Diet of the European Shag Gulosus aristotelis during the non-breeding season at a roost within a tidal channel

ABSTRACT

The diet of the European Shag Gulosus aristotelis was assessed at one of their most northerly roosts in the UK; Bluemull Sound, Shetland. One pellet and 40 faecal samples were collected during the non-breeding season. The most frequent prey was Velvet Swimming Crab Necora puber, while the highest number of otoliths were from Saithe Pollachius virens and estimated mean (±sd) fish length was 143.9 ± 66.9 mm (range 81.4–223.6 mm).

The European Shag Gulosus aristotelis (hereafter Shag) is a Bird of Conservation Concern (BoCC) in the UK, added to the Red List due to steep declines in breeding populations (BoCC, Eaton et al. 2015). It is presumed that high winter mortality of adults and declines in prey (especially Lesser Sandeel Ammodytes marinus) availability and/or quality are some of the driving mechanisms behind this decline (Frederiksen et al. 2008, Heubeck et al. 2015). Therefore, studying the diet composition of Shags specifically during the non-breeding season is important if we are to increase our understanding of these mechanisms. Due to logistical constraints, however, most diet studies of Shags, like those of many seabirds, are conducted during the breeding season (Barrett et al. 2007). Such studies most often make use of regurgitated pellets that are examined for fish otoliths and other prey items (Johnstone et al. 1990, Howells et al. 2018). Fish otoliths can often be identified down to species level and the age class of the fish determined from the length of the otolith (Härkönen 1986). While less commonly used in diet studies than pellets, faecal matter (also called excreta) can also be examined for prey remains (Lumsden & Haddow 1946, Barrett et al. 2007, Radhakrishnan et al. 2010).

Although long-term diet studies have been conducted (e.g. on the Isle of May; Howells et al. 2018), Shag diet may vary considerably between locations and across seasons, rendering such studies unlikely to be representative of other geographical regions (Michelot et al. 2017, Nascimento et al. 2021). It is, therefore, paramount to capture information on diet from multiple locations, including opportunistically in remote areas where longitudinal studies are challenging. This study reports on the diet composition of Shags at one of their most northerly roosts in the UK.

Bluemull Sound (60° 41′19.1″N 0° 58'51.5″W) is a coastal channel with strong tidal currents located between the islands of Yell and Unst, on the northernmost edge of the North Sea, Shetland, UK. The sound lies near (<8 km) the Hermaness, Saxa Vord, and Valla Field Species Protection Area (SPA) that supports internationally important numbers of breeding seabirds, including Shags (NatureScot 2009). Underwater tidal turbines have been operating in Bluemull Sound since 2016, with three active at the time of this study (2019–2020). Throughout the channel, there are roosting sites of Shags, including the selected roost of about 100 individuals at Hoga Ness near Belmont (60°41′16.5″N 0°58′51.4″W) (Langlois Lopez et al. 2022).

Fresh pellet, otolith, and faecal samples were opportunistically collected or scraped from the rocky and grassy substrate at the roost while performing trail camera maintenance as part of another study (Isaksson 2022). Samples were, therefore, collected on 6 and 9 September 2019 (N = 17 and 10), 12 November 2019 (N = 10), and 1 September 2020 (N = 4). All samples were frozen upon collection, then thawed prior to analysis, washed using water and a 60 µm mesh sieve, and subsequently dried (Kubetzki & Garthe 2003). Fish otoliths were paired based on matching appearance, wear, width, and length. Pairs were subsequently identified down to species, where possible, using a reference collection (C. Angus, University of the Highlands and Islands, Shetland) and an otolith identification guide (Härkönen 1986). Lengths were measured using a binocular microscope and callipers, and these were used to calculate fish length and body mass according to the equations provided in Härkönen (1986). Anything other than fish otoliths (e.g. crustaceans) was identified as far as possible using an identification guide for marine fauna (Hayward & Ryland 2017), a reference collection (D. Riley, University of the Highlands and Islands, Shetland) and expert opinion (D. Riley, pers. comm.). Finally, the frequency of occurrence of different prey items was calculated.

