer in guidance in how to address this at the impact assessment stage, would be helpful. There was discussion of the published mapping of seabird distributions in Waggitt et al. (2019)11. This will require further consideration in future guidance.

11 Waggitt, J. J., Evans , P. G. H., Andrade, J., Banks, A. N., Boisseau, O., Bolton, M., ... Hiddink, J. G. (2019). Distribution maps of cetacean and seabird populations in the North-East Atlantic. Journal of Applied Ecology. https://doi.org/10.1111/1365-2664.13525

Page 27: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

27 | P a g e

It was recognised that the long list of species is likely to get longer, but this was due to the bar set by the likely significant effect (LSE) test, and some of these sites / species could fall out as the baseline site characterisation analysis is carried out. Q2) What evidence for disturbance / displacement do we have for impacts on these species? In the discussion it was highlighted that for Manx shearwaters there should be evidence on how this species interacts with wind farms from sites in the Irish Sea and at Robin Rigg. It was suggested that for some of these species there may be information emerging from the east coast wind farms in North America, which may be appropriate to use. A key reference provided was Dierschke et al. (2016)12. Additionally, there has been some work funded through the BEIS SEA Offshore Energy Research Fund - tracking of whooper swans. It was identified that geese species, such as pale bellied brent, Greenland white-fronted and barnacle geese, have also been tracked. This information may be very useful in any review of the strategic CRM migratory work undertaken by WWT for Marine Scotland.13 A useful pointer - where there is key reference material, it will still be extremely useful for SNH to identify what we are expecting / asking developers to do. Q3) Can we rank these species in terms of concern or priority? There may be different methods to help prioritise our thinking on this depending on whether we approach this by considering:

Weight given to the importance of the population status

Perceived risk to species of offshore windfarms (flight heights (noting this is more relevant to CRM assessments), location of wind farms, whether species are likely to be attracted to wind farms etc.)

Although the breakout session was meant to be considering displacement, a query was raised that for these new species do we have agreed Avoidance Rates. For instance for skuas - are we likely to use the terrestrial agreed rates? There may be concern for some species depending on the location of the DPO areas, particularly with the sites on the west / north coast, due to the complexity of the coastline and linkages to breeding colonies and the non-breeding dispersal. For some of the east coast DPO areas where there is a high-level of identified ornithological constraints strategic studies are required – the priority should be:

To identify what species are present

In what numbers and

What are they doing – transiting, foraging etc.? This information should then be used to help identify priorities.

12 Dierschke, V., Furness, R.W. and Garthe, S. 2016. Seabirds and offshore wind farms in European waters: Avoidance and attraction. Biological Conservation. Volume 202, pages 59-68. 13 Strategic assessment of collision risk of Scottish offshore wind farms to migrating birds – MSS 2014 https://www2.gov.scot/Resource/0046/00461026.pdf

Page 28: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

28 | P a g e

Q4) Can the Furness and Wade vulnerability assessment be used to assist in answering for questions 2 and 3? If not what would help? There is a need to fundamentally review the original Furness et al. (2013)14 vulnerability assessment scores. These were based on reactions to vessels and helicopters, and now that there are more offshore wind farms, what evidence has been / is being collected to help inform these vulnerability scores. Furthermore, could these scores be applied to some of the species that we have not assessed previously? There was a question as to whether the confidence levels be improved or updated. For any new species, consider the factors used originally and apply these to what is known about these additional species. It was agreed that the Furness and Wade work warranted further investigation as to the usefulness in updating. Q5) Are current apportioning approaches, (breeding and non-breeding) appropriate for any new species identified in the DPO sites? There was concern that due to the long list lengthening due to the greater foraging ranges, that the apportioning may become much more complex. It is therefore important to define what is required as the long list is reviewed during baseline site characterisation to enable a proportionate and realistic apportioning approach to be applied. Do we know for other species if there are foraging territory segregations similar to gannet? What is known about this during the non-breeding season? It might be more helpful if this was considered more strategically, with strategic studies carried out to help address some of these knowledge gaps, taking into account the emerging and increasing knowledge from species tracking and the recent Woodward et al. (2019) work.

Action points for displacement subgroup 1

Action Who (suggestions)

1 Digital Aerial Surveys are unable to differentiate age classes in certain species – how should this be addressed? Will affect CRM and displacement calculations.

ScotMER / ORJIP / TCE Enabling Actions

2 Need to review and reflect on Waggit et al. (2019) seabird distribution maps. This may help in identifying which species should be considered further.

SNH

14 Furness, R.W., Wade, H.M. and Masden, E.A. 2013. Assessing Vulnerability of Marine Bird Populations to Offshore Wind Farms. Journal of Environmental Management, 119, 56-66.

Page 29: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

29 | P a g e

3 Review PCM evidence from Irish Sea windfarms and Robin Rigg for Manx shearwater interactions around wind farms.

SNH

4 Contact North American colleagues to see what species may be recorded at their sites and what evidence they may have of displacement being an issue to consider further in the context of Scottish wind farms.

SNH

5 Update the MSS Migratory non seabird report (WWT) SNH / Sectoral Plan steering / advisory group

6 Update displacement guidance with key reference material and what SNH expects developers and consultants to do when carrying out a displacement assessment.

SNH

7 Avoidance rates – need to agree avoidance rates for any additional species for which ARs have not yet been required

SNH

8 Review Furness et al. (2013) vulnerability scores-

All species - is there any post consent monitoring to support changes to confidence levels / scores?

Additional species – what scores would we suggest?

SNH

9 Identify research to help enable whether certain species have defined, segregated foraging patterns as opposed to wide ranging foraging patterns.

ScotMER / ORJIP /TCE Enabling actions

10 Need to consider advice on how to deal with cross border and transboundary effects

Regulators / SNCBs

Subgroup 2 - BDMPS Q1) Would the development of a framework to account for birds seen at sea that are from outside the UK, but have connectivity with UK wind farms, particularly outside the breeding season be useful? This could build upon the useful source of data from current projects funded by SEATRACK, Vattenfall and Equinor of year-round tracking data. How could this be developed? Attendees supported the development and use of such a framework, but then the question was raised as to what the suggested framework would be, and how it could be constructed? At present there is a loose collaboration ongoing for tracking studies but this needs to be more formal /managed. It was suggested that it could be built on (or around?) SEATRACK. Funding options were not discussed, other than contribution to the framework (evidence base) would be through the licence application to undertake the tracking work. It seems most appropriate to seek funds for this work from offshore development, but in the future there may not be funds if active offshore development declines. The evidence to be able to update current guidance and build any future framework is likely to be closely involved with the SEATRACK project. There was a discussion about what SEATRACK is and its current status.

