3.3. Assumptions and Limitations

3.3.1. Survey Timings

  1. Two historic surveys, namely TCE aerial surveys and Seagreen boat-based surveys, took place between 2009 and 2011 (section 3.2.2 provides further detail). These data are now more than ten years old, and it is possible that there may have been changes in the distribution and abundance of marine mammals in the wider Firth of Forth and Tay since these surveys were carried out. The most recent historic surveys for Berwick Bank (SSE Renewables, 2022) were conducted between March 2019 and April 2021 ( Table 3.3   Open ▸ ).
  2. The site-specific DAS for the Array marine mammal study area have been conducted monthly between March 2021 and February 2023. Data collected represent a snapshot over a single survey day in each calendar month. This is a standard survey method applied for marine mammal data collection to inform the baseline for offshore wind developments. The differences in sighting rates between months may be due to seasonal changes and it may be possible to investigate seasonality of sightings based on aerial survey data. The environmental conditions also have the potential to influence these results and data collected over a short time frame (one day) does not allow this to be explored further.

3.3.2. Marine Mammal Observers

  1. Boat-based surveys rely on marine mammal observers to record the number of marine mammals and accurately identify the individuals to species level. Ideally, a survey team, following a standard distance sampling approach, should consist of three people – one to monitor the track line, the second to monitor over distance and the third to write down sightings. The team is usually rotated to reduce the possibility of observer fatigue. The Seagreen boat-based survey adopted the use of only a single marine mammal observer which may have resulted in under-recording (Sparling, 2012).
  2. There is less potential for under-recording during aerial surveys, e.g. DAS, as all observations within the transect strip length are recorded and can be investigated during the analysis of data.

3.3.3. Weather Conditions

  1. Boat-based surveys are often carried out to collect bird and marine mammal data simultaneously. However, seabird surveys can be carried out in Douglas sea states of up to four, whilst marine mammals are surveyed only in Douglas sea states of up to three. If bird surveys were carried out simultaneously, as in the case of the Seagreen boat-based surveys (Sparling, 2012), there is a risk that encounter rates may be biased downwards if portions of the survey were carried out in sea states above three. Harbour porpoise in particular are difficult to record; detection probability of harbour porpoises decreases by 50% between Beaufort sea state 0 and Beaufort sea state 3 and continues to decrease substantially as Douglas sea state values increases (Palka 1996).
  2. Compared to boat-based surveys, sea state is less problematic for aerial surveys, e.g. DAS, as surveys can effectively be carried out in Douglas sea states of up to four for both marine mammals and birds (HiDef, pers. comm.).

3.3.4. Delayed Surveys

  1. Logistical issues and/or downtime due to unsuitable weather conditions prevented DAS from being flown in two months: May 2021 and February 2022 during the Array marine mammal study area surveys. Potential data gaps were avoided by flying additional surveys as close in time as possible to those missed. Subsequently, the May 2021 survey was replaced by a supplementary flight on 09 June 2021, and the survey for February 2022 was conducted on 06 March 2022.

3.3.5. Bias In Data

  1. Marine mammals spend most of their time underwater and therefore might be unavailable for detection at the surface and on the transect line. Undercounting as a result of this is known as ‘availability bias’ and is corrected for using an estimate of the probability that an animal is on the surface at any randomly chosen point in time. The resulting correction factor is used to estimate the total number of animals that may be present within the survey area. In the case of DAS, animals are available for detection if they are on the surface or just below the surface, as depth of detectability is dependent on water clarity. Data from historic surveys as well as DAS provide a count of the relative numbers of each species (or species group) within survey transects. However, sightings data from historic surveys as well as DAS do not allow for estimations of site-specific availability biases. Therefore, published correction factors, where considered to be appropriate, were applied to data to correct for availability bias to estimate absolute numbers. Correction factors applied to the site-specific DAS data in this Technical Report were consulted with stakeholders via the Array Marine Mammal Methodology Note (volume 3, appendix 5.1, annex B) and Array Marine Mammal Consultation Note 1 (volume 3, appendix 5.1, annex D) and are described in more detail in volume 3, appendix 10.2, annex A.
  2. If a group of animals on the transect line is at the surface, they may not be detected due to various factors such as observation conditions or observer fatigue, known as ‘perception bias’. Given that some marine mammal species are known to actively avoid vessels, either by moving away or by diving, unquantifiable bias may be introduced into the data collected during boat-based surveys (Palka and Hammond, 2001). Perception bias is less of a limiting factor for aerial surveys since high-definition video or still imagery captures all animals on the surface and their detection is less influenced by the ability of an observer to detect an animal. For both boat-based and aerial surveys it can be challenging to record wide ranging or cryptic species, especially when making the snapshot count.

