Production of Seabird and Marine Mammal Distribution Models for the East of Scotland

7. Discussion

7.1 General Comments

The results of the aerial surveys highlight some important regions across seabird and cetacean taxa, particularly in summer. These include the area of the North Sea east of Orkney southwards into the Moray Firth extending along the south side of the firth eastwards beyond Fraserburgh, and the Firth of Forth eastwards into the North Sea. North of Orkney in the northernmost North Sea, the coverage is less good but that may also be important. These important regions differed between groups with kittiwakes, guillemots and razorbill having highest abundance in the southeast and fulmar and skuas in the northern part of the study region.

For most species, the estimates of density and abundance derived from analysis of the APEM digital aerial surveys were comparable to those generated from collations of previous survey data which is encouraging. Some of the differences amongst cetaceans may be the result of different corrections applied to account for availability bias. Cetacean species estimated incorporated the availability of the animals at the surface. However uncertainty in surface availability has not been incorporated into the analysis. Some of the availability estimates used in this report are from old studies. More up-to-date estimates based on, for example, telemetry data or our knowledge of relationship between water visibility and species behaviour could be used in the future. A further discussion is also needed on the comparison of accounting for availability based on various methods such as visual and digital survey.

The apportioning of undifferentiated species groups to particular species may lead to error particularly if the proportions of those species has changed over time. Uncertainty associated with this process has not been incorporated into the analysis.

With only eight surveys undertaken across 14 months, it is possible that density distributions may change a lot in the interval between surveys such that the ones used here are unrepresentative. Less common species may also be missed altogether. The following species are known to regularly use the region but were scarcely recorded: amongst cetaceans, these included fin whale, humpback whale, bottlenose dolphin, Atlantic white-sided dolphin, Risso’s dolphin. long-finned pilot whale, and killer whale; and amongst seabirds, these included great northern diver, red-throated diver, great cormorant, European shag, European storm petrel, Manx shearwater, velvet scoter, long-tailed duck, common eider, arctic skua, arctic tern, sandwich tern, black guillemot, and little auk. Some of these are predominantly coastal species but others are pelagic. The surveys were targeting offshore areas beyond 12 nm, and so estimates of abundance for the overall region could be significant underestimates for some seabird species, particularly during the breeding season.

Finally, it is worth noting that two significant seabird mortality events have occurred after the survey work was completed. The first of these was a large seabird wreck in late summer of 2021, and the second an outbreak of highly pathogenic avian influenza during the seabird breeding season of 2022. These could lead to significant changes in the at-sea densities of some species.

There are a number of issues to take into account when interpreting the results of the analyses presented here. Inevitably, just eight aerial surveys across a period of 13 months can provide little more than a snapshot of occurrence. Marine mammals and birds are highly mobile, and their distribution and abundance may vary markedly between days so the results on a particular survey day may not be representative for that month. There were also some important spatial and temporal gaps in coverage. The months of May, August and December were not surveyed at all, and coverage was generally more evenly spread in the southern half of the study area than further north around the Northern Isles. Breeding seabirds are central place foragers during the period they are nesting. Unfortunately, there were some survey gaps east of Shetland during the main seabird breeding period (no surveys there between 17 April and 25 June) so birds from the large colonies in that region were not well sampled.

As noted earlier, digital aerial surveys can have their own limitations, mainly related to species identification. The APEM surveys were undertaken at a flight height of 2,000 ft in order to sample a wider area, at a cost of lower image resolution. Besides that issue, species identification may vary between analysts. This was particularly obvious during the first two surveys when no porpoises were specifically identified, and images were classified as porpoise/dolphin, despite the environmental conditions being no different to those on subsequent surveys. And, as noted in the Appendix, reviewing a selection of those images, we found that we could identify to species several of them. Although the relatively high flight height probably exacerbated the situation, aerial surveys inevitably lead to greater problems in differentiating species of similar appearance particularly when the distinguishing features cannot readily be seen from above. Consequently, this remained an issue throughout the survey period with much lower proportions of encounters assigned to a species than would be the case with visual vessel-based surveys. For this reason, we utilised MERP data sets to apply species ratios within species groups, and, for less common species, to improve accuracy of the abundance estimates. Note that some species in the region have predominantly coastal distributions, and so these offshore surveys will not adequately sample them. This applies to bottlenose dolphin amongst cetaceans, and diver, grebe and sea duck species, European shag and great cormorant, arctic skua, black-headed gull, common, arctic and sandwich terns, and black guillemot, amongst birds. Also we did not adjust for any changing ratios of species with time in the MERP data.

Modelling the data here proved challenging especially given the sparsity of some species, nevertheless plausible estimates were obtained. Regions of uncertainty are greatest typically in regions away from the tracklines and on the periphery of the survey region (see cv plots) where there was little effort and perhaps a strange combination of covariates not seen in the actual data.

Below, we first discuss the environmental covariates included in the modelling followed by discussion for each of the main species in the context of our current knowledge of their distribution and abundance. Comparisons are made with the results from previous at-sea surveys and in the case of seabirds with colony counts during the summer breeding season since that relates closely to local densities and abundance at sea. New analyses from updated collations of the MERP dataset were applied, as utilised in a European Commission Bycatch Risk Mapping Project (Evans et al. 2021) and for the ongoing Phase 2 of the ORJIP Project, led by the Centre of Ecology and Hydrology (CEH) for Marine Scotland.

7.1.1 Environmental Covariates

The modelling approaches used to predict the distribution and densities of animals from observational surveys are broadly divisible into two categories. In a species distribution model (SDM) approach, predictions are based entirely on establishing ecological associations between animal sightings and relevant environmental conditions with the study area, before using these associations to predict the probability of encounters or densities beyond the surveyed area (Elith and Leathwick 2009). For example, measurements of marine animal presence or densities from surveys are often combined with corresponding measurements of sea temperature, establishing which water-masses (e.g., cool, versus warm) are associated with an increased likelihood of encountering animals or higher numbers of animals (Waggitt et al 2020). In the density surface modelling (DSM) approach used here (Miller et al 2013), predictions are more dependent on identifying spatial and geographical descriptors of where animals were sighted, before using these descriptors to predict animal densities beyond the surveyed area. In most applications of DSM, coordinates are used to describe broad-scale aggregations of animals, and environmental conditions are used to describe fine-scale patterns within these aggregations (Gilles et al 2016). Therefore, whilst correlations between animal densities and environmental conditions are presented here, these associations do not necessarily represent habitat preferences of animals, and instead represent geographical descriptions of areas supporting animals. If SDM is going to be performed on this data, a range of additional environmental covariates can be considered such as SST in the previous years or difference from mean SST/salinity across whole area for given time period.

