NatureScot Research Report 1340 - Scapa Flow and North Orkney Special Protection Areas (SPAs) - Inshore wintering waterfowl surveys 2021/22 and 2022/2023

Published: 2026

Authors: Peters-Grundy, R., Thompson, K., Humphries, G., Harvey, J., Semple, M., Tyler, K., Harker, A.J., Pavat, D., Thomson, R., Olley, N. and Macleod, K. 

Cite as: Peters-Grundy, R., Thompson, K., Humphries, G., Harvey, J., Semple, M., Tyler, K., Harker, A.J., Pavat, D., Thomson, R., Olley, N. and Macleod, K. Scapa Flow and North Orkney Special Protection Areas (SPAs) - Inshore wintering waterfowl surveys 2021/22 and 2022/2023. NatureScot Research Report 1340.

Keywords

Scapa Flow SPA; North Orkney SPA; digital aerial survey; wintering waterfowl; great northern diver; black-throated diver; Slavonian grebe; red-breasted merganser; velvet scoter; long-tailed duck; common eider; European shag 

Background

The seas surrounding Orkney support several internationally important populations of wintering waterfowl (divers, grebes, sea ducks) and shag. In 2022, two areas of these waters were designated under The Birds Directive (2009/147/EC2) as marine Special Protection Areas (SPAs): Scapa Flow SPA and North Orkney SPA.

These seas are also of interest for current and future developments. As a part of the Orkney Harbours Masterplan Phase 1, Orkney Islands Council have proposed several developments, including a Scapa Flow deep water port, and expansion of Hatston pier (located within the North Orkney SPA) to extend the existing quay and create a facility for marine economic sectors (Orkney Islands Council, 2022).

In November 2021, NatureScot commissioned HiDef Aerial Surveying Ltd to undertake a programme of high-resolution digital aerial surveys (DAS) and shore-based counts of wintering waterfowl in the winters of 2021/22 and 2022/23. In particular, the surveys focused on eight target species, which together comprise the non-breeding qualifying features of the two SPAs: great northern diver (Gavia immer), black-throated diver (Gavia arctica), Slavonian grebe (Podiceps auratus), red-breasted merganser (Mergus serrator), velvet scoter (Melanitta fusca), long-tailed duck (Clangula hyemalis), common eider (Somateria mollissima) and European shag (Gulosus aristotelis).

The DAS surveys were designed to cover the extent of both SPAs, whilst the shore-based counts targeted a limited number of key inshore areas earmarked for development. The aim of the surveys was to supplement existing information on populations of wintering marine birds in these marine SPAs and provide finer-scale information on the distribution of birds within these key development areas. Results of the shore-based surveys would also assist interpretation of the DAS, for species which may pose challenges for broadscale aerial snapshot survey.

This report presents the results from the DAS which were conducted across both SPAs in February 2022 and January 2023 and eleven shore-based surveys targeted in nearshore areas conducted over the winters of 2021/22 and 2022/23. Population estimates with associated confidence and credible limits were calculated using both design-based strip transect analysis and model-based Bayesian point processing and are presented for the eight non-breeding qualifying features. The distributions and densities of target species across the SPAs have also been mapped. These density estimates between the 2021/22 and 2022/23 DAS have been compared with each other, as well as with previous surveys undertaken across the SPAs. Observations during the 20 January 2023 shore-based counts were also compared with the observations from the DAS conducted on the same day. 

Main findings

• Two Digital Aerial Surveys (DAS) were successfully completed of Scapa Flow SPA and North Orkney SPA on 21 February 2022 and 20 January 2023, as well as eleven shore-based surveys targeted in nearshore areas conducted over the winters of 2021/22 and 2022/23. This included simultaneous shore-based and DAS on 20 January 2023, in order to compare results from the two methods.
• Over 10,000 birds were detected during the February 2022 DAS, and 18,000 during the January 2023 DAS, with an overall identification rate to species level of 97.54 % and 93.8 %, respectively.
• For eight key wintering waterfowl species, which are qualifying features of one or both of the SPAs, population estimates for the SPAs were derived from the DAS using both design-based blocked bootstrapping (where the transect is the sampling unit) and model-based Bayesian point processing methods.
• Across the key species, more birds were detected during the January 2023 DAS compared with February 2022, which may be attributed to a variety of factors such as natural fluctuations in populations within or between winter seasons, or bird distribution being impacted by poor weather conditions observed prior to the 2022 DAS being conducted. The population estimates from the January 2023 DAS, as estimated using Bayesian point processing are therefore summarised here:
  • Great northern diver were recorded in both SPAs in both years, widely distributed across shallow and deep-water environments. Population estimates from the January 2023 survey were for Scapa Flow 1,072 birds (95 % credible interval 884 – 1,232) and for North Orkney 208 birds (95 % credible limits 145 – 282).
  • Black-throated diver were predominantly recorded in the west of Scapa Flow SPA (seven birds recorded in total across the whole area), with only one bird recorded at North Orkney SPA in February 2022 where it is not a qualifying feature and no birds in January 2023. Bayesian analysis was not undertaken for black-throated diver due to an insufficient number of records. Higher numbers of birds were detected in the shore-based counts within limited areas of Scapa Flow.
  • Slavonian grebe were recorded in both SPAs; primarily located close to shore. Population estimates from the January 2023 survey were for Scapa Flow 114 birds (95 % credible interval 69 – 175), and for North Orkney 103 birds (95 % credible interval 61 – 163).
  • Red-breasted merganser are a qualifying feature of the Scapa Flow SPA.  They were recorded in both SPAs, primarily located close to shore. Population estimates from the January 2023 survey were for Scapa Flow 338 birds (95 % credible interval 255 – 435) and for North Orkney 331 birds (95 % credible interval 236 – 521).
  • Velvet scoter were not recorded in the digital aerial survey in either SPA, across both surveys. The species is not a qualifying feature of the Scapa Flow SPA.
  • Long-tailed duck are a qualifying feature of the Scapa Flow SPA. They were the most abundant target species recorded in both SPAs, in both the 2022 and 2023 surveys and had a widespread distribution across both sites. Population estimates from the January 2023 survey were for Scapa Flow 3,678 birds (95 % credible interval 3,321 – 4,343) and for North Orkney 2,428 birds (95 % credible interval 2,199 – 2,654).
  • Common eider are a qualifying feature of the Scapa Flow SPA. They were recorded in both SPAs, predominantly distributed in shallow coastline waters. Population estimates from the January 2023 survey were for Scapa Flow 1,504 birds (95 % credible interval 1,322 – 1,706) and for North Orkney 2,134 birds (95 % credible interval 1,887 – 2,421).
  • European shag are a qualifying feature of the Scapa Flow SPA.  They were recorded in both SPAs, with a widespread distribution across both sites. Population estimates from the January survey were for Scapa Flow 872 birds (95 % credible limits 597 – 1,560) and for North Orkney 1,667 birds (95 % credible limits 1,356 – 2,189).
• A comparison of the DAS and shore-based survey data from 20 January 2023 concluded that method selection should be dependent both on the target species and the characteristics of the environment (i.e. open sea vs shallow coastal waters). For example, more great northern diver, eider and long-tailed duck were detected during the DAS compared with the shore-based counts, suggesting that DAS may be better suited for these species. Conversely, Slavonian grebe and European shag had higher detections during the shore-based counts, which may be attributed to their small size and dark plumage respectively, making detection during DAS more difficult. However, this comparison is likely to have been influenced by the temporal and spatial mismatch between the surveys. Both red-breasted merganser and black-throated diver were primarily recorded outwith the shore-based count survey area during the DAS and thus a comparison of the data is more complex. No velvet scoter were recorded during the DAS.  
• Inshore wintering waterfowl populations fluctuate naturally within and between seasons, such that single snapshot surveys cannot be used in isolation to assess possible underlying trends over time. This was evident in the monthly shore-based surveys of smaller areas, as detailed in the report, as well as high variation observed between the February 2022 and January 2023 DAS.
• Overall, DAS methods proved to be effective for detection and identification of populations of multiple waterfowl species, such as long-tailed duck and common eider. Supplementation of aerial data with shore-based surveys may be beneficial for cryptic species which form mixed flocks or which favour nearshore areas, or for species that spend time roosting on land or inter-tidal rocks such as velvet scoter, Slavonian grebe and European shag respectively. 

Abbreviations

Area of Search (AoS)

Coefficient of Variation (CV)

Confidence Interval (CI)

Confidence Limits (CLs) 

Digital Aerial Survey (DAS)

Gaussian Markov Random Field (GMRF)

Generalised Additive Model (GAM) 

Generalised Linear Model (GLM)

Global Positioning System (GPS)

Integrated Nested Laplace Approximation (INLA)

Kernel Density Estimation(KDE) 

Log Gaussian Cox Process (LGCP)

Markov Chain Monte Carlo (MCMC)

Mean Low Water Springs (MLWS)

National Planning Framework (NPF)

proposed Special Protection Area (pSPA)

Quality Assurance (QA)

Special Protection Area (SPA)

Stochastic Partial Differential Equation (SPDE) 

Visual Aerial Survey (VAS)

Introduction

The seas surrounding Orkney support several internationally important populations of wintering waterfowl (divers, grebes, sea ducks) and shag. In 2022, two areas of these waters were designated under The Birds Directive (2009/147/EC2) as marine Special Protection Areas (SPAs): Scapa Flow SPA and North Orkney SPA.

Scapa Flow SPA covers an area of 318.20 km² and is designated for the protection of eight qualifying features, including the Annex I species great northern diver (Gavia immer), black-throated diver (Gavia arctica) and Slavonian grebe (Podiceps auratus) and migratory red-breasted merganser (Mergus serrator), long-tailed duck (Clangula hyemalis), common eider (Somateria mollissima), and European shag (Gulosus aristotelis) in the non-breeding season (Table 1). Additionally, it is designated for Annex I red-throated diver (Gavia stellata) in the breeding season.

North Orkney SPA covers an area of 211.73 km² and is designated for the protection of four qualifying features, namely, Annex I great northern diver and Slavonian grebe together with migratory velvet scoter (Melanitta fusca) in the non-breeding season, and Annex I red-throated diver in the breeding season (Table 1).

Table 1. Qualifying features of Scapa Flow and North Orkney SPA listed in order of priority for surveys. Population sizes are the estimated populations within the SPA from the relevant SPA citation, derived from Lawson et al., (2015) and Black et al., (2015).  These population estimates are also derived from Lawson et al., 2015.

Species (season)	Not qualifying features of the designated SPA but which were included in the pSPA	Annex 1 or migratory	Population size in Scapa Flow SPA (%GB)	Population size in North Orkney SPA (%GB)
Great northern diver	Non-breeding	Annex 1	505 (20.2)	308 (12.3)
Black-throated diver	Non-breeding	Annex 1	57 (9.5)	N/A
Slavonian grebe	Non-breeding	Annex 1	135 (12.3)	120 (10.9)
Red-breasted merganser	Non-breeding	Migratory	539 (6.4)	344 (4.1)
Velvet scoter	Non-breeding	Migratory	N/A	147 (5.9)
Long-tailed duck	Non-breeding	Migratory	1,395 (12.7)	937 (8.5)
Common eider	Non-breeding	Migratory	1,997 (3.6)	1,453 (2.4)
European shag	Non-breeding	Migratory	2,927 (2.9)	1,742 (1.6)
Red-throated diver	Breeding	Annex 1	76 pairs (6.1)	47 pairs (3.7)

 

These seas are also of interest for current and future developments with the proposed Scapa Flow Deep Water Quay identified as a national development under Draft National Planning Framework (NPF4). 

As part of the Orkney Harbours Masterplan Phase 1, Orkney Islands Council have proposals in place for several developments, including development of a Scapa Flow deep water port, and expansion of Hatston pier (located within the North Orkney SPA) to extend the existing quay and create a facility for marine economic sectors (Orkney Islands Council, 2022). Successful management of protected areas requires monitoring. Not only is this important for guiding site management measures but also for informing planning and development. The potential impacts of developments on SPAs needs to be considered through a Habitats Regulation Appraisal (HRA). Robust data to carefully assess the potential impacts of development on the qualifying features of the sites is required and can further inform the development of appropriate management measures as necessary.

Previous surveys have been undertaken at both sites to quantify the number of inshore wintering waterfowl and have comprised a composite of survey methods. A mixture of visual aerial survey (VAS), shore-based counts and boat-based surveys underpinned the initial selection of proposed SPAs for wintering waterfowl across Scotland; termed the ‘Original Areas of Search’ (AoS) (Lawson et al., 2015). Further surveys were undertaken in Orkney in the winter of 2017/18 to assess the numbers of identified target species within the then proposed SPA (pSPA) boundaries, using a combination of shore-based and boat-based surveys across the Scapa Flow (Jackson, 2018) and shore-based counts only across North Orkney (Upton et al., 2018). Previously, no DAS have been undertaken to assess the populations of wintering birds within the finalised SPA boundaries. A visual comparison of the original AoS, pSPA and SPA boundaries is presented in Figure 1.

In November 2021, NatureScot, in partnership with Marine Scotland and Orkney Islands Council, commissioned HiDef Aerial Surveying Ltd (‘HiDef’) to undertake a programme of high-resolution DAS and shore-based counts of wintering waterfowl across winter 2021/22 and 2022/23. The DAS surveys were designed to cover the extent of the SPAs, whilst the shore-based counts targeted a limited number of inshore areas. 

The DAS, plus targeted shore-based vantage point (VP) surveys of selected areas, supplement existing information on populations of wintering marine birds in these marine SPAs and provide finer scale information on the distribution of birds within key areas targeted for development. The targeted shore-based VP surveys also assist interpretation of the DAS results. Some of the target species are particularly challenging for broadscale aerial snapshot surveys as they occur close inshore at low densities and spend substantial periods of time below the water surface.

This report presents the results from the DAS which were conducted across both SPAs in February 2022 and January 2023 and eleven shore-based surveys targeted in nearshore areas conducted over the winters of 2021/22 and 2022/23. Population estimates with associated confidence and credible limits were calculated using both design-based strip transect analysis and a model-based Bayesian point processing model and are presented for the eight non-breeding qualifying features. The distributions and densities of target species across the SPAs have also been mapped. Density estimates have been compared between the 2021/22 and 2022/23 DAS. Additionally, comparisons are made of density estimates within the targeted inshore areas from the DAS and shore-based surveys conducted within a similar time frame on 20 January 2023.

Methods

Shore-based surveys

The VP counts were undertaken on a selection of smaller inshore areas, hereafter referred to as ‘sectors’, utilised in previous surveys of the pSPAs (Jackson, 2018; Upton et al., 2018). These sectors were selected as being key areas targeted for potential development. Monthly shore-based VP counts were undertaken within the relevant sectors at both SPAs across the non-breeding season in 2021/22 and 2022/23.

The defined sectors and VPs used during the shore-based surveys are presented in Figure 2.

Five counts were undertaken at each sector in winter 2021/22, and six were undertaken at each sector in winter 2022/23, as presented in Table 2. In 2023, an additional shore-based count was undertaken on 20 January 2023 to coincide with the DAS in order to compare the abundance and densities of animals recorded using these two methods. However, due to difficulties with the terrain and access issues on 20 January 2023, the shore-based survey at sector 37 could not be undertaken and has not been included in the analysis for that month. 

Table 2:  Survey dates of Shore-based VP surveys undertaken across the winters of 2021/22 and 2022/23 in the sectors shown in Figure 2. Sector 14 refers to North Orkney SPA. 

Sector	November	December	January	February	March
9	30/11/2021	19/12/2021	09/01/2022	13/02/2022	05/03/2022
10	30/11/2021	19/12/2021	09/01/2022	13/02/2022	05/03/2022
11	30/11/2021	19/12/2021	09/01/2022	13/02/2022	05/03/2022
12	30/11/2021	19/12/2021	09/01/2022	13/02/2022	05/03/2022
14	23/11/2021	18/12/2021	15/01/2022	14/02/2022	06/03/2022
37	30/11/2021	19/12/2021	09/01/2022	13/02/2022	05/03/2022
38	30/11/2021	19/12/2021	09/01/2022	13/02/2022	05/03/2022
9	12/11/2022	13/12/2022	12/01/2023  and 20/01/2023	14/02/2023	04/03/2023
10	12/11/2022	13/12/2022	12/01/2023  and 20/01/2023	14/02/2023	04/03/2023
11	12/11/2022	13/12/2022	12/01/2023  and 20/01/2023	14/02/2023	04/03/2023
12	12/11/2022	13/12/2022	12/01/2023  and 20/01/2023	14/02/2023	04/03/2023
14	07/11/2022	05/12/2022	12/01/2023  and 20/01/2023	13/02/2023	04/03/2023
37	12/11/2022	13/12/2022	12/01/2023  and 20/01/2023 (this on 20 January 2023 was not completed due to difficulties with access, and this survey has not been considered further.)	14/02/2023	04/03/2023
38	12/11/2022	13/12/2022	12/01/2023  and 20/01/2023	14/02/2023	04/03/2023

 

The VPs were selected based on ease of access and view, assuming that if there was at least 5 m elevation above sea level that birds would be detected, identified and accurately counted up to 2 km (Jackson, 2018) with the aid of a spotting scope. 

Count protocols were based on those used by Jackson (2018). This is based on 15-20 minute counts at each site, including range and bearing of offshore birds. Each shore-based survey was conducted by an experienced ornithologist. All target species were searched for and recorded, including birds on the sea and roosting birds on the adjacent shoreline, islets and rocks. Flying birds were not included in count totals unless birds took off from or landed in the count area. Double counting was reduced as far as possible by avoiding spatial overlap and by minimising the time between recording adjacent areas by systematically working along the coast.

Digital aerial survey 

Two DAS of inshore wintering waterfowl per winter were undertaken, on the 21 February 2022 and 20 January 2023. The survey design (HiDef Aerial Surveying Ltd, 2023) placed 23 strip transects across both the Scapa Flow and North Orkney SPAs, spaced 2 km apart. The survey was originally designed for the previous pSPA boundaries, creating an overall survey area of 558.07 km². The final SPA boundaries, as designated in February 2022 cover a slightly smaller combined area of 530.29 km² (Figure 1).

The strip transects were orientated roughly north to south, perpendicular to the depth contours along the coast. Such a design ensured that each transect sampled a similar range of habitats (primarily relating to water depth) and would reduce the variation in bird and mammal abundance estimates between transects.

However, in the 2022 survey, a number of transects were rotated by roughly 90 degrees in the west of the Scapa Flow, shortly prior to the flight, due to the pilot having safety concerns with flying across the channel between Hoy and Orkney Mainland. To maintain consistency across both DAS, the survey in 2023 was conducted following the same transect pattern (Figure 3).

A comparison of the location of the DAS transects with the inshore sectors used for shore-based counts during winter 2021/22 and winter 2022/23 can be found in Figure 4.

Both DAS surveys were undertaken using a Diamond DA42 Twin Star aircraft equipped with four HiDef Gen II cameras with sensors set to a resolution of 2cm Ground Sample Distance (GSD). Each camera sampled a strip of 125 m width, separated from the next camera by ~20 m, thus providing a combined sampled width of 500 m within a ~560 m overall strip (Figure 5).

A minimum target of 24.3 % site coverage was agreed, with data from all four cameras being processed. This was to ensure a survey with sufficient coverage and number of transects for precise abundance estimation.

The survey was flown along the transect pattern shown in Figure 3 at a height of approximately 550 m above sea level (ASL; ~1,800’). Flying at this height ensured that there was reduced risk of flushing species that are easily disturbed by aircraft noise. Thaxter et al.(2015) recommend a minimum flight altitude of 460 m ASL.

Position data for the aircraft was captured from a Garmin GPSMap 296 receiver with differential Global Positioning System (GPS) enabled to give 1 m accuracy for the positions and recording updates in location at one second intervals for later matching to bird observations. 

Data review and object detection

Data were viewed by trained reviewers who marked any objects in the footage as requiring further analysis, as well as determining which were birds, marine megafauna or anthropogenic objects such as ships or buoys. 

As part of HiDef’s quality assurance (QA) process, an additional ‘blind’ review of 20 % of the raw data was carried out and the results compared with those of the original review. If 90 % agreement was not attained during the QA process, then the remaining data set was reviewed and where appropriate, the failed reviewer’s data discarded and the entire dataset re-reviewed. 

Objects were only recorded where they reached a reference line (known as ‘the red line’) which defines the true transect width of 125 m for each camera. By excluding objects that did not cross the red line, biases to abundance estimates caused by flux (movement of objects in the video footage relative to the aircraft, such as where the survey craft is buffeted by airflow) were eliminated.

Species identification

Images marked as requiring further analysis were reviewed by the ID Team of ornithologists (HiDef currently employs two of the ten current members of the British Birds Rarities Committee (‘BBRC’) as expert ornithologists) for identification to the lowest taxonomic level possible and for assessment of the approximate age and the sex of each animal, as well as any behaviour traits visible from the imagery. 

At least 20 % of all objects were selected at random and subjected to a separate ‘blind’ QA process. If less than 90 % agreement was attained then corrective action was initiated: if appropriate, the failed identifier’s data were discarded, and the data re-identified. Any disputed identifications were passed to a third-party expert ornithologist for a final decision¹. The level of agreement within the QA process is calculated as the final number of agreements as a percentage of all identifications subjected to QA for the entire survey. 

All objects were assigned to a species group and where possible, each of these then further identified to species level. The species identifications were given a confidence rating of ‘possible’, ‘probable’ or ‘definite’ (Definite: as certain as reasonably possible. Probable: very likely to be this species or species group. Possible: more likely to be this species or species group than anything else.). 

It is important to note that confidence ratings are not standardised. The likelihood of achieving a definite or probable identification is not consistent for all component members of a species group. For example, someone undertaking identification of a large auk will find it easier to be confident of guillemot (Uria aalge) identification than razorbill (Alca torda). Confidence scores should not be used to filter or weight the probability of ‘large auk’ being one species or another in any analysis, as this will lead to biased results, particularly if the identification rate is low.

Any animals that could not be identified to species level were assigned to a category ‘No ID’ and only identified to group level. If, on occasion, the unidentified bird is suspected of belonging to two possible genera, then a broader group category may be used. For example, a bird would usually be assigned to the group category ‘Shearwater species’ if identified as a Manx shearwater (Puffinus puffinus), or to ‘Large Auk species’ if identified as a guillemot. However, if the bird has the potential to be either, then it would be assigned to a wider group category ‘Shearwater / Auk species’ and the species level recorded as ‘No ID’.

Additional information was recorded on basic behaviour (i.e., whether the bird was sitting; loafing on land or other objects; flying; diving or taking off). Detail was recorded where possible on foraging behaviour, approximate age, sex and any other details of interest. Ageing of birds was based on moults and was conducted where possible on species which show seasonal variation in plumage.

Final processing

All data were geo-referenced, taking into account the offset from the transect line of the cameras, and compiled into a single output; Geographical Information System (GIS) files for the Observation and Track data are issued in ArcGIS shapefile format, using UTM30N projection, WGS84 datum. 

Data analysis

Data treatment - DAS

Survey data from the DAS were split between Scapa Flow SPA and North Orkney SPA. As many birds were recorded along the shoreline, just beyond the perimeter of the marine SPA boundaries, raw counts are presented for two scenarios: for the marine SPAs including birds just beyond the perimeter (‘untrimmed’ counts) and for the marine SPA boundaries directly (‘trimmed’ counts).  

No apportioning of ‘unidentified’ birds to species level was undertaken. All confidence levels of species identifications (i.e. Definite, Probable and Possible (as explained earlier)) were used in the analysis.

Population estimates: design-based strip transect approach

Population estimates were calculated for the Scapa Flow SPA and North Orkney SPA separately. Along the coasts, observations were trimmed to mean low water springs (MLWS) to match the SPA boundaries, and as such, roosting birds beyond these boundaries were not included in resulting estimates. 

Each strip transect was treated as a statistically independent random sample from the site. The length and breadth (i.e., the width of the field of view of the camera) of each transect were multiplied together to give the transect area; dividing the number of observations for each species on each transect by the transect area gives a point estimate of the density of that species for the transect. The density of animals at the site (and hence the population size by multiplying by the area of the site), the standard deviation, the 95 % confidence intervals (CIs) and coefficient of variation (CV) were then estimated using a non-parametric block bootstrap method with replacement (Buckland et al., 2001), to ensure equal transect effort was sampled across each bootstrap iteration. This was done by using transect ID as the sampling unit with replacement. A group of transects were randomly sampled until their total length equalled approximately the same length as the total survey length. 

To avoid autocorrelation, checks were built into the procedure. If autocorrelation in the randomly sampled group of transects was detected, then the distance between the sampled transects was adjusted to remove autocorrelation. No autocorrelation was detected during the generation of these estimates.

A total of 1,000 bootstrap iterations were performed from which we calculated the mean and standard deviation of the sampled means, as well as the relative standard deviation (or CV) as defined by the standard deviation divided by the mean. Data were processed in the R programming language (version 4.1.1).