3.4. Other Studies and Data Sources

3.4.1. SCANS Surveys

  1. The first SCANS survey was conducted in summer 1994 to provide estimates of abundance and density of small cetaceans in the North Sea and European Atlantic continental shelf waters. The SCANS II surveys were completed in July 2005 and SCANS III in July 2016. All surveys comprised of a combination of vessel and aerial surveys. Both aerial and boat-based survey methodologies were designed to correct for availability and detection bias and to allow the estimation of absolute abundance. The original SCANS III data was published in Hammond et al. (2017), which has been revised following the discovery of some analytical errors and the updated version Hammond et al. (2021) is used for the purpose of this baseline characterisation. SCANS IV was carried out in summer 2022 and the results of surveys are presented in Gilles et al. (2023)
  2. The Array is located in the SCANS II survey area V and SCANS III survey area R ( Figure 3.3   Open ▸ ), surveyed by boat and air, respectively. Due to the change in survey blocks used in the SCANS II and SCANS III surveys, direct comparison between the surveys for abundance and density estimation is not possible. Gilles et al. (2023) presented the North Sea divided in 13 blocks, NS-A to NS-M and the Array falls within the NS-D block. The boundaries and dimensions of SCANS III block R ( Figure 3.3   Open ▸ ) are closely aligned with SCANS IV block NS-D boundaries ( Figure 3.4   Open ▸ ) as the total surface areas differ by only 9 km2 (Hammond et al., 2021, Gilles et al., 2023).
  3. The most recent data from SCANS III and SCANS IV has been referred to in section 5 indicating trends in species abundance across survey years where relevant (Hammond et al., 2021, Gilles et al., 2023).

                        SCANS III density surfaces

  1. SCANS III data were used by Lacey et al. (2022) to provide information on summer distribution of recorded species by modelling the data in relation to spatially linked environmental features to generate density surface maps.
  2. Lacey et al. (2022) presents density surface modelling for harbour porpoise, bottlenose dolphin, short-beaked common dolphin Delphinus delphis and minke whale. Density surface modelling used environmental covariates (which were selected as having the potential to explain additional variability in cetacean density) including depth, slope, aspect, distance from the coast, topography, sea level anomaly (i.e. the difference between recorded sea level and mean sea level) and sea surface temperature. Consecutive records made along the aerial survey transects were combined into 10 km segments of search effort to allow density estimates to be predicted to a spatial grid of 10 km x 10 km resolution.
  3. Figures showing surfaces of predicted density and coefficient of variation (CV )of predicted density were produced for each species for SCANS-III, with patterns of predicted density influenced by model covariates, fitted smooth functions and spatial variation in the values of the covariates in the prediction grid (Lacey et al., 2022). To note, the density surfaces are for summer distributions only, as this is when SCANS-III was carried out. The figures allow density surfaces to be overlaid with the Array marine mammal study area for mean density outputs and are discussed in section 5 for relevant species.

