WESTERMOST ROUGH OFFSHORE WIND FARM: ORNITHOLOGICAL SURVEY ANNUAL REPORT, SEPTEMBER 2012 OCTOBER 2013

Transcription

1 WESTERMOST ROUGH OFFSHORE WIND FARM: ORNITHOLOGICAL SURVEY ANNUAL REPORT, SEPTEMBER 2012 OCTOBER 2013 Jon Ford (Gannet), Steve Percival (Puffin and Little gull) Dr Steve Percival and Jon Ford On behalf of DONG November 2013

2 CONTENTS Page 1 INTRODUCTION Background LICENSE CONDITIONS BOAT-BASED SURVEYS Survey Area Survey methods PRELIMINARY RESULTS Survey Count Totals and Densities MARINE MAMMALS SUMMARY AND CONCLUSION REFERENCES Westermost Rough ii Ornithological Monitoring

3 1 INTRODUCTION 1.1 Background This report presents the results of the bird monitoring programme that has been undertaken between September 2012 and October It forms part of the pre-construction phase monitoring of the Westermost Rough Offshore Wind Farm. The purpose of this report is to document the surveys that have been undertaken during this period, including the survey routes covered and present estimates of the bird populations present, and to discuss the main findings of the surveys including a comparison with the previous surveys results and assessment of any influence of weather conditions and other relevant information that may have affected species abundance and behaviour. The main aim of this phase of the work is to determine the distribution and abundance of seabirds using the Westermost Rough Offshore Wind Farm site and its surrounds before the construction phase of the wind farm. Standard survey methodologies have been used, following Camphuysen et al. (2004) and have remained consistent throughout the surveying undertaken. The Westermost Rough Offshore Wind Farm project is located in the Greater Wash Strategic Environmental Assessment (SEA) area, approximately 8 kilometres off the Holderness coast of East Yorkshire. The Westermost Rough Offshore Wind Farm extends over an area of 35km 2. An indicative site layout based on 35 turbines with a rotor diameter of 155m and tip height of 177m is shown in Figure 1. Figure 1. Indicative layout of 35 turbines and an offshore substation. Westermost Rough Ltd submitted the application for consents to the relevant authorities, accompanied with an Environmental Statement on 17 th November Section 36 consent was received from the Department of Energy and Climate Change (DECC) and the Marine Westermost Rough Ornithological Monitoring

4 Licence (Licence Number L/2011/00305) from the Marine Management Organisation (MMO) on 29 th November This permitted the wind farm to be developed with turbine dimensions comprising a maximum rotor diameter of 150m to a tip height of 177m. In March 2012, Westermost Rough Ltd submitted an Addendum to the original Environmental Statement to DECC and the MMO, requesting the consents be amended to accommodate for the larger proposed turbine. This included the re-assessment of a number of environmental studies, including the ornithology assessment. The new consent was received from DECC and the MMO in October The Scope of Works for the preconstruction ornithological surveys, which reflects the license conditions specified in the Marine License (see Section 2 below) has been developed in consultation with the Licensing Authorities. Offshore construction is due to commence in February 2014 with the installation of turbine foundations. The scope of ornithological surveys comprises the following: 14 x boat based ornithology surveys, September 2012 October 2013 (the originally 12 agreed surveys to August 2013 were extended with two further surveys in September and October 2013); Data analysis; and Pre-construction monitoring reporting (including quarterly reports, and this final annual report). 2 LICENSE CONDITIONS The conditions in the Marine Licence L/2011/00305/1 regarding ornithological monitoring are outlined below The Licence Holder shall carry out ornithological monitoring as outlined in Annex 2 attached to this Schedule. The Licence Holder must submit the Ornithological Monitoring Programme (OMP) at least two months prior to the commencement of pre-construction monitoring. The Licence Holder shall not commence construction until the full specification for the OMP has been agreed with the Licensing Authority following consultation with Natural England. The Ornithological Monitoring Programme must include a timetable for an annual and interim reporting mechanism The Licence Holder shall submit the reports required under the Ornithological Monitoring Programme to the Licensing Authority at the appropriate time in accordance with the agreed timetable. Each report must be forwarded to the Licensing Authority and Natural England by the date specified in the OMP. Annex 2: Monitoring will comprise a Before and After Control Impact (BACI) design and will be undertaken at the survey areas consisting of the wind farm site, a 1km and 2-4km buffer zone surrounding the wind farm and the selected reference site. The monitoring programme will include a 1 year contiguous pre-construction period and continue through the construction phase. This monitoring period will serve the dual purpose of identifying any possible changes in the ornithological value of the area (since ES surveys), and providing an immediate baseline for comparison with data collected during the construction phase. There is also a requirement to conduct Westermost Rough Ornithological Monitoring

