Changes to prey availability

858. Potential impacts on key prey species for razorbills breeding at St. Abb’s Head to Fast Castle SPA during the operation and maintenance phase have been assessed in volume 2, chapter 9 of the Offshore EIA Report using the appropriate maximum design scenarios for these receptors. Reduction or disruption to prey availability through temporary and long-term subtidal habitat loss/disturbance, increased SSC and deposition, reductions in water clarity, EMF from subsea electrical cabling, and colonisation of subsea structures, could affect razorbill survival and productivity in the St. Abb’s Head to Fast Castle SPA population.
859. As for guillemots, it is considered that there is relatively little potential for the St Abb’s Head to Fast Castle SPA razorbill population to be affected by changes to prey availability during the operation and maintenance phase, with any such effects being largely intermittent across a relatively small spatial extent, and with most effects temporary in nature. Consequently, it is considered that there is no potential for operational or maintenance related changes in prey availability to lead to an adverse effect on the St. Abb’s to Fast Castle SPA razorbill population. This conclusion is consistent with the outcome of the EIA which concluded that effects from changes in prey availability on razorbills during operation and maintenance were not significant in EIA terms (volume 2, chapter 11 of the Offshore EIA Report).

Project alone: population-level impacts

860. As determined above, the effects from the Proposed Development alone which could lead to an adverse effect on the St Abb’s Head to Fast Castle SPA razorbill population are limited to displacement (inclusive of barrier effects) during the operation and maintenance phase. For other effect pathways, there is considered to be no potential for an adverse effect on this population as a result of the Proposed Development alone, with any such effects likely to be small and of little, or no, consequence in terms of impacts at the population level.
861. Given this, PVA was undertaken on the mortality to the adult and immature age classes predicted due to the displacement effects associated with the Proposed Development, as determined by both the Scoping and Developer Approaches (see Table 5.37 above). The population model for the SPA population was a stochastic, density independent, matrix model, based upon the demographic parameters specified in Table 2.19 of volume 3, appendix 11.6 of the Offshore EIA Report. The starting population size was the 2016 – 2018 count for the SPA, with the projected population trends considered over a 35 year timescale (volume 3, appendix 11.5). The approach and methods to undertaking the PVA are as described for kittiwake above (with further details provided in the Offshore EIA Report, volume 3, appendix 11.6).
862. Outputs from the PVA are summarised according to the median predicted population-sizes at the end of the projection period, and the three metrics which the Scoping Opinion (volume 3, appendix 6.2 of the Offshore EIA Report) advised should be used for the interpretation of outputs and which have been shown to have relatively low sensitivity to factors such as varying population status and the mis-specification of the demographic rates underpinning the population model (Cook and Robinson 2015, Jitlal et al., 2017). These metrics are:

• The CPS – the median of the ratio of the end-point size of the impacted to un-impacted (or baseline) population, expressed as a proportion;
• The CPGR - the median of the ratio of the annual growth rate of the impacted to un-impacted population, expressed as a proportion; and
• The centile of the un-impacted population that matches the median (i.e. 50th centile) of the impacted population (based upon the distribution of the end-point population-sizes generated by the multiple replications of the model runs, the value should always be less than 50 because the median for the impacted population is not expected to exceed that for the un-impacted population).

Table 5.38: Projected 35 Year Population Sizes and Associated PVA Metrics for The St Abb’s Head to Fast Castle SPA Razorbill Population Under Different Impact Scenarios for The Proposed Development Alone

863. The PVA predicted that the St Abb’s Head to Fast Castle SPA razorbill population would increase over the 35 year projection period irrespective of the effects from the Proposed Development. Thus, the population is predicted to be six and five times larger than the current estimate of 3,928 adult birds under baseline conditions (i.e. no wind farm effects) and under the scenario of greatest annual mortality (i.e. Scoping Approach B), respectively (Table 5.38). Given that the PVAs are based on density independent models, which assume all mortality from the wind farm effects is additive and that there are no compensatory mechanisms operating within the population, the predicted increases are inevitably greatest for the baseline scenario and least for the scenario involving highest annual mortality (i.e. Scoping Approach B). Whilst the predicted levels of increase are unlikely to occur in reality (and are, in part, a consequence of the absence of any compensatory density dependence within the models – as discussed in the section on Project Alone: Population-level impacts for the SPA kittiwake population), the prediction for an increasing trend is broadly consistent with the documented, overall, long-term trend for this SPA population (Figure 5.16).
864. The PVA metrics suggest moderate differences in the predicted population-level impacts according to the Developer and Scoping Approaches. Thus, for the Developer Approach, the CPS value indicates that the displacement effects from the Proposed Development alone would result in a reduction of approximately 3% in the size of the SPA population after 35 years, relative to that in the absence of any wind farm effects (Table 5.38). The associated reduction in annual population growth rate (relative to that predicted under baseline conditions) is estimated to be 0.1%, whilst the centile value of 46.2 indicates a considerable overlap in the distributions of the predicted impacted and un-impacted population sizes and, hence, a high likelihood of the impacted population being of a similar size to the un-impacted population after 35 years.
865. For the Scoping Approach, the CPS values indicates a reduction of 8 – 14% in population size after 35 years, relative to that in the absence of any wind farm effects (Table 5.38). The reduction in annual population growth rate (relative to that predicted under baseline conditions) is estimated to be 0.2 – 0.4%. The centile metric indicates moderate to considerable overlap in the distribution of the predicted impacted and un-impacted population sizes, suggesting a reasonable likelihood of the impacted population being of a similar size to the un-impacted population after 35 years.
866. As for the assessment of the St Abb’s Head to Fast Castle SPA kittiwake population, the assessment of the SPA razorbill population incorporates high levels of precaution, which extend beyond the differences between the Developer and Scoping Approaches that are outlined above (and detailed in the Offshore EIA Report, volume 3, appendix 11.4). In the same way as for guillemot, of particular relevance to the razorbill assessment is the reliance on the seasonal mean peak abundances for the prediction of displacement mortality. For each of the defined seasonal periods, these estimates are 40 – 48% higher than the equivalent seasonal mean values (annex B of Offshore EIA Report, volume 3, appendix 11.4), whilst both of the breeding season peak counts coincide with the latter part of this period (i.e. July and August) and may be affected by post-breeding dispersal. Thus, the seasonal mean peak values are unlikely to be representative of the usage of the Proposed Development Array and two kilometre buffer by birds from nearby breeding colonies, and may grossly overestimate this. This is further supported by modelling of the foraging distributions of SPA razorbills (using data derived from tracking during the chick-rearing period) which indicates that the Proposed Development array area is beyond core areas of usage (Cleasby et al. 2018).
867. The reliance on PVAs which are based upon density independent populations models is also likely to cause overestimation of the population-level impacts and give overly precautionary outputs, for the same reasons as outlined above for kittiwake.