Page 30: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

30 | P a g e

SEATRACK

Using GLS recorders to follow birds year round

Expanding on the Isle of May

Plus adding other colonies

Now a European wide programme

Planning to use data to create at sea maps – but tracking is only for (successfully) breeding adults

A drawback is that while not difficult to tag chicks (of guillemot) recapture would be difficult as rarely come back to original site so no collection of tags

Some species easier to tag than others, so more data for them The approach to apportioning and BDMPS was suggested to be too focussed on UK birds. Although we do not assume all birds are from ‘UK’ SPA Populations but the impacts may be on ‘foreign’ birds – i.e. ones associated with SPA from outside UK are not followed through (or have not been until now). Our assessments only follow the impacts to UK birds. It was concluded that a framework should produce estimated numbers of the impact with a consistent approach to apportioning. To be able to construct a workable framework the need is for stakeholder agreement and engagement. The best approach to apportioning currently is using the existing SNH guidance but this could be improved. SNH guidance needs updating – current version is from 2018, but evidence is from 2017 and before. The group discussed what evidence would be required and what improvements to the current approach could be made. To update the SNH guidance we need an evidence base in which the focus should be on non-breeding foreign birds, understanding how they contribute to our counted populations. The framework should be done at UK scale, although it may need to be able to be split to devolved administrations. Apart from tracking studies, ringing was also suggested as good evidence. Conventional ringing provides useful information (and should not be overlooked) although there are more issues in interpreting ring recoveries. There was a call for a commitment to a regular update of the apportioning guidance. Advances in technology and techniques may make more sources of evidence available quite soon. This relates to advances in tag design, recording system, attachment method etc. The discussion on other questions was brief but we did touch on the approach to new evidence of movements. For specific species it was thought that some periods may need to be dealt with separately. For example, the postnatal dispersal period for guillemot and razorbill. Wintering and moulting areas need to be separated out. Partial populations may need to be dealt with. For example, kittiwake that leave colonies early may go to different places (more westerly) than those that leave later. Whilst considering additional tagging work that we were not currently aware of, the question was asked if SNH have a database of tagging projects. It was thought that this could be built from BTO permits to apply tags. It may be best if JNCC hosted this as they could cover all of UK, and also have links to European (international) government and research organisations. Perhaps EURING has a similar role to BTO?

Page 31: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

31 | P a g e

Starting with the SeaTrack group, could such a database be built? Also in future could contain MOTUS tracking projects. There is already a voluntary database of tracking information (e.g. http://www.seabirdtracking.org/) but this is only partial and not everyone registers. If it was required for licencing then it would be complete.

Action points for displacement subgroup 2

Actions Who (suggestions)

1 Develop the use of a framework, building on SEATRACK, to account for birds seen at sea but are from outside the UK.

ScotMER/ TCE Enabling actions

2 SNH guidance on apportioning needs updating. SNH

3 Explore the possibilities of establishing a tagging database, possibly hosted by JNCC.

ScotMER / TCE Enabling Actions / JNCC

Subgroup 3 – SEANSE Recommendations Q1/Q2) What scenarios should the SeabORD emulator consider i.e. what windfarm footprint, range of species and populations? Note that points highlighted in bold were taken as the top discussion points and summarised at the end of the session. The matrix approach does not have much evidence for mortality rates, therefore the emulator would give us a biological rational in data poor situations. CEH presentation suggested the following key parameters:

- Distance to colonies - Alignment to colonies - Density of footprint - Survival - Productivity

There seemed to be a general agreement on the need to look at a range of these based on realistic scenarios. Discussion points:

- What gives confidence in the output? As long as there are enough replications, there should be high confidence – however, this is only as good as the data and model which is used to run it.

- Sensitivity to some of the assumptions remains – new data from studies on the Isle of May has markedly affected the outputs, so it is still susceptible to refinement of assumptions.

Page 32: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

32 | P a g e

- It was suggested that distribution of prey data should be included in the model as they currently use tagging data for distribution. It was questioned whether estimates of sensitivities of prey could be provided. It was agreed that surveys should to include prey distributions but this is difficult especially with the move towards aerial survey methods.

- SeabORD doesn’t model based on individual turbines but it could do. It was agreed that

birds may go between rows of turbines depending on turbine densities. Turbine densities could be included, to look at impact on birds going between turbines.

- A question was raised about whether species are attracted to developments, e.g. roosting shags and how this was considered. It was explained that species that are thought to be positively attracted are not currently considered within the assessment. It was agreed that monitoring was important to better understand this.

- Within the emulator there is a need to consider non-breeding birds in the breeding season

i.e. sabbaticals or immature birds. Q3) Should Wakefield approach for GPS data be widened to other species or a different, novel approach taken? The outputs can be very different for either using at sea data or GPS data. This question related to how we understand the differences between these types.

- It highlights the importance of bird aging and understanding birds from different areas. This could be resolved by catching birds at sea, however this has not widely been undertaken in the UK and there is a need for a pilot project to establish the best methods on UK seabirds.

- It also highlights the importance of understanding where the birds come from and the

composition of birds on site, i.e. sabbatical birds etc. Action points for displacement subgroup 3

Action Who (suggestions)

1 Further development of SeabORD to include prey distribution data and turbine density/spacing.

ScotMER

2 Consider non-breeding birds (i.e. sabbaticals or immature birds) in the breeding season within the emulator.

ScotMER

3 Undertake feasibility study / pilot project for tagging seabirds at sea.

ScotMER / OWSMRF/ ORJIP / TCE Enabling Funds

Page 33: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

33 | P a g e

Group 4 - Knowledge gaps (based on SEANSE recommendations) Q1) How can uncertainty be better quantified by accounting for uncertainty in key model parameters (e.g. use of Bayesian calibration methods)? Bayesian approaches allow optimisation of input parameters by identifying key sources of uncertainty and can direct targeted data collection to reduce uncertainty; sensitivity analyses also helpful. How decision makers handle uncertainty (e.g. as inherent in the matrix approach) was discussed; concern among some developers that may lead to over-precaution. Conversely, a clear understanding of the uncertainties is important so that regulators are clear about level of certainty around outcomes of assessments, which means that decisions made are more likely to give adequate consideration to the precautionary principle. Quantifying uncertainty can also be helpful too, and can inform licencing conditions. A comment was made that there is no process by which previous decisions and associated uncertainties are revisited/reviewed. There are challenges for cumulative impact assessment. Q2) Incorporation of recent estimates of uncertainty in the mass-survival relations for species derived from the recent MS research project - Guillemot, Kittiwake, Puffin - some information on Razorbill. This could readily be applied – can pull them straight out of the reports. Up to SNCBs to decide! Q3) An adaptation to the model to allow foraging birds to increase their foraging range when encountering barrier or displacement effects from OSWFs at the edge of their foraging range. This would be an adaptation to the model. It currently assumes a fixed foraging range, but a more sophisticated model could allow birds to extend their foraging range if they are displaced. It was noted that species will vary considerably in extent to which they may vary foraging range/tracks (e.g. auks versus kittiwakes) Current studies (tracking and post-consent monitoring plus the Woodward et al 2019 foraging range review) should provide more information – but critical that we distinguish between displacement and barrier effects. It is an artefact of current model that foraging range is fixed – this could be modified. Q4) Within the SeabORD model, how could model simulation for flight paths and foraging trips be improved? Noted that flight patterns are different back and forth from the colony – generally being wavier on the way out and straight on the way back. Furthest point (auks) tends to be the last foraging location.

Page 34: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

34 | P a g e

See also previous question – agreed that GPS data could inform simulation of more realistic flight paths (e.g. for kittiwake) Q5) How could SeabORD (and potentially other simulation approaches) be expanded to include periods of the breeding season other than chick rearing (e.g. incubation) and to include the non-breeding season? CEH consider sufficient current information for year round model for guillemots, and possibly gannets. There may also be sufficient information from FAME data for kittiwakes during the incubation period. It was identified that it wouldn’t be a big step to go into the incubation period. Challenges in non-breeding season for species where there are major differences among individuals in wintering areas (e.g. kittiwakes) A question was raised about importance of barrier effects for migratory species – it was suggested that relative additional energy expenditure likely to be relatively low (cf foraging breeding seabirds). Overall conclusions from this group

It was generally agreed that it would be useful to better quantify overall model uncertainty by including variation and uncertainty around model parameters. This can help identify key sources of uncertainty where empirical data could be collected to reduce uncertainty. There was no clear agreement on how uncertainty should be used in decision making but several suggested that quantified uncertainty should be retained throughout the assessment process.

Further analysis of GPS data could improve how foraging trips are modelled within SeabORD, e.g. for kittiwake that make more tortuous foraging movements.