The choice between SDM and DSM depends upon data properties and study objectives. SDM have sometimes been favoured when analysing collations of surveys characterised by heterogeneous coverage, where the distribution of sightings is strongly influenced by effort, and a reliance on spatial and geographical descriptors could cause biased predictions (Waggitt et al 2020, Becker et al 2022). DSM are commonplace when analysing systematic surveys with relatively homogeneous coverage, where variation in effort is less problematic, and a combination of coordinates and geographical descriptors effectively predict areas of aggregations (Hammond et al 2013). Because they attempt to identify habitat preferences of animals, SDM could predict biogeographic ranges within a region well; because they provide a multi-dimensional description of sighting locations, DSM could better predict dense aggregations within this region (Waggitt et al, unpublished analyses). Here, the presence of a systematic survey and emphasis on identifying aggregations meant that DSM approaches were considered more suitable than SDM.

The breeding colony indices (detailed in Waggitt et al 2020) were not retained in seabird models. This is unexpected, as the distribution of seabirds is centred around their breeding colonies in summer months, when animals repeatedly commute between terrestrial nests and marine foraging grounds. Surveys suggest that densities of animals likely decline exponentially with distance from the breeding colony (Camphuysen 2011), with the highest densities of seabirds occurring <1km from the breeding colony (Gaston et al. 2004). Beyond their immediate vicinity, the location of breeding colonies could primarily influence animal presence rather than animal densities, and this influence likely occurs at a basin (100-1000 km) rather than a regional scale (10-100 km) (Stone et al 1994). Therefore, the absence of surveys <12 nm from coastlines, surveys occurring within foraging ranges of animals breeding within region (see Thaxter et al. 2010), and model focus on animal densities rather than presence may explain the omission of breeding colonies. Alternatively, coordinates could have explained aggregations around breeding colonies better than the breeding colony indices. However, the lack of clear aggregations around breeding colonies in density surfaces suggests that coordinates were not detecting this phenomenon.

7.2 Seabird species

7.2.1 Northern Fulmar

7.2.1.1 Distribution

The northern fulmar has a predominantly high latitude distribution with largest numbers in the Northeast Atlantic breeding in Iceland, the Faroes, Norway (including Svalbard) and northern Scotland (Mitchell et al., 2004). This is reflected in its at-sea distribution in the North Sea where greatest densities and abundance occur off north-east Scotland (Stone et al. 1995, Kober et al. 2010, Waggitt et al. 2020; Searle et al., in prep.). The APEM surveys also showed highest densities offshore in the northernmost part of the North Sea east of the Northern Isles south to the Moray Firth.

There is some seasonal variation in distribution patterns, with lower fulmar densities within the North Sea in winter (between December and March) compared with the rest of the year (Stone et al. 1995, Furness 2015, Waggitt et al. 2020, ORJIP Project unpublished data). In the north-western North Sea, the APEM surveys also showed seasonal variation with peak numbers in September declining through the winter to a low in March just before the start of the breeding season.

Highest densities of fulmars tend to occur near the edge of the continental shelf and around fishing vessels where birds may gather to feed on offal to supplement a diet of zooplankton and small fish (Mitchell et al. 2004). Changes in fishing practice and climate are thought to have influenced the marked changes in both distribution and abundance observed in fulmars over the last hundred years, resulting first in a marked increase and geographical spread, and then, since the mid-1990s, a protracted decline (Mitchell et al. 2004, Daunt & Mitchell 2013, JNCC 2021).

7.2.1.2 Abundance

Point estimates from the APEM surveys indicated 100,000-150,000 fulmars in the study area between January and July increasing to around 400,000 in September before declining again. Highest average densities reached 20 birds/km2 east of Shetland, Orkney and North-east Scotland, but most areas averaged c. 4 birds/km2 or less.

There are no direct comparisons of abundance. The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following counts of AOS (apparently occupied sites): Shetland (188,544), Orkney (90,846), Caithness (29,957), east coast from Ross & Cromarty south to Aberdeen (8,595), and from Aberdeen south to East Lothian (6,889). Although one cannot derive abundance estimates for the offshore region from these, they highlight the importance of the Northern Isles as a source of fulmars and reflect the lower numbers further south.

JNCC at-sea surveys yielded maximum mean densities of c. 5 birds/km2, mainly in the vicinity of Shetland (Stone et al. 1995). Analysis of the MERP wider data collation gave maximum mean densities of c. 5.5 birds/km² occurring in the same region (ORJIP Project Phase 2, unpublished data, using methods from Waggitt et al. 2020).

7.2.2 Northern Gannet

7.2.2.1 Distribution

The northern gannet experienced steady population growth at an overall rate of c. 2% per annum between the late 1960s and end of the twentieth century in the eastern North Atlantic, with several new colonies founded in Britain (Murray & Wanless 1997, Mitchell et al. 2004). Since then, colonies in the Northern Isles and East Scotland have continued to increase, most at annual rates of 3-6% (Murray et al. 2015, JNCC 2021). Greatest numbers breed in Shetland (Hermaness, Noss, Foula and Fair Isle) and the Firth of Forth (Bass Rock) but the density surfaces modelled from the APEM surveys suggest little variation in at-sea density from north to south.

The species forages over wide areas taking a variety of fish including mackerel, sandeel, sprat and herring, and often associating with fishing activities (Hamer et al. 2000, Bodey et al. 2014).

7.2.2.2 Abundance

Point estimates from the APEM surveys indicated a peak of c. 135,000 gannets in the study area across the breeding season (June to October) declining to c. 40,000 gannets for the rest of the year. Average densities ranged from 0 to 2 birds/km2, with densities generally increasing nearer the coast,

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following counts of AOS (apparently occupied sites): Shetland (26,249), Orkney (5,137), Banff & Buchan (1,085), and East Lothian (44,110). Not all colonies were counted during the Seabird 2000 survey period since there have been separate periodical national censuses. For those colonies not counted, numbers were extrapolated from the 1994-1995 census. By 2019, the number of AOS had increased further in all four regions: Shetland (44,782), Orkney (10,742), Banff & Buchan (4,825), and East Lothian (75,259) (Murray et al. 2015, JNCC 2021). It is worth noting that, in 2022, an outbreak of Highly Pathogenic Avian Influenza (HPAI) has killed large numbers of gannets from North Sea colonies.

JNCC at-sea surveys yielded maximum mean densities of c. 5 birds/km2 in small gridded areas east of Shetland and in the Firth of Forth between May and October, although most areas had densities of 1.0 bird/km2 or less (Stone et al. 1995). Analysis of the MERP wider data collation gave maximum mean densities of c. 1.5 birds/km2 during the same season but with most densities between 0.5 and 1.0 bird/km2 (Waggitt et al. 2020). Gannet densities in summer were greatest around Shetland and Orkney, and lower throughout the region between November and March (ORJIP Project Phase 2, unpublished data, using methods from Waggitt et al. 2020).

7.2.3 Great Skua

7.2.3.1 Distribution

The great skua is one of Britain’s rarest breeding seabirds, with 94% breeding in the Northern Isles, and Scotland accounting for c. 60% of the world population (Mitchell et al. 2004, JNCC 2021). Great skua numbers increased between 1969-70 and 1998-2002, by which time the UK population was estimated to number 9,634 AOT (apparently occupied territories). Since 2000, however, numbers have been decreasing in some areas. In Shetland, for example, three colonies (Hermaness, Noss, and Fair Isle) held 1,476 AOT in 2019, a decrease of 49% since 2007 (JNCC 2021). During 2021 and 2022, there have been outbreaks of Highly Pathogenic Avian Influenza (HPAI) which have caused further significant declines at several great skua breeding colonies, although the full impact is not yet known.