Project alone: conclusion

868. Based on the Developer Approach, the potential effects from the Proposed Development alone on the St Abb’s Head to Fast Castle SPA razorbill population are predicted to be small, as are the resultant population-level impacts. In addition, the PVA metrics indicate a high chance of the population being of a similar size to that which would occur in the absence of the Proposed Development after 35 years. Given that the SPA population has shown an overall, long-term, increase in size and is considered to be in ‘favourable maintained’ condition, it is concluded that the effects from the Proposed Development alone would not result in an adverse effect on this population.
869. As would be expected, the Scoping Approach predicts greater levels of effects and consequent population-level impacts than as predicted by the Developer Approach. However, the predicted levels of impact remain relatively small, with the PVA metrics for the Scoping Approach also indicating that there remains a reasonable likelihood of the impacted population being of a similar size to the un-impacted population after 35 years. When this is considered within the context of the current and longer-term status of this SPA population, as well as the highly precautionary basis for the assessment, it is concluded that the effects from the Proposed Development alone (as determined by the Scoping Approach) would not result in adverse effect on the population.

Effects in-combination

Effects of relevance to the in-combination assessment

870. As detailed above, any effects from the Proposed Development alone on the St Abb’s Head to Fast Castle SPA razorbill population during construction and decommissioning and resulting from disturbance and changes to prey availability during operation and maintenance will be small and highly localised. As such, there is considered to be no potential for these effect pathways to add to impacts at the population-level that might result from other effects pathways associated with the Proposed Development or from the effects due to other plans and projects.
871. Therefore, the potential for effects of the Proposed Development to act on the St Abb’s Head to Fast Castle SPA razorbill population in-combination with other plans and projects is limited to the displacement (inclusive of barrier effects) effect pathway during operation and maintenance. Following advice from NatureScot provided through the Ornithology Roadmap process (at meeting 3, 8th December 2021), the following sections consider these potential effects for (i) the Proposed Development in-combination with the other Forth and Tay wind farms and (ii) the Proposed Development in-combination with the other UK North Sea wind farms (noting that scenario (ii) includes those plans and projects which comprise scenario (i)).

Displacement/barrier effects – operation and maintenance

872. As described in volume 3, appendix 11.6, annex E of the Offshore EIA Report estimates of breeding season displacement mortality which had been attributed to the St Abb’s Head to Fast Castle SPA razorbill population were extracted from the existing assessments for offshore wind farms that are in planning, consented, under construction or in operation. As for the potential displacement mortality estimated for the Proposed Development, the mortality attributed to the SPA population from other offshore wind farms was estimated using the SNCB matrix approach, with details on the displacement and mortality rates that had been applied being available in each case. Thus, it was possible to adjust the estimated mortalities from each of the other projects to align with the displacement and mortality rates on which the Scoping and Developer Approaches are based.
873. For the non-breeding periods, razorbill numbers associated with other offshore wind farms that are in planning, consented, under construction or in operation were extracted for each of the relevant seasonal periods from the cumulative totals collated for the East Anglia TWO and East Anglia ONE North submissions (MacArthur Green and Royal HaskoningDHV 2021; see volume 3, appendix 11.6, annex E of the Offshore EIA Report for more details). The cumulative numbers for each of the non-breeding periods were apportioned to the St Abb’s Head to Fast Castle SPA razorbill population according to the BDMPS approach (Furness 2015), with the subsequent displacement mortality calculated according to the displacement and mortality rates appropriate to each of the Scoping and Developer Approaches (Table 5.37). This was done separately for all of the other UK North Sea wind farms and for the subset represented by the other Forth and Tay wind farms.

Table 5.39: Estimated Annual Mortality of St. Abb's Head to Fast Castle SPA Razorbills as a result of Displacement from the Proposed Development Array Area and 2 km Buffer as Determined by the Scoping Approach and Developer Approach, In-Combination with Other Forth and Tay Wind Farms and UK North Sea Wind Farms