SeabORD could be expanded beyond the chick-rearing period, especially into incubation when birds are central place foraging, for the non-breeding season this is more challenging but should be feasible for some species (e.g. guillemots).

Action points for displacement subgroup 4

Action Who (suggestions)

1 The quantification of uncertainty in displacement models must be achieved.

ScotMER

2 Use of more GPS tracking data to characterise foraging trips made by birds to use in SeabORD models.

RSPB/BTO – tracking

3 Expansion of SeabORD models to cover periods other than chick-rearing including non-breeding season.

ScotMER

4 Uncertainty in mass-survival relationships to be incorporated into SeabORD.

ScotMER

Page 35: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

35 | P a g e

Apportioning predicted impacts

1. Apportioning predicted impacts during the breeding season (Tom Evans – MS Science)

This presentation provided a summary of the SNH apportioning tool, the Marine Scotland apportioning tool, and the Seabird Sensitivity Mapping Tool for Scotland

Key messages:

The Marine Scotland apportioning tool is due to be published imminently, and can be used for apportioning impacts for shag, common guillemot, razorbill and kittiwake using the Wakefield et al. (2017) distribution maps

The SNH apportioning tool will still need to be used for other species not listed above.

The Seabird Sensitivity Mapping Tool is a fast, user-friendly tool to establish the sensitivities of key seabird species to offshore renewable developments. It incorporates the Marine Scotland apportioning tool as a module. Phase 1 is completed, and scoping for phase 2 is due to be tendered.

2. Apportioning predicted impacts during the non-breeding season (Glen Tyler – SNH)

This presentation provided a summary of SNH advice on the calculation of predicted impacts to SPA birds during the non-breeding season. It also provided information on how the movements of birds from breeding areas are assessed. Examples were provided for gannet, kittiwake and large auks. Key messages:

Different species have slightly different approaches, but the overriding principle is that birds mix evenly and freely within the defined non-breeding range.

The movements of birds is the most influential factor in the approach.

We encourage early discussion with developers on the non-breeding apportioning approach.

Question & answer session

Question 1. Apportioning approaches use Euclidean distance. Have the non-Euclidean also been investigated? Answer: To be correct then it would be best to use non-euclidean distances (i.e. non straight-line such as great circle / shortest distance on sphere), but for the relatively short distances used in most calculations it will not change the apportioning appreciably. As larger foraging ranges are incorporated using non-euclidean distances will need to be considered, to see what impact that has.

Page 36: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

36 | P a g e

Question 2. Will the Marine Scotland apportioning tool include differences between adult and juveniles? Answer: At present it does not; further data are needed to understand the distributions and movements of sub-adults. Currently gannet are the only species where there may be sufficient tracking data available (e.g. Grecian et al. 2018 - https://doi.org/10.1098/rsif.2018.0084). Question 3. When undertaking a cumulative impact assessment for kittiwake, it will include all colonies in the North Sea. Is this practical? Answer: This is a process issue as there are different ways in consenting between the different countries. This is about having data from all non-UK colonies and wind farms. Is the group that JNCC attend of European SNCBs / NGOs the place to discuss such a database set-up? Discussions on-going between countries but this is not easy. The regulators also need to have a discussion on how they will address this issue and provide advice. The wider SEANSE study (which the Scottish Case Study was a part of) was considering this. The reports for that are available online: https://northseaportal.eu/downloads/. Question 4. For ScotWind what account will be taken of the TCE headroom report or any future updates to this? Particularly with regard to consented and ‘as built’ individual development specifications. Answer: The project titled ‘Production of cumulative effects framework for key ecological receptors’, commissioned by MS Science, will consider cumulative effects and the assessment process. The objective is to develop a tool that allows recalculation for all UK wind farms ‘as built’ to inform future cumulative impact assessment requirements. Action points from apportioning question and answer session

Action Who?

1 Investigate impact of using non-straight-line distances in apportioning models

SNH

2 Determine what requirements would be to include pre-breeding ages in MS apportioning tool

MS

3 Investigate North Sea wide database of colonies and cumulative effects

Regulators / JNCC

Page 37: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

37 | P a g e

Panel session

Question 1. What are the greatest remaining uncertainties for collision risk modelling (CRM) and displacement assessment? Answer: For CRM the greatest areas of uncertainty are avoidance rates, model error (which can be up to 20%), flight height and speed. For displacement assessment, we need to establish displacement rates and how they change with time. Also need to know mortality rates. Question 2. For displacement models, do we know how turnover rates may affect them? Answer: Further analysis is required. We need to consider and better understand site faithfulness at sea at the species and individual level. Question 3. Now that Natural England have the PVA tool, will SNH and Marine Scotland provide updated advice for PVA? Answer: Yes, we will be updating our advice. It would be useful if developers and consultants could provide comments on their use of the PVA tool to enable any specific aspects of guidance to be provided. Question 4. Can the avoidance rates and empirical data gathered from the Thanet study be transferred to other offshore wind farms or is it site specific? Answer: Thanet study is very useful and has provided proof of methods and data collection. There have been delays in using the values derived from the study. Our advice is to use both the empirical avoidance rates derived from the Thanet study and avoidance rates as used in the Band model. Question 5. Has data similar to that gathered for the Thanet study been collected at Scottish wind farms? Answer: In Scotland, mostly the regional advisory groups have been advising on the monitoring requirements. Monitoring is still at an early stage with only two operational wind farms currently having monitoring equipment, one of which is at EOWDC Aberdeen and so data collection is still at an early stage. Question 6. Discrepancies have been found between the outputs from SeabORD and matrix approach to displacement assessment. Will there be any investigation into these discrepancies? Answer: We are aware of the discrepancy and whilst there has not been so much investigation, it is considered that we can use the SeabORD models to populate the matrix with a biologically based estimate of mortality (and productivity loss). The differences between the two are due to very fundamental differences in them. They both give an estimate of mortality due to displacement. Question 7. Are there plans for guidance on the outputs from PVA? Answer: SNH have a current commissioned research contract awarded to HiDef considering this. Guidance will not provide advice on specific thresholds, but will provide advice on a decision support system, which should provide a transparent and consistent platform on which we will base our decisions on the advice we provide to MS. The decision support system does not make the decision, it will however make the process clear and transparent, which is an aim of the project.

Page 38: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

38 | P a g e

Question 8. How is uncertainty incorporated into the decision support system for PVA? Answer: This is currently being considered, but it is likely that each element of the system will have a weighting assigned to it. Question 9. Will this tool / guidance be used by SNH and MS-LOT? Answer: Yes, by SNH to inform our advice. Our advice is the delivered to Marine Scotland. As the regulators Marine Scotland will consider our advice against its own polices and decision making processes. Question 10. Where does the remit for SNH stop and for JNCC start? Answer: In 2017 a delegation and transfer of responsibilities occurred and SNH now provides advice on all renewables projects within Scottish waters i.e. including and beyond 12nm. SNH will still seek advice on management measures in respect of Natura sites beyond 12nm. Question 11. Is it still appropriate to include +1 SD when using foraging ranges? Answer: Depends on the rationale. Including the +1 SD is a precautionary approach. We need to consider Woodward et al. (2019) in more detail. Using +1 SD is useful when considering colonies with poor data. Foraging range data is still patchy for some species. The likely significant effect test is high level, therefore the approach of using +1 SD is justified. Question 12. How useful are the colony specific foraging ranges? Answer: Colony specific foraging ranges and data is important. Colony specific foraging ranges can vary between years, possibly depending on prey distributions, etc. The SNCBs need to take this into account. Action points from panel session

Action Who?