The density estimate is expressed as the average number of birds per square km in the whole SPA boundary. The population estimate is expressed as the estimated number of birds within the whole SPA boundary. The upper and lower confidence limits (CLs) define the range that the population estimate falls within with 95 % certainty. The CV is a measure of the precision of the population and density estimates. Detailed definitions of statistical terms used are presented in Table 3.

Table 3a. Definition of statistical terms used in analysis for general statistics

Term	Definition
Density estimate (n/km2)	The mean number of birds per square km across the whole SPA.
Population estimate (number)	The mean number of birds estimated within the SPA.
Coefficient of variation (CV)	The coefficient of variation is a standard measure that describes the dispersion of data points around the mean. The lower the CV the more precise the estimate. It is calculated as the SD / mean.
Relative abundance	In the case of diving birds, this is the estimated population size based on birds recorded on or above the sea surface and does not account for any that may be diving and thus submerged at the time of survey.
Absolute abundance	The most accurate estimate of population size. In the case of diving birds, this includes an estimate for the number that are believed to be submerged at the time of survey.
Standard deviation (SD)	The amount of variation or dispersion of a set of values. A low SD indicates that the bootstrap values tend to be close to the mean of the set.
Coefficient of determination  (R2)	The coefficient of determination is the proportion of variance in the response variable that is explained by the predictor variable. This ranges between 0 and 1.
Root mean squared error	The standard deviation of the residuals, which are a measure of the variance that is unexplained by the model. A lower RMSE signals a better fitting model as there is less unexplained variance in the data.

 

Table 3b. Definition of statistical terms used in analysis for design-based statistics

Term	Definition
95 % confidence interval (CI)	A measure of uncertainty in the mean value. If the analysis was repeated, 95 % of the time the mean population estimate would fall within this range. The smaller the CI range the more confident we can be that the mean estimate is an accurate reflection of the true population size.
95 % confidence limits (CLs)	The upper and lower values that define the range of the 95 % confidence interval.

 

Table 3c. Definition of statistical terms used in analysis for Bayesian statistics 

Term	Definition
Credible interval	The interval within which 95 % of the estimated distribution of values fall.
Credible limit	The upper and lower values that define the range of the 95 % credible interval.
Deviance information criterion (DIC)	The DIC is used to compare the fit of a set of Bayesian models. It is used for model selection, with the lowest DIC indicating the best fitting model.

 

For most species these abundance estimates relate to absolute abundance, but for diving species such as shags and divers, these estimates relate to relative abundance due to an unquantified proportion of animals being submerged at the time of survey.

Population estimates: model-based Bayesian point processing

An alternative approach to density and abundance estimation is to use stochastic spatially explicit modelling of observations with associated covariates, and then predicting (extrapolating) into areas where we have information on the covariates, but not on observations. Using georeferenced observations as an input, extrapolation can be done using a point pattern process in a Bayesian framework. The Bayesian point process model offers an advantage over standard Kernel Density Estimates (KDE) approaches as it takes into account any potential unequal survey effort within transects (for example, should there be a change in the number of cameras processed along the transect) and between transects (differences in transect lengths). 

A point pattern records the occurrence of events in a study region where the locations of these observations depend on an underlying spatial process. This spatial process can be characterised using the Cox process, which is a Poisson process with intensity λ(s) that varies in space. This intensity function measures the average number of events per unit of space, and it can be modelled to depend on covariates and other effects (Diggle, 2014; Baddeley et al., 2015).

Under the log-Cox point process model assumption, log intensity of the Cox process is modelled with a Gaussian linear predictor. In this case, the log-Cox process is known as a log Gaussian Cox process (LGCP, Møller et al., 1998), and inference can be made using the Integrated Nested Laplace Approximation approach (INLA; Illian et al., 2012), which was developed as a computationally efficient alternative to Markov Chain Monte Carlo (MCMC) methods (Robert and Casella, 2004, Brooks et al., 2011). The log-Gaussian part of the LGCP name comes from modelling log(λ(s)) as a latent Gaussian (conditional on a set of hyper-parameters), in the typical Generalised Linear Model (GLM)/ Generalised Additive Model (GAM) framework.

To fit these models in INLA we used the stochastic partial differential equation (SPDE) approach (Simpson et al., 2016). The SPDE approach consists of representing a continuous spatial process (e.g., a latent stationary Gaussian Field) with the Matèrn covariance function as a discretely indexed spatial random process (e.g., a Gaussian Markov Random Field (GMRF); Rue and Held, 2005). This approach is computationally efficient while accounting for spatio-temporal interdependence and autocorrelation in the data and it considers a direct approximation of the log-Cox point process model likelihood, where observations are modelled considering their exact locations instead of binning them into cells. Along with the flexibility for defining a mesh, this approach can handle non-rectangular areas.

The DAS were modelled using only the spatial process (i.e. x and y coordinates of observations) as covariates which characterised the distribution and abundance of birds based on the location of individual observations. Environmental covariates were not used due to time constraints on data acquisition and tuning of hyperparameters that would be required to fit the model.

Species with a sample size of <10 observations were not included in the analysis, as model accuracy decreases sharply with sample sizes under this threshold (Stockwell & Peterson, 2002). Including species with a low sample size could lead to poor estimates of the species' distribution elsewhere in the study area.

The inlabru package (Bachl et al., 2019) automates much of the process to perform Bayesian inference in the INLA framework, making it a relatively lightweight option for performing density surface models. In this case, only the spatial component of the observations was modelled using the log Gaussian Cox process method. There are two hyper-parameters that need to be tuned here; the range (i.e., the mean distance at which we expect spatial autocorrelation to dissipate) and sigma (i.e., the standard deviation around the mean estimate of spatial autocorrelation) prior distributions, which are used for abundance estimation. The range parameter is determined from the mesh and the minimum nearest neighbour distance. It should not be smaller than the mesh resolution as the resolution of the mesh defines the minimum scale at which we can model the process. The range must be at least the minimum nearest neighbour distance between points as it uses the spatial smooth to model spatial correlation structures between the points. Therefore, it must be larger than this distance as the minimum distance across which points are correlated cannot be smaller than the distance between the points themselves. Using this information, the range parameter was chosen by fitting several models with different values for the range and identifying the model with the lowest DIC score.

Model predictions were output to a 1 x 1 km grid resolution with each grid of the output raster containing a predicted density of birds.

As with the design-based estimates, observations during Bayesian analysis were trimmed to each SPA boundary respectively, and as such, roosting birds above MLWS were not included in resulting estimates.

Detailed definitions of statistical terms used in the presentation of the results are presented in Table 3.

Shore-based counts

Counts of each target species within each sector per shore-based survey were recorded and summed. The mean percentage use per sector was also calculated.

Field maps were digitised in ArcMap (version 10.8.1). Dot maps were created to display the distribution of target species within the shore-based counts of the Scapa Flow SPA (sectors 9 – 12 and 37-38) and North Orkney SPA (sector 14), shown in Figure 2.  Dot size indicates ‘flock’ size. 

Comparison of shore-based counts and DAS

A qualitative comparison between the density estimates has been conducted for the simultaneous DAS and shore-based counts undertaken on 20 January 2023. Given the low coverage of the DAS, and low number of transects which intersect the shore-based count sectors, it was not possible to generate a robust population estimate for these sectors from the DAS or to conduct any other quantitative spatial analysis in order to compare the methods. Therefore, for each 500 m segment along the DAS within the shore-based count sectors, observations of birds across the four cameras were grouped and are represented by a single point where the larger the point, the more birds were observed in that 500 m segment. These have been mapped against the shore-based count observations for each of the key species for visual comparison, with the approximate location of the DAS camera swathe also indicated. However, it should be noted that these maps do not show the 20 m gap between cameras. Grey buffer marks 1.75 km from shore are also given, which indicated the distance at which shore-based detection is known to deteriorate (Webb and Reid, 2004)

Results

Survey effort

The survey effort undertaken during the DAS on 21 February 2022 and 20 January 2023 is shown in Table 4. The number of transects and the total length of transects presented for the SPAs are those used in subsequent population analysis (see Figure 6 for the aircraft flight patterns). 

Table 4a:  Survey effort across the Scapa Flow and North Orkney SPAs during the DAS undertaken on 21 February 2022 

Survey area	Whole	Scapa Flow SPA	North Orkney SPA
Number of transects analysed	20	16	14
Total length of transects analysed (km)	257.09	155.58	101.52
Area covered (km2)	128.55	77.79	50.76
Area covered (%)	24.26	24.43	23.96

 

4b: Survey effort across the Scapa Flow and North Orkney SPAs during the DAS undertaken on 20 January 2023

Survey area	Whole	Scapa Flow SPA	North Orkney SPA
Number of transects analysed	20	16	14
Total length of transects analysed (km)	253.49	158.82	107.77
Area covered (km2)	126.75	79.41	53.88
Area covered (%)	23.92	24.04	23.67

 

Environmental conditions during the surveys were good, with air clarity, glare and turbidity recorded as very low or slight. Across the survey area sea state varied, with scoring criteria on the World Meteorological Organisation (WMO) Sea State Code ranging between 0 ‘calm’ (0 m wave height) and 5 ‘rough’ (2.5 m to 4 m wave height) during the 2022 survey, and predominantly between 0 ‘calm’ and 3 ‘Slight’ (0.5 to 1.25 m wave height) across the 2023 survey. A map of sea states across the survey area for each survey recorded from footage is presented in Figure 7. During the 2022 survey sea states ranged between 0 and 5, with 4 being the most commonly observed value.  Sea state was generally calmest close to shore and within sheltered bays. The highest sea states (sea state 5) were generally recorded in the east, such as the eastern Scapa Flow, east of Burray and south and north of Shapinsay. In the January 2023 survey, sea states were generally substantially lower, with the highest sea state of 4 observed in the west of the survey area between north-east Hoy and Graemsay. 

Digital aerial survey count totals

The total number of birds identified to species level during the DAS of the Orkney SPAs are presented in Table 5 and Table 6 for the surveys conducted on 21 February and 20 January 2023, respectively. For the purposes of presenting raw counts, the ‘untrimmed’ records include shoreline birds just outside the perimeter of each SPA boundary and are presented for context as these birds are likely to use the SPA waters. The ‘trimmed’ records relate to birds found directly within the marine SPA boundary and these go on to inform the population analysis. The percentage difference between the trimmed and untrimmed counts is also presented. The overall identification rate of birds to species level, with possible, probable, and definite confidence (Definite: as certain as reasonably possible. Probable: very likely to be this species or species group. Possible: more likely to be this species or species group than anything else.) was 97.54 % in the 2022 survey, and 93.8 % in the 2023 survey. 

Table 5. Counts of wintering waterfowl from the DAS of Scapa Flow and North Orkney SPAs undertaken on 21 February 2022. Untrimmed counts include birds just beyond the perimeter of the marine SPA.

Target species/ Non-Target Species	Species	Scientific Name	Scapa Flow SPA Un-trimmed	Scapa Flow SPA Trimmed to SPA boundary	Scapa Flow SPA % Diff	North Orkney SPA Un-trimmed	North Orkney SPA Trimmed to SPA boundary	North Orkney SPA % Diff
Target species	Great northern diver	Gavia immer	66	64	3.03	28	19	32.14
Target species	Black-throated diver	Gavia arctica	3	3	0.00	1	1	0.00
Target species	Slavonian grebe	Podiceps auritus	18	18	0.00	9	9	0.00
Target species	Red-breasted merganser	Mergus serrator	29	19	34.48	65	51	21.54
Target species	Velvet scoter	Melanitta fusca	0	0	0.00	0	0	0.00
Target species	Long-tailed duck	Clangula hyemalis	359	336	6.41	377	374	0.80
Target species	Common eider	Somateria mollissima	192	139	27.60	224	201	10.27
Target species	European shag	Gulosus aristotelis	40	32	20.00	43	36	16.28
Target species	Red-throated diver	Gavia stellata	5	3	40.00	1	1	0.00
Non-target species	Barnacle goose	Branta leucopsis	1	1	0.00	0	0	0.00
Non-target species	Greylag goose	Anser anser	106	60	43.40	661	529	19.97
Non-target species	Pink-footed goose	Anser brachyrhynchus	0	0	0.00	17	13	23.53
Non-target species	Mute swan	Cygnus olor	0	0	0.00	2	0	100.00
Non-target species	Whooper swan	Cygnus cygnus	1	0	100.00	0	0	0.00
Non-target species	Shelduck	Tadorna tadorna	2	0	100.00	4	0	100.00
Non-target species	Wigeon	Mareca penelope	1,270	245	80.71	2,558	411	83.93
Non-target species	Mallard	Anas platyrhynchos	45	0	100.00	137	13	90.51
Non-target species	Teal	Anas crecca	14	9	35.71	143	0	100.00
Non-target species	Common scoter	Melanitta nigra	3	3	0.00	2	2	0.00
Non-target species	Goldeneye	Bucephala clangula	13	4	69.23	6	0	100.00
Non-target species	Oystercatcher	Haematopus ostralegus	159	7	95.60	50	0	100.00
Non-target species	Lapwing	Vanellus vanellus	122	122	0.00	40	39	2.50
Non-target species	Curlew	Numenius aquata	69	2	97.10	747	113	84.87
Non-target species	Bar-tailed godwit	Limosa lapponica	0	0	0.00	2	2	0.00
Non-target species	Turnstone	Arenaria interpres	0	0	0.00	19	0	100.00
Non-target species	Purple sandpiper	Calidris maritima	7	0	100.00	0	0	0.00
Non-target species	Redshank	Tringa totanus	3	3	0.00	3	3	0.00
Non-target species	Kittiwake	Rissa tridactyla	10	10	0.00	12	11	8.33
Non-target species	Black-headed gull	Chroicocephalus ridibundus	47	44	6.38	34	17	50.00
Non-target species	Common gull	Larus canus	479	330	31.11	629	240	61.84
Non-target species	Great black-backed gull	Larus marinus	40	35	12.50	38	23	39.47
Non-target species	Iceland gull	Larus glaucoides	1	1	0.00	0	0	0.00
Non-target species	Herring gull	Larus argentatus	105	55	47.6	96	43	55.21
Non-target species	Little auk	Alle alle	3	3	0.00	3	3	0.00
Non-target species	Guillemot	Uria aalge	59	59	0.00	17	16	5.88
Non-target species	Razorbill	Alca torda	25	25	0.00	8	7	12.50
Non-target species	Black guillemot	Cepphus grylle	54	46	14.81	74	69	6.76
Non-target species	White-billed diver	Gavia adamsii	1	1	0.00	0	0	0.00
Non-target species	Fulmar	Fulmarus glacialis	463	366	20.95	355	279	21.41
Non-target species	Cormorant	Phalacrocorax carbo	5	3	40.00	5	3	40.00
-	Total	Total	3,819	2,048	46.37	6,410	2,528	60.56

 

Table 6 Counts of wintering waterfowl from the DAS of Scapa Flow and North Orkney SPAs undertaken on 20 January 2023. Untrimmed counts include birds just beyond the perimeter of the marine SPA.

Target Species/ Non-target species	Species	Scientific Name	Scapa Flow SPA Un-trimmed	Scapa Flow SPA Trimmed to SPA boundary	Scapa Flow SPA % Diff	North Orkney SPA Un-trimmed	North Orkney SPA Trimmed to SPA boundary	North Orkney SPA % Diff
Target Species	Great northern diver	Gavia immer	221	220	0.45	45	45	0.00
Target Species	Black-throated diver	Gavia arctica	7	7	0.00	0	0	-
Target Species	Slavonian grebe	Podiceps auritus	29	24	17.24	22	22	0.00
Target Species	Red-breasted merganser	Mergus serrator	86	80	6.98	78	67	14.10
Target Species	Velvet scoter	Melanitta fusca	0	0	0.00	0	0	0.00
Target Species	Long-tailed duck	Clangula hyemalis	786	757	3.69	590	517	12.37
Target Species	Common eider	Somateria mollissima	354	351	0.85	556	543	2.34
Target Species	European shag	Gulosus aristotelis	198	184	7.07	329	318	3.34
Target Species	Red-throated diver	Gavia stellata	2	2	0.00	5	5	0.00
Non-target species	Barnacle goose	Branta leucopsis	13	11	15.38	5	5	0.00
Non-target species	Greylag goose	Anser anser	1,497	1,101	26.45	423	413	2.36
Non-target species	Pink-footed goose	Anser brachyrhynchus	63	63	0.00	555	554	0.18
Non-target species	Wigeon	Mareca penelope	1,146	1,015	11.43	3,366	1,538	54.31
Non-target species	Mallard	Anas platyrhynchos	72	28	61.11	104	55	47.12
Non-target species	Teal	Anas crecca	225	60	73.33	139	60	56.83
Non-target species	Common scoter	Melanitta nigra	4	4	0.00	0	0	-
Non-target species	Goldeneye	Bucephala clangula	11	9	18.18	14	0	100.00
Non-target species	Oystercatcher	Haematopus ostralegus	4	3	25.00	18	3	83.33
Non-target species	Lapwing	Vanellus vanellus	0	0	-	27	0	100.00
Non-target species	Golden plover	Pluvialis apricaria	0	0	-	367	0	100.00
Non-target species	Ringed plover	Charadrius hiaticula	0	0	-	6	0	100.00
Non-target species	Curlew	Numenius aquata	155	114	26.45	630	220	65.08
Non-target species	Bar-tailed godwit	Limosa lapponica	0	0	-	3	0	100.00
Non-target species	Dunlin	Calidris alpina	0	0	-	302	0	100.00
Non-target species	Purple sandpiper	Calidris maritima	0	0	-	13	0	100.00
Non-target species	Redshank	Tringa totanus	23	10	56.52	52	34	34.62
Non-target species	Kittiwake	Rissa tridactyla	15	15	0.00	1	0	100.00
Non-target species	Black-headed gull	Chroicocephalus ridibundus	7	5	28.57	185	25	86.49
Non-target species	Common gull	Larus canus	201	158	21.39	428	141	67.06
Non-target species	Great black-backed gull	Larus marinus	102	84	17.65	62	46	25.81
Non-target species	Glaucous gull	Larus hyperboreus	0	0	-	1	1	0.00
Non-target species	Herring gull	Larus argentatus	370	287	22.43	2,214	1,934	12.65
Non-target species	Little auk	Alle alle	4	4	0.00	5	5	0.00
Non-target species	Guillemot	Uria aalge	287	287	0.00	69	63	8.70
Non-target species	Razorbill	Alca torda	35	33	5.71	22	21	4.55
Non-target species	Black guillemot	Cepphus grylle	390	375	3.85	406	386	4.93
Non-target species	Fulmar	Fulmarus glacialis	708	647	8.62	386	325	15.80
Non-target species	Cormorant	Phalacrocorax carbo	0	0	-	2	2	0.00
Non-target species	Grey heron	Ardea cinerea	12	5	58.33	11	3	72.73
Non-target species	Total	Total	7,027	5,943	18.56	11,441	7,351	44.13

 

The total number of birds for which a species level identification could not be made, for both DAS, are presented in Table 7.

Table 7. Counts of unidentified wintering waterfowl from the DAS of Scapa Flow and North Orkney SPAs undertaken on 21 February 2022 and 20 January 2023, assigned to species groups. Untrimmed counts include birds just beyond the perimeter of the marine SPA.

Date	Species Group (Unidentified birds)	Scapa Flow SPA Untrimmed	Scapa Flow SPA Trimmed to SPA boundary	Scapa Flow SPA  % Diff	North Orkney SPA Untrimmed	North Orkney SPA Trimmed to SPA boundary	North Orkney SPA % Diff
21 February 2022 Survey	Duck species	95	18	81.05	18	6	66.67
21 February 2022 Survey	Wader species	0	0	0.00	51	1	98.04
21 February 2022 Survey	Small gull species	3	0	100.00	19	4	78.95
21 February 2022 Survey	Black-backed gull species	1	1	0.00	0	0	0.00
21 February 2022 Survey	Large gull species	1	1	0.00	2	1	50.00
21 February 2022 Survey	Gull species	5	3	40.00	3	3	0.00
21 February 2022 Survey	Large auk	8	7	12.50	0	0	0.00
21 February 2022 Survey	Auk species	14	11	21.43	10	9	10.00
21 February 2022 Survey	Auk / small gull	4	3	25.00	2	2	0.00
21 February 2022 Survey	Diver species	3	3	0.00	7	6	14.29
21 February 2022 Survey	Fulmar / gull species	3	2	33.33	8	5	37.50
21 February 2022 Survey	Total	137	49	64.23	120	37	69.17
20 January 2023 Survey	Goose species	2	2	0.00	0	-	-
20 January 2023 Survey	Duck species	67	48	28.36	132	89	32.58
20 January 2023 Survey	Wader species	81	72	11.11	325	29	91.08
20 January 2023 Survey	Small gull species	7	4	42.86	15	10	33.33
20 January 2023 Survey	Large gull species	6	5	16.67	6	6	0.00
20 January 2023 Survey	Gull species	52	41	21.15	175	25	85.71
20 January 2023 Survey	Large auk	52	51	1.92	19	18	5.26
20 January 2023 Survey	Auk species	79	76	3.80	78	72	7.69
20 January 2023 Survey	Auk / small gull	5	5	0.00	0	-	-
20 January 2023 Survey	Large auk / diver species	1	1	0.00	0	-	-
20 January 2023 Survey	Diver species	7	7	0.00	7	7	0.00
20 January 2023 Survey	Fulmar / gull species	20	17	15.00	17	14	17.65
	Total	433	329	24.02	774	270	65.12

 

The behaviours of target species recorded within the marine SPA boundaries are presented in Table 8 for both surveys. In general, very low proportions of birds were recorded flying through the SPA, with the exception of European shag with up to 25 % of birds observed flying during the February 2022 survey in North Orkney SPA.

Table 8. Behaviour of target species recorded within the Scapa Flow and North Orkney marine SPAs during the DAS undertaken on 21 February 2022 and 20 January 2023. 

Date	Species	Scapa Flow SPA - Flying	Scapa Flow SPA - Sitting	Scapa Flow SPA - % Flying	North Orkney SPA - Flying	North Orkney SPA - Sitting	North Orkney SPA – % Flying
21 February 2022 Survey	Great northern diver	0	64	0	0	19	0
21 February 2022 Survey	Black-throated diver	0	3	0	0	1	0
21 February 2022 Survey	Slavonian grebe	0	18	0	0	9	0
21 February 2022 Survey	Red-breasted merganser	0	19	0	0	51	0
21 February 2022 Survey	Velvet scoter	N/A	N/A	N/A	N/A	N/A	N/A
21 February 2022 Survey	Long-tailed duck	2	334	0.60	21	353	5.61
21 February 2022 Survey	Common eider	2	137	1.44	2	199	1.00
21 February 2022 Survey	European shag	5	27	15.63	9	27	25.00
20 January 2023	Great northern diver	1	219	0.46	1	44	2.27
20 January 2023	Black-throated diver	0	7	0.00	N/A	N/A	N/A
20 January 2023	Slavonian grebe	0	24	0.00	0	22	0.00
20 January 2023	Red-breasted merganser	1	79	1.27	1	66	1.52
20 January 2023	Velvet scoter	N/A	N/A	N/A	N/A	N/A	N/A
20 January 2023	Long-tailed duck	0	757	0.00	6	511	1.17
20 January 2023	Common eider	4	345	1.16	4	539	0.74
20 January 2023	European shag	7	162	4.38	19	297	6.46

 

Digital aerial survey population estimates (strip transect analysis)

The design-based density and population estimates of all species and species groups within the Scapa Flow SPA, calculated via strip transect analysis following the DAS, are presented in Tables 9 to 11 for target species and Tables 12 to 14 for all species groups. Estimates are for within the marine SPA boundary only and do not include birds roosting outside of this area. Estimates are presented for three scenarios: all birds within the SPA (flying and non-flying); records excluding flying birds, and flying birds only. 

