1. Introduction

1.1. Background

  1. When assessing the impact of a proposed offshore wind farm, it is crucial to determine the impact that such development will have on breeding seabird populations. Seabirds nest in colonies of variable sizes around the United Kingdom (UK) coastline (Burnell et al., 2023) and most species have large foraging ranges at sea (Woodward et al., 2019). Establishing the connectivity between marine renewable sites and colonies, which are often protected as Special Protected Areas (SPAs), is a key element of the assessment of impact. A theoretical approach has been developed by NatureScot (NatureScot, 2018) to determine the proportion of birds from SPA sites which use proposed development areas in the breeding season. In the non-breeding period, the standard approach to apportioning utilises the information presented in Furness (2015). These approaches allow the user to calculate apportioning values which are then used to ‘apportion’ the impact of a marine renewable project to multiple SPAs.
  2. This technical report presents the apportioning method and the seasonal apportioning values applicable to the Ossian Array (hereafter referred to as the “Array”) for SPAs that support qualifying species for which connectivity has been identified as part of the likely significant effect (LSE2) Screening Report (Part 1, appendix 1A; Ossian OWFL, 2023) (SPAs shown in Figure 1.1   Open ▸ ).
  3. Information and results from the following technical reports of the Array EIA Report (Ossian OWFL, 2024) were used to feed into the apportioning assessment:
  • volume 3, appendix 11.1: Offshore Ornithology Baseline Characterisation Technical Report;
  • volume 3, appendix 11.2: Offshore Ornithology Collision Risk Modelling (CRM) Technical Report; and
  • volume 3, appendix 11.3: Offshore Ornithology Displacement Technical Report.
  1. The resulting apportioning values are presented for each protected site and will be used in the Report to Inform Appropriate Assessment (RIAA) to support the assessment of potential for an Adverse Effect on Integrity (AEOI) (Part 3).


Figure 1.1:
SPAs Included in the Apportioning Assessment in Relation to the Offshore Array Area

Figure 1.1: SPAs Included in the Apportioning Assessment in Relation to the Offshore Array Area


2. Methodology

2.1. Approach

  1. Apportioning undertaken for the breeding season is based on the NatureScot ‘theoretical approach’ method for the breeding season (NatureScot, 2018). Apportioning for the non-breeding period (i.e. post-breeding and pre-breeding seasons and in winter) generally utilises the Biologically Defined Minimum Population Scales (BDMPS) approach developed by Furness (2015). Seasonal definitions are set out in volume 3, appendix 11.1 of the Array EIA Report.
  2. In Scotland BDMPS is applied to all species with one exception. For guillemot Uria aalge NatureScot (2023a) advises that the mean-maximum foraging range + one standard deviation (SD) (i.e. the mean average of the maximum foraging trips recorded) should be used instead to determine connectivity to SPAs. As the Array lies beyond this distance (section 2.4) apportioning during the non-breeding period is not required under the NatureScot approach. However, Natural England in their Array EIA Scoping Report consultee response advised the non-breeding season for guillemot of Flamborough and Filey Coast SPA should be based on the BDMPS apportioning approach (Furness, 2015) (refer to the Array EIA Report volume 2, chapter 11, section 11.5; volume 2, appendix 6.2). This approach has been included within the apportioning undertaken. The advice from NatureScot (2023a) is based on tracking data of movements in the non-breeding season, and therefore the inclusion of Flamborough and Filey Coast SPA is considered to be over precautionary. Therefore, whilst this approach is included to meet Natural England’s request, it should be noted that the main assessment in the RIAA uses the approach recommended by NatureScot.
  3. For four species (shag Gulosus aristotelis, black-legged kittiwake Rissa tridactyla (hereafter referred to as kittiwake), guillemot and razorbill Alca torda), the Marine Scotland Apportioning Tool (Butler et al., 2020) is available to attribute seabirds to their breeding colonies, the use of which relies on Seabird 2000 (1998 to 2002) colony count data. These colony count data have recently been updated following the fourth UK breeding seabird census, Seabirds Count (JNCC, 2023), which focuses on the period 2015 to 2021, but has yet to be incorporated into the Marine Scotland Apportioning Tool. It is the view of NIRAS that to rely on an apportioning tool that requires the ‘old’ colony count data (based on the Seabird 2000 count) would produce unrealistic outputs. This is because in the intervening period, significant changes in colony size and, importantly, proportionate spread of the Scottish population between colonies has occurred (and as evidenced by scrutiny of the latest seabird census data). Therefore, the Marine Scotland Apportioning Tool has not been applied, with apportioning following the theoretical approach for all species for the breeding season as previously described. This proposed approach was shared with NatureScot during consultation correspondence in November 2023. Feedback on this point was not provided by NatureScot, but no objections or concerns were raised.