1 Commission research on turnover rates ScotMER or ORJIP

Page 39: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

39 | P a g e

Conclusions

(Professor Bob Furness – Chair)

A significant amount of applied research has been undertaken and our understanding of bird interactions around offshore wind farms has increased. A large part of this has been the baseline data and post consent monitoring undertaken by developers.

There is a need for a database of tracking studies both within the UK and wider in Europe.

It would be extremely useful to have a single database of offshore-related research and monitoring that is kept updated.

There are still some significant knowledge gaps, and there are opportunities to fill these; aspects of this might best be undertaken collaboratively and strategically.

Surprised that cumulative impacts did not feature more heavily in the discussion. The MS Science Cumulative Effects Framework will help to address the difficult issues around cumulative impact assessment.

Environmental conditions are constantly changing (e.g. the change in the fortunes of great skua in Shetland in direct relationship to discards policy). We need to be aware environmental conditions are changing constantly and how this is taken into account.

Policy and decision making needs to be based on evidence. Actions generated from discussions in each of the workshop sessions are provided in the action plan below. For each action we provide information on who is best placed to lead on addressing the action, priority, timescale, and any constraints / dependences. We will review and update this action plan as work progresses.

Marine bird impact assessment guidance workshop - Action plan

Version 1 - May 2020

Action Who (suggestions)

Priority (H, M & L)

Timescale Short (i.e. this year, medium next year long term – 2 years hence)

Constraints / dependences: (e.g. staff time, stakeholder agreement , funding, data gaps, project / research completion, other actions, COVID-19, commercial confidentiality, & monitoring results)

DATA GATHERING

1 Develop guidance on the use of Woodward et al. (2019).

SNH H Short-term

2 Review guidance on seasonality / seasonal definitions.

SNH H Medium-term Other actions

COLLISION RISK MODELLING

3 It was highlighted that generic information on turbine pitch and

Marine Scotland

M Medium-term Commercial confidentiality,

Page 40: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

40 | P a g e

rotor speed is old and is generally based on onshore turbine information. There was agreement that better data is needed, however, it was acknowledged that non-disclosure agreements and commercial in confidence issues may hamper this. Marine Scotland to consider if confidential annexes within EIA would overcome these issues.

Stakeholder agreement

4 Update the blade shape parameter within the CRM model to make it more applicable to large offshore turbines.

Turbine manufacturers & Marine Renewables Ornithology Group (MROG)

M Medium-term Commercial confidentiality, Stakeholder agreement

5 Approach ORE Catapult to see if further modelling could better understand how blade-shape influences collision outputs. The group agreed that access to RPM information would be difficult but that blade tip speed information may be more readily accessible.

ORJIP / ScotMER

M Medium-term Commercial confidentiality

6 Nocturnal activity:

Investigate nocturnal activity including reactive behaviour through tags and thermal cameras.

Review Aberdeen [EWODC] study that has used thermal cameras to assess nocturnal activity, and micro-behaviour changes.

ScotMER H Medium-term Covid 19, Monitoring results

7 Age classification: it was suggested that boat-based surveys would improve age classifications due to disturbance rates being higher therefore higher proportion of birds in flight. Combinations must be used to improve sample sizes.

Develop study to explore the age structure of birds on sea vs in flight in archive digital imagery

ScotMER, ORJIP, and TCE Enabling Actions

M Medium to long term

Funding

Page 41: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

41 | P a g e

versus boat based survey data.

8 Sabbatical birds: Consider whether OWSMRF kittiwake and/or additional feasibility study (including wider range of species) on trapping birds at sea could be developed to gather data on sabbatical birds as well as pre-breeding birds.

ScotMER / OWSMRF

M Medium term Funding, Data gaps

9 New species: More information likely to be needed on skua, storm petrels, and shearwaters, with a focus on:

Consider the use of Motus tracking or nano-gps tags of storm petrels.

Identify what proportion of the shearwater population is attracted to light.

A key issue raised included shearwater displacement and disturbance from boats.

MSS/SNH H Medium Funding, Data gaps, Covid-19

10 Age of survey data: Update SNH guidance to include:

5 years is reasonable.

It was highlighted that older data tends to be limited by much smaller survey areas.

Power analysis could be useful but is quite hard to work out from historic data.

The seabird sensitivity index had a cut off of 20 years.

The age of colony density data and offshore populations should be matched.

SNH guidance M Short term

11 CRM outputs SNH guidance H Short-term Funding

Page 42: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

42 | P a g e

Ensure updated guidance continues to recommend monthly outputs. Ensure sCRM outputs need to be repeatable – this is key especially for advisers and regulators to improve transparency of calculations. Overall it was agreed that the outputs looked good and comprehensive (nothing seemed to be missing). Seek advice on how many simulations would be required. The general view that 1000 would be a minimum but statistical advice would be beneficial.

12 CRM presentation Update sCRM to produce a model log which should be presented in any EIA. It would provide a quick audit and better understanding of what is going into the model. The log should record which version of the model is being used and R-package versions also need to be recorded

MSS & SNH H Short-term Funding

13 Explore the possibility of standard report templates for sCRM (especially useful for cumulative assessments. Develop a data library and storing for easy access. MS/SNH guidance to be as prescriptive as possible.

MSS & SNH H Medium term Dependent on actions 11 & 12 above

14 Consideration of consistency across the UK for assessment of potential impact pathways of collision and displacement.

SNCBs / Regulators

H Short-term Project / research completion

15 Develop a seasonal output within the model, which could be defined by a slider on the species parameter tab.

MSS/SNH H Short-term Funding

Page 43: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

43 | P a g e

This needs to be flexible e.g. to account for different seasonal definitions applied in different countries / regions.

Any in-built flexibility needs to be transparent, i.e. it needs to obvious if and when any departure from a standard agreed approach (below) is undertaken.

A standard, consistent approach to determining non-breeding season collisions needs to be applied, with standard values / and parameters set by MS and SNH

16 Consider developing guidance on:

Proportions of non-breeding birds in any year

Proportions of birds from beyond Scotland

Quantification of immatures (flagged as especially important)

More tagging of birds at sea, i.e. in the DPOs, required to establish which, if any, SPAs they return to (and quantify the proportion of immatures staying at sea)

Recognised this is challenging and costly to deliver, especially to get a statistically significant number of records

Suggested that remotely operated gliders fitted with cameras could gather data more

SNH guidance H Short to medium term

Dependent on completion of research Data gaps Funding Covid-19

Page 44: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

44 | P a g e

economically (but what about costs of processing?)

17 Need to pay specific attention this time round to migrating birds because some DPOs may affect routes not previously affected by OWF development:

Of specific concern here will be any OWFs built in the west coast as some species migrations may occupy relatively narrow corridors.

There is a need for a strategic study to identify and define migration routes of greatest significance to Sectoral Plan (viz birds flying at turbine height).

Some species migrate in large flocks just over a few days (as opposed to weeks or longer). These can be predicted and, in such circumstances, if passage through an OWF seems likely, a practical approach may be to shut down the OWF for the short period concerned.

Develop a Migratory model that is within the sCRM app but also undertake a strategic project for use in this round.

Birds of specific concern: o Barnacle geese o Corncrake o Manx shearwaters o Geese and Swans o Whooper swan

MSS & SNH H Short to medium term

Funding

18 Upgrade sCRM app to include multiple windfarm scenarios. The project scope will need to consider which is the best approach: use of template

MSS/SNH H Short term Funding

Page 45: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

45 | P a g e

spreadsheets in folders (most user-friendly) to writing R code (greater functionality). It was highlighted that generally scenarios would be run in background and not presented.

19 Research on how array configuration influences impacts to bird species.

ScotMER / ORJIP / ORE Catapult

M Medium term Research & funding

DISPLACEMENT

20 Digital Aerial Surveys are unable to differentiate age classes in certain species – how should this be addressed? Will affect CRM and displacement calculations.