Table 9. Design-based estimates of density and population size for key bird species (both flying and non-flying birds) from the DAS of Scapa Flow SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Great northern diver	0.83	263	174	369	52	0.20
21 February 2022	Black-throated diver	0.04	13	0	32	9	0.71
21 February 2022	Slavonian grebe	0.23	73	29	123	24	0.32
21 February 2022	Red-breasted merganser	0.25	79	25	157	34	0.43
21 February 2022	Long-tailed duck	4.34	1,381	719	2,160	370	0.27
21 February 2022	Common eider	1.82	581	263	961	179	0.31
21 February 2022	European shag	0.41	130	58	227	44	0.34
20 January 2023	Great northern diver	2.78	919	572	1,310	189	0.21
20 January 2023	Black-throated diver	0.09	30	0	84	26	0.86
20 January 2023	Slavonian grebe	0.30	100	50	159	28	0.28
20 January 2023	Red-breasted merganser	1.00	330	147	569	109	0.33
20 January 2023	Long-tailed duck	9.53	3,149	625	7,559	2,025	0.64
20 January 2023	Common eider	4.41	1,456	454	2,938	677	0.46
20 January 2023	European shag	2.31	764	253	1,416	305	0.40

 

Table 10. Design-based estimates of density and population size for key bird species (excluding flying birds) from the DAS of Scapa Flow SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Great northern diver	0.82	263	166	365	52	0.20
21 February 2022	Black-throated diver	0.04	12	0	32	9	0.71
21 February 2022	Slavonian grebe	0.23	74	33	123	24	0.31
21 February 2022	Red-breasted merganser	0.25	79	25	146	32	0.40
21 February 2022	Long-tailed duck	4.37	1,391	729	2,117	376	0.27
21 February 2022	Common eider	1.78	566	249	937	179	0.32
21 February 2022	European shag	0.35	113	40	211	46	0.41
20 January 2023	Great northern diver	2.75	908	551	1,321	202	0.22
20 January 2023	Black-throated diver	0.08	29	0	84	25	0.87
20 January 2023	Slavonian grebe	0.31	102	52	162	28	0.27
20 January 2023	Red-breasted merganser	1.01	336	150	572	107	0.32
20 January 2023	Long-tailed duck	9.55	3,155	621	8,621	2,087	0.66.
20 January 2023	Common eider	4.46	1,475	512	2,984	666	0.45
20 January 2023	European shag	2.07	684	209	1,257	268	0.39

 

Table 11. Design-based estimates of density and population size for key bird species (flying birds only) from the DAS of Scapa Flow SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Long-tailed duck	0.03	9	0	21	6	0.64
21 February 2022	Common eider	0.03	9	0	25	9	0.96
21 February 2022	European shag	0.07	21	4	41	10	0.45
20 January 2023	Great northern diver	0.01	5	0	13	4	0.95
20 January 2023	Red-breasted merganser	0.01	5	0	13	5	0.97
20 January 2023	Common eider	0.05	17	0	50	17	0.96
20 January 2023	European shag	0.09	29	0	66	17	0.56

 

Table 12. Design-based estimates of density and population size for all bird species groups (both flying and non-flying birds) from the DAS of Scapa Flow SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species group	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Goose species	0.78	250	52	490	116	0.47
21 February 2022	Duck species	9.94	3,165	1,386	5,843	1,154	0.36
21 February 2022	Grebe species	0.23	73	33	120	23	0.31
21 February 2022	Wader species	1.71	546	8	1,408	392	0.72
21 February 2022	Small gull species	4.34	1,383	405	2,916	639	0.46
21 February 2022	Black-backed gull species	0.14	46	13	85	19	0.41
21 February 2022	Large gull species	0.85	271	187	377	51	0.19
21 February 2022	Gull species	0.77	246	49	570	139	0.56
21 February 2022	Large auk	1.01	322	236	409	46	0.14
21 February 2022	Auk species	0.9	286	166	420	68	0.24
21 February 2022	Auk / small gull	0.05	17	0	37	10	0.60
21 February 2022	Diver species	0.94	300	212	409	52	0.17
21 February 2022	Fulmar / gull species	4.74	1,511	1,047	1,955	227	0.15
21 February 2022	Cormorant / shag	0.46	146	69	241	46	0.31
20 January 2023	Goose species	15.17	5,013	1,198	10,585	2,531	0.50
20 January 2023	Duck species	30.14	9,957	3,270	19,423	4,266	0.42
20 January 2023	Grebe species	0.30	99	50	163	29	0.29
20 January 2023	Wader species	2.51	830	199	1,804	418	0.50
20 January 2023	Small gull species	1.72	570	180	1,142	260	0.46
20 January 2023	Large gull species	4.62	1,527	780	2,585	471	0.31
20 January 2023	Gull species	1.18	389	154	671	134	0.34
20 January 2023	Large auk	4.48	1,481	930	2,035	288	0.19
20 January 2023	Auk species	5.89	1,947	1,458	2,448	257	0.13
20 January 2023	Auk / small gull	0.08	25	8	45	10	0.39
20 January 2023	Large auk / diver species	0.03	9	0	21	7	0.70
20 January 2023	Diver species	2.98	985	628	1,394	196	0.20
20 January 2023	Fulmar / gull species	8.55	2,826	1,959	3,771	476	0.17
20 January 2023	Cormorant / shag	2.31	762	268	1,415	303	0.40
20 January 2023	Heron species	0.06	21	5	44	11	0.51

 

Table 13. Design-based estimates of density and population size for all bird species groups (excluding flying birds) from the DAS of Scapa Flow SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species group	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Goose species	0.03	9	0	25	8	0.98
21 February 2022	Duck species	9.47	3,015	1,261	5,544	1,187	0.39
21 February 2022	Grebe species	0.23	74	33	123	23	0.31
21 February 2022	Small gull species	2.63	839	36	2,235	594	0.71
21 February 2022	Black-backed gull species	0.06	21	0	49	13	0.62
21 February 2022	Large gull species	0.48	152	78	241	43	0.28
21 February 2022	Gull species	0.61	193	25	465	130	0.67
21 February 2022	Large auk	1	318	231	405	45	0.14
21 February 2022	Auk species	0.87	277	155	427	68	0.25
21 February 2022	Auk / small gull	0.05	17	0	37	10	0.60
21 February 2022	Diver species	0.96	307	202	432	58	0.19
21 February 2022	Fulmar / gull species	2.39	763	482	1,126	160	0.21
21 February 2022	Cormorant / shag	0.38	121	49	221	43	0.35
20 January 2023	Goose species	1.77	586	47	1,368	334	0.57
20 January 2023	Duck species	17.05	5,634	2,211	10,797	2,299	0.41
20 January 2023	Grebe species	0.30	101	54	160	28	0.27
20 January 2023	Wader species	0.13	44	0	134	40	0.91
20 January 2023	Small gull species	0.93	307	50	703	172	0.56
20 January 2023	Large gull species	2.13	703	335	1,217	233	0.33
20 January 2023	Gull species	0.70	231	75	434	93	0.40
20 January 2023	Large auk	4.34	1,435	898	2,002	283	0.20
20 January 2023	Auk species	5.89	1,946	1,488	2,459	251	0.13
20 January 2023	Auk / small gull	0.07	25	8	46	10	0.39
20 January 2023	Large auk / diver species	0.03	9	0	21	6	0.71
20 January 2023	Diver species	2.98	984	612	1,365	196	0.20
20 January 2023	Fulmar / gull species	4.94	1,632	890	2,477	413	0.25
20 January 2023	Cormorant / shag	2.09	689	214	1,307	286	0.41
20 January 2023	Heron species	0.01	5	0	13	5	1.05

 

Table 14. Design-based estimates of density and population size for all bird species groups (flying birds only) from the DAS of Scapa Flow SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species group	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Goose species	0.76	243	49	516	122	0.50
21 February 2022	Duck species	0.59	190	9	522	156	0.82
21 February 2022	Wader species	1.72	550	8	1,502	395	0.72
21 February 2022	Small gull species	1.76	561	196	1,026	210	0.37
21 February 2022	Black-backed gull species	0.05	17	0	37	10	0.56
21 February 2022	Large gull species	0.37	119	73	167	25	0.20
21 February 2022	Gull species	0.06	21	4	41	10	0.48
21 February 2022	Large auk	0.01	4	0	13	4	0.97
21 February 2022	Auk species	0.03	9	0	20	6	0.62
21 February 2022	Fulmar / gull species	2.35	748	495	1,030	141	0.19
21 February 2022	Cormorant / shag	0.06	21	4	41	11	0.49
20 January 2023	Goose species	13.04	4,307	693	9,867	2,454	0.57
20 January 2023	Duck species	11.10	3,666	262	10,287	2,870	0.78
20 January 2023	Wader species	0.49	163	49	361	83	0.51
20 January 2023	Small gull species	0.55	181	91	292	52	0.29
20 January 2023	Large gull species	1.45	481	233	830	153	0.32
20 January 2023	Gull species	0.08	26	9	46	10	0.38
20 January 2023	Large auk	0.12	41	0	100	27	0.65
20 January 2023	Auk species	0.01	5	0	13	4	0.98
20 January 2023	Diver species	0.01	5	0	13	4	0.95
20 January 2023	Fulmar / gull species	3.36	1,111	699	1,626	236	0.21
20 January 2023	Cormorant / shag	0.09	29	0	66	17	0.56

 

The density and population estimates of all species and species groups within the North Orkney SPA, calculated via strip transect analysis following the DAS, are presented in Tables 15 to 17 for the target species and  Tables 18 to 20 for all species groups. Estimates are for within the marine SPA boundary only and do not include birds roosting outside of this area. Estimates are presented for three scenarios: all birds within the SPA (flying and non-flying); records excluding flying birds, and flying birds only. 

Table 15. Design-based estimates of density and population size for key bird species (both flying and non-flying birds) from the DAS of North Orkney SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Great northern diver	0.37	79	30	140	29	0.37
21 February 2022	Black-throated diver	0.02	5	0	13	4	0.98
21 February 2022	Slavonian grebe	0.18	38	13	66	14	0.36
21 February 2022	Red-breasted merganser	1.01	214	37	506	127	0.59
21 February 2022	Long-tailed duck	7.41	1,570	554	3,170	684	0.44
21 February 2022	Common eider	3.96	840	328	1,535	317	0.38
21 February 2022	European shag	0.71	150	79	239	42	0.28
20 January 2023	Great northern diver	0.84	191	105	300	51	0.26
20 January 2023	Slavonian grebe	0.41	94	24	206	52	0.55
20 January 2023	Red-breasted merganser	1.25	284	68	663	159	0.56
20 January 2023	Long-tailed duck	9.62	2,190	1,051	3,331	576	0.26
20 January 2023	Common eider	9.95	2,266	285	5,842	1,586	0.70
20 January 2023	European shag	5.91	1,345	432	2,562	541	0.40

 

Table 16. Design-based estimates of density and population size for key bird species (excluding flying birds) from the DAS of North Orkney SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Great northern diver	0.38	81	37	143	29	0.35
21 February 2022	Black-throated diver	0.02	5	0	13	5	0.97
21 February 2022	Slavonian grebe	0.18	38	16	66	14	0.35
21 February 2022	Red-breasted merganser	1.02	217	34	537	137	0.63
21 February 2022	Long-tailed duck	7.01	1,486	540	2,911	629	0.42
21 February 2022	Common eider	3.95	837	307	1,481	310	0.37
21 February 2022	European shag	0.54	115	50	192	36	0.31
20 January 2023	Great northern diver	0.81	186	104	299	51	0.27
20 January 2023	Slavonian grebe	0.41	93	23	208	51	0.55
20 January 2023	Red-breasted merganser	1.23	281	62	673	168	0.60
20 January 2023	Long-tailed duck	9.31	2,119	1,038	3,205	576	0.27
20 January 2023	Common eider	10.06	2,291	254	5,782	1,578	0.69
20 January 2023	European shag	5.61	1,276	433	2,375	507	0.40

 

Table 17. Design-based estimates of density and population size for key bird species (flying birds only) from the DAS of North Orkney SPA undertaken on 21 February 2022 and 20 January 2023. 

Dae	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Long-tailed duck	0.41	88	0	254	80	0.91
21 February 2022	Common eider	0.04	9	0	26	9	0.94
21 February 2022	European shag	0.18	38	9	80	19	0.50
20 January 2023	Great northern diver	0.02	5	0	13	5	0.97
20 January 2023	Red-breasted merganser	0.02	5	0	13	5	0.94
20 January 2023	Long-tailed duck	0.11	26	5	54	14	0.51
20 January 2023	Common eider	0.08	18	0	48	13	0.75
20 January 2023	European shag	0.37	84	17	186	46	0.55

 

Table 18. Design-based estimates of density and population for all bird species groups (both flying and non-flying birds) from the DAS of North Orkney SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species group	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Goose species	10.54	2,233	474	4,657	1,089	0.49
21 February 2022	Duck species	20.99	4,449	2,400	7,317	12,51	0.28
21 February 2022	Grebe species	0.18	38	13	67	14	0.36
21 February 2022	Wader species	3.14	666	54	1,512	382	0.57
21 February 2022	Small gull species	5.21	1,104	403	2,099	454	0.41
21 February 2022	Black-backed gull species	0.24	51	17	90	19	0.37
21 February 2022	Large gull species	0.91	193	103	296	50	0.26
21 February 2022	Gull species	0.29	63	24	116	24	0.38
21 February 2022	Large auk	0.34	72	37	112	20	0.27
21 February 2022	Auk species	1.72	365	234	510	70	0.19
21 February 2022	Auk / small gull	0.04	9	0	21	6	0.70
21 February 2022	Diver species	0.54	116	57	184	33	0.28
21 February 2022	Fulmar / gull species	5.49	1,165	653	1,828	305	0.26
21 February 2022	Cormorant / shag	0.78	166	89	250	43	0.25
20 January 2023	Goose species	17.90	4,075	165	10,117	2,705	0.66
20 January 2023	Duck species	52.73	12,002	4,934	20,713	3,986	0.33
20 January 2023	Grebe species	0.42	96	24	212	52	0.55
20 January 2023	Wader species	5.40	1,229	240	2,457	576	0.47
20 January 2023	Small gull species	3.04	692	399	1,103	190	0.27
20 January 2023	Large gull species	35.5	8,080	672	22,448	6,172	0.76
20 January 2023	Gull species	0.98	223	112	355	63	0.28
20 January 2023	Large auk	1.64	375	274	484	55	0.15
20 January 2023	Auk species	8.79	2,002	1,713	2,279	145	0.07
20 January 2023	Auk / small gull	0.02	5	0	12	4	0.84
20 January 2023	Diver species	1.04	237	140	361	56	0.24
20 January 2023	Fulmar / gull species	6.50	1,481	737	2,519	458	0.31
20 January 2023	Cormorant / shag	5.95	1,356	475	2,684	564	0.42
20 January 2023	Heron species	0.05	13	0	34	10	0.74

 

Table 19. Design-based estimates of density and population for all bird species groups (excluding flying birds) from the DAS of North Orkney SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species group	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Duck species	14.21	3,011	1,651	4,715	799	0.27
21 February 2022	Grebe species	0.18	38	16	63	13	0.34
21 February 2022	Small gull species	1.46	309	67	675	163	0.52
21 February 2022	Black-backed gull species	0.08	17	0	41	11	0.60
21 February 2022	Large gull species	0.25	54	21	102	21	0.38
21 February 2022	Gull species	0.17	37	9	79	19	0.51
21 February 2022	Large auk	0.32	67	34	104	18	0.26
21 February 2022	Auk species	1.6	339	212	480	70	0.20
21 February 2022	Auk / small gull	0.04	9	0	21	7	0.70
21 February 2022	Diver species	0.53	113	57	187	33	0.29
21 February 2022	Fulmar / gull species	3.21	680	333	1,144	219	0.32
21 February 2022	Cormorant / shag	0.53	113	50	192	37	0.33
20 January 2023	Goose species	0.30	68	0	195	63	0.91
20 January 2023	Duck species	41.30	9,402	3,183	18,093	3,713	0.39
20 January 2023	Grebe species	0.42	96	22	219	54	0.56
20 January 2023	Small gull species	0.82	187	77	329	65	0.35
20 January 2023	Large gull species	33.53	7,634	252	21,306	6,061	0.79
20 January 2023	Gull species	0.51	117	51	192	36	0.30
20 January 2023	Large auk	1.35	308	188	442	67	0.22
20 January 2023	Auk species	8.69	1,978	1,682	2,258	146	0.07
20 January 2023	Auk / small gull	0.02	5	0	13	4	0.85
20 January 2023	Diver species	1.03	235	138	351	56	0.24
20 January 2023	Fulmar / gull species	4.62	1,051	366	1,971	430	0.41
20 January 2023	Cormorant / shag	5.58	1,272	395	2,412	525	0.41

 

Table 20. Design-based estimates of density and population for all bird species groups (flying birds only) from the DAS of North Orkney SPA undertaken on 21 February 2022 and 20 January 2023. 

Date	Species group	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
21 February 2022	Goose species	10.71	2,269	518	4,521	1,058	0.47
21 February 2022	Duck species	6.58	1,394	201	3,165	755	0.54
21 February 2022	Wader species	3.13	665	46	1,511	380	0.57
21 February 2022	Small gull species	3.77	799	304	1,467	314	0.39
21 February 2022	Black-backed gull species	0.16	34	13	64	14	0.41
21 February 2022	Large gull species	0.63	134	66	214	38	0.28
21 February 2022	Gull species	0.12	25	8	47	11	0.43
21 February 2022	Large auk	0.02	5	0	13	4	0.92
21 February 2022	Auk species	0.14	30	5	55	14	0.45
21 February 2022	Fulmar / gull species	2.3	488	299	696	101	0.21
21 February 2022	Cormorant / shag	0.24	52	21	95	20	0.37
20 January 2023	Goose species	17.98	4,093	163	10,594	2,883	0.70
20 January 2023	Duck species	11.13	2,535	467	5,458	1,314	0.52
20 January 2023	Wader species	5.24	1,192	214	2,452	580	0.49
20 January 2023	Small gull species	1.37	311	182	456	69	0.22
20 January 2023	Large gull species	2.53	575	199	1,112	248	0.43
20 January 2023	Gull species	0.13	30	4	63	16	0.53
20 January 2023	Large auk	0.30	70	5	149	38	0.54
20 January 2023	Auk species	0.09	22	4	47	12	0.54
20 January 2023	Diver species	0.02	5	0	13	5	0.92
20 January 2023	Fulmar / gull species	1.66	377	239	545	77	0.20
20 January 2023	Cormorant / shag	0.35	80	17	187	45	0.56

 

The density and population estimates of all non-target species within the Scapa Flow SPA and North Orkney SPA, calculated via strip transect analysis following the DAS are presented in Table A. 1 to Table A. 12 in the Appendix. Estimates are presented for three scenarios: all birds within the SPA (flying and non-flying); records excluding flying birds, and flying birds only.

Table A. 1. Design-based estimates of density and population size for all bird species (both flying and non-flying birds) from the DAS of Scapa Flow SPA undertaken on 21 February 2022. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	0.83	263	174	369	52	0.20
Target species	Black-throated diver	0.04	13	0	32	9	0.71
Target species	Slavonian grebe	0.23	73	29	123	24	0.32
Target species	Red-breasted merganser	0.25	79	25	157	34	0.43
Target species	Long-tailed duck	4.34	1,381	719	2,160	370	0.27
Target species	Common eider	1.82	581	263	961	179	0.31
Target species	European shag	0.41	130	58	227	44	0.34
Target species	Red-throated diver	0.04	13	0	28	7	0.54
Non-target species	Barnacle goose	0.01	5	0	16	5	0.98
Non-target species	Greylag goose	0.75	240	56	494	115	0.48
Non-target species	Wigeon	3.15	1,004	8	3,039	811	0.81
Non-target species	Teal	0.12	38	0	140	39	1.02
Non-target species	Common scoter	0.04	13	0	45	13	1.03
Non-target species	Goldeneye	0.05	17	0	50	17	0.98
Non-target species	Oystercatcher	0.09	29	0	82	26	0.90
Non-target species	Lapwing	1.53	487	0	1,414	383	0.79
Non-target species	Curlew	0.03	9	0	25	8	0.96
Non-target species	Redshank	0.04	13	0	37	13	0.98
Non-target species	Kittiwake	0.13	41	12	86	20	0.48
Non-target species	Black-headed gull	0.57	182	39	371	86	0.47
Non-target species	Common gull	4.32	1,377	324	2,991	702	0.51
Non-target species	Great black- backed gull	0.46	146	93	207	30	0.20
Non-target species	Iceland gull	0.01	5	0	13	4	0.89
Non-target species	Herring gull	0.70	224	128	334	54	0.24
Non-target species	Little auk	0.04	13	0	36	10	0.77
Non-target species	Guillemot	0.76	242	171	313	37	0.15
Non-target species	Razorbill	0.33	104	33	196	41	0.39
Non-target species	Black guillemot	0.58	184	112	263	40	0.21
Non-target species	White-billed diver	0.01	4	0	13	4	1.02
Non-target species	Fulmar	4.69	1,495	1,056	1,941	224	0.15
Non-target species	Cormorant	0.04	13	0	29	7	0.56

 

Table A. 2. Design-based estimates of density and population size for all bird species (both flying and non-flying birds) from the DAS of Scapa Flow SPA undertaken on 20 January 2023. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	2.78	919	572	1,310	189	0.21
Target species	Black-throated diver	0.09	30	0	84	26	0.86
Target species	Slavonian grebe	0.30	100	50	159	28	0.28
Target species	Red-breasted merganser	1.00	330	147	569	109	0.33
Target species	Long-tailed duck	9.53	3,149	625	7,559	2,025	0.64
Target species	Common eider	4.41	1,456	454	2,938	677	0.46
Target species	European shag	2.31	764	253	1,416	305	0.40
Non-target species	Barnacle goose	0.14	47	0	147	41	0.88
Non-target species	Greylag goose	13.94	4,606	967	10,342	2,468	0.54
Non-target species	Pink-footed goose	0.80	266	0	653	189	0.71
Non-target species	Wigeon	13.05	4,312	382	10,915	3,224	0.75
Non-target species	Mallard	0.34	113	5	299	82	0.72
Non-target species	Teal	0.80	264	0	829	222	0.84
Non-target species	Common scoter	0.05	17	0	50	17	1.03
Non-target species	Goldeneye	0.11	38	0	108	34	0.88
Non-target species	Oystercatcher	0.04	13	0	29	7	0.53
Non-target species	Curlew	1.41	466	119	1,007	230	0.49
Non-target species	Redshank	0.13	42	0	124	38	0.90
Non-target species	Kittiwake	0.19	63	25	106	21	0.32
Non-target species	Black-headed gull	0.06	21	5	41	10	0.47
Non-target species	Common gull	1.97	653	254	1,230	260	0.40
Non-target species	Great black-backed gull	1.06	350	194	565	96	0.27
Non-target species	Herring gull	3.57	1,181	567	1,998	372	0.31
Non-target species	Little auk	0.05	17	4	34	8	0.47
Non-target species	Guillemot	3.61	1,193	694	1,689	262	0.22
Non-target species	Razorbill	0.42	138	66	221	42	0.30
Non-target species	Black guillemot	4.71	1,556	1,173	2,040	217	0.14
Non-target species	Red-throated diver	0.03	9	0	21	6	0.61
Non-target species	Black-throated diver	0.09	30	0	84	26	0.86
Non-target species	Fulmar	8.18	2,702	1,883	3,655	457	0.17
Non-target species	Grey heron	0.06	21	5	46	11	0.52

 

Table A. 3. Design-based estimates of density and population size for all bird species (excluding flying birds) from the DAS of Scapa Flow SPA undertaken on 21 February 2022. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	0.82	263	166	365	52	0.20
Target species	Black-throated diver	0.04	12	0	32	9	0.71
Target species	Slavonian grebe	0.23	74	33	123	24	0.31
Target species	Red-breasted merganser	0.25	79	25	146	32	0.40
Target species	Long-tailed duck	4.37	1,391	729	2,117	376	0.27
Target species	Common eider	1.78	566	249	937	179	0.32
Target species	European shag	0.35	113	40	211	46	0.41
Target species	Red-throated diver	0.04	13	0	27	7	0.54
Non-target species	Greylag goose	0.02	8	0	25	8	1.00
Non-target species	Wigeon	2.65	845	8	2,385	761	0.90
Non-target species	Teal	0.12	38	0	111	38	1.01
Non-target species	Common scoter	0.04	13	0	37	13	0.96
Non-target species	Goldeneye	0.05	17	0	50	17	1.01
Non-target species	Kittiwake	0.01	5	0	16	5	0.98
Non-target species	Common gull	3.04	970	80	2,562	672	0.69
Non-target species	Great black-backed gull	0.28	90	44	149	27	0.29
Non-target species	Herring gull	0.37	118	48	202	42	0.35
Non-target species	Little auk	0.04	13	0	37	10	0.77
Non-target species	Guillemot	0.74	237	169	305	36	0.15
Non-target species	Razorbill	0.31	100	32	183	39	0.39
Non-target species	Black guillemot	0.54	173	101	260	42	0.24
Non-target species	White-billed diver	0.01	5	0	16	5	1.02
Non-target species	Fulmar	2.38	758	462	1,063	153	0.20
Non-target species	Cormorant	0.04	12	0	28	7	0.56

 