At-sea surveys during the 1980s and 1990s showed highest densities between April and October around the Northern Isles extending southwards into the Moray Firth and off the east coast of Scotland but at lower densities (Stone et al. 1995). Almost no great skuas were present in the study area between November and March (Stone et al. 1995). The more recent MERP collation of survey data generating modelled density distributions highlighted the comparative absence of the species in the region between December and March, with birds primarily recorded in the SW Channel Approaches and south-west of Ireland (Waggitt et al. 2020). However, densities start increasing around the Northern Isles in March, and between June and October are widely spread across the northern North Sea extending southwards as the summer progresses (Waggitt et al. 2020, ORJIP Project Phase 2, unpublished data). Throughout the main breeding season of May to September, highest densities are close to Shetland and Orkney.

The APEM offshore surveys also had highest densities close to Shetland and Orkney between April and October. During this period, there were small numbers of great skua encounters east of the Moray Firth and north-east Grampian coast but none recorded further south. No birds were recorded in the study area during surveys between November and March.

7.2.3.2 Abundance

Point estimates from the APEM surveys increased between April and October, reaching a peak in June of c. 1,800 great skuas in the study area. Care should be taken in interpreting these results because not all the model assumptions were met. Between November and March, numbers were very low, in the order of a few hundred individuals throughout the study area, reflecting the migratory nature of the species with most birds believed move south.

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following counts of AOT (apparently occupied territories): Shetland (6,846), Orkney (2,209), and Caithness (5), the rest of the UK population breeding in Sutherland, the west coast of Scotland and Hebrides.

Within the study area, JNCC at-sea surveys from the 1980s and 1990s recorded densities between April and June of 1 bird/km2 only in grid cells around Shetland and Orkney, but for the most part, densities were between 0.01 and 0.2 bird/km2 extending south into the Moray Firth (Stone et al. 1995). Between July and October, nearer shore densities in those regions largely ranged between 0.2 and 0.5 bird/km², after which birds were generally absent until the following April. Analysis of the MERP data collation yielded maximum predicted densities of c. 1 bird/km² close to the coast of Shetland between April and August, whilst further offshore densities averaged c. 0.05 to 0.5 bird/km² ((ORJIP Project Phase 2, unpublished data, using methods from Waggitt et al. 2020). Over the wider area to the east and south, densities averaged less than 0.05 bird/km2.

Average offshore densities from the APEM surveys between April and October were greatest east of Shetland mostly ranging from 0.05 to 0.5 bird/km2, with densities generally increasing nearer the coast and being highest in the north. From Orkney southwards and offshore, densities were at all times of the year less than 0.05 bird/km2. These results are very similar to the previous wider-scale survey analyses published.

7.2.4 Common Gull

7.2.4.1 Distribution

The common gull is a coastal and inland breeder largely confined in Britain to Scotland. Greatest numbers of coastal birds breed in Orkney and Shetland, with progressively smaller numbers southwards down the east coast of mainland Scotland (Mitchell et al. 2004). National censuses indicated an increase in the coastal-nesting population from around 13,000 pairs in 1969-70 to c. 15,500 pairs in 1985-88, but with little change in distribution (Cramp et al. 1974, Lloyd et al. 1991). Between 1985-88 and 1998-2002, the population increased by 37% in Orkney but remained fairly constant in Shetland (previously having almost doubled in size since 1969-70) (Mitchell et al. 2004). Over the last 20 years, the coastal common gull population appears to have declined; in 2019, 260 coastal sites held 821 AON, 75% fewer than during the Seabird 2000 census (JNCC 2021).

The APEM offshore surveys indicate a largely nearshore distribution from Orkney to the Firth of Forth between September and March, extending further offshore in a band east of Orkney to the Moray Firth. During April to July, birds are largely absent offshore. Because of its generally coastal and inland distribution, it has not been the focus of attention during analyses of previous at-sea survey collations (Stone et al. 1995, Waggitt et al. 2020, Searle et al., in prep.).

7.2.4.2 Abundance

Point estimates from the APEM surveys show very low numbers offshore during the breeding season (April to July) until September when numbers start to increase, peaking in October at c. 3,500 birds, before declining again through the winter. Density estimates were less than 0.01 bird/km2 throughout the study area between April and July, when birds were breeding. Between September and February, birds were recorded offshore at densities ranging from 0.01 to 1 bird/km2 east of Shetland, Orkney, the outer Moray Firth, and off the east Grampian coast south to the Firth of Forth, with densities highest in the south and nearest the coast.

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following counts of AON (apparently occupied nests) at coastal sites in the study area: Shetland (2,424), Orkney (11,141), Caithness (468), east coast Sutherland (124), east coast Ross & Cromarty (297), Inverness (135), Aberdeen (280), Kincardine & Deeside (22), Angus (19). They highlight the numerical importance of the Northern Isles for this species.

The numbers recorded at sea around the Northern Isles from the APEM surveys are clearly much lower than those breeding along adjacent coasts. This is probably largely due to a substantial part of the population foraging inland. Following the end of the breeding season, there appears to be a greater tendency for some birds to disperse offshore eastwards and southwards in the North Sea, persisting through the winter.

7.2.5 Lesser Black-backed Gull

7.2.5.1 Distribution

The lesser black-backed gull is a widespread coastal breeder around the British Isles, with at least some of the population migrating south in winter. In the study area, breeding numbers are greatest in Orkney and Shetland, with only small numbers on the east coast of Scotland (Mitchell et al. 2004). The species increased throughout much of its range during most of the twentieth century (Mitchell et al. 2004). Since the 1990s, however, numbers have declined although with some fluctuations (JNCC, 2021). The at-sea distribution of lesser black-backed gulls in the North Sea shows greatest densities and abundance along continental coasts of Germany and the Netherlands, with very low numbers offshore in the APEM survey area (Stone et al. 1995, Kober et al. 2010, Waggitt et al. 2020, ORJIP Project Phase 2, unpublished data).

7.2.5.2 Abundance

The best estimate that could be derived from the APEM offshore surveys was 640 lesser black-backed gulls across the study area, with highest numbers during the breeding season. Average densities were all less than 0.01 birds/km2, the maximum being 0.008 birds/km2 during the survey in June.

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following counts of AOS (apparently occupied sites): Shetland (341), Orkney (1,045), Caithness and east coast Ross & Cromarty (7), Inverness to Aberdeen (176), Aberdeen to Dundee (80), Dundee to Edinburgh (6,183), and East Lothian (1,470).