5 post-construction monitoring to provide a minimum of three years data from the operating phase. The detailed specification for the co-ordinated monitoring programme will be subject to separate written agreement with the Licensing Authority following consultation with Natural England four months prior to the proposed commencement of the monitoring work. These data will need to be empirically comparative with baseline data provided within the project's Environmental Statement. The need for additional ornithological monitoring, on-going during the lifetime of the wind farm's operation, will be determined, in consultation with Natural England and the Licensing Authority and reviewed at agreed periods. This will have regard to the magnitude of any change in bird populations observed during the initial three years operational monitoring period. The ornithological monitoring programme may have to be adapted and amended as new technologies and research findings become available, as determined by Natural England and the Licensing Authority. Ornithological monitoring reports will be provided to the Licensing Authority and Natural England on a quarterly basis as a draft report update and as a final annual report. This may be more frequent where the results of the data may trigger further, more intensive monitoring work. Monitoring of the agreed reference site will also continue parallel to the wind farm site and the 1km and 2-4km buffer zones surrounding the wind farm. Monitoring will need to fulfil the following objectives: 1. Determine whether there is change in bird distribution, use and passage, measured by species (with particular reference to Gannet, Guillemot, Lesser black-backed gull, Herring Gull, Kittiwake and Common Tern), abundance and behaviour, of the wind farm site, 1km and 2-4km buffer zones and the reference site. 3 BOAT-BASED SURVEYS Survey Area The surveys reported here cover the survey area as set out in the proposed monitoring programme, and include the proposed wind farm site plus 0-2km and 2-4km buffer zones. A reference area is located north of the wind farm site and covers approximately the same size as the wind farm site including buffer zones, but avoiding the shallow ground close to the coast to be as comparable as possible with the wind farm site. It was sought to have the reference area represent similar water depth as the wind farm site and buffer zones (Figure 2). The transect spacing used in is 2km with a total length of 160km. The total area surveyed was 308km 2, comprising the wind farm site plus a 4km buffer (185km 2 ) and a separate reference area of 123km 2 to the north (Figure 2). Westermost Rough Ornithological Monitoring

6 Figure 2. The Westermost Rough Offshore Wind Farm site and 4km buffer (green shades) and the reference area for the ornithological survey program. Distance between transects are 2km and the total line kilometres is ca. 156km. Numbers indicate transect/line numbers. A total of 14 surveys have been carried out during September October 2013, as scheduled. The GPS tracks showing the routes followed on each survey are shown in Appendix 1. The surveys were carried out on the following dates: 28 and 29 September 2012; 15 and 16 October 2012; 18 and 20 November 2012 (survey days not consecutive due to inclement weather on the 19 th ); 4-5 January 2013; 15 th April 2013; 30 th April 2013; 14 th May 2013; 29 th May 2013; 13 th June 2013; 8 th July 2012; 23 rd July 2013; 12 th August 2013; 24 th September 2013; and Westermost Rough Ornithological Monitoring

7 15 th /16 th October Survey methods The survey methods follow those detailed in the proposed Westermost Rough Offshore Wind Farm Bird Monitoring Protocol. The surveys comprise boat-based line transects of the study area, following the methodology recommended in Camphuysen et al. (2004) and as reviewed by Maclean et al. (2009). The Fruitful Harvest 3 is being used as the survey vessel. She is a robust 19m survey vessel with a sheltered and relatively stable observation platform giving 6.2m eye-height for surveyors. It is ideal for the work being of a size and a manoeuvrability (with an experienced local crew) to enable safe operation across the survey area and around busy shipping channels. A GPS record of the precise route was taken on each trip, so that the location at all times was known. The observation team on the surveys comprised Jon Ford, Trevor Charlton, Gary Elton, Peter Dodds, John Clarkson (with three surveyors on each survey), who were each involved in both observation and recording. All surveyors were JNCC ESAS qualified. Three surveyors were deployed in order to allow recording on both sides of the survey vessel simultaneously, rotation of duties and to enable one surveyor to be free to undertake continual forward scanning for the detection of species that may be flushed from the sea surface. The team are all highly experienced ornithologists, well able to identify all the species encountered accurately. All observers also have a good knowledge of the area and its ornithological interests, and are also trained Marine Mammal Observers. All birds encountered, their behaviour, flight height and approximate distance from the boat were recorded. Following the JNCC Seabirds at Sea recommendations, birds were recorded into five distance bands (0-50m, m, m, m and 300+m). Birds were recorded continuously, at a steady speed of approximately 10 knots, with the precise time of each observation recorded where possible to give as accurate a position as possible (linking to the GPS position information being recorded simultaneously). All records of birds observed flying as well as those on the sea were recorded. All sightings of marine mammals were also recorded during the surveys (and identified to species level when possible). The approximate height above the sea of all flying birds was recorded, estimated as accurately as possible (for later conversion to height bands for presentation and assessment as required). Flying birds were recorded using snapshot counts at two-minute intervals. Whilst all birds observed were recorded, a note of those in transect was made to facilitate later analysis. The weather conditions during the surveys were recorded, including sea state, wind speed and wind direction. Any specific conditions in the area that may affect bird abundance/behaviour (e.g. if a storm has passed the area in advance of a survey, many construction vessels etc.) were additionally noted. For each bird observation, the following is being recorded: Observation time; Latitude and longitude (WGS84 UTM30N); ; Numbers; age classes; Westermost Rough Ornithological Monitoring