874. The potential mortality resulting from the predicted displacement effects associated with the other plans and projects is smaller than that predicted for the Proposed Development alone but, nonetheless, represents a notable increase to the project alone estimates (Tables 5.37 and 5.39). Thus, inclusion of the other Forth and Tay wind farms increases the predicted displacement mortality of adult birds by approximately 17% for the Developer Approach and 20 - 23% for the Scoping Approach compared to the Proposed Development alone. The analogous increases for the Proposed Development in-combination with the other UK North Sea wind farms, relative to the Proposed Development alone, are 42% for the Developer Approach and 63 – 88% for the Scoping Approach.
875. For the Proposed Development in-combination with the other Forth and Tay wind farms, the additional annual mortality of adult razorbills from the St Abb’s Head to Fast Castle SPA population predicted due to displacement represents 0.08% of the current adult breeding population at this colony (i.e. 3,928 individuals – Table 3.3 in volume 3, appendix 11.5 of the Offshore EIA Report) as determined by the Developer Approach, and between approximately 0.2 – 0.4% of this population as determined by the lower and upper estimates from the Scoping Approach. In terms of percentage increases in the baseline annual adult mortality of the population (which is based on applying a mortality rate of 0.090 – see Table 2.19 in the volume 3, appendix 11.6 of the Offshore EIA Report), the estimates of adult displacement mortality equate to an increase of 0.9% for the Developer Approach and of 2.7 – 5.0% for the lower and upper estimates from the Scoping Approach.
876. For the Proposed Development in-combination with the other UK North Sea wind farms, the additional annual mortality of adult razorbills from the St Abb’s Head to Fast Castle SPA population predicted due to displacement represents 0.1% of the current adult breeding population at this colony as determined by the Developer Approach, and between approximately 0.3 – 0.6% of the current adult breeding population at this colony as determined by the lower and upper estimates from the Scoping Approach. In terms of percentage increases in the baseline annual adult mortality of the population, the estimates of adult displacement mortality equate to an increase of 1.4% for the Developer Approach and of 3.3 – 6.6% for the lower and upper estimates from the Scoping Approach.
877. The potential levels of impact on the St Abb’s Head to Fast Castle SPA razorbill population resulting from the mortality predicted from displacement and barrier effects associated with the Proposed Development in-combination with other wind farms in the Forth and Tay or in-combination with other wind farms in the UK North Sea during the operation and maintenance phase are considered in more detail below in the Effects In-Combination: Population-Level Impacts section. This presents the outputs from PVAs of the potential effects of predicted displacement mortality on the SPA population.

In-combination: population-level impacts

878. As for the Proposed Development alone, PVA was undertaken on the mortality to the adult and immature age classes predicted due to the combined displacement and collision effects associated with the Proposed Development in-combination with the other Forth and Tay wind farms and the Proposed Development in-combination with the other North Sea wind farms. This was on the basis of the potential mortality as determined by both the Scoping and Developer Approaches (see Table 5.39 above).
879. The approach to the PVA and the metrics used to summarise the PVA outputs are as described for the Proposed Development alone (see Project Alone: Population-Level Impacts section above).

Table 5.40: Projected 35 year Population Sizes and Associated PVA Metrics for the St Abb’s Head to Fast Castle SPA Razorbill Population Under Different Impact Scenarios for the Proposed Development In-Combination with the Other Forth and Tay Wind Farms

Table 5.41: Projected 35 year Population Sizes and Associated PVA Metrics for the St Abb’s Head to Fast Castle SPA Razorbill Population Under Different Impact Scenarios for The Proposed Development In-Combination with the Other UK North Sea Wind Farms

880. Given that the in-combination effects are inevitably greater than those for the Proposed Development alone, the PVA metrics for the Proposed Development in-combination with either the other Forth and Tay wind farms or the other UK North Sea wind farms suggest greater population-level impacts than as predicted for the Proposed Development alone (compare Table 5.38 with Tables 5.40 and 5.41). However, for the Proposed Development in-combination with the other Forth and Tay wind farms the changes in the values of the PVA metrics are small and it is considered that the conclusions reached for the Proposed Development alone are also applicable to this in-combination scenario.
881. For the Proposed Development in-combination with the other UK North Sea wind farms, the CPS value for the Developer Approach indicates that the SPA population size would be reduced by approximately 5% relative to that in the absence of any wind farm effects (Table 5.41). The reduction in annual population growth rate (relative to that predicted under baseline conditions) remains small, whilst the centile value continues to indicate a considerable overlap in the distributions of the predicted impacted and un-impacted population sizes and, hence, a high likelihood of the impacted population being of a similar size to the un-impacted population after 35 years.
882. The metrics associated with the Scoping Approach for the Proposed Development in-combination with the other UK North Sea wind farms inevitably suggest greater levels of effect. However, at the lower range of effects (i.e. Scoping Approach A) they continue to indicate a relatively small effect and a reasonable likelihood of the impacted population being of a similar size to the un-impacted population after 35 years. At the upper range of effects (i.e. Scoping Approach B), the CPS value indicates that the SPA population size would be reduced by approximately 21% relative to that in the absence of any wind farm effects (Table 5.41), whilst the reduction in annual population growth rate (relative to that predicted under baseline conditions) is estimated as 0.6%. The centile value of 22.4 suggests a reasonably high likelihood of the impacted population being smaller than the un-impacted population after 35 years.

In-combination: conclusion

883. On the basis of the Developer Approach, it is considered that the potential effects from the Proposed Development in-combination with the other Forth and Tay wind farms or the other UK North Sea wind farms would not result in an adverse effect on the St Abb’s Head to Fast Castle SPA razorbill population. The population-level impacts predicted to arise from these in-combination effects represent a small increase to those predicted due to the Proposed Development alone. As such, it is considered that the conclusions reached in relation to the Proposed Development alone are also valid for the in-combination scenarios.
884. For the Scoping Approach, the above conclusion is also considered to be valid for the Proposed Development in-combination with the other Forth and Tay wind farms. In terms of the Proposed Development in-combination with the other UK North Sea wind farms, it is considered that the lower range of the predicted impacts would not represent an adverse effect on the SPA population but that it is possible the upper range would. Consequently, it is concluded that the effects of the Proposed Development in-combination with the other UK North Sea wind farms could result in an adverse effect on the St Abb’s Head to Fast Castle SPA razorbill population.