2.2. Identification of Species

  1. Table 2.1   Open ▸ identifies the designated sites and associated features for which a LSE2 has been identified and therefore where apportioning values are required to apportion impacts from the Array. Although this table focuses on SPA populations, consideration has been given to all breeding colonies within the relevant foraging range of a species present within the Array survey area (as described in section 2.3, the approach to breeding season apportionment requires consideration of all colonies within foraging range, including those outside SPAs). Apportioning values for non-SPA colonies is provided in annex A.

Table 2.1:
SPAs and Associated Qualifying Features for Which Apportioning is Required for the Array

Table 2.1: SPAs and Associated Qualifying Features for Which Apportioning is Required for the Array

 

2.3. Apportioning of Potential Impacts During the Breeding Season

  1. In the breeding season, a population of birds in a given sea area is likely to comprise of breeding adult birds from breeding colonies, immature birds (i.e. birds that have not yet reached breeding age) and non-breeding birds (i.e. birds that have reached breeding age but have not yet started breeding or are skipping a breeding season (sabbatical birds)). The proportion of each group must be estimated to allow the proportion of breeding birds to be identified, as it is this group that is relevant for Habitats Regulations Appraisal (HRA).

2.3.1. Breeding Adult Birds

  1. The method followed NatureScot’s ‘Interim Guidance on Apportioning Impacts from Marine Renewable Developments to Breeding Seabird Populations in Special Protection Areas’ (NatureScot, 2018).
  2. To identify those breeding colonies (designated sites and other) for which there may be connectivity between breeding birds and the Array, the recommended foraging range given by NatureScot (2023b) has been used. In most cases, this is the mean-maximum foraging range + one SD as published by Woodward et al., (2019). However, a different foraging range, often informed by site-specific information, is recommended for certain species or protected sites. Guillemot is one such species, for which NatureScot advises use of a different foraging range under certain circumstances. For guillemot, NatureScot advise the use of the mean-maximum foraging range plus one standard for all Northern Isles SPAs (i.e. 153.7 km), and to use the mean-maximum foraging range plus one standard deviation calculated excluding data obtained from birds from Fair Isle for all designated sites south of the Pentland Firth (i.e. 95.2 km) (NatureScot 2023b).
  3. It should be noted that the Ossian Array Screening Report (Ossian OWFL, 2023) was drafted prior to the publication of NatureScot’s guidance and therefore used the mean-maximum foraging range + one SD as published by Woodward et al., (2019) for all species and sites. It should also be noted that the LSE2 Screening Report (Ossian OWFL, 2023) considered the shortest distance from the Array boundary to SPA boundaries when determining connectivity, while this report follows NatureScot (2018) in measuring distances from the centre of the Array to the centre of the colony. Therefore, the sites and species listed in this appendix will differ slightly from those identified in the LSE2 Screening Report (Ossian OWFL, 2023). Further details on changes since the LSE2 Screening Report (Ossian OWFL, 2023) are given in the main RIAA document (Parts 1 and 3).
  4. Following identification of breeding colonies with potential connectivity, three colony-specific weighting factors have been applied to each colony:
  •  colony size (with consistent count unit used between colonies for a species e.g. individuals, breeding pairs or apparently occupied sites);
  •  distance of colony from the development site; and
  •  sea area (the area extent of the open sea within the foraging range of the relevant species).
  1. Large colonies will contribute more individuals to the number of seabirds found at the Array, all other factors being equal. To account for this, a weighting factor based on colony size has been derived. Colony sizes for all species have been extracted directly from the recently published Seabirds Count (JNCC, 2023). The Seabirds Count is the fourth UK breeding seabird census, following on from the Seabird 2000 census. The Seabirds Count comprehensively surveyed seabird colonies, with surveys undertaken from 2015 to 2021[3]. As such, the Seabirds Count data represent the best available data on breeding seabird colony sizes (JNCC, 2023).
  2. Weighting by distance from the colony has been calculated using the measured distance between the geometric centre of the Array to the colony, using a by-sea route, as recommended by NatureScot (2018). The location of colonies was based on the coordinates given in Burnell et al. (2023). As birds move further away from a colony, density will decrease by a factor proportional to 1/distance2 as area increases proportionally by πr2. For the purposes of this assessment, a weighting factor based on 1/distance2 has been used as advised by NatureScot (2018).
  3. The available sea area for foraging has been measured by plotting a circle (defined by the species-specific foraging range around the colony) in ArcGIS and calculating the area of sea available to each seabird species. The fraction of the disc centred on the colony that is occupied by sea surface is then expressed as a decimal. As the density of birds is predicted to increase as the area of available foraging area decreases, this is used in the formula as 1/estimated area.
  4. The three weighting factors (weightings by colony size, distance from the colony and sea area) have been combined to produce an overall weighting for each colony. Each factor is given equal weight in the combined weighting. This calculation is provided below:

 

  1. The weighting was then used to calculate the proportion of birds attributed to each colony (“proportional weight of colony”) by calculating colony weight divided by sum of all colony weights.
  2. Where an SPA consists of multiple colonies, apportionment was carried out to each colony individually following the method described above. The total apportionment weight of the SPA population is then calculated as the sum of the weights of the individual colonies.  