ScotMer / ORJIP / TCE Enabling Actions

M Medium term Funding

21 Need to review and reflect on Waggit et al. (2019) seabird distribution maps. This may help in identifying which species should be considered further.

SNH M Short term

22 Review PCM evidence from Irish Sea windfarms and Robin Rigg for Manx shearwater interactions around wind farms.

SNH M Medium term

23 Contact North American colleagues to see what species may be recorded at their sites and what evidence they may have of displacement being an issue to consider further in the context of Scottish wind farms.

SNH M Medium

24 Update the MSS Migratory non seabird report (WWT). See action 17 above.

25 Update displacement guidance with key reference material and what SNH expects developers and consultants to do when carrying out a displacement assessment.

SNH H Short-term

26 Avoidance rates – need to agree avoidance rates for any additional species for which ARs have not yet been required

SNH M Medium term

27 Review Furness and Wade vulnerability scores-

All species - is there any post consent monitoring to support changes to

SNH M Medium term

Page 46: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

46 | P a g e

confidence levels / scores?

Additional species – what scores would we suggest?

28 Identify research to help enable whether certain species have defined, segregated foraging patterns as opposed to wide ranging foraging patterns.

ScotMER / ORJIP /TCE Enabling actions

H Medium term Funding

29 Need to consider advice on how to deal with cross border and transboundary effects

Regulators / SNCBs

H Short term Stakeholders

30 Develop the use of a framework, building on SEATRACK, to account for birds seen at sea but are from outside the UK.

ScotMER / TCE Enabling actions

M Long term Funding

31 SNH guidance on apportioning needs updating.

SNH H Short term

32 Explore the possibilities of establishing a tagging database, possibly hosted by JNCC.

ScotMER / TCE Enabling Actions / JNCC

M Medium term Funding Stakeholders

33 Further development of SeabORD to include prey distribution data and turbine density/spacing.

ScotMER H Long term Funding

34 Consider non-breeding birds (i.e. sabbaticals or immature birds) in the breeding season within the emulator.

ScotMER H Medium term Funding

35 Undertake feasibility study / pilot project for tagging seabirds at sea.

ScotMER / OWSMRF/ ORJIP / TCE Enabling Funds

H Short term H&S Funding Other research projects

36 The quantification of uncertainty in displacement models must be achieved.

ScotMER H Medium term Funding

37 Use of more GPS tracking data to characterise foraging trips made by birds to use in SeabORD models.

RSPB/BTO – tracking

H Short term Covid-19 Funding

38 Expansion of SeabORD models to cover periods other than chick-rearing including non-breeding season.

ScotMER H Short term Funding

39 Uncertainty in mass-survival relationships to be incorporated into SeabORD.

ScotMER H Short term Funding

APPORTIONING

Page 47: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

47 | P a g e

40 Investigate impact of using non-straight-line distances in apportioning models

SNH H Short term

41 Determine what requirements would be to include pre-breeding ages in MS apportioning tool

MS M Medium term Funding Data gaps

42 Investigate North Sea wide database of colonies and cumulative effects

Regulators / JNCC

H Long term Stakeholders Funding

OVERARCHING ACTIONS

43 Commission research on turnover rates

ScotMER or ORJIP

L Long term

44 Develop online impact assessment reference list

SNH H Short term

Page 48: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

48 | P a g e

Appendix 1 – questions from breakout sessions

Collision risk session Group 1 - TURBINE PARAMETERS 1) Pitch - how much does this vary. Variation is also seasonal, should monthly pitch mean and sd

be used in sCRM? 2) Do we want to include seasonal options or a monthly variation for rotor speed? 3) Turbine densities – how important are these and does it vary with turbine size? 4) Tidal offset. How should this be incorporated. Mean air gap or some other measure? (except for

floating wind when airgap is constant so sd = 0). 5) Blade tip speed – is this more useful than rotation speed? Based on fact that a larger diameter

turbine would have a faster tip speed for the same rotation speed. Is this reflected in modelling collision risk or just swept area? Any data on this?

6) Related to above, where do strikes occur? As most flights at rotor height are in lowest band of heights then would expect most strikes to be near tip of turbines. Is this the case?

Group 2 - BIRD INPUTS 1) How should measures of bird flight speed be obtained – laser rangefinder, radar tracks, GPS

tags? 2) How old can density data / colony counts be to inform assessments? 3) What source of age classification may be used to identify impacts to breeding birds? 4) Are sabbatical rates known for species of interest? Should estimates be obtained? 5) How do weather conditions affect flight speed, height or bird agility, and should this be

incorporated into the sCRM calculation? 6) Nocturnal activity – use civil or astronomical twilight? How to obtain estimates? 7) Do new DPO raise any knowledge gaps in terms of new species? Are current apportioning

approaches (breeding and non-breeding) are appropriate for any new species identified for the DPO sites?

Group 3 - PRESENTATION REPORTS/OUTPUTS 1) Which sCRM outputs do we require presentation for? (e.g. annual, monthly – which options, all

scenarios?) 2) What are your views on presenting the outputs from sCRM to aid SNH and MS in determining

an application? 3) Do we require presentation of a model log (e.g. pdf stating parameters used)? 4) To help manage future cumulative issues, can we use outputs such as standard report

templates from sCRM to help us? What should these standard templates include? 5) Do we want assessment to give impacts after periods other than 25 years? 6) Is it possible to separate out displacement/barrier and collision impacts or are we still needing

to assume all (flying) birds of relevant species are subject to both?

Page 49: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

49 | P a g e

Group 4 - NON-BREEDING SEASON CONSIDERATIONS

1) Should there be a seasonal calculation and output? E.g., inclusion of a slider with recommended

dates from guidance. 2) In some of our previous advice we advised on the need for 2 non breeding calculations e.g. for

gannet and kittiwake do account for spring and autumn passage movements - what new information is there, what advice do we need to consider for the next round of offshore wind applications?

3) Some aspects of assessment of non-breeding assessment was undertake on a qualitative basis previously e.g. kittiwake and puffin, what should we do now in terms of these species? Do we still undertake a qualitative analysis or move to a quantitative analysis, if so how? Can we discount any species from a non-breeding season assessment?

4) Migratory CRM – do we need an updated model built into the sCRM app/stand-alone – or should this be undertaken as a strategic exercise?

5) What updates do we need to consider in reviewing the migratory CRM (i.e. better evidence base for patterns of migration?)

6) Which species should this focus on? Does BDMPS cover migratory seabirds sufficiently to exclude them i.e. should this be a non-seabird exercise.

Group 5 OVERALL MODEL/CODE 1) Ability to run batches – input to include multiple different windfarm scenarios, and select which

scenarios and species to run. 2) Is there an alternative to the shiny app that we should be considering? Is it fit for mass roll out

and use? 3) What issues do we anticipate with in combination/cumulative comparing sCRM and CRM? 4) Is there a model log/id so that specific versions of the model can be run again – as per NE PVA?