Assessment for the breeding seabird assemblage

885. The breeding seabird assemblage for the St Abb’s Head to Fast Castle SPA is a qualifying feature on the basis of the SPA supporting 79,560 individual seabirds, including guillemot, razorbill, shag, kittiwake and herring gull.
886. Potential impacts of the Proposed Development alone and in-combination with either the other Forth and Tay or the other UK North Sea wind farms on the breeding seabird assemblage for the SPA could arise via effects on the individual species within the assemblage feature. For the Developer Approach, the assessments undertaken above identify the potential for an adverse effect on the SPA kittiwake population in relation to both the project alone and in-combination scenarios. For the Scoping Approach, the assessments undertaken above identify the potential for adverse effects on the SPA kittiwake and guillemot populations in relation to both the project alone and in-combination scenarios, and on the SPA razorbill population in relation to the Proposed Development in-combination with the other UK North Sea wind farms.
887. It is considered that the predicted impacts on the SPA kittiwake population (for both the Developer and Scoping Approaches) are sufficient to represent an increased risk of this population being lost from the breeding seabird assemblage. This is due to the relatively small size of this population combined with its long-term decline. For the Scoping Approach (but not the Developer Approach), it is also considered to be conceivable that the scale of the predicted impacts on the SPA kittiwake, guillemot and razorbill populations are such as to represent a risk of reducing the total number of individual seabirds present in the assemblage to a level that could represent an adverse effect on this qualifying feature. This is particularly relevant to the predicted impacts on the guillemot population because of the large size of this population (and hence its importance to maintaining the breeding seabird assemblage qualifying feature). This conclusion should be considered within the context of the high levels of precaution incorporated within the assessment, with these being outlined above in the sections on each of the named components of the St Abb’s Head to Fast Castle SPA breeding seabird assemblage.
888. Given the above, it is concluded that there is the potential for an adverse effect on the St Abb’s Head to Fast Castle SPA breeding seabird assemblage, both in relation to the Proposed Development alone and in-combination with (i) the other Forth and Tay wind farms and (ii) the other UK North Sea wind farms. This conclusion applies to the assessments undertaken according to both the Developer Approach and the Scoping Approach.

Site conclusion

Developer approach

889. It is concluded that the possibility of an adverse effect cannot be discounted for the St Abb’s Head to Fast Castle SPA population of breeding kittiwake (noting this species is a named component of the seabird assemblage feature only), as well as the breeding seabird assemblage qualifying feature (due to the impacts on the kittiwake component only). For the kittiwake population, the potential for an adverse effect arises from the Proposed Development alone and the Proposed Development in-combination with either (i) the other Forth and Tay wind farms or (ii) the other UK North Sea wind farms. The potential for an adverse effect on the breeding seabird assemblage is a direct consequence of the potential effects on the SPA kittiwake population, which is a named component of this assemblage.
890. Consequently, it is concluded that an Adverse Effects on Integrity of the St Abb’s Head to Fast Castle SPA cannot be excluded.

Scoping approach

891. It is concluded that the possibility of adverse effects cannot be discounted for the St Abb’s Head to Fast Castle SPA populations of breeding kittiwake, guillemot and razorbill (noting these species are named components of the seabird assemblage feature only), as well as the breeding seabird assemblage qualifying feature (due to the impacts on kittiwake, guillemot and razorbill components only). For the kittiwake and guillemot populations, the potential for adverse effects arises from the Proposed Development alone and the Proposed Development in-combination with either (i) the other Forth and Tay wind farms or (ii) the other UK North Sea wind farms. For the razorbill population, the potential for an adverse effect is in relation to the effects of the Proposed Development in-combination with the other UK North Sea wind farms (but not to the effects of the Proposed Development alone or the Proposed Development in-combination with the other Forth and Tay wind farms). The potential for an adverse effect on the seabird assemblage is a direct consequence of the potential effects on these SPA populations, all of which are named components of this assemblage.
892. Consequently, it is concluded that an Adverse Effects on Integrity of the St Abb’s Head to Fast Castle SPA cannot be excluded.

5.7.2.    Forth Islands SPA

European site information and conservation objectives

893. The Forth Islands SPA comprises seabird colonies on multiple islands in the Firth of Forth, southeast Scotland. The SPA is approximately 36 km from the Proposed Development array area and 14 km from the Proposed Development export cable corridor. The Isle of May is the closest of the islands within the SPA to the Proposed Development, with the other islands in the SPA being Inchmickery, Fidra, The Lamb, Craigleith, Bass Rock (all of which were classified in April 1990) and Long Craig (which was an extension to the site and was classified in February 2004). The SPA is underpinned by the following Sites of Special Scientific Interest (SSSIs): Long Craig, Inchmickery, Forth Islands, Bass Rock and the Isle of May. There is a seaward extension from each island of the SPA extending approximately 2 km into the marine environment.
894. There are four Annex I qualifying features and the site qualifies under Article 4.2 by regularly supporting four migratory seabird species and in excess of 20,000 breeding seabirds, including five named component species (Table 5.42). The potential for LSE has been identified in relation to nine of these 13 species, with the effect pathways associated with LSE for each of these detailed in Table 3.1 and set out in the assessment below.
895. The conservation objectives of this SPA (as determined from NatureScot’s SiteLink (SiteLink (nature.scot)) are:

• To avoid deterioration of the habitats of the qualifying species (listed below) or significant disturbance to the qualifying species, thus ensuring that the integrity of the site is maintained; and
• To ensure for the qualifying species that the following are maintained in the long term:

• Population of the species as a viable component of the site
• Distribution of the species within site
• Distribution and extent of habitats supporting the species
• Structure, function and supporting processes of habitats supporting the species
• No significant disturbance of the species
  896. Further information on this European site is presented in appendix 3A.

Table 5.42: Details on the Qualifying Features of the Forth Islands SPA

*Named components of the assemblage only.