Table A. 4. Design-based estimates of density and population size for all bird species (excluding flying birds) from the DAS of Scapa Flow SPA undertaken on 20 January 2023. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	2.75	908	551	1,321	202	0.22
Target species	Black-throated diver	0.08	29	0	84	25	0.87
Target species	Slavonian grebe	0.31	102	52	162	28	0.27
Target species	Red-breasted merganser	1.01	336	150	572	107	0.32
Target species	Long-tailed duck	9.55	3,155	621	8,621	2,087	0.66
Target species	Common eider	4.46	1,475	512	2,984	666	0.45
Target species	European shag	2.07	684	209	1,257	268	0.39
Target species	Red-throated diver	0.03	9	0	21	6	0.65
Non-target species	Barnacle goose	1.61	533	48	1,192	318	0.60
Non-target species	Greylag goose	0.16	52	0	162	52	1.00
Non-target species	Pink-footed goose	1.40	462	33	1,200	347	0.75
Non-target species	Wigeon	0.02	8	0	25	9	1.02
Non-target species	Mallard	0.01	5	0	13	5	1.00
Non-target species	Teal	1.61	533	48	1,192	318	0.60
Non-target species	Common scoter	0.05	17	0	63	18	1.04
Non-target species	Goldeneye	0.11	37	0	104	32	0.86
Non-target species	Redshank	0.13	43	0	124	39	0.90
Non-target species	Kittiwake	0.01	5	0	13	4	0.94
Non-target species	Black-headed gull	0.01	5	0	16	5	1.05
Non-target species	Common gull	1.28	423	125	786	178	0.42
Non-target species	Great black-backed gull	0.54	180	106	263	42	0.23
Non-target species	Herring gull	1.62	535	221	961	199	0.37
Non-target species	Little auk	0.05	17	4	32	8	0.46
Non-target species	Guillemot	3.54	1,171	708	1,711	257	0.22
Non-target species	Razorbill	0.38	126	53	203	39	0.30
Non-target species	Black guillemot	4.72	1,561	1,145	2,062	231	0.15
Non-target species	Fulmar	4.62	1,528	797	2,455	422	0.28
Non-target species	Grey heron	0.01	5	0	13	5	0.95

 

Table A. 5. Design-based estimates of density and population size for all bird species (flying birds only) from the DAS of Scapa Flow SPA undertaken on 21 February 2022. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Long-tailed duck	0.03	9	0	21	6	0.64
Target species	Common eider	0.03	9	0	25	9	0.96
Target species	European shag	0.07	21	4	41	10	0.45
Non-target species	Barnacle goose	0.01	5	0	13	5	1.02
Non-target species	Greylag goose	0.73	234	47	486	120	0.51
Non-target species	Wigeon	0.58	187	0	600	161	0.86
Non-target species	Oystercatcher	0.09	30	0	82	25	0.86
Non-target species	Lapwing	1.6	509	0	1,456	387	0.76
Non-target species	Curlew	0.03	9	0	25	8	0.96
Non-target species	Kittiwake	0.12	37	5	82	20	0.54
Non-target species	Black-headed gull	0.52	167	24	332	81	0.48
Non-target species	Common gull	1.09	347	119	629	133	0.38
Non-target species	Great black-backed gull	0.14	45	23	70	13	0.28
Non-target species	Iceland gull	0.01	5	0	13	4	0.97
Non-target species	Herring gull	0.32	102	61	148	23	0.22
Non-target species	Razorbill	0.01	5	0	13	4	0.92
Non-target species	Black guillemot	0.03	9	0	20	6	0.58
Non-target species	Fulmar	2.34	746	490	1,066	146	0.20

 

Table A. 6. Design-based estimates of density and population size for all bird species (flying birds only) from the DAS of Scapa Flow SPA undertaken on 20 January 2023. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	0.01	5	0	13	4	0.95
Target species	Red-breasted merganser	0.01	5	0	13	5	0.98
Target species	Common eider	0.05	17	0	50	17	0.96
Target species	European shag	0.09	29	0	66	17	0.56
Non-target species	Barnacle goose	0.14	46	0	125	38	0.81
Non-target species	Greylag goose	12.48	4,124	573	10,311	2,571	0.62
Non-target species	Pink-footed goose	0.38	125	0	468	128	1.03
Non-target species	Wigeon	9.82	3,245	217	10,631	2,844	0.88
Non-target species	Mallard	0.33	110	0	301	86	0.79
Non-target species	Teal	0.78	258	0	793	215	0.83
Non-target species	Oystercatcher	0.04	13	0	28	7	0.51
Non-target species	Curlew	0.45	150	37	330	83	0.55
Non-target species	Kittiwake	0.18	59	28	95	18	0.30
Non-target species	Black-headed gull	0.05	17	4	36	9	0.52
Non-target species	Common gull	0.37	124	50	213	43	0.34
Non-target species	Great black-backed gull	0.35	118	46	215	45	0.38
Non-target species	Herring gull	1.08	356	172	605	113	0.32
Non-target species	Guillemot	0.09	29	0	67	19	0.65
Non-target species	Razorbill	0.04	13	0	38	13	0.95
Non-target species	Black guillemot	0.01	5	0	16	5	1.02
Non-target species	Fulmar	0.01	5	0	13	4	0.95
Non-target species	Grey heron	3.39	1,120	703	1,617	237	0.21

 

Table A. 7. Design-based estimates of density and population size for all bird species (both flying and non-flying birds) from the DAS of North Orkney SPA undertaken on 21 February 2022. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	0.37	79	30	140	29	0.37
Target species	Black-throated diver	0.02	5	0	13	4	0.98
Target species	Slavonian grebe	0.18	38	13	66	14	0.36
Target species	Red-breasted merganser	1.01	214	37	506	127	0.59
Target species	Long-tailed duck	7.41	1,570	554	3,170	684	0.44
Target species	Common eider	3.96	840	328	1,535	317	0.38
Target species	European shag	0.71	150	79	239	42	0.28
Target species	Red-throated diver	0.02	4	0	13	5	1.03
Non-target species	Greylag goose	10.38	2,199	475	4,692	1,107	0.50
Non-target species	Pink-footed goose	0.26	56	0	152	45	0.80
Non-target species	Wigeon	8.21	1,740	208	3,674	882	0.51
Non-target species	Mallard	0.27	58	0	153	43	0.73
Non-target species	Common scoter	0.04	9	0	25	9	0.94
Non-target species	Lapwing	0.77	163	0	486	158	0.97
Non-target species	Curlew	2.28	484	0	1,193	332	0.69
Non-target species	Bar-tailed godwit	0.04	9	0	26	9	0.98
Non-target species	Redshank	0.06	13	0	33	10	0.73
Non-target species	Kittiwake	0.21	46	9	97	24	0.52
Non-target species	Black-headed gull	0.33	70	16	144	33	0.47
Non-target species	Common gull	4.65	986	338	1,978	436	0.44
Non-target species	Great black-backed gull	0.46	98	42	170	34	0.34
Non-target species	Herring gull	0.82	175	93	267	46	0.26
Non-target species	Little auk	0.06	13	0	37	10	0.74
Non-target species	Guillemot	0.32	67	30	113	21	0.31
Non-target species	Razorbill	0.14	29	5	62	16	0.54
Non-target species	Black guillemot	1.38	292	191	410	56	0.19
Non-target species	Fulmar	5.48	1,162	624	1,818	301	0.26
Non-target species	Cormorant	0.06	13	0	29	7	0.55

 

Table A. 8. Design-based estimates of density and population size for all bird species (both flying and non-flying birds) from the DAS of North Orkney SPA undertaken on 20 January 2023. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	0.84	191	105	300	51	0.26
Target species	Slavonian grebe	0.41	94	24	206	52	0.55
Target species	Red-breasted merganser	1.25	284	68	663	159	0.56
Target species	Long-tailed duck	9.62	2,190	1,051	3,331	576	0.26
Target species	Common eider	9.95	2,266	285	5,842	1,586	0.70
Target species	European shag	5.91	1,345	432	2,562	541	0.40
Target species	Red-throated diver	0.09	22	0	49	13	0.60
Non-target species	Barnacle goose	0.09	21	0	64	22	1.05
Non-target species	Greylag goose	7.30	1,661	135	4,607	1,338	0.81
Non-target species	Pink-footed goose	10.21	2,325	0	6,933	2,236	0.96
Non-target species	Wigeon	28.99	6,599	1,395	15,906	3,830	0.58
Non-target species	Mallard	1.02	232	71	440	98	0.42
Non-target species	Teal	1.11	253	58	527	119	0.47
Non-target species	Oystercatcher	0.05	13	0	39	13	1.03
Non-target species	Curlew	4.05	923	58	2,084	512	0.55
Non-target species	Redshank	0.66	151	0	472	135	0.89
Non-target species	Black-headed gull	0.47	108	13	253	66	0.61
Non-target species	Common gull	2.67	608	278	1,041	202	0.33
Non-target species	Great black-backed gull	0.86	195	128	277	40	0.20
Non-target species	Glaucous gull	0.02	5	0	13	4	0.83
Non-target species	Herring gull	34.73	7,905	523	22,218	5,999	0.76
Non-target species	Little auk	0.09	21	5	38	9	0.41
Non-target species	Guillemot	1.16	265	176	350	45	0.17
Non-target species	Razorbill	0.39	90	17	209	52	0.57
Non-target species	Black guillemot	7.16	1,629	1,364	1,911	142	0.09
Non-target species	Fulmar	6.10	1,389	632	2,438	472	0.34
Non-target species	Cormorant	0.04	9	0	22	7	0.73
Non-target species	Grey heron	0.05	13	0	37	10	0.78

 

Table A. 9. Design-based estimates of density and population size for all bird species (excluding flying birds) from the DAS of North Orkney SPA undertaken on 21 February 2022. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	0.38	81	37	143	29	0.35
Target species	Black-throated diver	0.02	5	0	13	5	0.97
Target species	Slavonian grebe	0.18	38	16	66	14	0.35
Target species	Red-breasted merganser	1.02	217	34	537	137	0.63
Target species	Long-tailed duck	7.01	1,486	540	2,911	629	0.42
Target species	Common eider	3.95	837	307	1,481	310	0.37
Target species	European shag	0.54	115	50	192	36	0.31
Target species	Red-throated diver	0.02	5	0	13	4	0.95
Non-target species	Wigeon	2.11	448	0	1,327	436	0.97
Non-target species	Mallard	0.16	35	0	96	26	0.72
Non-target species	Common scoter	0.04	9	0	25	8	0.96
Non-target species	Kittiwake	0.14	29	0	74	21	0.69
Non-target species	Common gull	1.34	284	67	638	153	0.54
Non-target species	Great black-backed gull	0.14	30	4	74	19	0.62
Non-target species	Herring gull	0.28	59	30	92	16	0.27
Non-target species	Little auk	0.06	13	0	33	9	0.70
Non-target species	Guillemot	0.29	62	29	104	20	0.31
Non-target species	Razorbill	0.14	31	5	67	16	0.53
Non-target species	Black guillemot	1.23	262	172	363	50	0.19
Non-target species	Fulmar	3.1	656	295	1,119	213	0.32

 

Table A. 10. Design-based estimates of density and population size for all bird species (excluding flying birds) from the DAS of North Orkney SPA undertaken on 20 January 2023.

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	0.81	186	104	299	51	0.27
Target species	Slavonian grebe	0.41	93	23	208	51	0.55
Target species	Red-breasted merganser	1.23	281	62	673	168	0.60
Target species	Long-tailed duck	9.31	2,119	1,038	3,205	576	0.27
Target species	Common eider	10.06	2,291	254	5,782	1,578	0.69
Target species	European shag	5.61	1,276	433	2,375	507	0.40
Target species	Red-throated diver	0.09	22	0	52	14	0.62
Non-target species	Greylag goose	0.29	66	0	198	62	0.94
Non-target species	Wigeon	17.35	3,949	253	10,637	3,218	0.81
Non-target species	Mallard	0.56	129	15	281	70	0.54
Non-target species	Teal	0.70	160	0	404	106	0.66
Non-target species	Black-headed gull	0.05	13	0	29	8	0.59
Non-target species	Common gull	0.73	167	66	310	65	0.39
Non-target species	Great black-backed gull	0.56	129	64	200	35	0.27
Non-target species	Glaucous gull	0.02	5	0	13	4	0.84
Non-target species	Herring gull	32.45	7,387	216	20,334	5,662	0.77
Non-target species	Little auk	0.09	22	9	42	9	0.42
Non-target species	Guillemot	0.87	197	135	265	33	0.17
Non-target species	Razorbill	0.39	90	17	199	50	0.55
Non-target species	Black guillemot	7.07	1,610	1,298	1,894	153	0.09
Non-target species	Fulmar	4.27	972	318	1,915	417	0.43
Non-target species	Cormorant	0.02	5	0	15	5	1.00
Non-target species	Grey heron	0.02	5	0	16	5	1.01

 

Table A. 11. Design-based estimates of density and population size for all bird species (flying birds only) from the DAS of North Orkney SPA undertaken on 21 February 2022. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Common eider	0.04	9	0	26	9	0.94
Target species	Long-tailed duck	0.41	88	0	254	80	0.91
Target species	European shag	0.18	38	9	80	19	0.50
Non-target species	Greylag goose	10.32	2,187	492	4,707	1,108	0.51
Non-target species	Pink-footed goose	0.26	56	0	159	47	0.84
Non-target species	Wigeon	6.03	1,279	130	2,986	733	0.57
Non-target species	Mallard	0.1	21	0	63	21	1.02
Non-target species	Lapwing	0.77	164	0	487	157	0.95
Non-target species	Curlew	2.28	485	0	1,259	343	0.71
Non-target species	Bar-tailed godwit	0.04	9	0	33	9	1.00
Non-target species	Redshank	0.02	5	0	13	5	1.00
Non-target species	Kittiwake	0.08	17	5	33	8	0.44
Non-target species	Black-headed gull	0.33	71	17	146	34	0.47
Non-target species	Common gull	3.42	724	245	1,385	297	0.41
Non-target species	Great black-backed gull	0.32	67	28	117	24	0.36
Non-target species	Herring gull	0.53	112	54	180	33	0.29
Non-target species	Guillemot	0.02	5	0	13	5	0.99
Non-target species	Black guillemot	0.14	30	5	58	14	0.46
Non-target species	Fulmar	2.31	489	294	699	106	0.22
Non-target species	Cormorant	0.06	13	0	25	7	0.52

 

Table A. 12. Design-based estimates of density and population size for all bird species (flying birds only) from the DAS of North Orkney SPA undertaken on 20 January 2023. 

Target species/ Non-target species	Species	Density estimate (n/km²)	Population estimate (number)	Lower 95 % confidence limit of population (number)	Upper 95 % confidence limit of population (number)	Standard deviation of population estimate (number)	CV
Target species	Great northern diver	0.02	5	0	13	5	0.97
Target species	Red-breasted merganser	0.02	5	0	13	5	0.94
Target species	Long-tailed duck	0.11	26	5	54	14	0.51
Target species	Common eider	0.08	18	0	48	13	0.75
Target species	European shag	0.37	84	17	186	46	0.55
Non-target species	Barnacle goose	0.09	21	0	63	21	0.97
Non-target species	Greylag goose	7.31	1,665	133	5,547	1,434	0.86
Non-target species	Pink-footed goose	10.28	2,341	0	6,932	2,243	0.96
Non-target species	Wigeon	9.79	2,230	342	5,055	1,223	0.55
Non-target species	Mallard	0.42	95	0	263	69	0.73
Non-target species	Teal	0.39	89	0	220	61	0.69
Non-target species	Oystercatcher	0.06	13	0	38	14	1.01
Non-target species	Curlew	4.01	914	66	2,165	524	0.57
Non-target species	Redshank	0.62	142	0	407	128	0.90
Non-target species	Black-headed gull	0.41	93	13	223	59	0.63
Non-target species	Common gull	1.01	231	118	380	70	0.30
Non-target species	Great black-backed gull	0.15	34	12	62	13	0.38
Non-target species	Herring gull	2.37	539	165	1,078	244	0.45
Non-target species	Guillemot	0.29	67	5	145	38	0.57
Non-target species	Black guillemot	0.09	21	4	46	12	0.53
Non-target species	Fulmar	1.59	363	232	523	75	0.21

 

Digital aerial survey population estimates (Bayesian point processing)

Population estimates of the eight target species within the Scapa Flow SPA, calculated via Bayesian point processing of the DAS data, are presented in Table 21. Estimates are derived from all birds, both flying and sitting, within the SPA boundary only and do not include birds roosting outside of this area. 

Table 21. Population estimates of target species within Scapa Flow SPA estimated through Bayesian point processing of the DAS data undertaken on 21 February 2022 and 20 January 2023. See Table 2 for definitions of statistical terms.

Date	Species	Population estimate (Mean)	Lower 95 % credible limits of population	Upper 95 % credible limits of population	R2	Root mean squared error	Deviance information criterion
21 February 2022 Survey	Great Northern diver	262	197	343	0.15	0.82	540.87
21 February 2022 Survey	Black-throated diver*	-	-	-	-	-	-
21 February 2022 Survey	Slavonian grebe	70	40	123	0.39	0.27	145.82
21 February 2022 Survey	Velvet scoter	N/A	N/A	N/A	N/A	N/A	N/A
21 February 2022 Survey	Red-breasted merganser	102	56	200	0.33	0.61	131.04
21 February 2022 Survey	Long-tailed duck	1,502	1,279	1,792	0.49	6.10	1,995.07
21 February 2022 Survey	Common eider	676	534	824	0.42	2.88	839.85
21 February 2022 Survey	European shag	134	88	193	0.22	0.47	281.52
20 January 2023 Survey	Great Northern diver	1,072	884	1,232	0.33	4.05	538.42
20 January 2023 Survey	Black-throated diver*	-	-	-	-	-	-
20 January 2023 Survey	Slavonian grebe	114	69	175	0.23	0.43	54.56
20 January 2023 Survey	Velvet scoter	N/A	N/A	N/A	N/A	N/A	N/A
20 January 2023 Survey	Red-breasted merganser	338	255	435	0.33	1.27	198.57
20 January 2023 Survey	Long-tailed duck	3,678	3,321	4,343	0.17	20.16	4,218.97
20 January 2023 Survey	Common eider	1,504	1,322	1,706	0.11	8.62	1,574.91
20 January 2023 Survey	European shag	872	597	1,560	0.18	3.77	837.84

 

Population estimates of the eight target species within the North Orkney SPA, calculated via Bayesian point processing of the DAS data, are presented in Table 22. Estimates are derived from all birds, both flying and sitting, within the SPA boundary only and do not include birds roosting outside of this area.

Table 22. Population estimates of target species within North Orkney SPA estimated through Bayesian point processing of the DAS data undertaken on 21 February 2022 and 20 January 2023. See Table 2 for definitions of statistical terms.

Date	Species	Population estimate (Mean)	Lower 95 % credible limits of population	Upper 95 % credible limits of population	R2	Root mean squared error	Deviance information criterion
21 February 2022 Survey	Great northern diver	76	43	122	0.14	0.53	142.68
21 February 2022 Survey	Black-throated diver*	-	-	-	-	-	-
21 February 2022 Survey	Slavonian grebe*	-	-	-	-	-	-
21 February 2022 Survey	Velvet scoter*	-	-	-	-	-	-
21 February 2022 Survey	Red-breasted merganser	214	135	305	0.78	2.26	179.13
21 February 2022 Survey	Long-tailed duck	1,915	1,668	2,734	0.73	14.79	1,636.94
21 February 2022 Survey	Common eider	905	759	1,580	0.56	5.56	1,150.25
21 February 2022 Survey	European shag	161	109	218	0.27	0.74	291.22
20 January 2023 Survey	Great northern diver	208	145	282	0.30	0.91	110.55
20 January 2023 Survey	Black-throated diver*	-	-	-	-	-	-
20 January 2023 Survey	Slavonian grebe	103	61	163	0.36	0.72	0.53
20 January 2023 Survey	Velvet scoter*	-	-	-	-	-	-
20 January 2023 Survey	Red-breasted merganser	331	236	521	0.56	2.08	37.17
20 January 2023 Survey	Long-tailed duck	2,428	2,199	2,654	0.38	11.03	2,307.24
20 January 2023 Survey	Common eider	2,134	1,887	2,421	0.44	20.45	1,459.23
20 January 2023 Survey	European shag	1,667	1,356	2,189	0.43	13.62	805.28

 

*Analysis not undertaken due to insufficient records (<10 birds)

Shore-based survey counts

Shore-based count totals are presented within each relevant species section. All surveys were conducted successfully, with the exception of sector 37 which could not be surveyed on 20 January 2023 due to access difficulties. Additionally in the 20 January 2023 survey, the shore-based counts at sectors 12 and 38 were conducted in suboptimal lighting conditions. Visibility during these surveys may have been hindered by the low sun and fading light, as a result of being conducted close to sunset (approximately 16:33), due to challenges in conducting all shore-based surveys on the same day in order to coincide with the DAS.

Species accounts – interpretation of maps

In the following species accounts for each of the two aerial surveys, maps are provided showing the locations of aerial observations of the species within the named SPA (e.g. Figure 8). In addition, where there were sufficient (10 or more) observations of a target species within an SPA, the following information generated through Bayesian point processing is presented in two figures:

• Maps (e.g. Figure 10) showing 1. the predicted mean densities of the species across the SPA and 2. the coefficient of variation (CV) in these estimates across the SPA. A lower CV indicates a more precise estimate.
• Maps (e.g. Figure 11) showing 1. the lower and 2. the upper credible density estimate limits across the SPA. The credible limits are a measure of uncertainty in the mean estimate. They define the range of the credible interval, within which 95 % of the estimated distribution of values fell, i.e. if the analysis was repeated, 95 % of the time the mean population estimate would fall within this range. The smaller this range the more confident we can be that the mean estimate is an accurate reflection of the true population size.  

Additional maps summarise the shore-based observations in the relevant survey sectors for the named SPA (e.g. Figure 14). In these maps, observations are represented by circles of increasing size relating to the number of birds seen. The circle markers are also coloured by survey month.

Scapa Flow SPA

Great northern diver

A total of 66 great northern diver were recorded in the February 2022 DAS, with an additional three birds identified as ‘diver sp’ which could not be identified to species level (Table 5). A greater number of great northern diver were observed in January 2023, with a total of 221 great northern diver recorded and an additional seven birds identified as 'diver sp.' (Table 6). The distribution of great northern diver observations in the 2022 and 2023 surveys is illustrated in Figure 8 and Figure 9, respectively. Birds were widely distributed across the Scapa Flow SPA, including in the deeper waters of central Scapa Flow in both surveys. 

Design-based strip transect analysis of the aerial survey data estimated the Scapa Flow SPA population at 263 birds (95 %CI 174 - 369) at the time of the February 2022 survey, and 919 birds (95 %CI 572 – 1,310) at the time of the January 2023 survey (Table 9), equating to a density of 0.83 birds/km² and 2.78 birds/km², respectively. This compares to an estimated 262 birds (95 % credible limits 197 – 343) in 2022 and 1,072 birds (95 % credible interval 884 – 1,232) in 2023 from Bayesian point processing (Table 21).

The predicted mean densities of great northern diver across the Scapa Flow SPA generated through model-based Bayesian point processing are presented for the February 2022 and January 2023 surveys in Figure 10 and Figure 12 respectively, alongside the CV in these estimates. The lower and upper credible limits are mapped in Figure 11 and Figure 13 respectively. 

During the February 2022 survey, the highest densities of 2 – 4.99 birds/km² were recorded in the west of the SPA area, particularly between the islands of Cava and Fara; the areas directly surrounding this area show density estimates of 1 – 1.99 birds/km², with the majority of the SPA having a density estimate of 0.5 - 0.99 birds/km². The lowest density estimates of 0.2 – 0.49 birds/km² occurred between Scapa Bay and St Mary’s (sub-area 3), to the east of Burray and to the south of Flotta. 

During the January 2023 survey, the highest densities of 10 – 19.9 birds/km² were observed across the central Scapa Flow, with lowest densities observed off to the east and south of the SPA.  The rest of the SPA predominantly had estimated densities between 2 – 4.99 birds/km².

Great northern diver were present in all sectors surveyed during the winter 2021/22 and 2022/23 shore-based counts. Total counts and mean percentage per sector are presented in Table 23 and Table 24 respectively for these surveys. During the winter 2021/22 surveys, the highest count of great northern diver across all shoreline sectors occurred in January 2022 (74 birds), followed by November 2021 (59 birds), whereas the highest count during the winter 2022/23 surveys was in November 2022 (105 birds). Markedly lower numbers were recorded in February during winter 2021/22 (6 birds), and in January (survey 2) in winter 2022/23 (7 birds). 

The distribution of great northern diver across sectors 9 – 12 and 37 – 38 is presented in Figure 14 and Figure 15 respectively for the winter 2021/22 and winter 2022/23 surveys. The majority of birds were recorded in the southerly sectors (sector 12 and 38) across both winter seasons. The largest flocks were generally recorded further offshore (>1 km).

