Scottish Marine and Freshwater Science Volume 5 Number 9: Strategic Surveys of Seabirds off the West Coast of Lewis to Determine Use of Seaspace in Areas of Potential Marine Renewable Energy Developments

5. Discussion

5.1 The last previous boat or aerial surveys of the north west coastal waters of Lewis were undertaken pre-1990, over 20 years ago. The aerial and vantage points surveys from seven surveys covering the most important periods of the year for Scottish seabirds presented here are therefore valuable in supporting our understanding of the distributions and abundances of seabirds in this area.

5.2 The combined approach of aerial surveys and ground counts enabled good spatial coverage of sampling across the area and more detailed data on selected coastal locations, in particular the whole of Loch Roag.

5.3 By accurate georeferencing of objects in video images the relative use by seabirds and marine mammals of proposed wave/tidal sites can be assessed, while the application of density surface modelling enables estimation of species densities within these sites. Of those species identified as most vulnerable to the impacts of tidal turbines (Wilson et al. 2007, Furness et al. 2012) great northern divers, red-throated divers, black-throated divers, cormorants, shags, guillemots, razorbill, puffins, minke whale, harbour porpoises, white-beaked dolphin and grey seals were all recorded using the proposed renewable development sites though in relatively small numbers compared to breeding site populations within foraging range. Diver species in winter were shown to exceed national 1% levels, though identification was generally to species groups rather than species. Black guillemots were also recorded in the coastal development sites from the ground-based counts.

5.4 Comparison between records from the Vantage Point counts and the aerial surveys was completed within the areas covered by both survey techniques. Fulmar and gannet distributions and numbers were similar between the two methods, though data were also collected on flying birds in the aerial surveys.

5.5 In all months there was little overlap in distributions or identifications of gull species. Sample sizes were too small to detect any trends in identification but it is likely that differences in records were due to the highly mobile nature of these species.

5.6 The ground counts recorded larger numbers of diver species (red-throated divers, black-throated divers and great northern divers), shags and black guillemots overall and in coastal areas higher numbers of guillemots, razorbills and puffins, though comparisons of these species were complicated by the use of species groups in the aerial survey data.

5.7 On a fine scale the overlap of records was too imprecise to allow verification of individual birds, which could move relatively large distances over very short spaces of time. The data were therefore summarised as densities in coastal distance bands which was more robust to these spatial and temporal differences. The results of these comparisons highlighted some interesting patterns with aerial surveys recording around 10% of guillemots and razorbills combined compared to Vantage Points in the summer around the Butt of Lewis, but the percentage had increased to 20% between all auk species combined in the summer along the west of Lewis and in the winter aerial surveys recorded twice the number of auks in these west coast 2km bands. For the former two results, where aerial surveys had recorded lower numbers of auks it is possible that availability of birds was a contributing factor, where diving birds could not be detected in a near instantaneous aerial survey but would surface during the vantage point count duration. Studies in the Baltic have shown that breeding guillemots spent around 34% of their time at sea under water in dives (Evans et al. 2013) and breeding razorbills around 16% (Benvenuti et al. 2001). From these data we may thus expect instantaneous surveys to record around 34% fewer guillemots and 16% fewer razorbills than longer scanning searches such as from VPs. Undoubtedly such availability has a contributing effect and will also help explain the greater numbers of shags and divers recorded during vantage point counts. However the even fewer auks recorded from aerial surveys than dive data would predict combined with the reversal of recorded auk densities between the two methods in the winter suggests further study is required to explore differences between the methodologies with greater sample sizes.

Future recommendations

Ground-based counts of Loch Roag

5.8 A single day does not give sufficient time to cover the inner Loch Roag as thoroughly as would be ideal. A single day results in gaps in coverage and probably exacerbates under-recording because surveyors are inevitably in a rush to move on to the next counting point. With a one-day single-observer method there is insufficient time to walk out to some additional vantage points that give views over the areas that were not covered. It is recommended that in future a visit is spread over two people days. One day to cover Great Bernera and the eastern sectors, and the other day to count Little Loch Roag, and the western sectors up to Valtos. The January and February counts were undertaken over two days and achieved better coverage as a result.

5.9 Seals were counted irrespective of tide state. It was good fortune that the low tide period coincided with counting on the April and July visits. Data collected to date suggests an uneven distribution of species of high conservation importance. However, at least one more year of data would be required to be confident that the observed patterns were consistent. Identifying areas of consistently high value to importance species is important for informing site designations and case work such as renewable energy and aquaculture proposals.

5.10 The 2012 surveys did not aim to link birds seen during the breeding season to breeding sites. A small amount of information on breeding sites was gathered incidentally in 2012, principally for terns and cormorants. Determining the size and location of colonies in Loch Roag of cormorants, shags, tern species and black guillemots during future fieldwork would be valuable.