2.3.2. Immature Birds

  1. A major part of any seabird population comprises immature birds. This is especially relevant for many of the species considered in this report, with some species not breeding until they reach six or more years of age (for example puffin, as taken from Horswill and Robinson, 2015). A proportion of immature birds return to natal waters during the breeding season, with the proportion of each immature age class increasing as individuals get closer to breeding age.
  2. To determine the proportion of immature birds present within the Array in the breeding season, data from the site-specific Digital Aerial Surveys (DAS) have been analysed ( Table 2.2   Open ▸ ). This approach can only be used for gannet, kittiwake and large gull species as they are easily distinguishable between immature and adult individuals. Identifying the age class of birds from other species, such as auks, using DAS is not feasible due to the similarity between immatures and adults. During site-specific surveys, no large gulls were assigned to any age class (volume 3, appendix 11.1, annex D of the Array EIA Report). Only those birds assigned to an age class have been included in the calculation in Table 2.2   Open ▸ . However, the number of birds for which an age class was not assigned is also provided. Those birds for which an age class can’t be determined are not included in the apportioning exercise. The data set out in Table 2.2   Open ▸ are used to inform the assessments undertaken in the RIAA.

 

Table 2.2:
Number of Birds Assigned to Different Age Class Categories During Site-Specific Surveys of the Array Survey Area

Table 2.2: Number of Birds Assigned to Different Age Class Categories During Site-Specific Surveys of the Array Survey Area

 

  1. The identification of kittiwake age classes at sea is difficult and in most cases impossible (with the exception of first summer of younger birds). Whilst one year old kittiwakes can be easily identified due to differences in plumage, second and third year old birds, which have not yet reached the age of first breeding, cannot (Coulson, 2011; Olsen and Larsson, 2003). Therefore data on age class collected during surveys will potentially represent a considerable overestimate of the proportion of breeding adults present at the Array.
  2. To calculate an apportioning value for the breeding season in respect to the number of two and three year old kittiwakes at the Array, the analysis uses survival rates for immature kittiwake from Horswill and Robinson (2015) ( Table 2.3   Open ▸ ). The apportioned values will likely remain an under-estimate for the second and third year immatures as proportionately those cohorts show a much greater affinity for natal waters than first year birds.

 

Table 2.3:
Estimated Breeding Season Contribution of Immature Kittiwake Predicted to be Present at the Array[4]

Table 2.3: Estimated Breeding Season Contribution of Immature Kittiwake Predicted to be Present at the Array[4]

 

  1. For other species, birds cannot be readily aged from DAS imagery. Therefore, the proportion of adults present at the array in the breeding season is assumed to equal the proportion of adults within the population as a whole. The proportion of age classes is calculated from the demographic data in Horswill and Robinson (2015) assuming a stable age structure. This is presented in full in volume 3, appendix 11.1 of the Array EIA Report (Ossian WOFL, 2024), and the key values used for apportionment are repeated in Table 2.4   Open ▸ .

 

Table 2.4:
Proportion of Age Classes of Other Species

Table 2.4: Proportion of Age Classes of Other Species

 

  1. In the non-breeding season, the proportion of adults within the population does not need to be calculated or specified separately, as it is already factored into the Furness (2015) data used (see section 2.4).

2.3.3. Sabbaticals

  1. Every breeding season a proportion of adults skip breeding and take a ‘sabbatical’. The inclusion of sabbatical birds within the birds apportioned to a colony  would likely lead to an overestimate of the effects to these species/populations (Marine Scotland 2017a; 2017b). This is because estimates of breeding colony population sizes used within the RIAA, do not include these sabbatical birds.

It is not possible to separate non-breeding adult birds from those that are breeding in a given sea area and therefore published estimates of sabbatical behaviour have been obtained ( Table 2.5   Open ▸ ). Consideration will be given in relevant assessments to the sabbatical values presented in

  1. Table 2.6   Open ▸ for each species, following the approach advised on other projects by the Scottish Ministers (2017a; 2017b, 2017c).

 

Table 2.5:
Proportion of Sabbatical Birds to be Considered in the RIAA

Table 2.5: Proportion of Sabbatical Birds to be Considered in the RIAA

 

Table 2.6:
Proportion of Sabbatical Birds Used for RIAA Apportionment

Table 2.6: Proportion of Sabbatical Birds Used for RIAA Apportionment