JNCC at-sea surveys indicated all densities in the study area were less than 1 bird/km2, but the scales used did not discriminate how much lower they were (Stone et al. 1995). Analysis of the MERP data collation yielded maximum predicted densities of 0.01 birds/km2 close to the Grampian coast between April and July (ORJIP Project Phase 2, unpublished data using methods from Waggitt et al. 2020).

7.2.6 Herring Gull

7.2.6.1 Distribution

Herring gulls are widespread around the coasts of Britain, with some of the largest concentrations breeding in northern and eastern Scotland (Mitchell et al. 2004). In many areas, herring gulls also nest on rooftops. This is particularly the case, for example in the city of Aberdeen where 3,350 apparently occupied nests were counted in 1999-2002 (Mitchell et al. 2004). Since 2000, all trends in numbers from the seabird monitoring programme have shown declines although with some regional variation (JNCC, 2021). This declining trend has been recorded since the first national seabird census in 1969-70, although some of the decline in numbers may have been compensated for by more inland breeding (Mitchell et al. 2004). On a wider scale, the at-sea distribution of herring gulls in the North Sea shows greatest densities and abundance in the south-eastern sector off the coasts of the Netherlands, Germany and Denmark, particularly in winter, with lowest numbers in the northernmost North Sea (Stone et al. 1995, Kober et al. 2010, Waggitt et al. 2020, Searle et al., in prep.).

The APEM surveys indicated highest numbers in the north-western North Sea occurring between Orkney and the Moray Firth from October to February, peaking in November.

7.2.6.2 Abundance

The APEM offshore surveys yielded an overall abundance estimate in the study area of c. 10,000 herring gulls between April and September rising to a peak of c. 55,000 in November before declining through the winter. Average offshore densities were generally between 0 and 1 bird/km2 but in the Moray Firth, during October and November, increased to between 1 and 5 birds/km2.

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following coastal counts of AON (apparently occupied nests): Shetland (4,027), Orkney (1,933), east coast Caithness (3,503), east coast Ross & Cromarty (1,345), Inverness to Aberdeen (12,063), Aberdeen to Dundee (5,582), Dundee to Edinburgh (7,579), and East Lothian (3,553). These figures do not include rooftop breeding birds for which a total count had not been made.

JNCC at-sea surveys indicated maximum densities in the study area occurring in the Moray Firth amounting to at least 5 birds/km2, although most grids had densities between 0.01 and 1 bird/km2 (Stone et al. 1995). There was little variation throughout the year. Analysis of the MERP data collation yielded maximum predicted overall densities of c. 1-2 birds/km2 in the Moray Firth and Firth of Forth, also with little variation through the year (ORJIP Project Phase 2, unpublished data using methods from Waggitt et al. 2020).

7.2.7 Great Black-backed Gull

7.2.7.1 Distribution

Most great black-backed gulls in Britain breed on the west coast and in the Northern Isles, with only a few small colonies on the east coast south of the Moray Firth (Mitchell et al. 2004). Although both the UK and Scottish population appeared to increase during the 1990s, since then there has been a prolonged decline (Mitchell et al. 2004, JNCC 2021). Greatest numbers breed in Shetland and Orkney, and this is where localised coastal at-sea densities are highest between April and July (Stone et al. 1995, Waggitt et al. 2020, Searle et al., in prep.).

The APEM offshore surveys indicate low numbers throughout the study area for most of the year although a small peak between October and January may reflect movement into the area, including birds from Fennoscandia (Stone et al. 1995, Furness 2015). Areas with higher densities were east of Orkney and around the Moray Firth, between October and February.

7.2.7.2 Abundance

Point estimates from the APEM offshore surveys indicated a peak of c. 25,000 great black-backed gulls in the study area in October-November declining to between 200 and 1,300 birds from March to July. However, note that high uncertainty in some peripheral regions of the survey region has led to high upper bounds in the confidence limits. Average densities ranged from 0.002 birds/km2 (June) to 0.256 birds/km2 (November).

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following counts of AOS (apparently occupied sites): Shetland (2,875), Orkney (5,505), Caithness (211), east coast Ross & Cromarty (220), Inverness to Aberdeen (66), Aberdeen to East Lothian (70).

JNCC at-sea densities indicated maximum densities of c. 5 birds/km2 east of Shetland and Orkney in March and April, and in the Moray Firth between August and February (Stone et al. 1995). Analysis of the MERP data collation yielded maximum predicted densities of c. 3.5 birds/km2 close to the coast of Orkney and Shetland between April and July (ORJIP Project Phase 2 unpublished data using methods from Waggitt et al. 2020). In both analyses, however, for most of the north-western North Sea, estimated great black-backed gull densities offshore were less than 1 bird/km2 (Stone et al. 1995, Waggitt et al. 2020).

7.2.8 Black-legged Kittiwake

7.2.8.1 Distribution

The black-legged kittiwake is believed to be the most abundant species of gull in the world (Mitchell et al. 2004). In Britain, the largest and most numerous colonies are in northwest Scotland, the Northern Isles and east Scotland south to north-east England (Mitchell et al. 2004). Kittiwake numbers in the UK increased by approximately 24% between the late-1960s and mid-1980s but since the early 1990s, have declined rapidly; by 2013, they had decreased to 70% below the 1986 baseline, although numbers may have now stabilised and be slowly increasing in some areas (JNCC 2021).

Kittiwakes in the North Sea are widely distributed, with highest offshore densities between October and March, whereas inshore, numbers are greatest around the major colonies in the north-west between March and September (Stone et al. 1995, Kober et al. 2010, Waggitt et al. 2020).

With improved GPS tracking technology, sample sizes of birds tracked from individual breeding colonies have increased, providing a wealth of individual bird data of known provenance. These have also been used to make predictions of at-sea distributions for some colonial seabird species (Wakefield et al. 2017). From a sample of 464 kittiwakes tracked from 20 colonies, Poisson point habitat use models were developed to predict the species distribution in the waters around Britain and Ireland. In East Scotland, these indicated greater offshore use by foraging kittiwakes with potential high-use areas north of Orkney, east of Caithness, Ross & Cromarty, offshore east of the Grampian Region, in coastal areas of the Scottish Borders, and offshore east of the Firth of Forth (Wakefield et al. 2017). Note that these are derived from breeding birds during a short period of the year; they do not include non-breeders, birds entering the region from populations outside the British Isles, or other periods in the annual cycle of the species.

The APEM surveys showed a slight increase between April and July but little change outside the breeding period. Greatest densities occurred towards the coast around the Moray Firth and Firth of Forth.

7.2.8.2 Abundance

The APEM offshore surveys yielded an overall abundance estimate in the study area of c. 115,000 kittiwakes over much of the year, rising to a peak of c. 150,000 during the main breeding season between March and July. Average offshore densities were greatest between Orkney and the Firth of Forth mostly ranging from 1 to 5 birds/km2, but with densities generally increasing nearer the coast, particularly between April and July.