8 Distance band from the vessel; Sitting/flight height; Flight direction; Behaviour; Association (e.g. with fishing vessels). In addition, fishing vessels and other vessels (e.g. construction vessels or ferries) are also recorded. For registration of behaviour, the standards outlined in Camphuysen and Garthe (2004) are being used. 3.3 Distance Modelling to Determine Population Estimates The data have been analysed in accordance with the standard principles of distance sampling, but the generally low numbers of records per species recorded on the sea during each survey meant that it was not possible to use the Distance 6 software (Thomas et al. 2009) to generate reliable distance correction factors for each survey. Instead therefore a simpler approach was adopted. The raw count data from the boat-based surveys were adjusted to take into account the fact that the likelihood of a bird being seen declines with distance from the observer (i.e. detectability is a function of distance from the transect line). Put simply, the chance of seeing a bird close to the observer would be higher than if it were at greater distance. The relationship between detectability and distance can be modelled using software packages such as Distance (Buckland et al. 2001), but for the purposes of this assessment a simpler approach was adopted (mainly because the limited number of distance bands makes modelling of the distance function difficult for many of the species encountered in this study, and the limited number of records on the sea of most species during each survey). The approach used here is similar to that used by JNCC in their Seabirds at Sea surveys (e.g. Stone et al. 1995), but correction factors have been calculated for each major species group specifically using the data collected from the boat survey, for each survey visit. were assigned to these groups on their similarly of likely detectability and pooled to give a robust sample size for each group. Group compositions are given in Table 1. The same process was used to correct the data for each survey visit, as detectability differed as a result of different observation conditions (apart from the seaduck category, the numbers of records of which on the sea were insufficient to provide robust estimates of visit-specific correction factors). Table 1. groups used in calculation of distance correction factors Group Auks Guillemot, Razorbill, Puffin, Little Auk Gannet Gannet Gulls Gulls, skuas, terns, shearwaters Seaduck, divers and cormorants All wildfowl, divers, grebes and cormorants. The process in calculating those correction factors was as follows: The total numbers of birds of each species group were calculated for each distance band during each of the surveys. Differences in the width of the distance bands were taken into account by dividing the total number by the band width, to give a standardised total (density index). It was assumed that bird detectability in the closest transect to the observer was 100% (a standard assumption of the Distance sampling methodology). Westermost Rough Ornithological Monitoring

9 As detectability of birds on the sea and flying were different from the boat survey data separate correction factors were used for each of these. In fact detectability of flying birds was so high that no correction factors were necessary for these birds effectively all of these birds were detected within the main transect. For each of the other bands, the percentage difference between that band s standardised total and the closest band to the observer were calculated. These differences were then applied as the correction factors, dividing each count by the appropriate factor. For example, auks in band C on Survey 1 were divided by 79%. Hence a count of 100 in that band on that survey would be corrected to 127 (=100/0.79). The correction factors used for each species group are shown in Table 2. Table 2a. Distance correction factors used for the boat survey data: auks. Survey Number Survey Date A [0-50m] B [50-100m] C [ m] D [ m] 1 28 & 29/09/12 100% 100% 79% 27% 2 15 & 16/10/12 100% 64% 51% 26% 3 18 & 20/11/12 100% 50% 38% 15% 4 04 & 05/01/13 100% 58% 55% 18% 5 15/04/13 100% 45% 26% 24% 6 30/04/13 100% 29% 18% 11% 7 14/05/13 100% 58% 38% 10% 8 29/05/13 100% 49% 32% 10% 9 13/06/13 100% 32% 15% 10% 10 08/07/13 100% 46% 43% 15% 11 23/07/13 100% 100% 63% 25% 12 12/08/13 100% 27% 10% 10% 13 24/09/13 100% 100% 100% 83% & 16/10/13 100% 50% 39% 10% Table 2b. Distance correction factors used for the boat survey data: gannet. Survey Number Survey Date A [0-50m] B [50-100m] C [ m] D [ m] 1 28 & 29/09/12 100% 79% 79% 79% 2 15 & 16/10/12 100% 73% 73% 33% 3 18 & 20/11/12 100% 100% 100% 100% 4 04 & 05/01/13 100% 100% 100% 100% 5 15/04/13 100% 90% 90% 90% 6 30/04/13 100% 30% 30% 30% 7 14/05/13 100% 75% 75% 50% 8 29/05/13 100% 100% 50% 30% 9 13/06/13 100% 34% 34% 27% 10 08/07/13 100% 48% 48% 23% 11 23/07/13 100% 81% 81% 44% 12 12/08/13 100% 68% 68% 68% 13 24/09/13 100% 100% 100% 58% & 16/10/13 100% 43% 25% 14% Westermost Rough Ornithological Monitoring