Assessment for the gannet population

897. Gannets only occur in the North Atlantic, nesting at relatively high latitudes and wintering south of their breeding sites. Most gannets nest in the eastern Atlantic, with the majority (60 – 70 per cent) of birds breeding in colonies around Great Britain. Other gannet colonies occur in France, Ireland, Norway, Faroe Islands and Iceland. Gannets forage entirely at sea on fish, including discards from fishing boats, and have large foraging ranges when breeding (Woodward et al. 2019). Gannets from the Forth Islands SPA forage across a large portion of the North Sea (Lane et al. 2020), though their foraging range shows little overlap with those of other colonies (Wakefield et al. 2013).
898. The largest gannet colony in the world occurs on Bass Rock, in the Forth Islands SPA (Murray et al. 2014). Gannet populations, including on the Bass Rock, have increased substantially through the 20th and 21st centuries, with expansion at existing colonies and the development of new colonies occurring (Mitchell et al. 2004, Murray et al. 2015). There are indications that the colony on Bass Rock is close to carrying capacity, with substantive increases having occurred over the last few decades and further increase likely limited by the availability of suitable nesting areas on the island (Figure 5.17, Murray et al. 2015). The population size has remained above the citation population size since designation.

Figure 517: Gannet Population Trend at the Forth Islands SPA for the Period 1985 - 2014. The Red Line Shows the Citation Population Size for the SPA (21,600 pairs). Data are from The Seabird Monitoring Programme Database (Seabird Monitoring Programme | JNCC (bto.org))

The potential for impacts on the gannet population

899. The Proposed Development and two kilometre buffer around the Proposed Development array area7 do not overlap with the Forth Islands SPA, so that potential impacts on its gannet population will only occur as a result of individuals from the colony occurring in the area (or vicinity) of the Proposed Development. Consequently, the main focus of the assessment for this SPA population is concerned with the Conservation Objective to maintain, in the long term, the population of the species as a viable component of the site because the other conservation objectives either apply to the site itself, and not to areas beyond the boundary, or are encompassed by the assessment of this first Conservation Objective (as for the maintain in the long term no significant disturbance of the species, because disturbance would only be considered significant if it caused an adverse effect on the population viability of the qualifying features).
900. From published information on gannet foraging ranges generally (Woodward et al. 2019) and tracking from the SPA specifically (Wakefield et al. 2013, Lane et al. 2020, volume 3, appendix 11.4, annex E of the Offshore EIA Report), it is apparent that during the breeding period gannets from the Forth Islands SPA occur within the area of the Proposed Development and of the two km buffer around the Proposed Development array area. This is reflected in the findings of the apportioning exercise, which estimates that 97% of the gannets occurring on the Proposed Development array area during the breeding season derive from this SPA colony (Offshore EIA Report, volume 3, appendix 11.5). The breeding period for gannet is defined as mid-March to September, following the NatureScot (2020) guidance.
901. Gannets from the Forth Islands SPA move south in autumn to winter at sea from the Bay of Biscay to the seas off west Africa, returning north in the spring (Fort et al. 2012), so that the non-breeding season is divided into autumn and spring passage periods (defined as October to November and December to mid-March, respectively, on the basis of applying the BDMPS defined periods within the context of the overall non-breeding period defined by NatureScot – Furness 2015, NatureScot 2020, volume 3, appendix 11.5 of the Offshore EIA Report). Given the above, the Proposed Development may have potential effects on the Forth Islands SPA gannet population during breeding and non-breeding periods.

Project alone: construction and decommissioning

Disturbance

902. Direct disturbance to gannets during the construction phase may arise within the Proposed Development array area (and its immediate vicinity) as a result of increased vessel movements and helicopter activity, as well as from other activities directly associated with the installation of the wind turbine foundations and other infrastructure, whilst there will also be increased vessel activity along the Proposed Development export cable corridor due to the cable laying activities. The levels of such activities that could arise are outlined in Table 4.1, with these activities occurring during construction campaigns within a construction period of at most eight years duration.
903. A total of up to 11,482 vessel round trips may occur over the construction phase, whilst it is estimated that a maximum of 134 vessels could occur within the area of the Proposed Development at any one time (Table 4.1). However, this is within the context of high baseline levels of vessel traffic within this area (e.g. surveys recorded an average of 14 vessels per day within a 10 nm buffer around the Proposed Development over a 14-day period in August 2022, whilst also showing an average of three to four vessels intersecting the Proposed Development array area per day over summer - Offshore EIA Report, volume 2, chapter 13).
904. When using the marine environment (and not at the breeding colony), gannets are considered to have a relatively low sensitivity to such sources of direct disturbance. Thus, reviews of the sensitivity of different seabird species to disturbance from vessels and helicopter traffic assign gannet as ‘2’ on a five-scale ranking system, where 1 indicates hardly any or limited escape/avoidance behaviour and very short flight distance when approached and 5 indicates strong escape/avoidance behaviour and a large response distance (Garthe and Hüppop 2004, Furness et al. 2013).
905. The total area to be affected by such disturbance over the full eight years of the construction phase also represents a small proportion of the total area of marine habitat available to gannets from the Forth Islands SPA. Thus, the Proposed Development array area encompasses 1,010 km2, whilst the Proposed Development export cable corridor encompasses 168 km2. Together these areas represent less than 0.5% of the total breeding season foraging area that is potentially available to the SPA gannet population, as defined by the generic measure of the species’ mean maximum breeding season foraging range plus 1 SD (i.e. 309.2±194.2 km - Woodward et al. 2019) and assuming that this range is represented by a semicircle to the main seaward side of the colony. Similarly, the Proposed Development Array and export cable corridor represent less than 1% of the breeding season foraging area if considering the mean maximum foraging range only. More detailed analyses based on the available tracking data for the Forth Islands SPA gannet population demonstrate that during the incubation and chick-rearing periods, the Proposed Development Array represents only 0.7% of the full home range used by the tracked birds, whilst only 26% of all tracks entered the Proposed Development array area (volume 3, appendix 11.4, annex E of the Offshore EIA Report).
906. During the autumn and spring passage periods, the potential for effects of construction-related disturbance is lower than during the breeding season because the SPA gannets are essentially transiting through the waters within which the Proposed Development is located.
907. In addition, it is important to consider that the construction activities will not occur simultaneously across the entirety of the Proposed Development array area or the Proposed Development export cable corridor but, rather, will be carried out in different areas at different times. Thus, the activities will be concentrated within discrete (often small) parts of these wider areas, and within such areas they will not extend over the full duration of the construction phase, so further reducing the potential to which birds may be subject to disturbance effects. For example, cable laying for the Proposed Development export cable will occur over a total of two years, whilst within the Proposed Development array area it is likely that construction activities would be confined largely to discrete areas at any one time.
908. The potential for disturbance effects during decommissioning is assumed to be the same (or less) as for construction, noting that the duration of the decommissioning phase will not exceed that of construction, and may be shorter.
909. Given the low sensitivity of gannet to disturbance effects, the relatively small areas that will be subject to activities with the potential to result in disturbance at any given time during the construction period and the fact that these potential effects will be temporary, it is considered that there is no potential for construction or decommissioning related disturbance to lead to an adverse effect on the Forth Islands SPA gannet population.