Page 50: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

50 | P a g e

Displacement session Group 1 – SCOTWIND SPECIES 1. ScotWind has identified 17 DPOs (maps of the locations will be on walls around the auditorium and in the Holly Room). Will these areas present new species for consideration in displacement assessment? 2. What evidence for disturbance or displacement do we have for impacts on these species? 3. Can we rank these species in terms of concern or priority? 4. Can the Furness and Wade vulnerability assessments be used to assist in answering for 2 and 3? If not, what would help? 5. Are current apportioning approaches (breeding and non-breeding) are appropriate for any new species identified for the DPO sites? Group 2 – BDMPS 1.Would the development of a framework to account for birds seen at sea that are from outside the UK, but have connectivity with UK wind farms, particularly outside the breeding season be useful? This could build upon the useful source of data from current projects funded by SEATRACK, Vattenfall and Equinor of year-round tracking data. How could this be developed? 2. What evidence from recent tracking studies should be used to update the BDMPS approach? 3. Would improvements to the calculation of apportioning proportions with both the breeding and non-breeding season be valuable?

o In the breeding season the use of real data to improve estimates of decay with distance from the colony, extending MS apportioning tool to cover more than the 4 species.

o In the non-breeding season it would be useful to improve upon BDMPS approach using long-term tracking of individuals with known-colonies of origin across NW Europe.

How best to gather the evidence required? 4. What is the most sensible approach to incorporating the new evidence of movements of non-breeding auks?

Change the seasons of assessment,

Calculate impacts on partial populations?

Maybe change for razorbills, but not guillemots? 5. Is there any additional tagging research being undertaken that could or should be taken into account? Which species/ where? Group 3 - SEANSE recommendations (related to displacement) 1. If you were using SeabORD to provide representative estimates to use within the displacement matrix as a basis for predicting mortality (alleviating the need to re-run SeabORD in more challenging data poor contexts), what windfarm foot print scenarios, range of species, and populations would you consider? 2. Would it be possible to calculate a single mortality rate for all sites, or would these depend on region, footprint size/shape, distance from coast etc? NB the SEANSE report suggested that there is unlikely to be a one-size-fits-all mortality rate. 3. The approach to analyse GPS tacking data is similar to Wakefield et al. (2017) should this be expanded upon to include additional species? Or should the use of novel statistical methodologies to

Page 51: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

51 | P a g e

better estimate uncertainty in bird distributions and densities be considered? For example the use of spatio-temporal movement models to describe trajectories of birds which have been developed but not applied to large multi-colony datasets. 4. How can we better understand the large differences between maps produced by GPS tracking data and at-sea survey data, which is currently difficult to interpret? Understanding and modelling these differences will help develop methods that integrate both sources of data. For example, the use of integrated population models to provide a general framework for estimating a common underlying process has the potential to be applied to this problem. Are there other methods we could consider? Group 4 - Knowledge gaps (based on SEANSE recommendations) 1. How can uncertainty be better quantified by accounting for uncertainty in key model parameters (e.g. use of Bayesian calibration methods)? 2. Incorporation of recent estimates of uncertainty in the mass-survival relations for species derived from the recent MSS research project - Guillemot, Kittiwake, Puffin - some information on Razorbill. 3. An adaptation to the model to allow foraging birds to increase their foraging range when encountering barrier or displacement effects from OSWFs at the edge of their foraging range. This would be an adaptation to the model. It currently assumes a fixed foraging range, but a more sophisticated model could allow birds to extend their foraging range if they are displaced. 4. Within the SeabORD model, how could model simulation for flight paths and foraging trips be improved? 5. How could SeabORD (and potentially other simulation approaches) be expanded to include periods of the breeding season other than chick rearing (e.g. incubation) and to include the non-breeding season? DISPLACEMENT RATE UPDATES - Discussion Wall 1. Has any research or monitoring been published that could update displacement rates to be used? 2. What improvements can we make to monitoring to enable displacement rates to be derived?

Page 52: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

52 | P a g e

Appendix 2 – feedback received post workshop

At the workshop, we asked for any further thoughts that occurred to delegates to be provided in writing and we received the following points. Thank you for this additional feedback, we have considered further and identified further actions from these points in the action plan (see Conclusions).

Baseline site characterisation – survey requirements 1. Time of day of surveys

There appears to be an increase in comments from key stakeholders, in Scotland at least, about a need to capture a wide range of different times of day in at-sea surveys of seabirds (and other marine mega-fauna). An aerial survey provider noted that the time of surveys is something that is considered carefully in survey design, within certain limits. Key elements that need considering are:

a. Timing of surveys of birds at sea need to be considered carefully for its impact on

abundance estimates in projects and how this might bias the relative abundance in

different parts of the survey area. We know from research around seabirds and sandeels

and early work by the NCC/JNCC Seabirds at Sea Team that seabird activity can be

matched to the diel patterns in prey fish availability at sea. Specifically, at seabird

colonies, auk attendance is lowest in the early morning while a lot of individuals take

advantage of high sandeel availability at that time. The numbers increase steadily in the

morning and at around 10:00 a.m., stabilise to relatively moderate numbers in the

middle of the day before peaking in the evening, before a nocturnal exodus to sea of all

non-breeding and off-duty partners. This pattern is universal and forms the basis of the

recommended survey methods for auk survey at colonies. Auk attendance at colonies

provides a good indicator for the activity of most piscivorous seabird species at sea.

The suggestion that surveys at sea should capture this variation is not wrong but comes

with a number of significant risks of biasing the data if done incorrectly. The surveys,

particularly of large sites can take several hours to complete. The risk is that in a survey

in the breeding season, some transects could be completed at times of the day when

densities at sea do not compare with those at other times of the day. Thus, any

differences in the relative abundance between different parts of the study area could be

purely an artefact of the time of day the site was surveyed. Similarly, overall population

estimates of birds on site could be heavily influence by the relative time proportion that

the survey took place inside v’s outside this window of stable numbers at sea. This bias

could have an important impact on decisions taken to mitigate the impacts of a

proposed development by excluding consistent hotspots within the study area.

As this diel variation in numbers at sea only comes into play when birds are in

attendance at their colonies, it is a limited time of year when this issue needs to be

considered (outwith this time, all birds are at sea near their food sources, so there

should be less need to adjust for the time of day effects).

The ability to account for time of day in the survey design is easier when multiple

Page 53: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

53 | P a g e

surveys take place per month or if the survey design allows multiple visits which gives

more flexibility to target different times of day without biasing the coverage.

b. Time should be relative to sunrise not GMT or BST, as these have no biological meaning.

c. Time of day will likely be a less important explanatory variable in seabird activity for any

one location at sea than the state of tide during a survey.

d. An aerial survey provider noted that surveys need to be at least 1.5 hours after sunrise

and 1.5 hours before sunset for suitable images to be captured. Beyond this, light levels

are typically too low to provide reliable bird detection and identification. They also limit

the timing of surveys to avoid sun glare at certain times of day when transects are not

oriented directly north to south.

One approach discussed with MS is to provide an impact report on the effect of time of day and state of tide on the findings of the survey on the population sizes and density estimates used in the site assessment. This would use the patterns described in sub-sections above to illustrate how they would influence the key metrics.