10 Table 2c. Distance correction factors used for the boat survey data: gulls. Survey Number Survey Date A [0-50m] B [50-100m] C [ m] D [ m] 1 28 & 29/09/12 100% 100% 86% 42% 2 15 & 16/10/12 100% 63% 28% 14% 3 18 & 20/11/12 100% 82% 82% 45% 4 04 & 05/01/13 100% 100% 74% 53% 5 15/04/13 100% 61% 61% 28% 6 30/04/13 100% 41% 31% 31% 7 14/05/13 100% 76% 62% 29% 8 29/05/13 100% 68% 32% 13% 9 13/06/13 100% 39% 25% 13% 10 08/07/13 100% 50% 44% 20% 11 23/07/13 100% 53% 38% 14% 12 12/08/13 100% 44% 25% 13% 13 24/09/13 100% 73% 49% 23% For the seaduck, divers and cormorants there were so few records (86 in total) that surveyspecific correction factors could not be reliably estimated so instead values over all of the surveys were pooled (100% for Band A and 65% for Bands B-D). 4 BIRD SURVEY NUMBERS AND DISTRIBUTIONS 4.1 Survey Count Totals and Densities The raw count totals for the surveys from all of the survey data (including out of transect observations) are summarised in Table 3. This gives the total (uncorrected) numbers of each species counted during each survey. The bird population estimates for the survey area for each survey, based on in-transect counts from the main survey transect sampling area (within 300m of the survey vessel) with a correction for distance sampling and survey coverage, are shown in Table 4. Table 5 gives the density of each recorded during each survey, again based on the main 300m transect data. Westermost Rough Ornithological Monitoring

11 Table 3. Survey area total raw bird counts during the September October 2013 surveys. Sep 12 Oct 12 Nov 12 Jan 13 Apr Apr May May Jun 13 Jul 13a Jul 13b Aug 12 Sep 13 Oct 13 13a 13b 13a 13b Pink-footed goose Greylag goose Mallard duck sp Tufted duck Common scoter Red-throated diver Black-throated diver Great northern diver diver sp Fulmar Sooty shearwater Manx shearwater Leach's Storm-petrel Gannet Cormorant Shag Red kite Oystercatcher Dunlin plover sp Bar-tailed godwit small wader sp wader sp Turnstone Pomarine Skua Arctic skua Great skua Common gull Lesser black-backed gull Herring gull Westermost Rough Ornithological Monitoring

12 Sep 12 Oct 12 Nov 12 Jan 13 Apr Apr May May Jun 13 Jul 13a Jul 13b Aug 12 Sep 13 Oct 13 13a 13b 13a 13b Great black-backed gull Little gull Black-headed gull Kittiwake black-backed gull sp gull sp large gull sp small gull sp Black tern Sandwich tern Common tern Arctic tern Common/Arctic tern Guillemot Razorbill auk sp Little auk Puffin Swift Skylark Sand martin Swallow House martin Meadow pipit Wren Wheatear Song thrush Lesser whitethroat Chiffchaff Goldcrest Starling Brambling Westermost Rough Ornithological Monitoring

13 Sep 12 Oct 12 Nov 12 Jan 13 Apr Apr May May Jun 13 Jul 13a Jul 13b Aug 12 Sep 13 Oct 13 13a 13b 13a 13b Goldfinch finch sp small passerine sp Table 4. Survey area total population estimates corrected for distance sampling and survey coverage, September October Note: estimates based on in-transect data only. Sep 12 Oct 12 Nov 12 Jan 13 Apr 13a Apr 13b May 13a May 13b Jun 13 Jul Jul 13b Aug 12 Sep 13 Oct 13 13a Common scoter duck sp Red-throated diver diver sp Fulmar Sooty shearwater Manx shearwater Leach's Petrel Gannet , Cormorant Shag Red kite Oystercatcher Dunlin Pomarine Skua Arctic skua Great skua Common gull Lesser black-backed gull Herring gull Great black-backed gull Little gull 9, , Black-headed gull Westermost Rough Ornithological Monitoring