Displacement

910. As detailed above, gannet is considered to have a low sensitivity to disturbance, whilst potential effects of disturbance during the construction and decommissioning phases will only extend across a small part of the wider foraging areas used by the Forth Islands SPA gannet population and be limited to, at most, an eight year period during construction (and a likely similar or shorter period during decommissioning). Furthermore, as detailed above, potential effects of disturbance will not occur simultaneously across the entirety of the Proposed Development array area or Proposed Development export cable corridor but, rather, will be carried out in different areas at different times. Thus, at any given time the potential for disturbance effects that could lead to displacement of gannet from this SPA will be limited to relatively small areas, with the potential effects also being of a temporary nature.
911. Therefore, based upon the above, it is considered that there is relatively little potential for the Forth Islands SPA gannet population to be affected by displacement during the construction or decommissioning phases, with any such effects only extending across relatively small areas and tending to be temporary in nature. Consequently, it is considered that there is no potential for construction or decommissioning related displacement to lead to an adverse effect on the Forth Islands SPA gannet population.

Changes to prey availability

912. Gannets predominantly prey upon fish including herring, mackerel, sprat and sandeel, as well as fishery discards (del Hoyo et al., 1996). Indirect effects on gannets may arise as a result of changes in the availability, distribution, or abundance of these species during the construction and decommissioning phases of the Proposed Development. Reduction or disruption to prey availability may cause displacement from foraging grounds or reduced energy intake, affecting survival rates or productivity in the Forth Islands SPA gannet population in the short-term.
913. During construction there are a number of ways in which effects on gannet prey species could occur, which are for the same reasons as outlined in the section on Project Alone: Construction and Decommissioning – Changes to prey availability for the St Abb’s Head to Fast Castle SPA kittiwake population and in volume 2, chapter 9 of the Offshore EIA Report. However, the Proposed Development array area and export cable corridor represent less than 0.5% of the total breeding season foraging area that is potentially available to the SPA gannet population, as defined by the species’ mean-maximum breeding season foraging range plus 1 SD (i.e. 309.2±194.2 km; Woodward et al., 2019) and assuming that this range is represented by a semicircle to the main seaward side of the colony. Effects during the autumn and spring passage periods are considered to be lower than during the breeding season given that birds disperse widely through UK waters to their wintering grounds (Kubetski et al., 2009; Furness 2015).
914. During decommissioning, the effects from changes in prey availability are considered to be the same (or less) as for construction. It is currently unclear as to how the presence, and subsequent removal of, subsea structures may affect gannet prey species (Birchenough and Degrae 2020; Scott, 2022). It is possible that prey abundance could decline from the levels present during the operation and maintenance period. This could occur if the sub-surface structures associated with the Proposed Development in the marine environment lead to an increase in key prey abundance within the Proposed Development array area and export cable corridor via the provision of artificial reef habitats. However, some infrastructure (such as scour and cable protection) is assumed to be left in situ with the impact of colonisation of infrastructure continuing in perpetuity following decommissioning. Thus, any reduction in prey abundance through removal of foundations is likely to be very small relative to the area over which breeding and non-breeding gannets forage.
915. Given their wide-ranging foraging behaviour and degree of plasticity in diet (del Hoyo et al., 1996), together with any effects being intermittent, spatially-restricted and temporary in nature, it is considered that there is no potential for construction or decommissioning related changes in prey availability to lead to an adverse effect on the Forth Islands SPA gannet population. This conclusion is consistent with the outcome of the EIA which concluded that effects from changes in prey availability on gannets during construction and decommissioning were not significant in EIA terms (volume 2, chapter 11 of the Offshore EIA Report).