Figure 3.3:
SCANS III Survey Blocks

Figure 3.3: SCANS III Survey Blocks

Figure 3.4:
SCANS IV Survey Blocks

Figure 3.4: SCANS IV Survey Blocks


3.4.2. Joint Cetacean Protocol (JCP) Phase III Analysis

  1. The JCP Phase III analysis combined datasets from 38 sources, collected from boat-based and aerial platforms between 1994 and 2010. The total survey effort of over 1.05 million km was undertaken to estimate spatial and temporal patterns of abundance for seven species of cetaceans (paragraph 32) (Paxton et al., 2016). Developer areas were chosen based on best available information at this time of the areas of interest for renewable energy developments and referred to as “areas of commercial interest” (Paxton et al., 2016).
  2. The following species were included in the analysis: harbour porpoise, minke whale, bottlenose dolphin, short-beaked common dolphin, Risso’s dolphin Grampus griseus, white-beaked dolphin and Atlantic white-sided dolphin Lagenorhynchus acutus. Density surface models were used to predict species density over a fine scale grid of 25 km2 resolution for one day in each season in each survey year. The data were divided into regions for which seasonal abundance in winter (January to March), spring (April to June), summer (July to September) and autumn (October to December) was estimated. The site boundary is situated just outside the “Firth of Forth area of commercial interest” ( Figure 3.5   Open ▸ ), however, this study is considered to provide context within the regional marine mammal study area.
  3. It should be noted that, as stated by Paxton et al. (2016), the abundance estimates produced by the JCP Phase III modelling will be less reliable than those obtained from a well-designed dedicated abundance survey given the assumptions made when standardising the data and the spatial and temporal patchiness of the data available.

Figure 3.5:
JCP Phase III Developers Areas

Figure 3.5: JCP Phase III Developers Areas


3.4.3. JNCC Report 544: Harbour Porpoise Density

  1. Heinänen and Skov (2015) conducted a detailed analysis of the majority of the standardised JCP data resources to identify “discrete and persistent areas of high density” that might be considered important for harbour porpoise, with the goal of determining SACs for the species. Outputs of the analysis included distribution maps of density estimates for the waters around the UK, and the results are summarised in paragraphs 89, 90 and 98. The analysis grouped data into three subsets: 1994 to 1999, 2000 to 2005 and 2006 to 2011 to account for patchy survey effort. To explore whether distribution patterns differed between seasons, the study analysed summer (April to September) and winter (October to March) data separately. The analysis presented in Heinänen and Skov (2015) relied on extensive extrapolation of survey data over space and time. Any such extrapolation is sensitive to the covariates used in models and makes the assumption that these relationships hold true outside of the surveyed areas. Given the uneven survey effort over the modelled period, there was a large degree of uncertainty in modelled distributions.

3.4.4. SCOS

  1. Natural Environment Research Council (NERC) provides scientific advice to government on matters related to the management of seal populations (under the Conservation of Seals Act 1970 and the Marine (Scotland) Act 2010). NERC has appointed SCOS to formulate this advice which is provided by SMRU through a series of scientific briefing papers and meetings and an annual report is produced. The annual report includes advice on matters related to the management of seal populations, including general information on British seals and information on their current status. Upfront sections of the report often address specific questions raised by regulators and stakeholders. The most recent publicly available SCOS report is SCOS (2023) which presents data collected and population estimates up to and including 2022.

3.4.5. SMRU Seal Surveys

  1. SMRU carries out surveys of harbour and grey seals in Scotland and on the east coast of England to contribute to the NERC’s statutory obligation under the Conservation of Seals Act 1970 through provision of scientific advice on matters related to the management of seal populations to the UK Government. SMRU surveys form the routine monitoring of seal populations around the UK. Most surveys are carried out in August from the air by either light aircraft or helicopter and record seals that are hauled out on shore. Although both species are surveyed during the month of August, on account of differences in the breeding behaviour of harbour and grey seals, these surveys correspond to different points in the two species’ annual cycles.
  2. A SMRU report was commissioned to support the baseline assessment for the Array and associated Array marine mammal study area. The report provided a detailed account of grey and harbour seal haul outs and telemetry tracks within the vicinity of the Array as well as East Scotland and Northeast England seal MUs (Stevens, 2023) ( Figure 3.6   Open ▸ ).

                        Harbour seal

  1. Surveys of harbour seals are carried out during the summer and early autumn months. There are two types of surveys conducted: breeding season counts and August moult counts. Given that there are no harbour seal breeding surveys conducted in the East Scotland or Northeast England seal MUs, these are not considered further in this report (Stevens, 2023). The main population surveys are carried out when harbour seals are moulting, during the first three weeks of August. The frequency of surveys differs by area (Stevens, 2023). In general, moult surveys are conducted annually in Lincolnshire and Norfolk (England), and in the Moray Firth and the Firth of Tay (Scotland). The remainder of the Scottish coast is surveyed approximately every four to five years, although there is considerable variation between areas. The most recent data available for the East Scotland and Northeast England seal MUs are from 2021 (Stevens, 2023).