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following coastal counts of AON (apparently occupied nests): Shetland (16,732), Orkney (57,668), Caithness (49,533), east coast Ross & Cromarty (749), Moray (488), Banff & Buchan (30,599), Gordon (3,560), Aberdeen (1,695), Kincardine & Deeside (34,501), Angus (2,926), North East Fife (6,639), Kirkcaldy (3,249), Dunfermline (146), and East Lothian (3,349). These highlight the importance numerically of the Northern Isles and Caithness, Banff & Buchan, and Kincardine & Deeside. JNCC at-sea surveys indicated maximum densities of at least 5 birds/km2 throughout the coastal region of the study area, whereas most offshore grids had densities between 0.01 and 1 bird/km2 (Stone et al. 1995). There was little variation throughout the year. Analysis of the MERP data collation yielded maximum predicted densities of c. 3-5 birds/km2 in the Moray Firth and Firth of Forth areas, throughout the year, with densities offshore rarely below 1 bird/km2 (Seale et al. in prep, using methods from Waggitt et al. 2020).

7.2.9 Common Guillemot

7.2.9.1 Distribution

The common guillemot is the most abundant of the auks in Britain, with the largest colonies in the north and west (Mitchell et al. 2004). National census results show that the UK guillemot population increased by over 130% between the Operation Seafarer (1969-70) and Seabird 2000 (1998–2002) censuses (Mitchell et al. 2004). National census data showed an increase of 82% in the Scottish guillemot breeding population between 1969-70 and 1985-88, with a further increase of 24% up to the time of Seabird 2000. The population trend for guillemot in Scotland was stable during the early 1990s, after which it increased slightly over a few years before levelling off (JNCC 2021). After Seabird 2000 (1998–2002), the index declined and was lower than the 1986 baseline from 2004 to 2016. Since then, it has risen and, in 2019, was 18% above the baseline (JNCC 2021). However, over the last 20 years there has been much regional variation, with study plots in mainland Shetland showing strong declines whereas Fowlsheugh in Kincardine, for example, has increased (JNCC 2021). These varying temporal trends are important to bear in mind when comparing the APEM survey results with previous analyses of at-sea survey data from the region.

Guillemots in the North Sea are widely distributed. Analysis of the ESAS survey data in the 1990s showed high densities particularly nearshore from March through to October, declining between November and February (Stone et al. 1995). A more recent analysis generating modelled density predictions indicated highest densities offshore post-breeding, between July and November, with nearshore densities highest over a wider period between May and September (Waggitt et al. 2020). Areas with greatest densities were between Shetland and Caithness and around the East Grampian coast southward to the Firth of Forth (Waggitt et al. 2020).

Predictions of at-sea distributions in the waters around Britain and Ireland from GPS tracking of 178 breeding guillemots from 12 colonies along with habitat modelling (Wakefield et al. 2017) indicated In East Scotland potential high-use areas around Orkney, east of Caithness, along the south side of the Moray Firth, and around the East Grampian coast southward to the Firth of Forth.

As with the MERP and tracking data, the APEM surveys showed greatest densities in coastal regions, particularly between June and September from east of Orkney to the Moray Firth, and down the East Grampian coast as far as the Firth of Forth. No surveys were undertaken in May so the seasonal increase in densities may occur earlier coinciding with the start of breeding. During September, densities increase further offshore suggesting a post-breeding dispersal eastwards as indicated also by the MERP data collation.

7.2.9.2 Abundance

The APEM offshore surveys yielded an overall abundance estimate in the study area of somewhere between 200,000 and 300,000 guillemots during summer, peaking in September after the end of the breeding season. Average summer densities were greatest between Orkney and the Moray Firth and down the east Grampian coast to the Firth of Forth mostly ranging from 2 to 10 birds/km2, and generally increasing nearer the coast, exceeding 10 birds/km2 in some localised areas. Between October and March, except for localised areas in the Moray Firth and Firths of Forth and Tay where densities ranged between 1 and 5 birds/km2, densities were generally less than 1 bird/km2.

It should be noted, however, that no correction was made to account for availability bias resulting from guillemots being underwater when the plane flew over. On foraging trips, guillemots rearing small chicks tracked in the North Sea spent 28.8 ±9.5% (mean ±sd) of their time underwater (Thaxter et al. 2010). Dive times averaged 46.4 ±27.4 and 50.4 ±7.4 seconds for guillemots during long and short dives respectively. Dunn et al. (2020) found that guillemots spent on average 4.1 ±0.23 hours per day (i.e. 17%) diving, with proportionately more time underwater during the breeding season than outside this period.

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following counts of individuals: Shetland (172,681), Orkney (181,026), Caithness (226,254), east coast Ross & Cromarty (1,944), Banff & Buchan (73,970), Gordon (3,345), Aberdeen (395), Kincardine & Deeside (72,179), Angus (1,002), Northeast Fife (28,103), Kirkcaldy (48), and East Lothian (8,266). These emphasise the numerical importance of the Northern Isles and north-east mainland coast of Scotland, although since that census, several colonies in the Northern Isles have suffered declines (JNCC 2021).

JNCC at-sea surveys indicated densities of at least 5 birds/km2 over large parts of the study area between March and October, particularly around Orkney, off the Caithness coast and throughout the Moray Firth, as well as down the East Grampian coast to the Firth of Forth (Stone et al. 1995). These persisted over a large part of the year with grid densities of between 2 and 5 birds/km2 becoming more prevalent off the east coast of mainland Scotland between November and February, when densities around the Northern Isles were around 1 bird/km2 (Stone et al. 1995). Analysis of the MERP data collation showed a similar spatial and seasonal pattern, and indicated densities in some areas (e.g. south Shetland to Caithness and East Grampian) reaching densities of 10-20 birds/km2 between July and September (ORJIP Project Phase 2, unpublished data using methods from Waggitt et al. 2020). However, those analyses used survey data largely collected in the 1990s and/or 2000s since when guillemots in Shetland have experienced declines which may partly explain the relatively low numbers found in that region during the APEM surveys.

7.2.10 Razorbill

7.2.10.1 Distribution

Razorbill numbers in Britain are greatest in Scotland, the largest colonies being in the Hebrides, on Handa island in Sutherland, in the Northern Isles, in east Caithness, Banff and Buchan, and at Fowlsheugh (Kincardine) (Mitchell et al. 2004). Census results showed steady increases in the UK population between 1969-70 and 1998-2002 (Mitchell et al. 2004). Seabird monitoring at sample sites since 2000 indicate major declines in Shetland and Orkney, but increases further south (Fowlsheugh, islands in the Firth of Forth, and St Abbs Head) (JNCC 2021).

At-sea surveys highlight the north-westerly distribution of razorbills within the North Sea, between April and September with distributions more spread out across the southern North Sea between October and March (Stone et al. 1995). The more recent analysis generating modelled density predictions indicated dispersal starting earlier, in July, and remaining over a wider area spanning the central and southern North Sea until March/April (Waggitt et al. 2020). This difference may be due to wider survey coverage of the North Sea region in later years. Areas with greatest densities were between Shetland and Caithness and around the East Grampian coast southward to the Firth of Forth, from May to October (Waggitt et al. 2020, Searle et al., in prep.).