14 Sep 12 Oct 12 Nov 12 Jan 13 Apr 13a Apr 13b May 13a May 13b Jun 13 Jul Jul 13b Aug 12 Sep 13 Oct 13 13a Kittiwake 6, , , black-backed gull sp large gull sp gull sp Black tern Common tern Arctic tern Guillemot 157 4, ,969 1, ,238 1,085 2, Razorbill 8, , Little auk Puffin auk sp , Domestic pigeon House martin Meadow pipit Wren Lesser whitethroat Chiffchaff Starling Table 5. Survey area population densities corrected for distance sampling and survey coverage, September October Note: as in Table 4 estimates based on in-transect data only. Sep 12 Oct 12 Nov 12 Jan 13 Apr 13a Apr 13b May 13a May 13b Jun 13 Jul Jul 13b Aug 12 Sep 13 Oct 13 13a Common scoter duck sp Red-throated diver diver sp Fulmar Sooty shearwater Manx shearwater Westermost Rough Ornithological Monitoring

15 Sep 12 Oct 12 Nov 12 Jan 13 Apr 13a Apr 13b May 13a May 13b Jun 13 Jul Jul 13b Aug 12 Sep 13 Oct 13 13a Leach's Petrel Gannet Cormorant Shag Red kite Oystercatcher Dunlin Pomarine Skua Arctic skua Great skua Common gull Lesser black-backed gull Herring gull Great black-backed gull Little gull Black-headed gull Kittiwake black-backed gull sp large gull sp gull sp Black tern Common tern Arctic tern Guillemot Razorbill Little auk Puffin auk sp Domestic pigeon House martin Meadow pipit Wren Westermost Rough Ornithological Monitoring

16 Sep 12 Oct 12 Nov 12 Jan 13 Apr 13a Apr 13b May 13a May 13b Jun 13 Jul Jul 13b Aug 12 Sep 13 Oct 13 13a Lesser whitethroat Chiffchaff Starling Westermost Rough Ornithological Monitoring

17 4.2 Key Distributions The distribution of the key species present in notable numbers during the September 2012 October 2013 surveys (fulmar, gannet, herring gull, great black-backed gull, little gull, kittiwake, guillemot, razorbill and puffin) are shown in Figures All were generally widespread over the whole survey area though for little gull, kittiwake and razorbill, densities appeared to be higher in the southern part of the survey area (the wind farm plus buffers), whilst for guillemot and puffin numbers were higher in the northern (reference) area. Further analysis of the spatial distribution of these birds in relation to the wind farm site is presented in section 6 below. Westermost Rough Ornithological Monitoring

18 Westermost Rough Ornithological Monitoring

19 Westermost Rough Ornithological Monitoring

20 Westermost Rough Ornithological Monitoring

21 4.3 Bird Flight Activity within the proposed wind farm site The bird flight activity within the collision risk zone (taken as the footprint of the Offshore Project site see Figure 2) is summarised in Table 6. This gives the mean count within this Westermost Rough Ornithological Monitoring

22 zone, the percentage of birds flying (the number of birds recorded as flying during the boat surveys divided by the total number observed during those surveys), the percentage of those observed at rotor height (again derived from the boat survey data), which are combined to give the estimated mean numbers flying at rotor height. The latter is calculated as a mean value for each species for each month for input into the collision risk modelling see below). To maximise the explanatory power of the modelling the percentage of flying birds recorded at rotor height, this included all records of flying birds where height was recorded. The overall percentage value (being based on the largest possible sample size) was then be applied to the number of flying birds in transect within the wind farm area. Table 6 also shows flight heights published by Cook et al. (2012) in a review of data from 40 wind farm sites. The local data have been used in preference in the collision modelling, as a reasonable sample size was obtained for all of these species (>20 flocks measured). Table 6. Bird numbers and flight behaviour within the WMR site from the boat survey data, and the number flying at risk height. Mean count flying in wind farm [A] Mean density flying in wind farm [B]=[A]/Ar ea of wind farm % of flying birds at rotor height [C] Sample size (flocks) Cook et al % flights at rotor height Mean density flying at collision height in wind farm = B x C Red-throated diver % % 0 Fulmar % % 0 Gannet % % Common gull % % Lesser blackbacked gull % 93 25% Herring gull % % Great blackbacked gull % % Little gull % % Kittiwake % % Guillemot % % 0 Razorbill % % 0 Puffin % % 0 auk sp % House martin % 3-0 Starling % Key species at risk of collision (i.e. those observed flying through the wind farm site at rotor height) comprised Gannet, Common gull, Lesser black-backed gull, Herring gull, Great Blackbacked gull, Little gull, and Kittiwake. Collision risk modelling was therefore undertaken for all of these species and is presented below. Unidentified species were allocated to species for collision modelling according to the proportions of identified birds of each main taxonomic group on each survey. Westermost Rough Ornithological Monitoring