Table 23. Shore-based count totals of great northern diver and the mean percentage of observations within each sector within the Scapa Flow SPA between November 2021 and March 2022, calculated as the mean of the monthly percentage of observations within each sector. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22	Mean %
9	0	5	1	1	2	9.69 %
10	0	1	13	0	0	4.16 %
11	1	0	2	0	0	0.88 %
12	14	8	28	2	5	31.28 %
37	0	0	6	0	0	1.62 %
38	44	17	24	3	7	52.37 %
Total	59	31	74	6	14	100 %

 

Table 24. Shore-based count totals of great northern diver and the mean percentage of observations within each sector within the Scapa Flow SPA between November 2022 and March 2023. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23	Mean %
9	5	2	2	0	2	3	6.06 %
10	6	4	3	0	5	2	10.85 %
11	20	2	1	2	1	13	15.38 %
12	30	1	8	1	4	3	14.63 %
37	1	4	10	-	2	1	10.22 %
38	43	4	50	4	4	28	44.57 %
Total	105	17	74	7	18	50	100.00 %

 

Black-throated diver

Only three and seven black-throated diver were recorded in the February 2022 and January 2023 DAS, respectively (Table 5 and Table 6). An additional three and seven birds respectively were identified as ‘diver sp’ but could not be identified to species level in these surveys. The distribution of black-throated diver is illustrated in Figure 16 and Figure 17. During the February 2022 survey, birds were present in the north-west of the SPA, closer to the coast than their great northern counterparts; with one individual in Hoy Sound, one individual in Burra Sound and one individual east of Orphir on Mainland Orkney. During the January 2023 survey, birds were only observed in the waters to the east of Hoy, near Farra and Flotta. 

Design-based strip transect analysis of the aerial survey data estimated the Scapa Flow SPA population at 13 birds (95 %CI 0 – 32) and 30 birds (95 %CI 0 – 84) in the February 2022 and January 2023 surveys respectively (Table 9), equating to a density of 0.04 birds/km² and 0.09 birds/km².

Density surface mapping was not undertaken for black-throated diver, as fewer than 10 observations were recorded in both surveys. 

Black-throated diver were only present in the January survey of the winter 2021/22 shore-based counts and birds were present in sectors 11 (14 birds) and 38 (12 birds) only. In the winter 2022/23 surveys, birds were present in all surveys except February 2023, but were absent from sectors 10,11 and 38 across all surveys. Observations of black-throated diver peaked in March 2023 with 19 observations followed by December 2022 with 14 birds recorded.  Total counts and mean percentage per sector for the winter 2021/22 and 2022/23 surveys are presented in Table 25 and Table 26, respectively. The distribution of black-throated diver across sectors 9 – 12 and 37 – 38 is presented in Figure 18 and Figure 19 respectively for the winter 2021/22 and winter 2022/23 surveys.

Table 25. Shore-based count totals of black-throated diver and mean percentage of observations in each sector within the Scapa Flow SPA between November 2021 and March 2022. The mean percentage has been calculated across months where the species has been observed.   

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22	Mean %
9	0	0	0	0	0	0.00 %
10	0	0	0	0	0	0.00 %
11	0	0	14	0	0	53.85 %
12	0	0	0	0	0	0.00 %
37	0	0	0	0	0	0.00 %
38	0	0	12	0	0	46.15 %
Total	0	0	26	0	0	100.00 %

 

Table 26. Shore-based count totals of black-throated diver and mean percentage of observations in each sector within the Scapa Flow SPA between November 2022 and March 2023. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23	Mean %
9	0	14	0	1	0	0	33.33 %
10	0	0	0	0	0	0	0.00 %
11	0	0	0	0	0	0	0.00 %
12	12	0	7	0	0	0	33.33 %
37	0	0	0	-	0	19	20.00 %
38	0	0	0	0	0	0	0.00 %
Total	12	14	7	1	0	19	83.33 %

 

Slavonian grebe

A total of 18 and 24 Slavonian grebe were recorded in the DAS in February 2022 and January 2023 respectively (Table 5 and Table 6). The distribution of birds for these surveys is illustrated in Figure 20 and Figure 21. Slavonian grebe are a coastal species, and this is evident with individuals found close to shore such as surrounding North Bay, Fara, Flotta, and South Ronaldsay in the south; Burray and Lamb Holm in the east and within Scapa Bay and Swanbister Bay in the north in both surveys. 

Design-based strip transect analysis of the aerial survey data estimated the Scapa Flow SPA population at 73 birds (95 %CI 29 - 123) and 100 birds (95 % CI 50 – 159) in the February 2022 and January 2023 surveys respectively (Table 9), equating to densities of 0.23 birds/km² and 0.3 birds/km². This compares to a marginally lower estimate of 70 birds (95 % credible limits 40 - 123) during February 2022 and a marginally higher estimate of 114 birds (95 % credible interval 69 – 175) during January 2023 from model-based Bayesian point processing (Table 21).

The predicted mean densities of Slavonian grebe across the Scapa Flow SPA generated through Bayesian point processing for the February 2022 and January 2023 surveys are presented in Figure 22 and Figure 24 respectively, alongside the CV. Bayesian point process analysis estimated birds at a density of up to 0.2 – 0.49 birds/km² across the majority of the SPA during both surveys. In 2023 there are areas of higher density (2 – 4.99 birds/km²) at the south-west of the SPA, between Hoy and Flotta, as well as in the Bay of Ireland to the north-west of the SPA. In the February 2022 surveys, similar densities were recorded around Burray and within Scapa Bay. The lower and upper credible limits are mapped in Figure 23 and Figure 25 for the February 2022 and January 2023 surveys respectively.  

Slavonian grebe were present in all monthly shore-based surveys throughout winter 2021/22 and 2022/23 surveys. In winter 2021/22, increased numbers were observed in December 2021 (17 birds) and January 2022 (21 birds), whilst in winter 2022/23 low numbers of birds (2 – 5 birds) were observed in all months except the second January survey (35 birds). Total counts and mean percentage per sector are presented in Table 27 and Table 28 respectively for the winter 2021/22 and 2022/23 surveys. Birds were observed in all sectors except sector 37 across the winter 2021/22 surveys, and only one bird was observed in this sector across the winter 2022/23 surveys. In both winters, the majority of observations occurred in sectors 9 (inner Scapa Bay; 2022: ~25 %; 2023 ~50 % mean percentage occurrence) and 38 (St. Mary’s Bay; 2022: ~44 %; 2023: ~27 % mean percentage occurrence).

The distribution of Slavonian grebe across sectors 9 – 12 and 37 – 38 is presented in Figure 26 and Figure 27 respectively for the winter 2021/22 and winter 2022/23 surveys. As expected, birds were recorded relatively close to shore, preferring shallow inshore waters in both surveys.

Table 27. Shore-based count totals of Slavonian grebe and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2021 and March 2022. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22	Mean %
9	0	11	6	1	0	25.32 %
10	0	0	7	0	0	6.67 %
11	0	0	2	0	0	1.90 %
12	3	0	2	2	0	21.90 %
37	0	0	0	0	0	0.00 %
38	6	6	4	0	2	44.20 %
Total	9	17	21	3	2	100.00 %

 

Table 28. Shore-based count totals of Slavonian grebe and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2022 and March 2023.The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23	Mean %
9	0	1	3	6	2	2	49.52 %
10	0	0	0	4	0	0	1.90 %
11	1	1	0	0	0	0	11.67 %
12	0	2	0	0	0	0	6.67 %
37	0	0	1	-	0	0	4.00 %
38	1	1	1	25	0	0	26.90 %
Total	2	5	5	35	2	2	100.00 %

 

Red-breasted merganser

More than twice as many red-breasted merganser were recorded in the January 2023 survey (86 birds) than the February 2022 survey (29 birds) (Table 5 and Table 6). An additional 95 birds during February 2022 and 67 birds during January 2023 were identified as ‘duck sp.’ but could not be identified to species level. The distribution of red-breasted merganser is illustrated in Figure 28 and Figure 29 for the February 2022 and January 2023 surveys respectively. In both surveys, birds were distributed close to shore. In the February 2022 survey birds were primarily observed in the south and east of the SPA, including North Bay, Flotta, South Ronaldsay, Burray and St. Mary’s Bay. In the January 2023 survey, birds were also observed in the northwest of the site with high numbers in Clestrain sound.

Design-based strip transect analysis of the aerial survey data estimated the Scapa Flow SPA population at 79 birds (95 %CI 25 - 157) during the February 2022 survey compared with 330 birds (95 %CI 147 – 569) during the January 2023 survey (Table 9), equating to densities of 0.25 birds/km² and 1 bird/km² respectively. This compares to a higher estimate of 102 birds (95 % credible limits 56 - 200) in February 2022, and slightly higher estimate of 338 birds (95 % credible interval 255 – 435), from Bayesian point processing (Table 21).

The predicted mean densities of red-breasted merganser across the Scapa Flow SPA generated through Bayesian point processing are presented for the February 2022 and January 2023 surveys in Figure 30 and Figure 32 respectively, alongside the CV. 

Bayesian model-based analysis estimated highest densities of red-breasted merganser 5 – 9.99 birds/km² in both surveys in nearshore waters around the Scapa Flow SPA. Estimated density reduced out from these points, with 0 – 0.09 birds/km² estimated across the majority of the SPA in the February 2022 survey, and around 0.5 – 1.99 birds/km² in the January 2023 survey. Lowest densities were observed across the deeper waters in the central Scapa Flow in both surveys. The lower and upper credible limits are mapped in Figure 31 and Figure 33 for February 2022 and January 2023 respectively.

Red-breasted merganser were present in all monthly shore-based surveys throughout both winter survey periods, with the highest numbers occurring in January (2022: 24 birds; 2023: 15 birds). Total counts and mean percentage per sector are presented in Table 29 and Table 30 respectively for the winter 2021/22 and 2022/23 surveys. Birds were observed in all sectors except sector 37 in both winters. The majority of observations occurred in sector 9 (inner Scapa Bay) during the winter 2021/22 surveys (~61 % mean percentage occurrence), while in the winter 2022/23 surveys, most observations occurred in sector 38 (St Mary’s Bay) (~51 % mean percentage occurrence).

The distribution of red-breasted merganser across sectors 9 – 12 and 37 – 38 is presented in Figure 34 and Figure 35 for the winter 2021/22 and winter 2022/23 surveys respectively. As expected, birds were mostly recorded close to shore; preferring sheltered coastlines during both surveys.

Table 29. Shore-based count totals of red-breasted merganser and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2021 and March 2022. The mean percentage has been calculated across months where the species has been observed. 

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22	Mean %
9	8	12	12	6	8	61.03 %
10	0	0	4	0	0	3.33 %
11	0	0	2	0	0	1.67 %
12	2	2	2	5	0	16.28 %
37	0	0	0	0	0	0.00 %
38	5	0	4	0	5	17.69 %
Total	15	14	24	11	13	100.00 %

 

Table 30. Shore-based count totals of red-breasted merganser and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2022 and March 2023. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23	Mean %
9	2	0	3	1	2	1	36.03 %
10	0	2	0	2	0	0	5.93 %
11	0	2	0	0	0	0	3.70 %
12	0	2	0	0	0	0	3.70 %
37	0	0	0	-	0	0	0.00 %
38	4	3	4	12	0	2	50.63 %
Total	6	9	7	15	2	3	100.00 %

 

Velvet scoter

Velvet scoter is not a qualifying feature of the Scapa Flow SPA. 

No velvet scoter were recorded in the aerial surveys of Scapa Flow in February 2022 or January 2023. Consequently, no design-based strip transect or model-based Bayesian point process analyses were undertaken. 

A flock of seven birds were recorded in sector 38 (St. Mary’s Bay) in the January 2022 shore-based count (Table 31) at a distance of ~850 m offshore from Glimps Holm (Figure 36). No velvet scoter were observed in any other shore-based count across the winter 2021/22 and winter 2022/23 surveys.

Table 31. Shore-based count totals of velvet scoter and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2021 and March 2022. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22	Mean %
9	0	0	0	0	0	0.00 %
10	0	0	0	0	0	0.00 %
11	0	0	0	0	0	0.00 %
12	0	0	0	0	0	0.00 %
37	0	0	0	0	0	0.00 %
38	0	0	7	0	0	100.00 %
Total	0	0	7	0	0	100.00 %

 

Long-tailed duck

Long-tailed duck was the most abundant target species recorded in both aerial surveys of Scapa Flow SPA; although, more birds were observed during the January 2023 (786 birds) survey compared with the February 2022 survey (359 birds) (Table 5 and Table 6). An additional 95 and 67 birds were identified as ‘duck sp.’ but could not be identified to species level during the February 2022 and January 2023 surveys respectively (Table 7). The distribution of long-tailed duck is illustrated in Figure 37 and Figure 38. Birds were widely distributed across the Scapa Flow SPA during both surveys, with birds preferring waters closer to the coastlines and islands and avoiding the deeper waters of the central Scapa Flow. 

Design-based strip transect analysis of the aerial survey data estimated the Scapa Flow SPA population at 1,381 birds (95 %CI 719 – 2,160) during the February 2022 survey, and 3,149 birds (95 %CI 625 – 7,559) during the January 2023 survey (Table 9), equating to densities of 4.34 birds/km² and 9.53 birds/km², respectively. This compares to higher estimates of 1,502 birds (95 % credible limits 1,279 – 1,792) during the February 2022 survey and 3,678 birds (95 % credible interval 3,321 – 4,343) during the January 2023 survey from Bayesian point processing, although there is overlap in the confidence intervals of both approaches (Table 21).

 

 

The predicted mean density of long-tailed duck across the Scapa Flow SPA generated through model-based Bayesian point processing is presented in Figure 39 and Figure 41 respectively for the February 2022 and January 2023 surveys, alongside the CV in these estimates. 

The highest estimated density during the February 2022 survey was 50 – 99.9 birds/km², with generally higher densities observed around the shallow inlets and along the coast. In the January 2023 surveys, the highest densities of over 500 birds/km² were observed within the western parts of the SPA. Birds have been recorded frequenting the eastern areas of Hoy and Fara and were commonly found in waters around fish farms (Jackson, 2018), although detailed analysis on the spatial distribution of marine birds in relation to fish farms was not carried out. Birds appeared to avoid the deeper, central parts of the Scapa Flow during the February 2022 survey, where estimated density reduced to between 0 – 0.49 birds/km² in these areas, whereas moderate densities (2 – 9.99 birds/km²) were observed across this area during the February 2023 survey. The lower and upper credible limits are mapped in Figure 40 and Figure 42.

Long-tailed duck were present in all monthly shore-based surveys throughout winter 2021/22 and 2022/23. The highest numbers occurred in December 2022 (265 birds), compared with a peak in the previous winter of 168 birds in January 2022. Total counts and mean percentage per sector are presented in Table 32 and Table 33 respectively for the winter 2021/22 and 2022/23 surveys. As with other species, birds were observed in all sectors except sector 37 across the winter 2021/22 surveys, and only three birds were observed in this sector across the winter 2022/23 surveys. The majority of winter observations occurred in sector 12 (Bay of Sandoyne) across both the 2021/22 (~48 % mean percentage occurrence) and 2022/23 (~57 % mean percentage occurrence) surveys.

The distribution of long-tailed duck across sectors 9 – 12 and 37 – 38 is presented in Figure 43 and Figure 44 respectively for the winter 2021/22 and winter 2022/23 surveys. Birds were present in varying flock size, generally along the shallowest inshore waters during the winter 2021/22 surveys, compared with slightly further offshore during the winter 2022/23 surveys. 

Table 32. Shore-based count totals of long-tailed duck and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2021 and March 2022. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22	Mean %
9	34	41	45	8	33	29.31 %
10	3	4	6	0	7	3.37 %
11	4	0	7	2	0	2.74 %
12	58	93	85	14	28	48.13 %
37	0	0	0	0	0	0.00 %
38	11	5	25	10	22	16.45 %
Total	110	143	168	34	90	100.00 %

 

Table 33. Shore-based count totals of long-tailed duck and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2022 and March 2023. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23	Mean %
9	29	26	5	20	10	16	18.39 %
10	15	19	15	1	0	0	4.30 %
11	7	10	3	4	0	0	3.59 %
12	113	200	132	0	62	118	56.51 %
37	0	0	0	-	0	3	0.40 %
38	57	10	19	6	36	12	16.87 %
Total	221	265	174	31	108	149	100.00 %

 

Common eider

A total of 192 common eider were recorded in the February 2022 DAS, compared with 354 common eider during the January 2023 survey (Table 5 and Table 6). An additional 95 birds and 67 birds were identified as ‘duck sp.’ but could not be identified to species level in the February 2022 and January 2023 surveys respectively. The distribution of common eider is illustrated in Figure 45 and Figure 46. Birds were widely distributed along the shallow coastline waters surrounding the Scapa Flow SPA, with birds completely avoiding the central Scapa Flow in both surveys.

Design-based strip transect analysis of the February 2022 aerial survey data estimated the Scapa Flow SPA population at 581 birds (95 %CI 263 - 961) at the time of survey (Table 9), equating to a density of 1.82 birds/km². In the January 2023 survey, the estimate was much higher at 1,456 birds (95 %CI 454 – 2,938). This compares to higher estimates, of 676 birds (95 % credible limits 534 - 824) during the February 2022 survey and 1,504 birds (95 % credible interval 1,322 – 1,706) during the January 2023 survey, estimated from model-based Bayesian point processing (Table 21).

The predicted mean density of common eider, alongside the CV, across the Scapa Flow SPA generated through Bayesian point processing is presented in Figure 47 and Figure 49 respectively for February 2022 and January 2023. 

Bayesian analysis estimated birds at higher densities of between 2 – 49.9 birds/km² in the west of the SPA around Graemsay and Hoy, and in the east, between South Ronaldsay and Burray, and around Lamb Holm in both surveys. In February 2022, the highest density of between 50 – 99.9 birds/km² was estimated in inner Scapa Bay. In the January 2023 survey, birds were more evenly distributed across the SPA, although still with lower densities (0.1 – 0.99 birds/km²) in the deeper Scapa Flow waters and to the east of the SPA. The highest densities (between 2 – 49.9 birds/km²) of common eider were observed along the coast and towards the west of the SPA, with low densities (0 – 0.19 birds/km²) in the deeper waters in the middle and towards the east of Scapa Flow observed in both surveys. The lower and upper credible limits are mapped in Figure 48 and Figure 50.

Common eider were present in all monthly shore-based surveys throughout both the winter 2021/22 and winter 2022/23 surveys, with the highest numbers occurring in January for both seasons (2022: 100 birds; 2023:107 birds). Total counts and mean percentage per sector are presented in Table 34 and Table 35 respectively. As with other species, birds were observed in all sectors except sector 37 across the winter 2021/22 surveys, and only one bird was observed in this sector across the winter 2022/23 surveys. In both winter periods, most observations (calculated as mean percentage occurrence) occurred in sector 12 (Bay of Sandoyne; 2022: ~31 %; 2023 ~30 %) and sector 38 (St. Mary’s Bay; 2022: ~26 %; 2023: ~38 %), in addition to sector 9 (inner Scapa Bay; 2022: ~33 %), during the winter 2021/22 surveys. 

The distribution of common eider across sectors 9 – 12 and 37 – 38 is presented in Figure 51 and Figure 52 respectively for the winter 2021/22 and winter 2022/23 surveys. Birds were present in varying flock sizes generally along the shallowest inshore waters during the winter 2021/22 surveys, compared with slightly further offshore during the winter 2022/23 surveys.

Table 34. Shore-based count totals of common eider and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2021 and March 2022. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22	Mean %
9	23	22	17	8	27	32.73 %
10	4	0	1	0	0	1.51 %
11	5	0	26	3	0	8.98 %
12	23	19	30	9	14	30.64 %
37	0	0	0	0	0	0.00 %
38	6	24	26	8	17	26.13 %
Total	61	65	100	28	58	100.00 %

 

Table 35. Shore-based count totals of common eider and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2022 and March 2023. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23	Mean %
9	2	10	5	5	4	9	13.35 %
10	18	0	40	5	0	1	17.44 %
11	0	0	0	2	0	0	1.39 %
12	6	10	48	3	1	56	29.95 %
37	0	0	0	-	0	1	0.29 %
38	14	24	14	9	24	2	37.63 %
Total	40	44	107	24	29	69	100.00 %

 

European shag

Many more European shag were observed in the January 2023 DAS (198 birds) compared with February 2022 (40 birds) (Table 5 and Table 6). The distribution of European shag is illustrated in Figure 53 and Figure 54 respectively for the February 2022 and January 2023 surveys. Birds were widely distributed along the coastlines primarily in the east, south and west of the Scapa Flow SPA, with birds avoiding the north and central Scapa Flow in both surveys.

Design-based strip transect analysis of the aerial survey data estimated the Scapa Flow SPA population at 130 birds (95 %CI 58 - 227) at the time of the February 2022 survey, compared with 764 birds (95 %CI 253 – 1,416) during the January 2023 survey (Table 9), equating to densities of 0.41 birds/km² and 2.31 birds/km² respectively. Model-based Bayesian point processing resulted in similar estimates during the February 2022 survey (134 birds; 95 % credible limits 88 - 193), and slightly higher estimates during the January 2023 survey (872 birds; 95 % credible limits 597 – 1,560), as presented in Table 21. It should be noted that observations were trimmed to the marine SPA boundary to derive both the design-based and model-based population estimates. European shags frequently roost and wing-dry onshore or on skerries and tidally exposed rocks. As such, roosting birds that fall beyond this SPA boundary (i.e. above MLWS) are not included in such estimates.

The predicted mean density of European shag, alongside the CV, across the Scapa Flow SPA generated through Bayesian point processing is presented in Figure 55 and Figure 57 respectively for the February 2022 and January 2023 surveys. Bayesian analysis estimated birds at a uniform mean density of 0.2 – 0.99 birds/km² across much of the SPA in the February 2022 survey. Much higher densities of up to 20 – 49.9 birds/km² were observed around the edges of Scapa Flow and nearshore waters in January 2023 most notably around Graemsay and between Burray and South Ronaldsay. The lower and upper credible limits are mapped in Figure 56 and Figure 58 respectively for the February 2022 and January 2023 surveys. 

European shag were present in all shore-based counts throughout winter 2021/22 and winter 2022/23 surveys. During the winter 2021/22 surveys, relatively consistent numbers were observed, ranging between 31 - 53 birds, whereas between five (March) and 57 (February) birds were observed during the winter 2022/23 surveys. Total counts and mean percentage per sector are presented in Table 36 and Table 37 respectively for the winter 2021/22 and 2022/23 surveys. Birds were found in all sectors, with relatively even mean occurrence during the winter 2021/22 surveys, whilst more variation was observed between in the winter 2022/23 surveys. Sector 38 (St. Mary’s Bay) had the highest mean percentage occurrence for both surveys, with ~24 % and ~53 % in the winter 2021/22 and 2022/23 surveys, respectively.

The distribution of European shag across sectors 9 – 12 and 37 – 38 is presented in Figure 59 and Figure 60 respectively for the winter 2021/22 and winter 2022/23 surveys. More birds were observed along the shallowest inshore waters during the winter 2021/22 surveys, compared with slightly further offshore during the winter 2022/23 surveys.

Table 36. Shore-based count totals of European shag and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2021 and March 2022. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22	Mean %
9	5	17	0	5	7	15.37 %
10	4	7	3	3	5	10.02 %
11	2	5	24	2	8	17.62 %
12	15	7	10	5	7	19.60 %
37	3	8	2	6	8	13.11 %
38	20	9	12	10	3	24.30 %
Total	49	53	51	31	38	100.00 %

 

Table 37. Shore-based count totals of European shag and mean percentage of observations in each sector mean within the Scapa Flow SPA between November 2022 and March 2023. The mean percentage has been calculated across months where the species has been observed.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23	Mean %
9	1	4	0	19	1	0	14.80 %
10	0	3	0	6	0	0	6.07 %
11	4	0	0	2	3	1	8.03 %
12	14	3	0	0	1	1	16.29 %
37	0	2	0	-	0	0	2.35 %
38	5	5	19	5	52	3	52.85 %
Total	24	17	19	32	57	5	100.00 %

 

Non-target species

A number of non-target species were recorded in the DAS of Scapa Flow SPA (Table 5; Table 6), the distributions of which are presented in Figure 61 for the February 2022 survey, and Figure 62 for the January 2023 survey. 

North Orkney SPA

Great northern diver

A total of 28 great northern diver were recorded in the February 2022 DAS, compared with 45 birds in the January 2023 survey (Table 5 and Table 6). An additional seven birds were identified as ‘diver sp.’ but could not be identified to species level in each survey. The distribution of great-northern diver is illustrated in Figure 63 and Figure 64 respectively for the February 2022 and January 2023 surveys. Birds were predominantly distributed on the northeast and southeast of North Orkney SPA during the February 2022 survey and were distributed across the SPA during the January 2023 survey.