Predictions of at-sea distributions in the waters around Britain and Ireland from GPS tracking of 281 breeding guillemots from 14 colonies along with habitat modelling (Wakefield et al. 2017) indicated In East Scotland potential high-use areas around Orkney, east of Caithness, along the south side of the Moray Firth, off the north Grampian coast and along the coast of the Scottish Borders.

The APEM surveys showed greatest densities in coastal regions, between April and October from east of Orkney to the Moray Firth, and down the East Grampian coast as far as the Firth of Forth. This coincided well with the results from the collated MERP data. During June and July, higher densities extend further offshore in the southern part of the survey area from East Grampian to the Firth of Forth.

7.2.10.2 Abundance

Point estimates from the APEM offshore surveys yielded peak abundance of c. 180,000 razorbills (but with wide upper bounds) in the study area in June, declining thereafter. Average summer densities were greatest between Orkney and the Moray Firth and down the east Grampian coast to the Firth of Forth mostly ranging from 1 to 5 birds/km2, and generally increasing towards the south and nearer the coast, where in the region of the Firth of Forth densities exceeded 10 birds/km2. Between October and March, except for localised areas in and around the Moray Firth and Firths of Forth and Tay where densities ranged between 1 and 5 birds/km2, they were generally less than 1 bird/km2.

As was the case for guillemots, no correction was made to account for availability bias resulting from razorbills being underwater when the plane flew over. On foraging trips, razorbills rearing small chicks tracked in the North Sea spent 17.5 ±6.6% (mean ±sd) of their time underwater (Thaxter et al. 2010). Dive times averaged 23.1 ±14.9 seconds for razorbills during their dives.

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following counts of individuals: Shetland (9,492), Orkney (10,194), Caithness (21,657), east coast Ross & Cromarty (251), Banff & Buchan (7,606), Gordon (547), Aberdeen (157), Kincardine & Deeside (9,760), Angus (562), Northeast Fife (4,114), Kirkcaldy and Dunfermline (91), and East Lothian (566). These emphasise the numerical importance of the Northern Isles and north-east mainland coast of Scotland, although as with the guillemot, since that census, several colonies in the Northern Isles have suffered declines (JNCC 2021).

JNCC at-sea surveys indicated densities of at least 5 birds/km2 in coastal areas from Orkney to the Moray Firth, and East Grampian south to the Firth of Tay, particularly between June and September (Stone et al. 1995). Over most of the study area, however, gridded average densities were less than 1 bird/km2. Densities tended to be lower during October to March, with only localised higher values in the inner Moray Firth and in the Firths of Forth and Tay (Stone et al. 1995). Analysis of the MERP data collation showed similar spatial and seasonal patterns, highest densities occurring in coastal areas of east Caithness, east Grampian, and the Firths of Forth and Tay) reaching densities of between 1 and 10 birds/km2 between April and September (ORJIP Project Phase 2, unpublished data using methods from Waggitt et al. 2020). During the breeding season, these were very localised around the main colonies but became more dispersed in August and September at an average density of c. 5 birds/km2 (Waggitt et al. 2020, Searle et al., in prep.).

7.2.11 Atlantic Puffin

7.2.11.1 Distribution

The Atlantic puffin is the second most numerous breeding seabird in Britain, with the largest colonies mainly in the Hebrides, Shetland and Orkney, and around the Firth of Forth (Mitchell et al. 2004). Whereas national censuses indicate that puffin numbers in Britain increased between 1969-70 and 1998-2002, and possibly beyond, there have been marked declines since then in Shetland (for example on Fair Isle), and on the Isle of May (Firth of Forth) (Mitchell et al. 2004, JNCC 2021).

At-sea surveys during the 1980s and 1990s had highest densities of puffins during the breeding season (April to August) in Shetland and Orkney, and the Firth of Forth, after which densities declined reflecting the wide dispersal of puffins during the winter (Stone et al. 1995).

The more recent analysis generating modelled density predictions also indicated highest densities in the north-western sector of the North Sea, particularly between April and September, concentrated around the Northern Isles and east Grampian region, and decreasing further offshore (Waggitt et al 2020, ORJIP Project Phase 2, unpublished data). The results indicated dispersal starting earlier, in July, and remaining over a wider area spanning the central and southern North Sea until March/April (Waggitt et al. 2020, ORJIP Project Phase 2, unpublished data). This difference may be due to wider survey coverage of the North Sea region in later years. Areas with greatest densities were between Shetland and Caithness and around the East Grampian coast southward to the Firth of Forth, from May to October (Waggitt et al. 2020, Searle et al., in prep.).

The APEM surveys showed greatest densities in coastal regions, between April and June east of Shetland, from east of Orkney to the Moray Firth, and down the East Grampian coast as far as the Firth of Forth. Densities were generally greater further south. From July onwards, puffin densities declined and remained very low from October to March.

7.2.11.2 Abundance

Point estimates from the APEM offshore surveys yielded peak abundance of c. 20,000 puffins in the study area in June, declining thereafter to between 3,000 and 5,000 birds. Average summer densities were greatest east of Shetland, south of Orkney and down the east Grampian coast to the Firth of Forth mostly ranging from 0.2 to 1 bird/km2, and generally increasing towards the south and nearer the coast. From July until March, densities were largely 0.1 bird/km2 or less. In the absence of local surveys it is not known whether or not the apparent persistent winter hotspots close to the coast off north-east Caithness and at the north-east corner of the Grampian coast are genuine.

The national census of breeding colonies in 1998-2002 (Mitchell et al. 2004) yielded the following counts of AOB (apparently occupied burrows): Shetland (107,676), Orkney (61,758), Caithness (1,278), Banff & Buchan (1,026), Gordon (619), Aberdeen (75), Kincardine & Deeside (768), Angus (190), Northeast Fife (42,000), Kirkcaldy and Dunfermline (1,701), Edinburgh (22), and East Lothian (28,412). These highlight the numerical importance of the Northern Isles as well as colonies in and around the Firth of Forth.

At-sea surveys in the 1980s and 1990s showed densities of 5 birds/km2 or more only around known colonies during the breeding season between April and August (Stone et al. 1995). From October to March, densities are low, for the most part less than 1 bird/km2 and spread evenly across the study area (Stone et al. 1995). Analysis of the MERP data collation showed similar patterns, with highest densities of between 1 and 5 birds/km2 occurring in coastal areas around the Northern Isles and East Grampian coast south to the Firth of Forth between April and September (Searle et al, in prep., using methods from Waggitt et al. 2020). Between October and March, densities were for the most part less than 1 bird/km2 (Waggitt et al. 2020, Searle et al., in prep.). Numbers estimated from the APEM surveys across the study area during summer were much lower than one might expect from the neighbouring colony counts. Part of this may be due to recent declines in the region but could also be because Shetland is on the periphery of the prediction region so if there are relatively large numbers there, they were excluded from the overall abundance estimates. On the other hand, there are also lower densities further south than expected from the colony counts, and it may be that not accounting for potential availability bias is the main issue. There is limited telemetry information on puffin dive durations, but birds from the Isle of May spent an average of 7.8h of the day underwater (Harris & Wanless 2012) whilst birds from Skomer Island spent an average of 4.6h per day with fewer dives but of longer mean dive duration (Shoji e al. 2015).