Aerial survey and ground count comparisons

5.11 This project identified differences in the numbers of birds and hence densities of particularly diving birds recorded from vantage point counts and digital aerial surveys. Though attempt was made to explore these differences the small sample sizes precluded conclusive trends to be determined. Similar studies on waters supporting higher densities of auks and/or shags would yield larger sample sizes with which to compare the methods and determine sources and scales of discrepancies. It should be borne in mind that camera technologies are continually being improved and it is possible that auk identification and possibly detection rates may improve with this.

Ground-based counts - Loch Roag priority bird species

Red-throated diver

5.12 Inner Loch Roag was shown to have high importance for summering red-throated divers, many of which are likely to be breeding locally ( Tables 14, Appendix II Figure 88, Appendix I). At least 39 were present in April and at least 26 in July. Allowing for gaps in survey coverage, possible under recording and that around half of breeding individuals are likely to be at breeding sites at the time of survey work, it is likely that inner Loch Roag provides feeding for in the region of 80 individuals through the breeding season. These are likely to comprise a mix of breeding birds and immature non-breeding birds, with the former likely to form the majority. The UK breeding population is approximately 1,255 pairs. If Loch Roag provided feeding for 25 pairs, this would represent in the region of 2% of the UK breeding population.

Black-throated diver

5.13 The regular occurrence of black-throated divers in the spring and summer is of note as this species has a small population size ( Table 15, Appendix II and Figure 89, Appendix I). It is likely that Loch Roag is an important feeding site for a small number of pairs, probably 1-2% of the UK population. There are at least two regular breeding pairs close to Loch Roag and the birds seen are likely to have been breeding locally and visiting Loch Roag to feed. Numbers increased in the winter with the arrival of wintering birds. The 12 recorded in February represents around 2% of the British overwintering population (Musgrove et al. 2011).

Great northern diver

5.14 Inner Loch Roag was shown to have high importance for wintering/passage great northern divers, with at least 95 individuals present in February ( Table 16, Appendix II, Figure 90, Appendix I). The low numbers recorded in October were to be expected as this species does not typically return to wintering areas (from sub-Arctic and Arctic breeding grounds) until late October and November.

5.15 There are thought to be around 2500 great northern divers overwintering in British and Irish waters (Musgrove et al. 2011), though this is likely to be an underestimate because of poor coverage by national surveys. Allowing for survey coverage and some under recording, it is likely that the numbers using inner Loch Roag in the winter comfortably exceeds 1% of the UK wintering population.

Slavonian grebe

5.16 The occurrence of up to 16 Slavonian grebes in the winter surveys ( Table 18, Appendix II and Figure 91, Appendix I) is notable as the size of the overwintering population of this species is small. It is estimated that 339 individuals overwinter in Scotland (Forrester and Andrews 2007) and around 1,100 in Britain (Musgrove et al. 2011), though these figures may be an underestimate as the species is easily overlooked. The numbers overwintering or occurring on passage in inner Loch Roag is in the region of 1% and 5% of the British and Scottish overwintering populations respectively and is therefore of importance.

Cormorant

5.17 Up to 14 cormorants were seen in the breeding season in inner Loch Roag ( Table 22, Appendix II, Figure 92, Appendix I). An additional 18 cormorants were seen incidentally on the July visit approximately 2 km north of the survey area, in the outer part of Loch Roag. These 18 birds were on or close to the breeding colony situated at Stac an Tuill, a small isolated stack north of Great Bernera. The numbers seen were almost certainly an underestimate of the total number breeding there as the site is too far offshore to be readily counted from Great Bernera. This is the only cormorant colony on Lewis. The sheltered waters of Loch Roag are likely to provide most of the feeding for birds from this colony. The breeding colony on Stac an Tuill represents approximately 5% of the Western Isles population (Mitchell et al. 2004).

Eider

5.18 Moderate numbers (maximum count 275) of eiders over-summer in Loch Roag ( Table 28, Appendix II). Females with 18 dependent ducklings estimated to six different broods were found in July, showing that the Loch Roag supports a small breeding population. These are the only eiders known to breed in Lewis.

Arctic tern and Common tern

5.19 Arctic and common terns were present in small numbers (up to 47 and 33 individuals recorded on one visit, respectively) and almost certainly breed on skerries in the 'East Loch Roag' and 'Kyles Pabbay and Valtos' sectors ( Tables 40 and 41, Appendix II). The numbers of Arctic terns present are well below 1% of the Western Isles population (4146 pairs). The numbers of common terns present, were they all breeding, would represent around 3% of the Western Isles population (502 pairs). Common terns were not recorded breeding on Lewis during the Seabird 2000 seabird colony census, so inner Loch Roag would be a new breeding site for this species. Establishing the size and location breeding colonies in Loch Roag should be an aim of future survey work.

Black guillemot

5.20 Up to 55 black guillemots were recorded on a single visit, and these are all likely to have been breeding locally ( Table 44, and Figure 94, Appendix I). Allowing for survey coverage and some under recording, it is likely that inner Loch Roag supports in the region of 40-60 breeding pairs. This would represent approximately 2% of the Western Isles breeding population, but well below 1% of the UK breeding population. The survey work did not determine the location of breeding sites.