Design-based strip transect analysis of the February 2022 aerial survey data estimated the North Orkney SPA population at 79 birds (95 %CI 30 – 140) at the time of survey (Table 15), equating to a density of 0.37 birds/km². A higher abundance of 191 birds (95 %CI 105 – 300; Table 15) was estimated during the January 2023 survey, equating to a density of 0.84 birds/km². This compares to similar estimates from Bayesian point processing of 76 birds (95 % credible limits 43 – 122) during the February 2022 survey, and 208 birds (95 % credible interval 145 – 282) during the January 2023 survey (Table 22).

The predicted mean density of great northern diver alongside the CV in these estimates, generated through model-based Bayesian point processing, is presented in Figure 65 for the February 2022 survey and in Figure 67 for the January 2023 survey. The highest densities (1 - 1.99 birds/km² in February 2022; 10 – 19.9 in January 2023) were recorded around Rerwick Head, in the southeast of the SPA in both surveys, in addition to around the Bay of Firth in the January 2023 survey. The rest of the SPA is predominantly estimated to have had low densities of birds of 0.1 - 0.49 birds/km² in the February 2022 survey and 0.5 – 0.99 birds/km² during the January 2023 survey. Slightly higher densities of 1 – 1.99 birds/km2 were also estimated in the deeper waters in the north of the SPA in the January 2023 survey, which were not observed in the February 2022 surveys. The lower and upper credible limits are mapped in Figure 66 for the February 2022 survey and Figure 68 for the January 2023 survey. 

Great northern diver were observed in all months during the winter 2021/22 shore-based counts, and in all months except March in the 2022/23 shore-based counts. The total monthly counts for the single sector 14 are presented in Table 38 and Table 39 respectively for the winter 2021/22 and 2022/23 surveys. The highest counts of great northern divers occurred in January 2022 (18 birds) during the winter 2021/22 surveys and February 2023 (20 birds) during the winter 2022/23 surveys. Low numbers of great northern divers (3 - 7 birds) were recorded in the other months surveyed across both winters.

The distribution of great northern diver across sector 14 during the shore-based counts is presented in Figure 69 and Figure 70 respectively for the winter 2021/22 and winter 2022/23 surveys, with most birds recorded at the mouth of the Bay of Kirkwall across both surveys. The majority of birds were recorded as individuals, except for two small groups of four birds during the winter 2021/22 surveys, and a flock of 18 during the February 2023 survey. 

Table 38. Shore-based count totals of great northern diver within sector 14 of the North Orkney SPA between November 2021 and March 2022.       

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22
14	6	4	18	3	3

 

Table 39. Shore-based count totals of great northern diver within sector 14 of the North Orkney SPA between November 2022 and March 2023.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23
14	4	4	7	4	20	0

 

Black-throated diver

Black-throated diver is not a qualifying feature of the North Orkney SPA. 

A single bird was recorded in the February 2022 DAS (Figure 71). Strip transect analysis of the aerial survey data estimated the North Orkney SPA population at 5 birds (95 %CI 0 – 13) at the time of survey (Table 15), equating to a density of 0.02 birds/km². No black-throated diver were observed during the January 2023 DAS.

Density surface mapping was not undertaken for black-throated diver, as fewer than 10 observations were recorded. 

No black-throated diver were recorded within sector 14 during any of the shore-based counts conducted over the winter of 2021/22 or 2022/23. 

Slavonian grebe

A total of nine Slavonian grebe were recorded in the February 2022 DAS, compared with 22 in the January 2023 survey (Table 5 and Table 6). The distribution of Slavonian grebe is illustrated in Figure 72 and Figure 73 respectively for the February 2022 and January 2023 surveys. Slavonian grebe are a coastal species, and this is evident here, with individuals predominantly found close to shore across both surveys.

Design-based strip transect analysis of the aerial survey data estimated the North Orkney SPA population at 38 birds (95 %CI 13 – 66) during the February 2022 survey, compared with 94 birds (95 %CI 24 – 206) during the January 2023 survey (Table 15), equating to a density of 0.18 birds/km² and 0.41 birds/km², respectively. Model-based Bayesian point processing was only possible for the January 2023 data, generating an estimated 103 birds (95 % credible interval 61 – 163; Table 22).

The predicted mean density of Slavonian grebe across the North Orkney SPA in the January 2023 survey is presented in Figure 74, alongside the CV in these estimates. Low densities (up to 0.2 – 0.49 birds/km²) were estimated across the majority of North Orkney SPA, with higher densities (up to 5 – 19.9 birds/km²) in some shallower waters in the west and northwest of the SPA (Figure 74). The lower and upper credible intervals are mapped in Figure 75.

Slavonian grebes were only observed in November 2021 (one bird) and January 2022 (five birds) during the winter 2021/22 shore-based counts. Whilst birds were observed in all survey months except November 2022 during the 2022/23 surveys, in both seasons they were present in relatively low numbers, peaking at 8 birds in the first survey in January 2023. The total monthly counts for sector 14 are presented in Table 40 and Table 41 respectively for the winter 2021/22 and 2022/23 surveys. 

The distribution of Slavonian grebe across sector 14 is presented in Figure 76 and Figure 77 respectively for the winter 2021/22 and winter 2022/23 surveys. Individual birds and small flocks (up to three birds) were recorded across the Bay of Kirkwall. 

Table 40. Shore-based count totals of Slavonian grebe within sector 14 of the North Orkney SPA between November 2021 and March 2022.     

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22
14	1	0	5	0	0

 

Table 41. Shore-based count totals of Slavonian grebe within sector 14 of the North Orkney SPA between November 2022 and March 2023.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23
14	0	3	8	1	6	2

 

Red-breasted merganser

Red-breasted merganser is not a qualifying feature of the North Orkney SPA. 

A similar number of red breasted merganser were observed in the February 2022 and January 2023 DAS, with 65 and 78 birds, respectively (Table 5 and Table 6). However, the January 2023 surveys recorded a higher number of birds (89) than in February 2022 (18) which could only be identified as 'duck sp.' and were not identified to species level. The distribution of red-breasted merganser is illustrated in Figure 78 for the February 2022 survey and Figure 79 for the January 2023 survey. Birds were recorded close to shore, or in shallow near-shore waters and are distributed across the North Orkney SPA in both surveys.

Design-based strip transect analysis of the aerial survey data estimated the North Orkney SPA population at 214 birds (95 %CI 37 – 506) during the February 2022 survey and 284 birds (95CI 68 – 663) during the January 2023 survey (Table 15), equating to a densities of 1.01 birds/km² and 1.25 birds/km² respectively.  Bayesian point processing generated estimates of 214 birds (95 % credible limits 135 – 305) for February 2022 and 331 birds (95 % credible interval 236 – 521) for January 2023 (Table 22).

Figure 78

The predicted mean density of red-breasted merganser across the North Orkney SPA generated through Bayesian point processing is presented in Figure 80 for the February 2022 survey and Figure 82 for the January 2023 survey, alongside the CV. The distribution of birds was similar between the 2022 and 2023 surveys. The highest densities of 20 - 49.9 birds/km² were recorded in the shallow coastal waters at the Bay of Firth, in the west of the SPA during both surveys, in addition to a few small areas at the Bay of Suckquoy, Inganess Bay and Rousay Sound in the January 2023 survey. The rest of North Orkney SPA had relatively low estimated densities, from 0 - 0.99 birds/km² in both surveys. The lower and upper credible limits are mapped in Figure 81 and Figure 83 respectively for the February 2022 and January 2023 surveys.

Red-breasted merganser were observed in all months during both the winter 2021/22 and winter 2022/23 shore-based counts. The total monthly counts for sector 14 are presented in Table 42 and Table 43 respectively for the winter 2021/22 and 2022/23 surveys. Higher counts occurred in January in both seasons; January 2022 (41 birds) and January (survey 1) 2023 survey (39 birds), although the peak from the winter 2022/23 surveys was in November 2022 (44 birds). Lower numbers of red-breasted merganser (2022: 3 - 14 birds; 2023: 2 – 22 birds) were recorded in the other months surveyed.

The distribution of red-breasted merganser across sector 14 is presented in Figure 84 and Figure 85 respectively for the winter 2021/22 and winter 2022/23 surveys. As expected, birds were mostly recorded close to shore, preferring sheltered coastlines across both winter seasons.

Table 42. Shore-based count totals of red-breasted merganser within sector 14 of the North Orkney SPA between November 2021 and March 2022.       

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22
14	6	14	41	3	6

 

Table 43. Shore-based count totals of red-breasted merganser within sector 14 of the North Orkney SPA between November 2022 and March 2023.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23
14	44	22	39	2	10	8

 

Velvet scoter

No velvet scoter were recorded within the North Orkney SPA during either of the DAS in February 2022 and January 2023, nor within sector 14 during any of the shore-based counts conducted over the winter of 2021/22. In the winter 2022/23 shore-based counts, ten birds were observed in December 2022 and 19 birds in January 2023 (survey 1) at the mouth of the Bay of Kirkwall (Table 44; Figure 86). 

Table 44. Shore-based count totals of velvet scoter within sector 14 of the North Orkney SPA between November 2022 and March 2023.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23
14	0	10	19	0	0	0

Long-tailed duck

Long-tailed duck is not a qualifying feature of the North Orkney SPA. 

A total of 377 long-tailed duck were recorded in the February 2022 DAS, compared with 590 recorded in the January 2023 survey (Table 5 and Table 6). An additional 18 birds in February 2022, and 89 in January 2023 were identified as 'duck sp.' but could not be identified to species level. The distribution of long-tailed duck is illustrated in Figure 87 and Figure 88 respectively for the February 2022 and January 2023 surveys. Birds were widely distributed across the North Orkney SPA during both surveys.

Design-based strip transect analysis of the aerial survey data estimated the North Orkney SPA population at 1,570 birds (95 %CI 554 – 3,170) during the February 2022 survey, compared with 2,190 birds (95 %CI 1,051 – 3,331) during the January 2023 survey (Table 15), equating to densities of 7.41 birds/km² and 9.62 birds/km² respectively. Model-based Bayesian point processing generated somewhat higher estimates of 1,915 birds (95 % credible limits 1,668 – 2,734) from in the February 2022 survey, and 2,428 birds (95 % credible interval 2,199 – 2,654) in the January 2023 survey (Table 22).

The predicted mean density of long-tailed duck across the North Orkney SPA generated through Bayesian point processing is presented in Figure 89 for the February 2022 survey and Figure 91 for the January 2023 survey, alongside the CV in these estimates. The highest estimated densities in both surveys (200 - 499.9 birds/km² in February 2022 and 100 - 199 birds/km² in the January 2023 survey) were recorded at the headland west of the Sands of Evie, off the north coast of Orkney Mainland, in addition to west of the Bay of Kirkwall in the January 2023 survey. During the February 2022 surveys, high densities (50 - 99.9 birds/km²) were also estimated off the east coast of Gairsay Island, in the centre of North Orkney SPA. Relatively high densities up to 20 - 49.99 birds/km² were estimated across the remainder of the SPA, with low densities in the northeast of the site and in Shapinsay Sound during both surveys. The lower and upper credible limits are mapped in Figure 90 and Figure 92 respectively for the February 2022 and January 2023 surveys. 

Long-tailed duck were observed in all months during both the winter 2021/22 and winter 2022/23 shore-based counts, with many more birds observed over the winter 2022/23 surveys, as shown in Tables 45 and 46. In the winter 2021/22 surveys, the highest count of long-tailed duck occurred in January 2022 (72 birds), compared with a much higher peak in the first January 2023 survey (471 birds) during the winter 2022/23 surveys. Low numbers of long-tailed duck (1 - 10 birds) were recorded in the other months surveyed during winter 2021/22, whereas during the winter 2022/23 surveys relatively high numbers were also recorded in December 2022 (145 birds), February 2023 (379 birds), and March 2023 (185 birds).

The distribution of long-tailed duck across sector 14 is presented in Figure 93 and Figure 94 respectively for the winter 2021/22 and winter 2022/23 surveys. Birds were present across the Bay of Kirkwall.  In 2021/22 the largest flock of up to 18 individuals was recorded in January 2022, but much larger numbers and flocks of birds were recorded across the Bay of Kirkwall during the winter 2022/23 surveys.

Table 45. Shore-based count totals of long-tailed duck within sector 14 of the North Orkney SPA between November 2021 and March 2022.     

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22
14	2	8	72	10	1

 

Table 46. Shore-based count totals of long-tailed duck within sector 14 of the North Orkney SPA between November 2022 and March 2023.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23
14	15	145	471	9	379	185

 

Common eider

Common eider is not a qualifying feature of the North Orkney SPA. 

A total of 224 common eider were recorded in the February 2022 DAS, and more than twice as many (556 birds) in the January 2023 survey (Table 5 and Table 6). An additional 18 birds during the February 2022 survey, and 89 birds during the January 2023 survey were identified as 'duck sp.' but could not be identified to species level. The distribution of common eider is illustrated in Figure 95 for the February 2022 survey and Figure 96 for the January 2023 survey. Birds were widely distributed across the North Orkney SPA, except for the deepest waters in Gairsay and Shapinsay Sounds.

Design-based strip transect analysis of the February 2022 aerial survey data estimated the North Orkney SPA population at 840 birds (95 %CI 328 – 1,535) at the time of survey, equating to a density of 3.96 birds/km², compared with an estimated population of 2,266 birds (95 %CI 285 – 5,842) in January 2023, equating to a density of 9.95 birds/km² (Table 15). This compares to an estimated 905 birds (95 % credible limits 759 – 1,580) and 2,134 birds (95 % credible interval 1,887 – 2,421) from model-based Bayesian point processing for the February 2022 and January 2023 surveys respectively (Table 22).

The predicted mean density of common eider, alongside the CV in these estimates, across the North Orkney SPA generated through Bayesian point processing is presented in Figure 97 and Figure 99 respectively for the February 2022 and January 2023 surveys. The highest densities (100 - 199.9 birds/km²) were recorded in Deer Sound during the February 2022 survey. Relatively high densities were also estimated at the mouth of Wyre Sound, south of Rousay. Wide Firth, in the centre of North Orkney SPA had an estimated density of 2 - 9.99 birds/km². Highest densities in the January 2023 surveys were observed to the west of the Bay of Kirkwall with peak of 100 – 499.9 birds/km². In both surveys, the lowest densities were in Shapinsay Sound, and in the north-east of the SPA, with 0.5 - 0.99 birds/km² estimated in the February 2022 survey and 0 – 0.09 birds/km² in the January 2023 survey. The distribution of birds was more even across the SPA in the February 2022 survey compared with the January 2023 survey. The lower and upper credible limits are mapped in Figure 98 and Figure 100 respectively for the February 2022 and January 2023 surveys. 

Common eider were observed in all months during the both the winter 2021/22 and winter 2022/23 shore-based counts, with significantly higher numbers of birds observed during winter 2022/23. The total monthly counts for sector 14 are presented in Table 47 and Table 48 respectively for the winter 2021/22 and 2022/23 surveys. The highest count of common eider during the winter 2021/22 surveys occurred in January 2022 (44 birds), followed by 23 birds observed in March 2022, with lower numbers of birds (11 - 18 birds) recorded in the other months surveyed. During the winter 2022/23 surveys, much higher counts of birds were observed with the peak in February 2023 (1,155 birds). 

The distribution of common eider across sector 14 is presented in Figure 101 and Figure 102 respectively for the winter 2021/22 and winter 2022/23 surveys. Birds were most abundant in the west of the Bay of Kirkwall but were distributed across the sector during both the winter 2021/22 and 2022/23 surveys. 

Table 47. Shore-based count totals of common eider within sector 14 of the North Orkney SPA between November 2021 and March 2022.     

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22
14	12	18	44	11	23

 

Table 48. Shore-based count totals of common eider within sector 14 of the North Orkney SPA between November 2022 and March 2023.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23
14	19	295	774	7	1155	539

 

European shag

European shag is not a qualifying feature of the North Orkney SPA. 

A total of 43 European shag were recorded in the February 2022 DAS (Table 5), and were widely distributed across the North Orkney SPA, as shown in Figure 103. More European shag were recorded in the January 2023 survey (Table 6), with 318 birds, the majority being observed on the northwest of the North Orkney SPA, and to the south-east, notably in Deer Sound (Figure 104). 

Design-based strip transect analysis of the aerial survey data estimated the North Orkney SPA population at 150 birds (95 %CI 79 – 239) equating to a density of 0.71 birds/km² during the February 2022 survey, compared with 1,345 birds (95 %CI 432 – 2,562) during the January 2023 survey, equating to a density of 5.91birds/km² (Table 15). This compares to a similar estimate of 161 birds (95 % credible limits 109 – 218) from Bayesian point processing during the February 2022 survey, and a higher estimate of 1,667 birds (95 % credible interval 1,356 – 2,189) during the January 2023 survey (Table 22). It should be noted that observations were trimmed to the marine SPA boundary to derive both the design-based strip transect and model-based Bayesian population estimates. European shags frequently roost and wing-dry onshore or on skerries and tidally exposed rocks. As such, roosting birds that fall beyond this SPA boundary (i.e. above MLWS) were not included in such estimates.

The predicted mean density of European shag, alongside the CV in these estimates, across the North Orkney SPA generated through model-based Bayesian point processing is presented in Figure 105 for the February 2022 survey and Figure 107 for the January 2023 survey. In the February 2022 survey, the estimated density of European shag across the site was mostly uniform at 0.5 - 0.99 birds/km² but with areas of slightly higher estimated density (1 - 1.99 birds/km²) in Deer Sound, at the western end of Eynhallow Sound, and south of Gairsay Island. During the January 2023 survey, the number of birds recorded was much greater, with high densities occurring along the Orkney Mainland east coast and Deer Sound and highest densities (100 – 199.9 birds/km²) in Eynhallow Sound off the coast at Evie. Low densities were recorded in the deeper waters in the Shapinsay Sound and in the northeast of the SPA. The lower and upper credible limits are mapped in Figure 106 and Figure 108 respectively for the February 2022 and January 2023 surveys. 

European shag were observed in all months during the winter 2021/22 and winter 2022/23 shore-based counts (Table 49 and Table 50 respectively). The highest count of European shag occurred in November during both the 2021/22 (24 birds) and 2022/23 (38 birds) surveys, followed by 17 birds observed in March 2022 and 22 birds in December 2022. Low numbers of European shag (7 - 14 birds) were recorded in the other months surveyed across both surveys.

The distribution of European shag across sector 14 is presented in Figure 109 and Figure 110 respectively for the winter 2021/22 and winter 2022/23 surveys. Birds were recorded across the sector in all months during the winter 2021/22 surveys, with one larger flock (10 individuals) observed in the middle of the Bay of Kirkwall in January 2022. During the winter 2022/23 surveys, birds were primarily distributed along the coast to the eastern side of the Bay of Kirkwall.

Table 49. Shore-based count totals of European shag within sector 14 of the North Orkney SPA between November 2021 and March 2022.    

Sector	Nov ‘21	Dec ‘21	Jan ‘22	Feb ‘22	Mar ‘22
14	24	14	14	7	17

 

Table 50. Shore-based count totals of European shag within sector 14 of the North Orkney SPA between November 2022 and March 2023.

Sector	Nov ‘22	Dec ‘22	Jan ‘23 (survey 1)	Jan ’23 (survey 2)	Feb ‘23	Mar ‘23
14	38	22	12	9	8	9

 

Non-target species

A number of non-target species were recorded in the DAS of North Orkney SPA (Table 5; Table 6), the distributions of which are presented in Figure 111 for the February 2022 survey, and Figure 112 for the January 2023 survey.

Comparison of the results of DAS and shore-based surveys

On the 20 January 2023, DAS were undertaken on the same day as the shore-based counts in order to compare the results of the two survey methods. For Scapa Flow, DAS transects which crossed the shore-based count sectors were flown between approximately 11:30 and 12:07, whilst the shore-based counts were conducted between approximately 12:00 and 16:35 (Table 51). There was a direct temporal overlap in the surveys in sector 9, where the shore-based survey was conducted between 12:00 and 12:25, and the DAS transect across this sector was flown at 12.07. All other shore-based counts surveys within the Scapa Flow count sectors were conducted within approximately 4.5 hours of the DAS. At sector 14, within North Orkney SPA, the shore-based survey was conducted between 12:30 and 13:20 and the DAS transect at approximately 12:06.

Table 51. Survey timings of each sector for the DAS and shore-based surveys conducted on 20th January 2023. DAS timings are approximate. The survey in sector 37 on 20 January 2023 was not completed due to difficulties with access.

Sector	DAS approximate timing	Shore-based survey timing
9	12:07	12:00 – 12:25
10	11:52	14:00 – 14:30
11	11:43	15:40 – 16:00
12	11:44	16:20 – 16:35
14	12:06	12:30 – 13:20
37	11:53	N/A
38	11:30	16:00 – 16:20

Scapa Flow Comparison

Figure 113 shows the DAS and shore-based counts recorded across the Scapa Flow shore-based count sectors. All birds from the DAS that were observed within the shore-based count sectors were identified to species level. For great northern diver two birds were observed in sector 9 (inner Scapa Bay) during the DAS, but not during the shore-based surveys. In sectors 11 and 38, a similar number of birds were observed by both survey methods in approximately the same location. 

Similarly for red-breasted merganser, four birds were observed during the DAS in sector 9, whilst no birds were observed near this location in the shore-based counts. However, in the southernmost sector 38, two red-breasted merganser were observed by the DAS which may have been part of the adjacent flock of 12 birds observed during the associated shore-based count. 

There were no Slavonian grebe observed during the DAS, whilst three observations within the DAS transect area were observed in sector 9 during the shore-based counts. However, it should also be noted that the mapped swathe does not show the gaps between cameras (~25 m), which could mean birds were missed during the DAS within this area. Slavonian grebe were also detected in two other sectors by the shore-based counts, including a flock of 28 birds in sector 38.

For common eider, there were more birds observed in sector 9 in the DAS compared with the shore-based counts. In sector 11, lower numbers of birds were observed within the DAS swathe during the shore-based counts; however, several groups of birds were observed nearby and in sector 10 during the shore-based counts which could represent the same individuals recorded several hours later in the day. 

Large numbers of long-tailed duck were observed in sector 11 during the DAS, whereas many fewer were recorded during the shore-based counts in the same area. In sector 9, the DAS recorded a similar number of individuals to shore-based counts within the DAS swathe, but birds were recorded further offshore in the DAS. Similar observations for this species between the two survey methods were also observed in sector 38. 

In sector 9, the DAS recorded slightly more European shag in the same camera swathe area than recorded during the shore-based counts, but the shore-based counts recorded a greater number of European shag overall. In sector 38 there were birds recorded during the shore-based counts where there were no observations during the DAS. 

For black-throated diver, there were no observations in the DAS, and only one individual was recorded during the shore-based survey, which was observed outside of the DAS camera swathe in sector 12. Therefore, this species has not been mapped. 

 

North Orkney Comparison

Figure 114 shows the DAS and shore-based counts recorded across the North Orkney shore-based count sectors. All birds from the DAS that were observed within the shore-based count sectors were identified to species level. No great northern diver or Slavonian grebe were recorded in sector 14 during the DAS surveys. However, few birds of either species were observed during the shore-based counts; with only a single great northern diver overlapping the DAS transects. 

For common eider, a large flock of birds was observed in the DAS in the northwest of sector 14, whilst far fewer birds were recorded during the shore-based counts in the same location. Large numbers of long-tailed duck were also observed in the northwest of sector 14, with similar distributions observed in the shore-based counts. For both species, several observations were also made in similar locations in the east of the sector.

For red-breasted merganser, a single bird was observed in the east of the sector during the shore-based counts. The individuals recorded during the DAS may represent some of the individuals recorded during the shore-based counts. 

During the shore-based counts, 21 European shag were recorded, whilst there were no observations made of this species in sector 14 during the DAS. For black-throated diver, no birds were observed in either the DAS or shore-based counts. 

Discussion

Target species estimates

Estimates of target species within the SPAs from the DAS are discussed alongside previous survey estimates of the region for comparison. This includes the original ‘Areas of Search’ that underpinned the pSPA site selection using a mixture of VAS and land-based counts (Lawson et al., 2015) and the 2017/18 surveys of the pSPA boundaries which utilised both land-based and boat-based counts (Jackson, 2018 and Upton et al., 2018). It is important to note that previous surveys covered larger areas than that of the current SPA boundaries (Figure 1) and so population estimates are not directly comparable. In particular, waters around South Ronaldsay included in the Scapa Flow pSPA are not included in the site as classified. In addition, the results reported previously were derived from multiple surveys within a single winter season (Jackson, 2018 and Upton et al, 2018) or across several winter seasons (Lawson et al, 2015) and as such are not directly comparable with snapshot DAS, as reported here. The various surveys also differed in their methodology as stated above. 