Project alone: operation and maintenance

Disturbance

916. Vessel use within the Proposed Development array area and export cable corridor during the operation and maintenance phase may lead to direct disturbance of gannets. The maximum design scenario is for up to 3,393 vessel round trips per year over the operational lifetime of the project (Table 4.1).
917. Based on information presented in the Offshore EIA Report, volume 2, chapter 13, baseline levels of vessel traffic in the Offshore Ornithology study area are relatively high (see section on Project Alone: Operation and Maintenance – Disturbance for the St Abb’s Head to Fast Castle SPA kittiwake population). In the context of the baseline levels of vessel traffic across the Offshore Ornithology study area, the increase during the operation and maintenance phase is considered to be relatively small. Vessel movements will be within the Proposed Development array area and export cable corridor and will follow existing shipping routes to/from ports. In addition, Project Codes of Conduct included as a part of the NSVMP (Offshore EIA Report, volume 4, appendix 25) will be issued to all project vessel operators to avoid sudden changes in course or speed which will minimise the potential for disturbance.
918. The size and noise outputs from vessels during the operation and maintenance phase will be similar to those used in the construction phase. However, the number of vessel return trips per year and their frequency will be much lower than during the construction phase. In addition, activities during the operation and maintenance phase will not occur simultaneously across the entirety of the Proposed Development array area and offshore export cable corridor but intermittently within discrete (often small) parts of these wider areas.
919. Given the low sensitivity of gannet to disturbance effects at sea (Garthe and Hüppop 2004; Furness et al., 2013), the relatively small areas relative to the species’ foraging range that will be subject intermittently to potentially disturbing activities (Woodward et al., 2019), and the fact that these potential effects will be reduced compared to the construction and decommissioning phases, it is considered that there is no potential for disturbance during operation and maintenance to lead to an adverse effect on the Forth Islands SPA gannet population. This conclusion is consistent with the outcome of the EIA which ‘screened’ out gannet as a species for which detailed consideration of the effects of construction disturbance was required (volume 2, chapter 11 of the Offshore EIA Report).

Displacement/barrier effects

920. As outlined above, displacement effects on the Forth Islands SPA gannet population are estimated using the SNCB matrix approach, as applied to the Proposed Development Array and two kilometre buffer (SNCBs 2022, Offshore EIA Report, volume 3, appendix 11.4). Thus, throughout this section, mortality from displacement is assumed to refer to that which results from both displacement and barrier effects. The approach used to derive predicted levels of mortality is as described in the section on Project Alone: Operation and Maintenance – Displacement/Barrier Effects for the St Abb’s Head to Fast Castle SPA kittiwake population above (and in the Offshore EIA Report, volume 3, appendix 11.4).
921. On the basis of the advice provided in the Scoping Opinion (volume 3, appendix 6.2 of the Offshore EIA Report), displacement effects on gannet are estimated for the breeding period and each of the autumn and spring passage periods. The displacement and associated mortality rates advised in the Scoping Opinion (subsequently termed the Scoping Approach) for gannet are:

• Breeding period: 70% displacement with lower and upper mortality rates of 1% and 3%.
• Non-breeding periods: 70% displacement with lower and upper mortality rates of 1% and 3%.
  922. As with other species for which displacement effects are assessed (see above), the approach to estimating gannet displacement effects advocated by the Scoping Opinion was considered overly precautionary. For gannet, this was specifically concerned with the upper range of the proposed mortality rates, and the evidence available to support this (volume 3, appendix 11.4, annex G of the Offshore EIA Report). Furthermore, the mortality rates advocated by the Scoping Opinion represented a marked change from the assumptions applied in assessments for other recent Scottish offshore wind farms (Marine Scotland 2017a,b,c) with no clear evidence apparently being available to justify such a change. Thus, based on a consideration of the available evidence for gannet displacement, the extent of the species’ ranging behaviour, previous precedent and a need to incorporate precaution within the assessment, an alternative Developer Approach to estimating displacement effects was determined (volume 3, appendix 11.4, annex G of the Offshore EIA Report). The rates adopted by the Developer Approach (for both displacement and consequent mortality) are as for the lower range of the Scoping Approach (i.e. 70% displacement and 1% mortality in for all seasonal periods).
  923. Estimates of gannet mortality were produced using the SNCB matrix on the basis of both the Scoping Approach and the Developer Approach (Offshore EIA Report, volume 3, appendix 11.4), with these estimates then apportioned to the Forth Islands SPA gannet population during the breeding and non-breeding periods according to the NatureScot (2018) approach and the BDMPS approach (Furness 2015), respectively (Offshore EIA Report, volume 3, appendix 11.5, Table 5.43). The resulting mortality estimates for the breeding period were apportioned to age classes on the basis of the plumage characteristics of gannets recorded during the breeding period in the baseline surveys (Offshore EIA Report, volume 3, appendix 11.1), whilst for the non-breeding periods age classes were apportioned according to the stable age distributions of the population model used in Furness (2015). Based on advice provided by NatureScot and Marine Scotland Science following Roadmap Meeting 4 (G. Holland, email 26/01/2022), it was also assumed that 10% of the breeding adults in the SPA population miss breeding in any given year (i.e. sabbatical birds) so that the number of estimated adult deaths during the breeding season was adjusted accordingly.

Table 5.43: The Mean Peak Abundance Estimates of Gannet in the Proposed Development Array Area and 2 km Buffer for Each Seasonal Period, Together with the Proportion of Birds Estimated to Belong to the Breeding Adult Age Class and to be from the Forth Islands SPA Population in Each Period. The Proportion of Adults Assumed to be Sabbaticals During the Breeding Season is also Presented

924. Based upon the estimates and assumptions detailed above, the potential annual mortality amongst the SPA gannet population as a result of displacement is estimated as 32 adult and two immature birds based on the Developer Approach and the lower mortality rates for the Scoping Approach (i.e. Scoping Approach A) and as 94 adult and six immature birds based upon the higher mortality rates for the Scoping Approach (i.e. Scoping Approach B) (Table 5.44). As expected on the basis that gannets from this breeding colony SPA use the waters within the vicinity of the Proposed Development array area during the breeding season (and as reflected by the seasonally-specific apportioning rates), the predicted displacement effects are largely attributable to the breeding season (with the potential breeding season mortality accounting for almost 90% of the overall annual mortality, irrespective of whether this is determined by the Developer or Scoping Approach – Table 5.44).