7.3 Marine mammals

7.3.1 Common Minke Whale

7.3.1.1 Distribution

The North Sea, particularly the north-western sector, has long been known as an important region for minke whales in Europe (Evans 1992, Northridge et al. 1995, Hammond et al. 2002, Reid et al. 2003, Hammond et al. 2013, Paxton et al. 2016, Evans & Waggitt 2020b, Hague et al., 2020, Waggitt et al. 2020, Evans et al. 2021). In the north-western North Sea, whereas some coastal areas around the Northern Isles (Evans & Baines 2010), in the Moray Firth (Robinson et al. 2009), and East Grampian region (Anderwald & Evans 2010) have been identified as local hotspots (see also Paxton et al. 2014), minke whale distribution further offshore has been less well examined, and has relied largely upon synoptic snapshot surveys of abundance such as SCANS (Hammond et al. 2002, 2013, 2021) and the Norwegian minke whale surveys that have extended southwards into the North Sea (Solvang et al. 2015).

Minke whales show general seasonal off-shelf/on-shelf movements, with some evidence of a southwards migration in autumn, returning north in spring (Risch et al. 2014), although a portion of the population remains around the British Isles through the winter (Anderwald & Evans 2007, Waggitt et al. 2020).

The APEM surveys show broadly similar results, with all but the last two surveys (in February and March 2021) having minke whale detections, and most occurring in June. In coastal regions within the study area, peak sighting rates occur in July and August (Evans et al. 2003, Robinson et al. 2009, Anderwald & Evans 2010).

At least in summer, minke whales favour shelf seas, feeding around banks and in areas of upwelling or strong currents around headlands or small islands (Tetley et al. 2008, Robinson et al. 2009, Anderwald et al. 2012).

7.3.1.2 Abundance

The SCANS III survey in July 2016 yielded an estimate of minke whale abundance in the North Sea of around 10,000 animals (Hammond et al. 2021). Within the broad bounds of the APEM survey area, SCANS estimated 2,498 (95% CI: 604-6,791) minke whales in block R east of Grampian region south to east of Tyne & Wear in NE England, 383 (95% CI: 0-1,364) in block S spanning the Moray Firth, Orkney and west of Shetland, and 2,068 (95% CI: 290-6,960) in block T covering the Shetland Isles south to offshore east of the Moray Firth. The boundaries of these blocks do not coincide at all with the APEM survey area so it is not possible to make direct comparisons. In SCANS, minke whale densities per block were estimated at 0.0387/ km2 (block R), 0.0095/km2 (block S), and 0.0316/km2 (block T).

The APEM surveys yielded an overall abundance estimate of a little under 2,000 animals for the survey area in the corresponding month of July with point estimates of densities over the region varying from 0 to 0.1 animal/km2, the areas with higher densities being between 0.05 and 0.1 animal/km2. SCANS surveys used a correction of 0.106 to account for availability bias whereas in this study, the correction applied was 0.04. This discrepancy is due in part to the instantaneous nature of the digital survey. It is based on surfacing rates based on visual shipboard survey which may not be entirely appropriate for a digital survey which dependent on water conditions can see further into the water column.

7.3.2 Common Dolphin

7.3.2.1 Distribution

The common dolphin has a predominantly westerly and southerly distribution around the British Isles, the species being uncommon in the North Sea (Hammond et al. 2002, Reid et al. 2003, Hammond et al. 2013, Paxton et al. 2016, Evans & Waggitt 2020b, Waggitt et al. 2020, Hammond et al. 2021). However, in recent years the species has been seen regularly in small numbers in the northern North Sea, and this has been attributed to the influence of climate change on some of their prey species (Evans & Waggitt 2020a, b).

Common dolphin detections during the APEM surveys occurred on one out of two survey days in March 2020, and three out of four days in June 2020. Most detections were far offshore in the middle of the North Sea. However, the total number of encounters across the eight surveys remains low.

Common dolphins are abundant and widely distributed in the eastern North Atlantic, occurring mainly in oceanic and shelf edge temperate seas from the Iberian Peninsula north to approximately 65°N latitude (though rare north of 62°N), west of Norway and the Faroe Islands (Reid et al. 2003, Murphy et al. 2013). In the offshore North Atlantic it seems to favour waters over 15°C SST and shelf edge features at depths of 400-1,000 m between 49°-55°N, especially between 20°-30°W (Cañadas et al. 2009). In shelf waters off the west coasts of Ireland and Scotland, and in the Irish Sea, common dolphin abundance tends to be greatest in the summer months at depths of 50-150 m (Evans et al. 2003).

7.3.2.2 Abundance

During the SCANS 3 survey in July 2016, there were no encounters with common dolphins in the North Sea and so no abundance estimates for this region. However, the presence of many casual sightings in the region (Evans & Waggitt 2020b) and the APEM aerial survey results support the presence of the species here but in small numbers, and possibly largely seasonally. The point estimate from these surveys is 4,110 common dolphins, assuming an availability at the surface of 0.05. Again, the low availability figure is generated by the instantaneous nature of the digital survey.

7.3.3 White-beaked Dolphin

7.3.3.1 Distribution

The white-beaked dolphin is a cool temperate to arctic species of the North Atlantic, occurring also widely across the North Sea, mainly in the northern and central sectors (Hammond et al. 2002, Reid et al. 2003, Hammond et al. 2013, Paxton et al. 2016, Evans & Waggitt 2020b, Waggitt et al. 2020, Hammond et al. 2021). There is some indication that the species is contracting its range northwards (Lambert et al. 2011, Evans & Waggitt 2020a); during the 1980s and 1990s, the species was common in summer around the north of Scotland and Northern Isles where now it is scarce (Evans & Baines, 2010, Evans & Waggitt 2020a, b).

Most sightings in British waters occur in summer, particularly in July and August (Evans et al. 2003, Evans & Waggitt 2020b), and this applies also to north Scotland (Evans & Baines 2010), and Grampian regions (Anderwald & Evans 2010). The APEM surveys also showed a strong seasonal peak in sightings across the study area in July (there were no surveys in August). No obvious hotspots were identified from these surveys, which was also the conclusion of an earlier wider analysis by Paxton et al. (2014).

White-beaked dolphins occur over a large part of the North-West European continental shelf, mainly in waters of 50–100 m depth, and almost entirely within the 200 m isobath (Reid et al. 2003, Evans et al. 2020b). However, in west Greenland, it can be found in much deeper waters of 300–1000 m (Hansen & Heide-Jørgensen, 2013), and, in the Barents Sea, commonly at 150–200 m and 400 m depths (Fall & Skern-Mauritzen, 2014).