Great northern diver

In Scotland, wintering great northern diver populations are generally located around the west and north coasts, with the highest winter population estimates being recorded in Orkney and Shetland (Booth et al., 1984; Williams, 2000; Lawson et al., 2015). The Orkney population is recognised as internationally important, since it comprises over 1 % of the biogeographic population; Scapa Flow was identified as supporting the third largest population in Scotland (Lawson et al., 2015). 

The SPA population estimates generated from the DAS are compared with counts from previous surveys of the region in Table 52. The Bayesian model-based population estimate of great northern diver within the Scapa Flow SPA estimated from the 2022 DAS data (262 birds (95 %CI 197 - 343)) was comparable to the VAS estimate presented in Lawson et al. (2015) for a larger Area of Search (236 birds). The DAS estimates from the February 2022 survey were however lower than land-based counts presented in the same report (506 birds and 317 birds for Scapa Flow and North Orkney, respectively) and substantially lower than counts from the 2017/18 land and boat-based surveys (Jackson 2018, Upton et al 2018). In addition, there are apparent differences in distributions, with fewer birds detected in the central Scapa Flow and closer to shore in the north and east of the Scapa Flow than in the 2017/18 surveys. It should however, be noted that sea conditions were rougher during the 2022 survey, potentially affected detectability of great northern diver. 

In comparison, the Bayesian model-based population estimate of great northern diver within the Scapa Flow SPA estimated from the January 2023 DAS survey (1,072 birds (95 % CI 884 – 1,232) was akin to the upper count of the original pSPA boundary from land and boat-based surveys in 2017/18 (Jackson, 2018) and higher than the Lawson et al. (2015) estimates. As with the 2022 survey, estimates in the North Orkney SPA from the 2023 DAS survey were lower than both previous reports.

Great northern divers are large birds and unlikely to have been missed along aerial transects if present at the sea surface. Whilst the species is known to be susceptible to flushing and/or diving in the presence of overhead aircraft, the high flight altitude of the DAS (550 m ASL) is beyond that known to flush birds or cause disturbance (Thaxter and Burton, 2009), so this is considered unlikely to have been a factor in the low population estimates recorded in the 2022 DAS survey. In addition, no signs of flushing of other species were noted during the survey. 

In the case of North Orkney SPA, a proportionally high number of birds in the February 2022 survey were located just beyond the perimeter of the marine SPA boundary, potentially related to the higher sea-states observed across the SPA during this survey. The trimming of records to the marine SPA boundary resulted in a 32 % reduction in estimated numbers, partially accounting for the lower-than-expected population estimate. 

In addition, some of this variation may be partially attributed to availability bias, where in shore-based and boat-based surveys, observers have a longer duration to survey a given area compared with DAS, increasing the likelihood that diving birds would return to the surface and be captured in these surveys. This applies particularly in the central Scapa Flow where deeper waters may mean birds are submerged for longer durations. The DAS estimates within this study have not been adjusted for availability bias. 

Divers are visual predators, with increased dive duration strongly correlated with increased water clarity (Thompson and Price, 2006). A higher sea state is likely to increase water turbidity. However, when comparing the distribution of birds recorded in the February 2022 DAS with sea state, the highest densities were recorded between Cava and Fara, where sea state was ‘moderate’ (4) and birds were also recorded to the southwest of Burray and between Hoy and Graemsay, where sea state reached ‘rough’ (5). However, there were areas of calm water where birds were not recorded. It could be that birds were more likely to be recorded in these areas of higher sea state, due to them taking shorter foraging dives than in calmer areas.

Weather, and its associated sea state, was better in the January 2023 DAS and likely affected the natural variation in abundance and distribution of birds, which in turn may contribute to the substantial difference in the great northern diver population estimates between the two years. However, as reported in Jackson (2018), bird numbers can range greatly between different counts.

Typically, shore-based detection of birds begins to deteriorate ~1.75 km offshore (Webb and Reid, 2004). However, the relatively close proximity of surveyors to birds in land-based surveys under 2 km means more individuals are likely to be identified to species level compared to aerial surveys, which is particularly important for species groups with similar plumage (e.g. divers). In February 2022, the respective number of unidentified diver records from the DAS was 4.5 % of the number of great northern diver recorded in Scapa Flow and 25 % of the number recorded in North Orkney. This was 3.2 % and 15.6 % respectively in January 2023. Therefore, identification rate is unlikely to be a measurable factor in the low estimates for Scapa Flow SPA and does not account for the full difference in the North Orkney estimates. 

The lowest numbers of great northern diver recorded in the 2021/22 shore-based counts occurred in February (6 birds compared to between 14 - 74 in Scapa Flow in other months, and 3 birds compared to between 3 – 18 in North Orkney).  This corroborates the low abundances recorded in the February 2022 DAS, when not a single diver was recorded in the inshore sector areas. It may be that bird abundance was truly low in that month, supported by suggestion that from February onward birds start to move to other sites (Jackson, 2018). 

Table 52 Comparison of SPA population estimates derived from DAS in this study with counts from previous surveys of the region for great northern diver

Survey	Method	Scapa Flow	North Orkney
Original Areas of Search (Lawson et al., 2015)	Mixture of visual aerial and land-based counts over multiple seasons and months	LBC - 506 Aerial - 236 WeBS -0	LBC - 317 WeBS - 8
pSPAs boundaries (Jackson, 2018; Upton et al., 2018)	Mixture of land-based and boat-based counts at roughly monthly intervals in 2017/18	571 – 1,016	441 - 777
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Strip transect analysis	263 (95 %CI 174 - 369)	79 (95 %CI 30 - 140)
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Bayesian point processing	262 (95 %CL 197 - 343)	76 (95 %CL 42 - 122)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Strip transect analysis	919 (95 %CI 572 – 1,310)	191 (95 %CI 105 - 300)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Bayesian point processing	1,072 (95 %CL 884 – 1,232)	208 (95 %CL 145 - 282)

 

Black-throated diver

During the winter non-breeding period, black-throated divers are generally located in a few key areas in Scotland, such as Scapa Flow (Williams, 2000), Caithness and the Moray Firth (Evans, 1998). At wintering sites, birds often congregate in small groups away from coast during which time they undergo a full moult (Forrester et al., 2012), making it difficult to accurately identify to species level regardless of the survey methodology utilised. 

Single snapshot surveys, such as large-scale DAS, may result in few detections of relatively scarce species; consequently, Bayesian model-based analysis of the February 2022 and January 2023 DAS data could not be undertaken for this species due to the low number of birds detected. However, design-based population estimates for black-throated diver in Scapa Flow both DAS exceed the 1 % threshold of the GB population. 

The Scapa Flow population estimated from the February 2022 DAS data using design-based methods (13 birds, 95 %CI 3 – 32) was within the range of birds recorded during different shore-based count rounds by Jackson (2018) (9 - 39 birds). The estimate from the January 2023 DAS (30 birds (95 % CI 0 – 84)) was closer to the higher estimates from the 2017/18 survey counts (Table 53). This is despite the difference in survey area, as no black-throated diver were observed in the additional area included in the pSPA surveyed by Jackson (2018).

In addition, the North Orkney population estimated from the February 2022 DAS data using design-based methods (5 birds, 95 %CI 0 - 13) is consistent with the number reported from shore-based counts by Upton et al. (2018) (4 birds). 

The species is highly mobile, making it difficult to prevent double counting using traditional shore-based surveying methods (Williams, 1999). The relatively high flight speed (~120KN) of DAS surveys alleviates this problem, although distinguishing black-throated divers from other diver species can prove more challenging. This is evident in the present study, where a total of 11 black-throated diver were observed across both SPAs and both years, but a further 24 divers were recorded which could not be identified to species level (Table 5; Table 7). However, this number is likely to include great northern divers or even red-throated divers.

Shore-based counts in the present study recorded multiple small flocks, indicative of the wintering groups as described in Forrester et al. (2012). Such flocks may be missed by DAS if outside survey swathe.  However, the use of aerial survey methods to monitor black-throated divers is also potentially beneficial since it can capture birds outwith the ~1.75 km detection range of shore-based surveys (Webb and Reid, 2004) and individuals are often located between 1 – 3 km offshore (Jackson, 2018).  This is evident in the detections of birds during both DAS surveys (Figure 16; Figure 17).  

Black-throated diver were only recorded in January 2022 during the winter 2021/22 shore-based counts while birds were recorded in all months apart from February 2023 in the winter 2022/23 shore-based counts. 

Table 53 Comparison of SPA population estimates derived from DAS in this study with counts from previous surveys of the region for black-throated diver

Survey	Method	Scapa Flow	North Orkney
Original Areas of Search (Lawson et al., 2015)	Mixture of visual aerial and land-based counts over multiple seasons and months	LBC - 57 Aerial - 0 WeBS -1	LBC - 4 Aerial - 0 WeBS - 0
pSPAs boundaries (Jackson, 2018; Upton et al., 2018)	Mixture of land-based and boat-based counts at roughly monthly intervals in 2017/18	9 - 39	LBC - 4
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Strip transect analysis	13 (95 %CI 0 - 32)	5 (95 %CI 0 - 13)
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Bayesian point processing	-	-
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Strip transect analysis	30 (95 %CI 0 - 84)	-
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Bayesian point processing	-	-

 

Slavonian grebe

Slavonian grebe are distributed throughout Scotland during the breeding season, primarily concentrated around inland freshwater lochs, after which they move towards shallow, coastal areas (Forrester et al., 2012). Although the breeding population of Scotland was estimated at around 50 birds in the 1990s, wintering populations were calculated at 339 – 499 birds, indicating the presence of birds migrating from other areas such as Iceland and north Norway (Forrester et al., 2012). Relatively high numbers of the species are known to overwinter at Scapa Flow and Orkney (Evans, 2000), with Scapa Flow considered to be the most important site in Scotland for the species (Scottish Natural Heritage, 2016a). 

VAS of Scapa Flow in 2006/07 by Lewis (2008) did not record any Slavonian grebe despite there being a known population in the region (Pollitt et al., 2003). In contrast, the February 2022 DAS in the present study did detect Slavonian grebe and generated population estimates of 70 birds (95 %CI 40 - 123) in Scapa Flow SPA through Bayesian analysis and 38 birds (95 %CI 13 - 66) for North Orkney SPA, through design-based strip transect analysis (Bayesian could not be undertaken due to low counts). Surveys in 2023 detected higher numbers with population estimates of 114 birds (95 % CI 69 – 175) in Scapa Flow and 103 birds (95 % CI 61-163) in the North Orkney SPA through Bayesian analysis. This likely indicates reduced disturbance through increased flight speed and altitude for DAS as compared to VAS, which may be beneficial when recording this species.

However, population estimates from shore and boat-based surveys in winter 2017/18 were higher, comprising 27.5 % and 14.6 % of the GB population in Scapa Flow pSPA (peak count: 161 birds) and North Orkney pSPA (peak count: 303 birds), respectively (Jackson, 2018; Upton et al., 2018) (Table 54). This is despite the difference in survey area, as Slavonian grebe were not observed in the additional areas within the Scapa Flow and North Orkney pSPAs as surveyed by Jackson (2018) and Upton et al. (2018), respectively. DAS may not fully capture Slavonian grebe populations due to their small size and preference for sheltered areas which means shore-based counts are likely to be more reliable.  This is also supported by the relatively high number of birds recorded from the shore-based VP surveys in the present study. Weather and associated sea state may also have affected the DAS results, whether by natural variation in abundance and distribution of birds, or by a reduced ability to detect grebe in higher sea states, with markedly higher estimates in both SPAs in the 2023 DAS when sea state was calmer.

However, the mean population estimate generated from the February 2022 DAS fell within the individual count range for Scapa Flow pSPA using shore and boat-based surveys (67 – 161 birds; as reported by Jackson, 2018). Slavonian grebe numbers may simply have been low during the February 2022 DAS. Only 3 birds were recorded in the inshore sectors in that month (with numbers ranging between 2 and 21 birds in other months). Population estimates generated from the January 2023 DAS were higher, particularly in the North Orkney SPA, and the highest shore-based counts were recorded in January in both years, suggesting there may be an increase of birds in this area at this time of year.

Table 54 Comparison of SPA population estimates derived from DAS in this study with counts from previous surveys of the region for Slavonian grebe

Survey	Method	Scapa Flow	North Orkney
Original Areas of Search (Lawson et al., 2015)	Mixture of visual aerial and land-based counts over multiple seasons and months	LBC - 135 Aerial - 0 WeBS - 11	LBC - 120 Aerial - 0 WeBS - 1
pSPAs boundaries (Jackson, 2018; Upton et al., 2018)	Mixture of land-based and boat-based counts at roughly monthly intervals in 2017/18	67 - 161	200 - 303
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Strip transect analysis	73 (95 %CI 29 - 132)	38 (95 %CI 13 - 66)
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Bayesian point processing	70 (95 %CL 40 - 123)	-
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Strip transect analysis	100 (95 %CI 50 - 159)	94 (95 %CI 24 - 206)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Bayesian point processing	114 (95 %CL 69 - 175)	103 (95 %CL 61 - 163)

 

Red-breasted merganser 

Red-breasted merganser are found year-round along the Scottish coast, typically breeding in sea lochs and estuaries and remaining around coastal areas during the winter non-breeding period (Forrester et al., 2012; Scottish Natural Heritage, 2016b). The wintering population is likely to be boosted by migrant birds from northern latitudes, such as those from Iceland and Fennoscandia (Forrester et al., 2012). 

Shore-based counts presented in Lawson et al. (2015) indicated populations of 539 birds and 279 birds in the Scapa Flow and North Orkney Areas of Search, respectively (Table 55). Peak counts for Scapa Flow and North Orkney pSPAs in 2017/18, generated through shore and boat-based surveys, suggest the two areas may support approximately 4.4 % and 8.5 % of the GB population, respectively (equivalent to 370 and 717 birds; Jackson, 2018; Upton et al., 2018). However, the February 2022 DAS generated lower model-based population estimates of 102 birds (95 %CI 56 - 200) in Scapa Flow SPA and 214 birds (95 %CI 135 – 305) in North Orkney SPA. The January 2023 DAS generated higher estimates; 338 birds (95 % CI 255 – 435) in Scapa Flow SPA, similar to the peak count made by Jackson (2018).  However, the estimated 331 birds (95 % CI 236 – 521) in North Orkney SPA were below the range recorded by Upton et al. (2018). Fewer than 10 red-breasted mergansers were observed in the additional pSPA area surveyed in Jackson (2018) and Upton et al. (2018), so the difference in counts in the 2022 DAS cannot be attributed to the reduced survey area in the DAS. 

Despite this, the population size for North Orkney estimated by both DAS exceeded the 1 % threshold of the GB population (84 birds). The distribution of birds seen across the SPAs also match hotspots presented in Jackson (2018) and Upton et al. (2018); however, no birds were found near Stromness and Stenness in 2021/22, nor in some of the most northern Orkney bays such as Veantrow in both years.  

Lower population sizes from DAS may be partly attributed to the ability of DAS to effectively distinguish species plumage from others such as goosander (as suggested by the relatively high number of ‘duck sp.’ recorded), due to overlap in distribution and habitat preference. Whilst red-breasted merganser are known to be sensitive to disturbance from boats and aircraft, with particularly high sensitivity to noise (Jarrett et al., 2018), and have been observed to flush from survey sites when surveyed from low flying aircraft or boat, the DAS surveys were flown well beyond the minimum height recommended and so are considered unlikely to have caused disturbance (Thaxter and Burton, 2009).  

Sample snapshot surveys, as used within DAS, may not result in sufficient detections of relatively scarce species. Shore-based methods are likely to be more effective to monitor sparsely populated species typically found in inshore coastal areas, and more easily allow detection of birds located on intertidal rocks or land. Red-breasted merganser typically occur close inshore, in small groups or pairs, almost exclusively in the most sheltered coastlines, as is evident from the DAS (Figure 28 and Figure 29) and shore-based surveys (Figure 34 and Figure 35) in the present study. The species is known to roost on land, with birds recorded on brackish lochans during the 2017/18 shore-based counts. The production of the SPA estimates included records within the marine SPA boundary only and so some additional roosting birds may have been omitted from the estimates. For both SPAs, a proportionally high number of birds were located just beyond the perimeter of the marine SPA boundaries, with the trimming of records to the marine SPAs resulting in a 34.48 % and 21.54 % reduction for Scapa Flow and North Orkney during the winter 2021/22 surveys, respectively, partially accounting for some of the lower-than-expected population estimates. This had a lower proportional effect during the winter 2022/23 surveys, with the trimming of records resulting in only 6.98 % and 14.10 % reductions respectively for each SPA. Rougher sea conditions during the February 2022 DAS may also have impacted the detectability of the species. 

There is also the possibility that bird abundance was truly low in February 2022. The lowest numbers of birds recorded in the 2021/22 shore-based VP surveys occurred in February; with 3 birds observed in the North Orkney inshore sector (compared to between 6 – 41 in other months) and 11 observed in the Scapa Flow inshore sector (compared to between 13 – 24 in other months). This was also the case in the 2022/23 shore-based suveys in Scapa Flow, where the lowest number of birds was recorded in February 2023 (2 birds compared with between 3 and 15 in other months). Observations of red-breasted merganser were higher in the North Orkney SPA with the lowest number of birds recorded in the second January 2023 survey, the same month as the January 2023 DAS (2 birds compared with between 8 and 44 in other months).

One of the disadvantages to sample-snapshot surveys, such as DAS, is that they may coincide with periods of low bird abundance, as is expected with natural variation. Undertaking repeated surveys can counteract this but may be less cost effective than other methods.

Table 55 Comparison of SPA population estimates derived from DAS in this study with counts from previous surveys of the region for red-breasted merganser 

Survey	Method	Scapa Flow	North Orkney
Original Areas of Search (Lawson et al., 2015)	Mixture of visual aerial and land-based counts over multiple seasons and months	LBC - 539 Aerial - 101 WeBS - 0	LBC - 279 WeBS - 35
pSPAs boundaries (Jackson, 2018; Upton et al., 2018)	Mixture of land-based and boat-based counts at roughly monthly intervals in 2017/18	200 - 370	546 - 717
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Strip transect analysis	79 (95 %CI 25 - 157)	214 (95 %CI 37 - 506)
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Bayesian point processing	102 (95 %CL 56 - 200)	214 (95 %CL 135 - 305)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Strip transect analysis	330 (95 %CI 417 - 569)	284 (95 %CI 68 - 663)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Bayesian point processing	338 (95 %CL 255 - 435)	331 (95 %CL 236 - 521)

 

Velvet scoter

During the winter period, velvet scoters are typically found in coastal waters, generally distributed along the east coast and north coast around Orkney and Caithness (Forrester et al., 2012). 

No velvet scoters were recorded during either DAS surveys of Scapa Flow or North Orkney, nor in the original VAS conducted by Lawson et al. (2015) (Table 56).

No birds were recorded in shore-based surveys in North Orkney SPA during winter 2021/22; however, a total of 29 birds were recorded in December 2022 and the first January 2023 survey in winter 2022/23. This is despite the shore-based surveys being conducted in sectors in which no velvet scoter were recorded in previous counts (Upton et al., 2018). Existing evidence suggests North Orkney supports between approximately 6-10 % of the GB population (Scottish Natural Heritage, 2016a; Upton et al., 2018), however this was not supported by the present study.

A small flock of seven birds was recorded in the shore-based surveys of Scapa Flow in January 2022, however, no velvet scoter were recorded in any other survey during the winter 2021/22 or winter 2022/23 shore-based surveys. In Jackson (2018), no velvet scoter were recorded in this area across the surveys, and peak counts summed across the wider pSPA area equated to eight birds. 

Velvet scoter are relatively obvious birds to identify at close range and can form substantial flocks. However, these flocks often occur beyond 2 km offshore and as mixed-species flocks which can make identification challenging. Sitting birds can be mistaken for common scoter in both aerial and shore-based surveys, particularly if the characteristic white wing bar is concealed which is more likely to occur when looking from above. HiDef cameras are angled to avoid this problem somewhat and size and shape are still useful indicators to determine species.   However, few common scoter were observed during both the DAS and shore-based surveys, making misidentification an unlikely cause of the low estimates. 

Typically, aerial surveys are not recommended for the species, with shore-based surveys generally yielding higher detection rates. Population estimates from strip transect survey designs, as used in DAS, may underestimate true population sizes if large flocks happen to fall outside of transects. However, shore-based methods may also be insufficient for this species, as flocks often occur further than ~1.75 km offshore, which may account for low counts or absence of velvet scoter during the shore-based counts.  

Fluctuations in estimated populations have been attributed to the challenges in accurately surveying the species and have been suggested to not accurately reflect true population change (Forrester et al., 2012).

Table 56 Comparison of SPA population estimates derived from DAS in this study with counts from previous surveys of the region for velvet scoter

Survey	Method	Scapa Flow	North Orkney
Original Areas of Search (Lawson et al., 2015)	Mixture of visual aerial and land-based counts over multiple seasons and months	LBC - 13 Aerial - 0 WeBS - 0	LBC - 147 Aerial - 0 WeBS - 2
pSPAs boundaries (Jackson, 2018; Upton et al., 2018)	Mixture of land-based and boat-based counts at roughly monthly intervals in 2017/18	8	131 - 244
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Strip transect analysis	-	-
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Bayesian point processing	-	-
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Strip transect analysis	-	-
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Bayesian point processing	-	-

 

Long-tailed duck

Wintering populations of long-tailed duck in Scotland are primarily located in sheltered, coastal areas, although birds have been observed offshore during this period (Forrester et al., 2012). They typically have high site fidelity, possibly due to habitat requirements, with wintering hotspots located in the Moray Firth, Shetland and Orkney (Forrester et al., 2012). 

Long-tailed duck was the most abundant species recorded in both DAS surveys of Scapa Flow; abundance was also relatively high in North Orkney. Model-based estimates from Bayesian point processing varied slightly to those from design-based but were inside the 95 % CIs. Model-based population estimates from the 2021/22 DAS across Scapa Flow and North Orkney SPA were calculated at 1,502 birds (95 % CI 1,279 – 1,792) and 1,915 birds (95 % CI 1,668 – 2,734), respectively. This is compared to the 2022/23 DAS survey where 3,678 birds (95 % CI 3,321 – 4,343) and 2,428 birds (95 % CI 3,199 – 2,654) were estimated in Scapa Flow and North Orkney SPA, respectively.

The 2022 DAS estimates were higher than both VAS and land-based counts undertaken by Lawson et al. (2015) but lower than the 2017/18 peak counts for the pSPA areas acquired through shore and boat-based surveys (Table 57). However, these figures were still well within the individual count ranges from the 2017/18 surveys; with between 893 and 1,996 birds recorded in the Scapa Flow pSPA and 1,271 - 2,651 birds recorded in the North Orkney pSPA (Jackson,2018; Upton et al., 2018). This is despite the difference in survey area, as very few long-tailed ducks were observed in the additional pSPA areas surveyed by Jackson (2018) and Upton et al. (2018), such as the southern sectors around South Ronaldsay. In contrast, the 2023 DAS estimate for Scapa Flow was 84 % higher than the highest count by Jackson (2018), whilst the North Orkney estimate was similar to the highest count from Upton et al. (2018).

As with other species, timing of the snapshot DAS may be relevant. February had relatively low numbers of long-tailed duck recorded in the 2021/22 shore-based surveys compared to other months. Shore-based counts in the present study recorded 34 birds in the Scapa Flow inshore sectors (compared to between 90 and 168 in other months) and 10 birds in the North Orkney inshore sector (compared to between 1 and 72 in other months). It was a similar case in the January 2023 DAS where shore-based counts in this month had the lowest number of long-tailed ducks recorded (31 birds compared to between 108 and 265 in other months and 9 birds compared to between 15 and 471 in other months in Scapa Flow and North Orkney, respectively). This may be indicative of lower abundances of long-tailed duck during February 2022 and second January 2023 shore-based surveys and shows the inter-monthly variability of populations over the winter period.

As with any diving species, snapshot surveys such as aerial surveys may miss birds diving at the time of survey leading to underestimated population size, and no correction factor was applied to the DAS estimates to account for this. In shore-based counts observers have a longer period of time to count a given area increasing the likelihood that diving birds would return to the surface and be captured in the shore-based surveys. 

For transect surveys, flocks of birds present between transects can be missed if they are beyond the range of sampled area. Long-tailed duck exploit a variety of prey species and can dive up to 60 m, therefore shore-based counts may not be as effective at capturing true population size if individuals are distributed in deeper, offshore waters. Numbers of long-tailed duck were markedly higher in the 2023 DAS, so it may be that the 2022 DAS was a reflection of the poorer weather and sea state at that time. Again, this may suggest a reduction in the ability to detect long-tailed duck in high sea states, or truly reflect a lower abundance of birds in the area at this time. 

The distribution of long -tailed duck in Scapa Flow is likely to be influenced by attraction to fish farms, such as those located along the east coast of Hoy, the north coast of South Ronaldsay and within sector 12 of the Scapa Flow inshore sectors; which had the highest mean percentage use of any sector in the present study (48.13 % in February 2022; 56.51 % in January 2023). 