                        Grey seal

  1. In the UK, grey seals are surveyed during their breeding season (August to December), wherein pup counts are conducted at known breeding colonies. Most breeding colonies are surveyed by SMRU by fixed wing aerial vertical photography (Hebrides, Orkney, north Scotland the north-east Scotland and most of the Firth of Forth) while other colonies are surveyed by ground count by other organisations (including NatureScot, Natural England, Natural Resources Wales, National Trust, and Lincolnshire Wildlife Trust) (Stevens 2023). The grey seal pup production database contains data from 1996 to 2021 and includes 74 breeding colonies, 70 of which are in Scotland and one of which is in north-east England (though not all colonies have been surveyed consistently since 1989 and some smaller colonies are surveyed more sporadically than others). The most recent complete grey seal pup production survey (covering Orkney, Inner and Outer Hebrides and the North Sea colonies) was conducted in 2019. It should be noted that grey seal distribution during the breeding season is very different to their distribution at other times of the year.
  2. Grey seals are also counted during SMRU’s harbour seal August moult surveys, however, counts of grey seals during the summer months can be highly variable and, although these counts are not used as a population index, they provide useful information on the summer and non-breeding season distribution of grey seals. The most recent data available for the East Scotland and Northeast England seal MUs are from 2021.

3.4.6. Designated Seal Haul Out Sites

  1. Seal haul out sites are locations on land where seals come ashore to rest, moult or breed. In Scotland, seal haul out sites are designated under section 117 of the Marine (Scotland) Act 2010. The Protection of Seals (Designation of Haul-out Sites) (Scotland) Order 2014 laid in the Scottish Parliament on 26 June 2014 which, from 30 September 2014, makes it an offence to harass seals at these sites. Harassment involves any activity that “pesters, torments, troubles or attacks a seal on a designated haul-out site. In particular, it would include any action that causes a significant proportion of seals on a haul out site to leave that site either more than once or repeatedly or, in the worst cases, to abandon it permanently” (Marine Scotland, 2014).
  2. The closest designated haul out site to the Array marine mammal study area, Kinghorn Rocks, is located approximately 157 km to the south-west ( Figure 3.6   Open ▸ ).

Figure 3.6:
Seal MUs and Designated Sites

Figure 3.6: Seal MUs and Designated Sites


3.4.7. Seal Telemetry Data

  1. SMRU has deployed telemetry tags on grey seals and harbour seals in the UK since 1988 and 2001, respectively. Tags are glued to the fur on the back of the seal’s neck and fall off with the fur during the annual moult, if not before. These tags transmit data on seal locations with the tag duration (number of days) varying between individual deployments. Data obtained during telemetry studies provide information on seal movement patterns away from their haul out sites, as well as data on the foraging behaviour of seals at sea, and demonstrate connectivity between areas.
  2. Telemetry data presented in this report for harbour and grey seal (sections 5.3.1 and 5.3.2, respectively) draws on the SMRU commissioned study (Stevens, 2023), which presents an analysis of existing satellite data to describe the movements of harbour and grey seal within or in the vicinity of the Array marine mammal study area.

3.4.8. Seal Usage Maps

  1. Carter et al. (2022) presents the most up-to-date seal usage maps for UK waters. The study utilised a high-resolution GPS tracking dataset (114 grey and 239 harbour seals) and wide spatial coverage to model habitat preference and generate at-sea distribution estimates for the entire UK and Ireland populations of both species of seals. Additionally, the study provides SAC-specific estimates of at-sea distribution, demonstrating that hotspots of at-sea density cannot always be apportioned to the nearest SAC. The at-sea usage maps represent the number of grey and harbour seals estimated to be in the water in each 5 km x 5 km grid cell at any one time. Values in the Carter et al. (2022) report are presented as spatial predictions of relative density. Absolute densities were provided by the author (Carter, pers. comm.).

3.4.9. Distribution Maps Of Cetacean and Seabird Populations In The North-East Atlantic

  1. Waggitt et al. (2020) collated and standardised data from 2.68 million km of cetacean and seabird surveys carried out in the north-east Atlantic between 1980 and 2018. The study consisted of three stages – collating survey data, linking differences among surveys with various parameters (platform type, transect design, observation method and weather) to calculate the variations in the surface area covered, and generating species distribution models. As a result, distribution maps were provided for 12 cetacean and 12 bird species at 10 km resolution and monthly frequency in the north-east Atlantic.