Table 5.44: Estimated Potential Annual Mortality of Forth Islands SPA Gannets as a Result of Displacement from the Proposed Development Array Area and 2 km Buffer as Determined by the Scoping Approach and Developer Approach

925. The additional annual mortality of adult gannets from the Forth Islands SPA population predicted due to displacement as a result of the Proposed Development array represents 0.02% of the current adult breeding population at this colony (i.e. 150,518 individuals – Table 3.3 in volume 3, appendix 11.5 of the Offshore EIA Report) as determined by the Developer Approach, and between approximately 0.02 – 0.06% of this population as determined by the lower and upper estimates from the Scoping Approach. In terms of percentage increases in the baseline annual adult mortality of the population (which is based on applying a mortality rate of 0.046 – see Table 2.7 in volume 3, appendix 11.6 of the Offshore EIA Report), the estimates of adult mortality equate to an increase of approximately 0.5% for the Developer Approach and of 0.5 – 1.4% for the lower and upper estimates from the Scoping Approach.
926. The potential levels of impact on the Forth Islands SPA gannet population resulting from the mortality predicted from displacement and barrier effects associated with the Proposed Development array during the operation and maintenance phase are considered in more detail below in the Project Alone: Population-Level Impacts section. This presents the outputs from PVAs of the combined effects of predicted displacement and collision mortality on the SPA population.

Collision risk

927. Predictions of the number of gannets at risk from collisions due to the Proposed Development were calculated using the deterministic version of the SOSS offshore collision risk model (Band 2012, Offshore EIA Report, volume 3, appendix 11.3). Following the Scoping Opinion (volume 3, appendix 6.2 of the Offshore EIA Report), the assessment is based on option 2 of the CRM, which uses the generic flight height data from Johnston et al. (2014a,b) and assumes a uniform distribution of flight heights across the rotor swept zone (as opposed to using the modelled flight height distribution) (Band 2012). An avoidance rate of 98.9% was applied to these CRM outputs, as recommended for gannet (SNCBs 2014) and as advised by the Scoping Opinion.
928. As outlined for St Abb’s Head to Fast Castle SPA kittiwake population above, guidance on the use of the CRM suggests that model predictions should be based upon the mean monthly densities of flying birds estimated within the array area (Band 2012)8 and, to the best of the Applicant’s knowledge, this approach has been applied in all recent UK offshore wind farm assessments. Despite this, the Scoping Opinion advised that the CRMs for the Proposed Development should use the maximum monthly densities of flying birds within the array area. Further details on this are provided above in the Project Alone: Operation and Maintenance - Collision Risk section for St Abb’s Head to Fast Castle SPA kittiwakes and in volume 3. appendix 11.3 of the Offshore EIA Report, but as a result of this overly precautionary approach (which does not follow previous precedent) the CRMs for gannet were undertaken following:

• The Scoping Approach of using the maximum monthly densities, and
• The Developer Approach of using the mean monthly densities.
  929. In addition to the above, collision estimates for gannets were also calculated using option 2 of the stochastic version of the CRM (McGregor et al. 2018) with avoidance rates as derived from the bird collision-avoidance study undertaken at the Thanet offshore wind farm (Bowgen and Cook 2018). These additional collision estimates are not used as the basis of the assessments on the SPA gannet populations but, instead, are used solely to illustrate the consequences of applying these alternative avoidance rates which have been derived from studies at an actual offshore wind farm. Details of these additional CRMs are provided in annex C of the Offshore EIA Report, volume 3, appendix 11.3.
  930. As for the predicted displacement effects, gannet collision estimates are calculated for the breeding and non-breeding periods, with the latter separated into autumn and spring passage periods (Offshore EIA Report, volume 3, appendix 11.5). Estimates were apportioned to the gannet SPA population during the breeding and non-breeding periods according to the NatureScot (2018) approach and the BDMPS approach (Furness 2015), respectively (Offshore EIA Report, volume 3, appendix 11.5, Table 5.43). The age class proportions and assumptions on sabbatical rates are also as detailed above in relation to displacement effects (Table 5.43).
  931. Based upon option 2 of the deterministic CRM with a 98.9% avoidance rate applied, and in conjunction with the estimates and assumptions detailed above, the annual collision mortality of gannets from the Forth Islands SPA is predicted to be 151 adults and 5 immatures as determined by the Scoping Approach, and 123 adults and 3 immatures as determined by the Developer Approach (Table 5.45). As for displacement, the vast majority of this mortality (i.e. 95% overall) is predicted to occur during the breeding season.

Table 5.45: Predicted Collision Effects from the Proposed Development on the Forth Islands SPA Gannet Population, as Determined by the Scoping Approach and Developer Approach. Estimates are for the Maximum Design Scenario and are Based on Option 2 of the Deterministic CRM using a 98.9% Avoidance Rate (see text)

932. The additional annual mortality of adult gannets from the Forth Islands SPA population predicted due to collisions with wind turbines in the Proposed Development array represents approximately 0.08% of the number of adults currently estimated to breed at this colony (i.e. 150,518 individuals – Table 3.3 in volume 3, appendix 11.5 of the Offshore EIA Report) as determined by the Developer Approach and approximately 0.10% as determined by the Scoping Approach. In terms of percentage increases in the baseline annual adult mortality of the population (which is based on applying a mortality rate of 0.046 – see Table 2.7 in volume 3, appendix 11.6 of the Offshore EIA Report), the predicted adult collision mortality equates to increases of 1.8% and 2.2% for the Developer and Scoping Approaches, respectively.
933. The collision estimates produced using option 2 of the stochastic CRM with the Bowgen and Cook (2018) avoidance rates applied are 54% lower than those presented in Table 5.45 (for both the Scoping and Developer Approaches).
934. The potential levels of impact on the Forth Islands SPA gannet population resulting from the mortality predicted from collisions associated with the Proposed Development during the operation and maintenance phase are considered in more below in the Project Alone: Population-Level Impacts section. This presents the outputs from PVAs of the combined effects of predicted displacement and collision mortality on the SPA population.