Table 57 Comparison of SPA population estimates derived from DAS in this study with counts from previous surveys of the region for long-tailed duck 

Survey	Method	Scapa Flow	North Orkney
Original Areas of Search (Lawson et al., 2015)	Mixture of visual aerial and land-based counts over multiple seasons and months	LBC – 1393  Aerial - 631 WeBS - 0	Aerial – 952 WeBS - 81
pSPAs boundaries (Jackson, 2018; Upton et al., 2018)	Mixture of land-based and boat-based counts at roughly monthly intervals in 2017/18	893 – 1,996	1,274 – 2,651
Final SPA boundaries from the current study	DAS – Strip transect analysis	1,381 (95 %CI 719 – 2,160)	1,570 (95 %CI 554 – 3,170)
Final SPA boundaries from the current study	DAS – Bayesian point processing	1,502 (95 %CL 1,279 – 1,792)	1,915 (95 %CL 1,668 – 2,734)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Strip transect analysis	3,149 (95 %CI 625 – 7,559)	2,190 (95 %CI 1,051 – 3,331)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Bayesian point processing	3,678 (95 %CL 3,321 – 4,343)	2,428 (95 %CL 2,199 – 2,654)

 

Common eider

Common eider are widely distributed across the UK but are predominantly found in northern England and Scotland. They are a qualifying feature of Scapa Flow only, which supports the 4^th largest population in Scotland. 

In 2017/18, a peak of 2,324 birds were recorded in Scapa Flow pSPA (3.9 % of the GB wintering population; Jackson, 2018) and 3,806 birds were recorded in North Orkney pSPA (6.3 % of the GB wintering population; Upton et al., 2018), in shore-based surveys, with additional boat-based surveys of the central Scapa Flow (Table 58). The model-based population estimates derived from the February 2022 DAS were considerably lower, with 676 birds (95 % CI 534 - 824) and 905 birds (95 % CI 759 – 1,580) estimated for Scapa Flow SPA and North Orkney SPA respectively. Estimates from the January 2023 DAS (Scapa Flow: 1,504 birds (95 % CI 1,322 – 1,706); North Orkney: 2,134 birds (95 % CI 1,887 – 2,421)) were much higher and similar to the lowest respective counts from the 2017/18 surveys (Jackson, 2018; Upton et al., 2018). However, estimates from the February 2022 DAS do overlap with previous VAS for North Orkney (Lawson et al., 2015). Despite common eider being a qualifying feature of the Scapa Flow SPA only, higher population estimates were estimated for North Orkney in both the February 2022 DAS and the 2017/18 pSPA surveys. 

Common eiders are a relatively large (50-60cm), conspicuous bird with contrasting plumage, making them relatively easy to detect in DAS surveys. However, as a flocking species, there is potential for eider in dense aggregations to be underestimated during both aerial and shore-based surveys. February produced the lowest counts in the 2021/22 shore-based surveys for both the Scapa Flow inshore sectors (28 birds compared to between 58 – 100 in other months) and the North Orkney inshore sector (11 birds compared to between 12 and 44 in other months). The same was true for the 2022/23 shore-based surveys where the month with both the DAS and shore-based counts (January Survey 02 2023) produced the lowest counts (2 birds compared to between 29 and 109 in Scapa Flow sectors and 7 birds compared to between 19 – 1,155 in North Orkney). As such, the DAS population estimates may be a reflection of natural variation and low bird abundance in those months.

During the February 2022 DAS, a proportionally high number of birds were located just beyond the perimeter of the marine SPA boundaries, with the trimming of records to the marine SPAs resulting in a 27.60 % and 10.27 % reduction for Scapa Flow and North Orkney respectively. This may be due to birds taking refuge following the poor weather conditions observed prior to this survey and may partially account for some of the lower-than-expected population estimates. The percentage difference in the January 2023 survey was much lower (Scapa Flow: 0.85 %; North Orkney: 2.34 %)

Common eider predominantly occurred around the coasts and shallow inshore waters around the SPAs. The eastern part of the central Scapa Flow is generally too deep for foraging, as common eider dive to the seabed for their prey. The distribution of birds in Scapa Flow is also likely influenced by attraction to fish farms, such as those located along the east coast of Hoy, the north coast of South Ronaldsay and within sector 12 of the Scapa Flow inshore sectors; which had a relatively high mean percentage use compared to other sectors in the present study (30.64 % and 29.95 % during the 2021/22 and 2022/23 surveys, respectively). Birds did not appear to use sector 37 in either year (with the exception of one observation in March 2023), which was consistent with findings presented in Jackson (2018). This may be a result of the seabed dropping off quickly in this part of the coastline such that this area is not optimal for foraging.

Table 58 Comparison of SPA population estimates derived from DAS in this study with counts from previous surveys of the region for common eider 

Survey	Method	Scapa Flow	North Orkney
Original Areas of Search (Lawson et al., 2015)	Mixture of visual aerial and land-based counts over multiple seasons and months	LBC – 1,994 Aerial – 1,850 WeBS - 0	Aerial – 1,471 WeBS - 49
pSPAs boundaries (Jackson, 2018; Upton et al., 2018)	Mixture of land-based and boat-based counts at roughly monthly intervals in 2017/18	1,696 – 2,324	2,102 – 3,806
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Strip transect analysis	581 (95 %CI 263 - 961)	840 (95 %CI 328 – 1,535)
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Bayesian point processing	676 (95 %CL 534 - 824)	905 (95 %CL 759 – 1,580)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Strip transect analysis	1,456 (95 %CI 454 – 2,938)	2,266 (95 %CI 285 – 5,842)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Bayesian point processing	1,504 (95 %CL 1,322 – 1,706)	2,134 (95 %CL 1,887 – 2,421)

 

European shag

European shag is a common and widespread resident of coastal regions throughout Scotland. They are partially migratory, with some individuals remaining at the breeding colonies in Scotland throughout the year, and others migrating south into England in winter (Forrester et al., 2012). Scapa Flow SPA supports the largest concentration of wintering European shag in Great Britain, and North Orkney is one of only three areas of Scotland where non-breeding aggregations of European shag have been identified (Scottish Natural Heritage, 2016a; 2016b). 

In 2017/18, a peak count of 3,726 birds was recorded in shore and boat-based surveys of Scapa Flow (3.4 % of the GB wintering population; Jackson, 2018) and 2,968 birds in shore-based surveys of North Orkney (2.7 % of the GB wintering population; Upton et al., 2018). European shag were consistently recorded throughout the 2021/22 shore-based counts for both sites. Despite a marked increase in the January 2023 DAS, the Bayesian model-based population estimates derived from both the February 2022 and January 2023 DAS in this report were considerably lower than the 2017/18 peak estimates, at 134 birds (95 % CI 88 - 193) and 161 birds (95 % CI 109 - 218) in February 2022 and 872 birds (95 % CI 597 – 1,560) and 1,667 birds (95 % CI 1,356 – 2,189) in January 2023 for Scapa Flow SPA and North Orkney SPA respectively (Table 59).

Whilst there were European shag observed throughout the pSPA areas surveyed by Jackson (2018) that were not subsequently within the final SPA boundaries, the vast majority of observations were within the Scapa Flow SPA as surveyed by DAS in the present study. For North Orkney SPA in 2017/18 (Upton et al., 2018), 294 and 59 individuals respectively were observed in sectors 1 and 5, which were not surveyed in the present study.  However, the reduction in survey area to match the SPA boundary does not fully account for the lower DAS estimates.

An additional number of factors are likely to explain the differences between the 2017/18 and 2021/22 and 2022/23 population estimates.

The appearance and behaviour of European shag can elicit challenges for multiple survey methods. European shag can be missed from the shore during VP surveys because of their dark plumage, especially in rough seas or at a distance of more than 1 km from shore. Whilst the densest aggregations occur close inshore, European shags also favour exposed coastlines where sea states are frequently between 3 and 5. Birds are often present on the outer limits of shore-based visibility, and even occur in the central Scapa Flow, despite much of this area being too deep for foraging. 

As with shore-based surveys, the dark plumage of European shags can make detection hard during DAS, due to a lack of contrast against the sea surface, especially if the sea state is high. Birds also roost on land, generally within 100 m of the mean low water springs (MLWS) boundary, making concurrent land-based counts a necessity.

The number of detections of European shags was markedly lower in the 2022 DAS and one factor could be that weather and sea state were not optimal for their detection. Mean sea state recorded from the 2022 DAS footage (Figure 7) can be compared with the distributions and detections of European shag. In Scapa Flow, birds were recorded within sea state 0 ‘Calm (glassy)’ to sea state 5 ‘Rough’. In North Orkney, birds were recorded in sea state 0 ‘Calm (glassy)’ to sea state 4 ‘Moderate’; however, no birds were detected in the few sections of sea state 5.  

Whilst the species is known to be susceptible to flushing and/or diving in the presence of overhead aircraft, the high flight altitude undertaken during the DAS (550 m ASL) is beyond that known to flush birds or cause disturbance (Thaxter and Burton, 2009), so this is considered unlikely to have been a factor influencing detection. However, European shags do spend a significant proportion of time diving, with mean dive durations of 85 - 97 seconds recorded (Watanuki et al., 2005; De Rijcke, 2015). As such, snapshot surveys such as DAS are likely to underestimate their numbers compared with shore-based and boat-based surveys where observations are made over a longer duration, during which birds may return to the surface. Population estimates within this report have not been corrected for availability bias of diving birds and this is likely to account for a large proportion of the discrepancy between population estimates. 

In addition, the species is known to regularly occur on land, roosting and wing-drying along the shoreline, on skerries and on tidally exposed rocks. Identification of birds within the DAS was only undertaken to the high-tide mark, and so will not have recorded any potential birds occurring on rocks above this. Population estimates were also generated to the marine SPA boundary specifically, meaning any observations of birds beyond the SPA shapefile, above MLWS, would have been cropped out.  For both SPAs within the February 2022 DAS, a relatively high number of birds were located just beyond the perimeter of the marine SPA boundaries, with the trimming of records to the marine SPA boundaries resulting in a 20.00 % and 16.28 % reduction for Scapa Flow and North Orkney respectively. Trimming of records from the January 2023 DAS to the marine SPA boundaries resulted in only a 7.07 % and 3.34 % reduction for Scapa Flow and North Orkney respectively; suggesting that perhaps the rougher sea state prior to the 2022 DAS led to more birds taking refuge onshore. 

Lastly, there is always an element of seasonal variation in bird abundance. Counts recorded in the winter 2021/22 shore-based surveys were lowest in February (although not considerably), suggesting that fewer birds may have been present in the region in this month when the DAS was flown. A total of 31 birds were recorded in the Scapa Flow inshore sectors (compared to between 38 and 53 in other months) and only 7 birds were recorded in the North Orkney inshore sector (compared to between 14 and 24 in other months). High variability in shore-based counts was also observed during the winter 2022/23 surveys, with between five (March 2023) and 57 (February 2023) birds recorded across the sectors. 

Table 59 Comparison of SPA population estimates derived from DAS in this study with counts from previous surveys of the region for European shag 

Survey	Method	Scapa Flow	North Orkney
Original Areas of Search (Lawson et al., 2015)	Mixture of visual aerial and land-based counts over multiple seasons and months	LBC – 2.929 Aerial - 0 WeBS - 190	LBC – 1,742 Aerial – 0 WeBS - 23
pSPAs boundaries (Jackson, 2018; Upton et al., 2018)	Mixture of land-based and boat-based counts at roughly monthly intervals in 2017/18	582 – 3,726	1,758 – 2,968
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Strip transect analysis	130 (95 %CI 58 - 227)	150 (95 %CI 79 - 239)
Final SPA boundaries from the current study	DAS single survey Feb 2022 – Bayesian point processing	134 (95 %CL 88 - 193)	161 (95 %CL109 - 218)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Strip transect analysis	764 (95 %CI 253 – 1,416)	1,345 (95 %CI 432 – 2,562)
Final SPA boundaries from the current study	DAS single survey Jan 2023 – Bayesian point processing	872 (95 %CL 597 – 1,560)	1,667 (95 %CL 1,356 – 2,189)

 

Comparison of DAS and shore-based surveys

On 20 January 2023, an additional shore-based count was undertaken on the same day as the DAS in order to compare the abundance and densities of birds recorded using these two methods. However, direct comparisons were only feasible in one sector, given timing of counts. The results were visually compared by plotting the observations of birds from these methods against each other. Across the species, there was no consistent pattern in the comparison of the results from the two methods. 

In the Scapa Flow SPA surveys, great northern diver, red-breasted merganser, eider and long-tailed duck were observed within the DAS swathe areas compared with the shore-based surveys, whilst more Slavonian grebe and European shag were recorded in the shore-based surveys. In the North Orkney surveys, substantially more common eider and long-tailed duck were detected during the DAS compared with the shore-based surveys. However, these surveys were conducted approximately 24 minutes apart, during which time the birds detected during the DAS could have moved from the area. As in the Scapa Flow SPA survey, a large number of European shag were detected during the shore-based surveys, where no birds were detected during the DAS. 

Great northern diver are large, conspicuous birds which are relatively easy to detect during DAS. The higher number of birds recorded during the Scapa Flow SPA DAS compared with the shore-based counts, particularly in sector 9, where there was a direct temporal overlap in the DAS and shore-based counts, suggests that DAS may be the preferred method for this species. In the North Orkney SPA surveys, birds were observed just inside the DAS swathe area during the shore-based surveys, whereas no birds were detected during the DAS, although these birds may have moved outside of the DAS swathe between surveys. 

Across both the Scapa Flow SPA and North Orkney SPA shore-based surveys, fewer common eider and long-tailed duck were recorded within the DAS swathe area compared with numbers detected by the DAS, despite these being large and conspicuous birds and being observed less than 1 km offshore. However, these comparisons are constrained by the temporal mismatch between the surveys. More red-breasted merganser were also observed during the Scapa Flow SPA DAS compared with shore-based counts. However, where species are generally rare or sparsely distributed, such as red-breasted merganser, and the area of interest is primarily inshore waters, the species may still benefit from the increased survey coverage usually obtained with shore-based counts. Alternatively, DAS may be supplemented by shore-based counts for this species.

Due to their dark plumage, it can be challenging to detect European shag especially during higher sea states, such as those along exposed coastlines which these species tends to favour. Similarly, Slavonian grebe are small in size, and prefer sheltered inshore areas which can make it more challenging to detect these birds during DAS. Given that more European shag and grebe were recorded during the shore-based counts compared with the DAS, the current study suggests that it might be advantageous to supplement DAS with shore-based surveys for these species.

There were insufficient observations in either/both the DAS and shore-based surveys to allow a comparison of methods for black-throated diver and velvet scoter. 

Survey limitations

As touched upon in the previous sections, there are a number of factors, additional to the reduced extent of the SPAs from the previous pSPAs and Areas of Search, which may contribute to the lower population estimates produced for some species in the 2021/22 and 2022/23 DAS in comparison to previous surveys. 

Weather conditions may have had an influence on the presence of birds within the survey area, especially during the 2021/ 2022 winter DAS. Feedback from the shore-based observer on the day of the February 2022 DAS reported that the number of birds present was far lower than what would normally be expected, and that this was almost certainly due to the very stormy weather that occurred prior to the survey date (Higson, P., pers. comm., 21 February and 25 November 2022). These low abundances were also reflected in the 2021/22 shore-based counts, where February was consistently the month with the lowest number of records for all target species. The higher number of birds recorded in the January 2023 DAS (19,740 birds in total compared to 10,504 birds in the February 2022 DAS) supports this theory. There is natural variation in the distributions of marine bird species, which may not be captured in snapshot surveys, as the survey may fall on a period of natural low abundance. 

Additionally, weather conditions may have resulted in sub-optimal conditions for bird detection and identification, particularly during the February 2022 DAS when weather conditions were poorer. Dark birds, such as European shags, can be particularly hard to detect in rocks and crags in aerial footage and detection may be further challenged offshore in higher sea states, where there is little contrast between their plumage and the surrounding darkened waters. However, both DAS were flown within suitable weather windows with varied sea states  across the SPA survey areas, ranging between sea state 0 ‘Calm (Glassy)’ and 5 ‘Rough’ (Figure 7). These sea states are deemed viable for detection and identification by digital video aerial standards, which captures multiple frames per detection for scrutiny.

Some of the target species are known to frequently roost on land, including European shag and red-breasted merganser. In this DAS study, birds were identified up to the high tide mark. Therefore, there is the possibility that birds may have been roosting beyond this point and therefore not included. Additionally, in poorer weather conditions, such as those observed prior to the February 2022 DAS, birds may take refuge on the coast, where they may be missed above the high tide mark. This cut off point could be re-evaluated in future DAS, to include more birds found onshore but which will forage within the SPA boundaries.

Whilst there are many advantages to aerial survey methods, transect designs are not always suitable for species which form large, unevenly distributed flocks. In these instances, if large flocks fall between surveyed transects then analysis can lead to an underestimated population size. In contrast, if large flocks fall within transects it may lead to an overestimation. This is especially true for single surveys, and repeated surveys across the winter period would enable more robust estimates of abundance.

In addition, all of the target species spend a measurably amount of time under the water surface diving to forage. Whilst the height of the plane in the present study is not considered likely to be a factor in increasing disturbance or diving behaviour (Thaxter & Burton, 2009), estimates were not corrected for the availability of birds at the sea surface. As such, they are likely to be measurably lower than shore or boat-based counts where observers have a longer duration to witness any birds returning to the sea surface after diving.

The survey was designed with a minimum target of 24.3 % site coverage, as this was deemed sufficient to ensure precise abundance estimation of target species. The CV is a measure of the precision of the population and density estimates. Whilst this coverage was achieved, the CV of the some of the target species population estimates generated through design-based strip transect analysis were relatively high, ranging between 0.20 and 0.97 in Scapa Flow and 0.26 and 0.98 in North Orkney across the two surveys. The CV was highest for red-breasted merganser in Scapa Flow and black-throated diver in North Orkney, whilst the CV was lowest for great northern diver and long-tailed duck in both SPAs across the two surveys. However, the relationship between survey coverage and resultant CVs is not straightforward and increasing coverage will not necessarily result in increased precision of estimates. 

In terms of comparing the results of the simultaneous DAS and shore-based surveys, it is challenging to make a direct comparison between survey methodologies. This is for a variety of reasons such as variation in survey coverage, temporal mismatch and observer bias. It is possible that in the time between surveys, birds could have been disturbed or moved to other areas which would result in a discrepancy between the survey results. The location of birds from the shore-based surveys is also visually estimated, which makes it difficult to accurately compare the locations of birds between the two methods. Additionally, the comparisons do not take into consideration the 20 m gap between the cameras of the HiDef four-camera system (shown in Figure 5), between which birds may have been missed. A gapless rig may be more suitable for future comparison studies. 

Additionally, shore-based counts take considerably longer to conduct than DAS, due to logistical constraints with moving between VP locations, as well as the time taken to conduct the survey itself. In the 20 January 2023 survey, the shore-based counts at sectors 12 and 38 were conducted in suboptimal lighting conditions due to challenges in conducting all shore-based surveys on the same day in order to coincide with the DAS, meaning that surveys were conducted around sunset. Reduced visibility due to the low sun and fading light may have affected the observations of birds in these sectors.

Due to the low number of transects which cross the shore-based count sectors, it was also not possible to estimate the absolute abundance of birds within these areas for comparison with the shore-based counts, which had 100 % coverage. This meant that the study was limited to a qualitative comparison of the locations of birds observed between the two methods. However, this reflects the advantages and limitations of the two survey methods in general, whereby shore-based counts may be more suitable for small, inshore areas, whilst DAS may be more suitable for larger-scale and offshore surveys, thus making it difficult to compare results from the two methods.

Data analysis

In order to generate population estimates for the marine SPAs within this study, records of birds were trimmed directly to the SPA boundary at MLWS. As a result, any records of roosting or other birds that fell beyond this perimeter would not have contributed to the final population estimates. This would need consideration in future analysis if it was deemed appropriate that these birds are included in the SPA estimates. 

Due to a lack of readily accessible environmental covariates at an appropriate resolution, the model-based estimates generated through Bayesian point processing were run using only the spatial process. This inherently leads to more uncertainty in areas where no observations are recorded as evidenced by figures of the spatial distribution of the coefficient of variation. The reasoning for this is that although we are certain of observations, in areas where there are no observations, we are not completely certain that no birds would ever occur there (e.g. pseudo absence). Environmental covariates could help explain some of this variation, but would not be expected to affect the overall population estimate. This is because the sample population (i.e. the number of observations) would be the same no matter which environmental covariates are used, and those observations are extrapolated through the boundary area based on the relationships between the environmental variables where animals were detected. Spatial patterns and spatial coefficient of variation could be affected as there would be additional certainty associated with areas where no observations were recorded. An assessment of predicted versus observed densities demonstrated this further as higher R² values were associated with species that had more observations. This coefficient of determination (R²) is the proportion of variance in the response variable that is explained by the predictor variables, ranging between 0 and 1 (see Table 3 for definitions of statistical terms used).

Conclusion

Orkney supports multiple inshore wintering waterfowl species, occurring in internationally important numbers. In particular, divers, grebes, sea ducks and shag are relatively common compared to other parts of the UK. In 2022, Scapa Flow SPA and North Orkney SPA were classified under the Birds Directive (2009/147/EC2) to protect waterfowl species during the non-breeding period. 

The seas are also of great interest for current and future developments and successful implementation and management of these will require monitoring, especially considering the potential for impact on species protected by SPAs. Robust data is required in order to meet these objectives. To quantify wintering waterfowl and shag populations around Orkney a variety of survey methods have previously been adopted, collecting information using shore, boat and visual aerial-based methodologies. However, prior to the winter 2021/22 and 2022/23 DAS surveys reported here, no data have been collected solely within and outwith the final SPA boundaries. Additional shore-based VP counts of nearshore areas were also undertaken. 

The most abundant target species recorded in Scapa Flow and North Orkney SPAs were long-tailed duck and common eider during the winter 2022/23 surveys.  Population estimates derived from DAS data for some additional waterfowl species, including great northern diver, red-breasted merganser, black-throated diver and Slavonian grebe, also surpassed the 1 % threshold of GB or biogeographic populations for either or both Scapa Flow SPA and North Orkney SPA. However, this was not the case for species such as European shag and velvet scoter, where population estimates from DAS data were considerably lower than previous estimates for the two SPAs (produced using data collected via different survey methods). 

In North Orkney SPA, the DAS surveys indicated Slavonian grebe and red-breasted merganser were generally distributed to the south and west, compared to European shag, common eider and long-tailed duck which were distributed throughout the SPA. At Scapa Flow SPA, most species had a widespread distribution throughout the site although few birds were seen in the centre, where water is generally deeper and the seabed may be too deep for profitable foraging compared with elsewhere in the SPA. 

The Bayesian point process models were able to generate predicted distribution maps using only the spatial process, where sample numbers were sufficient (>10 birds). The uncertainty in the spatial distribution increased with fewer observations, however this did not greatly affect the population estimates that are calculated, as point process modelling can give estimates with robust confidence limits.

Population estimates for waterfowl species derived from DAS and shore-based survey data may be affected by suitability of some species to different survey techniques. For example, larger numbers of great northern diver, eider and long-tailed duck were detected during the DAS compared with the shore-based counts within the DAS swathe, suggesting that DAS may be a more suitable survey method for these species. This is likely due to the increased flight altitude and speed offered by DAS as compared with previous VAS methods through decreased disturbance, reduced flushing from areas and larger area coverage in shorter timescales. In addition, DAS data can be easily audited and stored for future use. Conversely, Slavonian grebe and European shag had higher detections during the shore-based counts, which may be attributed to their small size and/or dark plumage making detection more difficult during DAS. However, these results may be constrained by the temporal and spatial mismatch between the surveys, and additionally shore-based counts and DAS methods may also have different advantages and disadvantages depending on the size and location of the area to be surveyed.   

Naturally occurring variation in marine bird populations is difficult to capture with snapshot surveys, evidenced by the count differences in 2022 and 2023 DAS as well as monthly shore-based surveys of smaller areas. Further, sub-optimal weather conditions may hamper detection and identification of certain species, particularly those with darker plumage or those taking refuge above the high tide mark. 

However, while there are limitations to DAS for some species, relying on population estimates only derived from shore-based counts for these species would not be sufficient, since flocks of birds sitting on the water surface may be present beyond the visual detection range of shore-based surveyors. This can be overcome by using boat-based surveys, as utilised by Jackson (2018) in the central Scapa Flow, however this method is also known to increase disturbance. 

Ultimately, it is likely that different methods will be selected depending on the target species and the characteristics of the area (i.e. small, nearshore vs large, deep water) to be surveyed. Therefore, when monitoring multiple waterfowl species,  a combination of survey techniques is likely to be beneficial.