2. Surveys across multiple days

Developers are being advised to carry out surveys on a single day snapshot for characterisation surveys on a monthly basis. This is correct in terms of good survey design requiring that the spatial extent of the survey area is covered in a single day. However, this does not mean that surveys can only take place on one day in the month. This is creating considerable logistical problems for aerial survey companies who have to provide multiple aircraft to ensure the survey design is executed in a single day at larger sites. Four aircraft in the middle of the winter, with shrinking weather windows is a recipe for incomplete coverage. An aerial survey provider has had to spend a lot of money taking aircraft from different airfields to the study area, only to have to abort surveys when the weather changes, there are technical faults or even if one of the airplanes cannot leave its home base. It only takes one aircraft to have a problem for the entire survey to be compromised and a replacement of the entire survey required. It is difficult enough to find suitable weather windows in winter without this additional, wasteful and un-necessary requirement. A much better solution would be to use an inter-leaved transect design, in which a single aircraft can survey the full spatial extent of the study area using broad transect spacing, then subsequent surveys on other days repeat the survey pattern, slightly offset from the original design, so that gaps are filled. The advantage of this approach is that it obtains the spatial detail of the design but is more representative of the patterns within the month and is logistically easier to effect, given the lack of dependency on other aircraft. This approach was deployed successfully for surveys of Dogger Bank and resulted in very little loss of data. This survey approach for large sites gives much better representation of the patterns within a study area by spreading coverage over multiple days, increasing the chances of capturing e.g. migration events, variation in the exploitation of prey hotspots on different days and would also lend itself to exploring time of day effects. 3. Survey coverage

There has been recent discussion by aerial survey providers with Natural England about the level of survey coverage that is adequate. The level of coverage needs to be a trade-off across the potential densities of the likely key species. These can be difficult to know with a high degree of certainty before a site is surveyed, but experience has shown that a minimum survey coverage of about 10% is usually sufficient. One aerial survey provider noted their surveys are parallel strip transects using four cameras within each transect allowing a range of coverage from 5% to 25% to be collected and analysed later, depending on the survey design. While data from the individual cameras within each

Page 54: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

54 | P a g e

transect cannot be treated as being statistically independent, varying the number of cameras used in the analysis for each survey can be adjusted without affecting the statistical independence of the transects within the overall survey design. Much of the discussion has centred around how to determine when survey coverage may be sufficient. Care has to be taken when selecting and assessing summary statistics from at sea survey data. Comments from NE recently suggested that a lower coefficient of variations (CV) would be the main aim of increased survey coverage. Since the CV is a summary statistic (calculated as the ratio of the mean to the standard deviation) describing the concentration of the data around the mean (i.e. the more concentrated the data are around the mean the lower the CV) there is no biological reason to expect that more survey coverage (or other forms of survey effort) would result in a lower CV. The correct CV for a sample would be one that is as close as possible to the population. When considering the spatial and temporal scale of relevance to an offshore wind farm development, there would seem little expectation that the CV should necessarily be low, as seabirds are often highly mobile and range at scale much larger than a single wind farm and its buffer. So, increasing the survey coverage may either reduce or even increase the CV depending on what happens to the distribution of the data with a change in survey coverage. Foraging range estimates 4. Use of mean max +1SD foraging range for determining connectivity: As discussed at the

workshop, this may run the risk of needlessly including many colony populations (at the cost of adding to workloads). Whilst there may be justification for doing this with rarer species for which there are no, or few, colonies within mean max foraging range of a development site, for commoner species which already had several colonies within mean max foraging range of the site there would be little, if any, chance of effects occurring on those colonies that are further than this from the site (and therefore no point in including them at the screening stage).

5. Foraging range - is it a flawed metric?

There is inconsistent use of the term ‘mean foraging range’. The term is not defined in Thaxter or Woodward. Furthermore I don’t think we are necessarily using the best summary parameter(s) of the tracking data to best inform apportionment and impact assessment questions. Ultimately the questions of highest relevance to OFW projects comes down to what is the likelihood that an individual at a OWF site is from a particular colony (population), and what is the relative importance of one location relative to another location (i.e., spatial sensitivity).

The concept of a seabird foraging range is easy to understand though it needs to be carefully defined.

It could be defined as ‘the distance travelled by a seabird between its nest site and where it chooses to forage’ or it could be defined as the ‘area over which a seabird forages, and relative to its nest site’ – using the word ‘range’ in an analogous way to the term ‘home-range’ as used widely in vertebrate behavioural studies. Seems wrong and inconsistent that the term ‘foraging range’ should be a measure of distance if ‘home range’ is a measure of area. It should be made clear that the OWF guidance is actually discussing distance metrics of the foraging range.

Page 55: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

55 | P a g e

The maximum foraging range metric is a very crude as it’s potentially influenced by rare events, atypically behaving individuals and sample size. So clearly is not a clever use of the data.

Mean Maximum Foraging Range (MMFR) is almost crude for the same reasons, but not quite as bad. So neither of these are particularly robust or clever metrics to be using in impact assessment and apportionment methods.

Using ‘mean foraging range makes better use of data’, as do other potential summarising measures not considered. But we need to be very clear what the measures are that we are taking a mean off, and apply this consistently between studies. I doubt they have been in the literature.

Mean foraging range could be any one of the following (here I assume we are talking of the mean foraging range for a particular colony based on tagging several individuals in that colony):

The mean across individuals in the study of the mean distance from nest site of all tag locations of a bird collected in a tag deployment. But this would be nonsense if it included records when bird at the colony.

As above but with the at-the-colony tag locations (i.e. when the bird is at the nest site) excluded. But this does not necessarily give an accurate measure of foraging distances because birds may be engaged in other activities when away from the nest site, e.g. loafing and feather maintenance time on the sea.

The mean across individuals in the study of individual’s maximum distance from nest site of all tag locations collected in a tag deployment. But this is not in any sense a mean foraging range, it’s the mean of individuals’ maximum foraging distances in the study. Furthermore it’s affected by sample size in terms of the duration that tags remain attached.

The mean across individuals in the study of the maximum distance from the nest site attained for each foraging ‘trip’ in a tag deployment. But this approach is problematical because of the difficulty of defining a forging trip, e.g., some short trips might not actually be foraging trips, rather they may be to loaf or bath at a local site. It is also complicated by the fact that trips may be of different value in terms of calories of food attained and ideally this needs to be taken into account.

And there are other variants to the list above. The point is, however, the ‘mean’ is defined, it results in a value that is unsatisfactory for reasons of bias, is rather meaningless, is difficult to interpret, and generally is not particularly useful to inform the needs of impact assessment and apportionment. This raises the question of whether there is a better metric that overcomes these shortcomings.

It seems to me there is an alternative approach that is more satisfactory, easier to understand and more useful to EIA and apportionment studies. This would be to quantify the proportion of at-sea time (e.g., proportion of GPS tag records) that a bird spends in each of a series of distance bands, say of 10-km wide zones out from the colony (the width of bands chosen should be appropriate for the species being considered). This method has the advantage (over the obvious alternative of percentile based kernel summaries of distance metrics that one sometimes sees in publications) that it is easy to go on to combine results across individual birds, several colonies, and across different people’s studies. Thus in this

Page 56: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

56 | P a g e

way one could easily produce (at a colony, regional or national level) summary metrics on % marine habitat utilisation (i.e. % of at-sea time) versus distance from the colony.

My suggestion has some similarities (but is not the same!) to the habitat utilisation modelling approach used by RSPB (Cleasby et al., 2018). Incidentally, the RSPB habitat utilisation modelling study results (as presented) appear not to help much with the question of apportioning birds between colonies, except in as much that the results indicate that most birds spend most of their time rather close to the colony.

6. Use of MMFR+1SD

As a basis for high level screening for HRA, MMFR+1SD is a reasonable starting point. But it is so high level and cautious as to produce rather meaningless results – it’s being driven by the exceptional bird and not the typical birds. What matters more is deciding in an intelligent way to how apportion birds to individual SPAs. Although a first level screening on MMFR+1SD is cautious and therefore good as no sites slip through the net, at the next stage this could lead to people going on apportion an impact to too many colonies. If one ‘over dilutes’ an impact between many colonies then this would result in non-cautious assessments for individual colonies. Thus we need to be wary of unintended consequences.

My suggested method of measuring mean (either across colonies or colony specific) utilisation of at-sea distance zones might be a good approach to apportionment, i.e. essentially using measures of % utilization of an area by birds in a particular colony as a surrogate for strength of connectivity.

7. Use of colony-specific foraging ranges

I think that using colony-specific foraging ranges is an excellent idea, or at least regionally specific ranges – if the data justify this.