23 There were no surveys carried out during February, March and December, as had been agreed with NE and MMO in the design of the survey protocol, so for the collision modelling the mean densities from the previous and subsequent months were used as available. 5 MARINE MAMMALS The raw numbers of marine mammals recorded during each survey are shown in Table 7. There were occasional sightings of harbour porpoise, common seal and grey seal, but all were only seen in low numbers. Peak counts were 8 harbour porpoises, 2 common seals and 24 grey seals. Table 7. Survey area marine mammal counts during each of the September October 2013 surveys. Sep 12 Oct 12 Nov 12 Jan 13 Apr 13a Apr 13b Common seal Grey seal seal sp Harbour porpoise May 13a May 13b Jun 13 Jul 13a Jul 13b Aug 12 Sep 13 Oct 13 6 SPATIAL ANALYSIS OF KEY SPECIES DISTRIBUTIONS The survey data have been used to estimate bird numbers within the proposed wind farm site, 2km and 4km buffers around that site and for the reference area. Table 8 gives the peak populations estimated in each of these zones during the surveys, together with the overall mean density of each species in each zone. Table 8. Seabird population estimates within the proposed Westermost Rough wind farm footprint, 2km and 4km buffers and the reference area. Peak count Mean density (birds/km 2 ) WF WF +2km WF +4km Refere nce WF WF +2km WF +4km Refere nce Common scoter Red-throated diver diver sp Fulmar Sooty shearwater Manx shearwater Leach's Storm-petrel Gannet Cormorant Shag Pomarine Skua Arctic skua Great skua Common gull Westermost Rough Ornithological Monitoring

24 Peak count Mean density (birds/km 2 ) WF WF +2km WF +4km Refere nce WF WF +2km WF +4km Refere nce Lesser black-backed gull Herring gull Great black-backed gull Little gull 1,525 3,522 7,673 1, Black-headed gull Kittiwake 521 1,209 5, black-backed gull sp large gull sp gull sp Black tern Common tern Arctic tern Guillemot ,054 3, Razorbill 685 2,593 5,819 2, Little auk Puffin auk sp , COMPARISON OF BIRD NUMBERS WITH PREVIOUS SURVEYS Different sampling regimes makes direct comparison of the results of the surveys with those carried out for the ES difficult and no population estimates (that accounted for survey coverage and variation on detectability with distance) were reported in the ES, but further analysis has been carried out of the raw survey data to facilitate comparison between the surveys. As the extent of the surveys areas changed (with the surveys covering a wider 4km area around the wind farm rather than the 2km of the ES surveys, and a larger reference site), these comparisons are focussed on the wind farm plus 2km buffer and the reference area. The ES data were re-analysed following the same principles as described above for the data, applying correction factors to take into account survey coverage and variation in detectability of birds on the sea with distance from the survey vessel. As the data were more sparse from the ES surveys it was not possible to estimate reliable distance correction factors for each survey but rather the data were pooled for each main seabird group. The correction factors used (derived from the survey data) are shown in Table 9. Table 9. Distance correction factors used for the boat survey data. group A [0-50m] B [50-100m] C [ m] D [ m] Auks 100% 57% 13% 7% Gannet 100% 100% 25% 19% Divers 100% 100% 80% 60% Gulls 100% 75% 18% 16% As the factor for Band D was so low for auks (and hence unreliable) the analysis was restricted to Bands A-C (i.e. within 200m of the survey vessel) for birds on the sea. Westermost Rough Ornithological Monitoring