7.3.3.2 Abundance

The SCANS III survey in July 2016 generated an estimate of white-beaked dolphin abundance in the North Sea of around 20,000 animals (Hammond et al. 2021). Within the broad boundaries of the APEM survey area, SCANS estimated 15,694 (95% CI: 3,022-33,340) white-beaked dolphins in block R east of Grampian region south to east of Tyne & Wear in NE England, 868 (95% CI: 0-2,258) in block S spanning the Moray Firth, Orkney and west of Shetland, and 2,417 (95% CI: 593-5,091) in block T covering the Shetland Isles south to offshore east of the Moray Firth. As noted earlier, the boundaries of these blocks do not coincide at all with the APEM survey area so it is difficult to make direct comparisons. In SCANS, white-beaked dolphin densities per block were estimated at 0.243/km2 (block R), 0.021/km2 (block S), and 0.037/km2 (block T).

The APEM surveys yielded an overall abundance estimate of around 150,000 animals for the survey area in the corresponding month of July. Point estimates of densities over the region varied from 0 to 5 animal/km2, with progressively higher densities occurring further offshore. These values are obviously much higher than the SCANS estimates. SCANS surveys used a correction of 0.676 to account for availability bias whereas in this study the correction applied was 0.06. Again, this was due to the instantaneous nature of the survey.

7.3.4 Harbour Porpoise

7.3.4.1 Distribution

Harbour porpoises are widely distributed throughout the shelf seas of the United Kingdom, with the North Sea a persistent important region for the species in Europe (Evans 1992, Northridge et al. 1995, Hammond et al. 2002, Reid et al. 2003, Hammond et al. 2013, Paxton et al. 2016, Evans & Waggitt 2020b, Waggitt et al. 2020, Hammond et al. 2021). The APEM offshore surveys show highest densities east of the Moray Firth and Grampian region. During the 1990s, numbers around the Northern Isles appear to have been much larger than they are today (Evans et al. 1997, Hammond et al. 2002, 2013), attributed to the marked decline in sandeel stocks in the region over that period (Evans & Borges 1996).

In coastal regions of North and East Scotland, porpoise sighting rates and numbers peak between July and October (Evans et al. 2003, Anderwald & Evans 2010, Evans & Baines 2010). The APEM surveys show an offshore peak in June, which is when most UK porpoises are born (Lockyer 1995), suggesting that some animals may then make a seasonal movement inshore.

Harbour porpoises mainly inhabit temperate and sub-arctic (11-14o C SST) shelf seas in depths of 20-200 metres, although some populations (e.g. West Greenland) may seasonally migrate into deep waters of the central North Atlantic (Nielsen et al. 2018). In coastal regions, the species frequently uses tidal conditions for foraging (Johnston et al. 2005, Pierpoint 2008, Marubini et al. 2009, Isojunno et al. 2012, Jones et al. 2014, Waggitt et al. 2017).

7.3.4.2 Abundance

The SCANS III survey in July 2016 yielded an estimate of harbour porpoise abundance in the North Sea of a little over 300,000 animals (Hammond et al. 2021), with greatest numbers in the central and southern North Sea following a southward shift in the 1990s (Hammond et al. 2002, 2013). Within the broad boundaries of the APEM survey area, SCANS estimated 38,646 (95% CI: 20,584-66,524) porpoises in block R east of Grampian region south to east of Tyne & Wear in NE England, 6,147 (95% CI: 3,401-10,065) in block S spanning the Moray Firth, Orkney and west of Shetland, and 26,309 (95% CI: 14,219-45,280) in block T covering the Shetland Isles south to offshore east of the Moray Firth. Again, the boundaries of these blocks do not coincide at all with the APEM survey area so it is difficult to make direct comparisons. In SCANS, porpoise densities per block were estimated at 0.599/km2 (block R), 0.152/km2 (block S), and 0.402/km2 (block T).

The APEM surveys yielded an overall abundance estimate of around 55,000 animals for the survey area in the corresponding month of July, with a similar number through most of the year except April to June when it increased to c. 120,000 animals. Point estimates of densities over the region varied from 0 to 5 animal/km2, with progressively higher densities occurring in the south of the survey area. We used an instantaneous availability of 0.123.

7.4 Recommendations for monitoring and future data analysis

Due to uncertainties in species identification in many cases from digital aerial images, we used additional information from the survey collation undertaken recently within the NERC/Defra funded Marine Ecosystem Research Programme (MERP) (Waggitt et al. 2020). These included visual survey data both from ships and planes where there was much better species discrimination. This will always be the case where distinguishing features require viewing from the side rather than above. One recommendation for future monitoring therefore would be to include some visual surveys particularly from vessels. Aerial surveys using digital video may also provide useful supplementary information and in some circumstances, is favoured over still imagery alone as a continuous sequence may allow a higher number of identifications at least for marine mammals. The digital aerial surveys were undertaken with relatively low image resolution (high GSD – flight height of 2000 ft) as a trade-off to achieve higher coverage. Consideration should be given on whether to use a lower GSD (i.e. higher image resolution) although only if this will improve species identification rates (e.g. auks & gulls amongst seabirds, and porpoises & dolphins amongst cetaceans). There is another issue of how best to treat subsurface animals, particularly cetaceans, whilst incorporating a correction for availability bias. APEM attempted to identify to species whether or not the animal was at the surface. Unsurprisingly, a higher proportion of species discrimination was made for animals at the surface. However, this will vary according to the levels of turbidity at the time. This was graded 0, 1, 2 or 3 (clear, slightly turbid, moderately turbid, highly turbid). This information may need to be incorporated in future analyses.

The correction for availability for the two dolphin species and the minke whales come from estimates based on visual survey (Table 4). As digital surveys also include animals which are visible beneath the water surface, using corrections based on visual survey may lead to overestimation of the animals. Correction for availability based on telemetry data (as in case for porpoises) would be the most appropriate to use, however to our knowledge, such numbers are not available for the two dolphin species and minke whales.

Some areas and periods of the year were not surveyed. The most important gap was east of Shetland, particularly between mid-April and the end of June, when there are large breeding seabird concentrations in the area as well as marine mammals such as killer whale. This was due to survey interruption arising from restrictions during the early period of the Covid-19 pandemic, and so could not be avoided, but it would beneficial if surveys at those times could take place.

Although the focus was on offshore areas (>12NM from the coast), the modelling would have benefitted from inclusion of results from nearshore surveys to avoid spurious results at the periphery of the target survey area,

The Northern Isles in general would merit more at-sea survey effort. They hold some of the largest populations in Britain of several seabird species and have the richest marine mammal fauna in Britain. They also have experienced significant declines in numbers for several species. And yet this is generally the region of Britain that has been least well surveyed. It would also be advisable to combine the data with those from nearshore surveys in the Northern Isles since depth gradients are more pronounced here than many areas within the North Sea further south off the east coast. Some species have habitat preferences that are within the 50-metre isobath, and therefore will be poorly sampled by the sawtooth line transect design used for the APEM surveys.

Future work could consider whether given the considerable autocorrelation in the data for almost all species, it could be worth taking photos at greater intervals allowing more widely spaced spatial coverage with little gain in uncertainty.