It is not surprising perhaps that isolated colonies tend to have larger foraging ranges than ones with close neighbours. Isolated colonies usually reflect geography and the lack of other potential breeding sites, e.g. along soft coasts. For example contrast the situation for auks and kittiwakes breeding at the isolated Bempton Cliffs colony in east England with those at the multiple colonies and near continuous availability of suitable nesting habitat off the north coast of Scotland and around Orkney.

Data analysis 8. MRSea and using co-variates with baseline surveys: An implication in the presentation on

baseline site characterisation appeared to be that there is potential to improve on-site bird abundance estimates via the use model-based estimates (ie by using co-variates and applying MRSea). Whilst there is undoubtedly some potential for this I think that this would need to be trialled on a number of data sets to determine how well such approaches perform before they are recommended. Such approaches may work well for relatively inshore species such as divers and seaducks but I’m not aware that they have been shown to offer benefits over design-based approaches for more offshore species. Without some fairly rigorous testing, there would be a risk of incurring a lot of additional effort and work with little (or perhaps no) gain.

Page 57: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

57 | P a g e

Collision risk assessment 9. sCRM: It seems likely that there will be a need to have tight guideline produced on how the

sCRM should be populated and used, particularly with regard to identifying the parameters for which it is likely to be feasible, and useful, to incorporate measures of variability. Without such guidance there is clearly scope for multiple approaches to be taken, resulting in many permutations of potential outputs.

Displacement assessment 10. SeabORD: The SeabORD model for estimating the effects of displacement / barrier effects on

seabird demographics clearly has much greater biological basis than the existing matrix approach currently used for assessments. However, I think that there is still considerable potential for ‘error’ in the outputs because it is (of course) a model and the model involves a number of key assumptions around which there is still much uncertainty. Our experience of using SeabORD was that the magnitude of the discrepancies with the matrix estimates varied between SPAs, and for some SPA populations the levels of mortality predicted by SeabORD did not seem credible (based on sense-checks which used the estimated on-site densities, albeit with an approximate apportioning applied as well).

Such experience suggests that caution is still required in applying SeabORD and that consideration should be given to such issues before deciding that this should be the advised approach for estimating impacts from displacement/barrier effects. Further investigation of the discrepancies between the two approaches is probably warranted.

Apportioning predicted impacts 11. To some extent (a large extent in some places) the problems of apportioning breeding birds at

sea to individual colonies is a problem of our own making caused by focusing the impact assessment on an arguably inappropriately defined (biologically speaking) receptor. Whereas SNH chooses for reasons connected to legislation compliance to focus assessment at the SPA level, in many places in Scotland where there are multiple colonies in relatively close proximity, indeed the very places where apportioning is so tricky, it is biologically inappropriate to be considering population processes at the individual SPA level because in reality the population processes are operating at a larger scale. If the assessment was based on a more biologically realistic wider scale (as appropriate for a particular species and region) then we would not need to obsess so much about apportionment to a particular colony within that region. An SPA based focus may be inappropriate for discontinuous SPAs (e.g., North Caithness Cliffs), that are in reality several separate colonies. This particular so when some constituent parts of the SPA are further from each other than from other SPAs (e.g. I’m thinking here of Hoy SPA and the components of North Caithness Cliffs SPA) .

Personally I would favour assessments and modelling to more realistically reflect the spatial scale of a species’ population processes, i.e. something akin to regional scale is likely to be appropriate in most cases, rather than the individual colony/SPA scale (‘mega’ colonies excepted perhaps, and also species like gannet and Manx shearwater excepted), and then apply the assumption that a SPA within the area examined is expected to be affected in the same way as the population as a hole in the wider region examined.

Page 58: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

58 | P a g e

Displacement breakout session – Group 1 – ScotWind Species 12. Relating to discussions about identifying new species for consideration for ScotWind sites.

To me it seems like the following groups might need considering slightly differently:

Breeding seabirds

Non-breeding seabirds

Pelagic migrant seabirds

Migrant waterfowl

Other migrants (terrestrial / passerines)

My thoughts (below) focus on the first 3 groups

For group 1 I think degree of overlap of Scotwind areas with foraging ranges (mean max + SD and mean) around SPAs (and vice versa) might be useful to inform discussions.

Imagine Scotwind outlines plonked on below map (bonxie) for Woodward et al. 2019

Page 59: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

59 | P a g e

And resultant tables for all relevant SPA species For example (just made up from looking at some maps in the Woodward appendix):

Species % foraging range within ScotWind areas

% ScotWind areas within foraging range

Mean max + sd mean Mean max + sd mean

Great skua 1 0.1 100 40

Storm petrel 3 ? 100 ?

Leach’s petrel ? 2 ? 100

Puffin 2 20 100 80

Etc. (do this for all sp in Woodward?)

And then for groups 1 to 3 – particularly 2 and 3 – I think consideration of the site outlines in terms of Waggit et al., (2019) data would be a useful avenue to go down.

Page 60: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

60 | P a g e

Appendix 3 – list of delegates

First name Surname Organisation

Frank Fortune AECOM

Laura Jervis APEM Ltd

Sean Sweeney APEM Ltd

Digger Jackson Atlantic Ecology

Aonghais Cook BTO

Daniel Johnston BTO

Hannah Whyte Crown Estate Scotland

Francis Daunt CEH

Kate Searle CEH

Marc Murray Cierco

Claire Macnamara DAERA

Carl Donovan DMP Stats

Nancy McLean EDF

Pierre-Louis Chopin EOLFI

Monica Fundingsland Equinor

Sarah Edwards EDPR

Andrew Hamilton EDPR

Richard Wakefield Floatation

Fraser Carter GoBe

Ross McGregor Hi Def

Andy Webb Hi Def

Andrew Logie Innogy

Julie Black JNCC

Orea Anderson JNCC

Charlie Whyte KOWL

Sue King King Consulting Ltd

Rafe Dewar MacArthur Green

Nicola Goodship MacArthur Green

Ciorstaidh Couston Mainstream

Tom Evans Marine Scotland

Gayle Holland Marine Scotland

Zoe Crutchfield Marine Scotland

Kerry Bell Marine Scotland

Fiona MacKintosh Marine Scotland

Janelle Braithwaite Marine Scotland

Mike Meadows Natural England

Graeme Cook Natural Power

Ewan Walker Neart na Gaoithe

Polly Tarrant Neart na Gaoithe

Claire Gilchrist Neart na Gaoithe

Matthew Hazleton NIRAS

Page 61: Marine Bird Impact Assessment Guidance Workshop Report bird impa… · Marine bird impact assessment guidance workshop report 5 | P a g e Summary The purpose of the workshop was to

Marine bird impact assessment guidance workshop report

61 | P a g e

Liz Morgan NIRAS

Phil Bloor Pelagica

Nathalie Stevenson Quaybridge

Ben King Red Rock Power

Sarah Arthur Red Rock Power

Jack Farnham Ridge Power

Helen Riley Royal Haskoning

Murray Grant Royal Haskoning

Robert Iredale Royal Haskoning

Simon Zisman RPS

Mike Armitage RPS

Aly McCluskie RSPB

Darren Jameson Scottish Power

Bob Furness SNH Board

Alex Robbins SNH

Emma Philip SNH

Erica Knott SNH

Chris Eastham SNH

Glen Tyler SNH

Kate Thompson SNH

George Lees SNH

Catriona Gall SNH

Karen Taylor SNH

Kirstie Dearing SNH

Cathy Tilbrook SNH

Jen Graham SNH

Emma Ahart SSE

Nick Brockie SSE

Martha Lovatt SSE

Beth Scott University of Aberdeen

James Chapman University of Aberdeen

Morgane Declerck University of Aberdeen

Mailys Billet Xodus

David Bloxsom Xodus

MacNeill Ferguson Xodus