25 Table 10 compares the peak population estimates and mean densities recorded in the wind farm plus 2km buffer and the reference areas in the ES and the surveys. As in the surveys, there was high variability in the surveys, though generally rather lower bird numbers were recorded in , particularly for fulmar, gannet, kittiwake, guillemot and razorbill, but there were higher peak population estimates in for divers, Manx shearwater, great black-backed gull and puffin. Table 10. Seabird peak population estimates and mean densities in the wind farm site plus 2km buffer and the reference area in the ES surveys and in the surveys. Peak population Mean density WF Referen WF Referen WF Referen WF Referen +2km ce +2km ce +2km ce +2km ce Common scoter Red-throated Diver Diver sp Fulmar Sooty shearwater Manx Shearwater Gannet Cormorant Shag Arctic Skua Great Skua Common Gull Lesser Black-backed Gull Herring Gull Great Black-backed Gull 320 1, Little Gull 119 3,417 3,522 1, Black-headed Gull Kittiwake , Gull sp Black tern Sandwich Tern Common Tern Arctic tern Guillemot 796 1, , Razorbill 373 2,011 2,593 2, Little Auk Puffin 362 7, Auk sp , The high degree of inherent variability in the numbers of birds using the survey area will make detection of any effects of the wind farm difficult, as even major changes in numbers would Westermost Rough Ornithological Monitoring

26 be within the observed variation in numbers pre-construction. Gradient analysis may provide a more powerful tool to investigate such effects rather than just before/after (BACI) comparisons. 8 COLLISION RISK MODELLING 8.1 Collison Risk Modelling Methods Collision risk modelling has been undertaken for all of the very high, high and medium sensitivity species that have been recorded flying through the collision risk zone at rotor height. The collision risk model used in this assessment is the one developed by SNH and BWEA (Percival et al., 1999; Band, 2001, Band et al., 2007), recently updated for specific use for offshore wind farm assessments (Band, 2012). Details of the model are given in these publications. The model runs as a two-stage process. Firstly the risk is calculated making the assumption that flight patterns are unaffected by the presence of the wind turbines, i.e. that no avoidance action is taken. This is essentially a mechanistic calculation, with the collision risk calculated as the product of (i) the probability of a bird flying through the rotor swept area, and (ii) the probability of a bird colliding if it does so. This probability is then multiplied by the estimated numbers of bird movements through the wind farm rotors at the risk height (i.e. the height of the rotating rotor blades) in order to estimate the theoretical numbers at risk of collision if they take no avoiding action. The second stage then incorporates the probability that the birds, rather than flying blindly into the turbines, will actually take a degree of avoiding action, as has been shown to occur in all studies of birds at existing wind farms. SNH has recommended a precautionary approach, using a value of 98% as an avoidance rate for all of the species modelled here (Urquhart, 2010). Maclean et al. (2009) however recommended the use of more realistic rates (99%-99.9%) in their review for COWRIE. Results for a range of avoidance rates are presented here, as recommended by Band (2012). These rates relate to avoidance exhibited by birds once they are in proximity to the wind turbines, as they have been primarily derived from studies that have only been carried out post-construction, avoidance typically termed micro-avoidance. Avoidance of the wind farm site altogether through displacement, or macro-avoidance would therefore act in addition to these micro-rates for species that exhibited such behaviour. The collision model requires data on bird body size and flight speed. Body sizes and baseline mortality rates were taken from Robinson (2005), and flight speeds from Alerstam et al. (2007). Bird flight heights are taken from the field data where there was a sufficient number of records to generate a reliable estimate, but values from the SOSS-02 review project (Cook et al. 2012) are also considered to make the collision predictions as robust as possible. Details of the collision risk model input data and an example set of Band model spreadsheets (for gannet) are given in Appendix Collision Risk Modelling Results Table 11 summarises the collision risk analysis for each of the key species for the current 35- turbine layout. The Table gives the number of collisions predicted per year based on a range of avoidances rates (from the collision risk model), the percentage increase that each avoidance rate would represent over the baseline mortality, the magnitude of that effect and whether such an effect would be significant. Westermost Rough Ornithological Monitoring

27 Table 11: Collision risk modelling predictions for the Westermost Rough Offshore wind farm: 35 x 6MW turbines. Predicted number of collisions per year applying the following avoidance rates: 98% 99% 99.5% Magnitude of effect Gannet Negligible No Kittiwake Negligible No Common Gull Negligible No LBB Gull Negligible No Herring Gull Negligible No GBB Gull Negligible No Little Gull Negligible No Likely significant effect? For all species the predicted collision risk was so low that it would be clearly of negligible magnitude and not significant (as concluded in the ES and Ornithological Assessment Addendum). 8.3 Comparison with previous collision modelling results Table 12 compares the previous collision risk predictions presented in the ES (for a 50-turbine layout) and in the 2012 Ornithology Addendum (for the same 35-turbine layout as the current proposal). The updated collision risk predictions are higher than previous values for all species except common tern, though these differences are not sufficient to make any material differences to the conclusions reached in the previous assessments, i.e. that there would not be any significant collision risk resulting from the scheme. Table 12. Comparison of the ES and 2012 Addendum Collision Risk Modelling Results with those based on survey data. ES (50T) 2012 addendum (35T) This report Gannet Kittiwake Common Gull LBB Gull Herring Gull GBB Gull Common tern Little Gull Westermost Rough Ornithological Monitoring

28 9 SUMMARY AND CONCLUSION Fourteen further bird surveys of the Westermost Rough offshore wind farm site have been completed during the September 2012 October 2013 period, over a larger area than that surveyed previously for the ES baseline. Seabird populations within the survey area have continued to be highly variable, as was found in the ES surveys. Overall bird numbers were higher in , with a particularly notable peak of little gulls and more gannets recorded than previously. The predicted collision risks were as a result generally higher, but these differences were not sufficient to change the conclusion reached in the ES that there would not be any significant collision risk resulting from the scheme. The high degree of inherent variability in the numbers of birds using the survey area will make detection of any effects of the wind farm difficult, as even major changes in numbers would be within the observed variation in numbers pre-construction. Gradient analysis may provide a more powerful tool to investigate such effects rather than just before/after (BACI) comparisons. 10 REFERENCES Alerstam, T., Rosén, M., Bäckman, J., Ericson, P. & Hellgren, O. (2007) Flight speeds among bird species: allometric and phylogenetic effects. PLoS biology, 5. Band, W Estimating collision risks of birds with wind turbines. SNH Research Advisory Note. Band, W., M. Madders, and D. P. Whitfield. (2007). Developing field and analytical methods to assess avian collision risk at wind farms. In M. Lucas, de, G. F. E. Janss, and M. Ferrer, editors. Birds and Wind Farms. Quercus, Madrid. Band, W Using a Collision Risk Model to Assess Bird Collision Risks for Offshore Wind Farms. SOSS report. Buckland, S.T., Anderson, D.R., Burnham, K.P., Laake, J.L., Borchers, D.L., and Thomas, L. (2001). Introduction to Distance Sampling - Estimating abundance of biological populations Oxford University Press. Camphuysen, C. J., A. D. Fox, M. F. Leopold, and I. K. Petersen Towards standardised seabirds at sea census techniques in connection with environmental impact assessments for offshore wind farms in the UK: A comparison of ship and aerial sampling methods for marine birds, and their applicability to offshore wind farm assessments. COWRIE Report:3 9pp. Camphuysen, K.C.J. and Garthe, S. 2004: Recording foraging seabirds at sea. Standardised recording and coding of foraging behaviour and multi-species foraging associations. Atlantic Seabirds 6, Cook, A.S.C.P., Johnston, A., Wright, L.J. & Burton, N.H.K. (2012) Strategic Ornithological Support Services Project SOSS-02. A review of flight heights and avoidance rates of birds in relation to offshore wind farms. BTO Research Report Number 618. King, S., I. M. D. Maclean, T. Norman, and A. Prior Developing guidance on ornithological cumulative impact assessment for Offshore wind farm developers. COWRIE Ltd. Maclean, I. M. D., L. J. Wright, D. A. Showler, and M. M. Rehfisch A Review of Assessment Methodologies for Offshore Windfarms. British Trust for Ornithology report to COWRIE Ltd. Percival, S.M., Band, B. and Leeming, T Assessing the ornithological effects of wind farms: developing a standard methodology. Proceedings of the 21st British Wind Energy Association Conference Westermost Rough Ornithological Monitoring

29 Percival, S. M. (2007). Predicting the effects of wind farms on birds in the UK: the development of an objective assessment methodology. In M. de Lucas, Janss, G.F.E. and Ferrer, M., (eds). Birds and Wind Farms: risk assessment and mitigation. Quercus, Madrid. Robinson, R.A BirdFacts: profiles of birds occurring in Britain & Ireland (v1.1, Jan 2006). BTO Research Report 407, BTO, Thetford ( Stone, C.J., Webb, A., Barton, C., Ratcliffe, N., Reed, T.C., Tasker, M.L., Camphuysen, C.J. & Pienkowski, M.W. (1995). An atlas of seabird distribution in north-west European waters. JNCC. Thomas, L., Laake, J.L., Rexstad, E., Strindberg, S., Marques, F.F.C., Buckland, S.T., Borchers, D.L., Anderson, D.R., Burnham, K.P., Burt, M.L., Hedley, S.L., Pollard, J.H., Bishop, J.R.B. and Marques, T.A. (2009). Distance 6.0. Release 2. Research Unit for Wildlife Population Assessment, University of St. Andrews, UK. Urquhart, B Use of Avoidance Rates in the SNH Wind Farm Collision Risk Model. SNH Guidance Note. Westermost Rough Ornithological Monitoring