A total of 41 samples were recovered from the roost site. Almost all samples were faecal matter (N = 40). The singular pellet contained six otoliths: from three Saithe Pollachius virens, one Atlantic Cod Gadus morhua, one Norway Pout Trisopterus esmarkii, and one unidentified gadoid. A solitary Haddock Melanogrammus aeglefinus otolith was also found at the site (Tables 1 and 2, Figure 1(a)). The majority of faecal samples contained crustaceans, predominantly from Velvet Swimming Crabs Necora puber and Edible Crabs Cancer pagurus and unidentified lobster (most likely Rugose Squat Lobster Munida rugosa, D. Riley pers. comm.; Table 1, Figure 1(b)). Additionally, 14 samples contained fish remains, either bones, plates, scales, or vertebrae (Figure 1(b), Table 1).

Figure 1. Identified fish otoliths (a) and examples of prey (b) from European Shag roost diet samples: (a) (i) Atlantic Cod, (ii)–(iv) Saithe, (v) Haddock, (vi) Norway Pout; (b) (i) Edible Crab Cancer pagurus, (ii) Velvet Swimming Crab Necora puber, (iii) Risso’s Crab Xantho pilipes (iv) unidentified lobster, likely Munida rugosa, (v) Common Sea Urchin Echinus esculentus, (vi)–-(vii) unidentified fish bones, (viii) fish scales, (ix) molluscs, likely Cingula trifasciata (top) and Spiralinella spiralis (bottom), and (x) unidentified bivalve.

Table 1. Frequency of occurrence of prey species in diet samples (pellet N = 1, faecal N = 40) from European Shags roosting at Bluemull Sound, 2019–2020.

Species/family/taxon	%	N
Mollusca	4.9	2
Edible Crab Cancer pagurus	49.0	20
Velvet Swimming Crab Necora puber	76.0	31
Risso’s Crab Xantho pilipes	17.0	7
Unidentified lobster	34.0	14
Common Sea Urchin Echinus esculentus	27.0	11
Atlantic Cod Gadus morhua	2.4	1
Haddock Melanogrammus aeglefinus	2.4	1
Saithe Pollachius virens	2.4	1
Norway Pout Trisopterus esmarkii	2.4	1
Unidentified gadidae	2.4	1
Unidentified fish	34.0	14
Total number of samples		41

Table 2. Details on otoliths collected, including length, mass, fish length, and weight calculated from equations in Härkönen (1986), and estimated age class as per Hillersøy & Lorentsen (2012).

Species	Otolith length (mm)	Otolith mass (g)	Fish length (mm)	Fish weight (g)	Estimated age class
Atlantic Cod	8.8	0.039	223.6	105.9	2
Haddock	10.7	0.079	231.4	108.6	2
Saithe	5.8	0.024	132.1	19.9	1
Saithe	4.8	0.015	98.9	8.5	0
Saithe	4.7	0.013	95.8	7.72	0
Norway Pout	4.2	0.003	81.4	2.48	0
Unidentified gadidae	2.2	0.001	–	–
Mean ± sd	5.9 ± 2.9	0.025 ± 0.03	143.9 ± 66.9	42.2 ± 50.7

This study reports on Shag prey during the non-breeding season at one of the most northerly roost sites in the UK. The high frequency of occurrence of crustacean species, especially Velvet Swimming Crabs, suggests that such species are an important component of Shag diet at this site (Figure 1(a), Table 1). This finding is consistent with fish trap studies within Bluemull Sound and close to our study site where Edible and Velvet Swimming Crabs were among the most frequent species caught (Fraser & Waggitt 2022). As these crabs are epibenthic (bottom-dwelling), the high occurrence of these species in Shag diet at the roost suggests that Shags, as elsewhere, are primarily foraging at the seabed (Neal & Wilson 2008, Watanuki et al. 2008, Dehnhard et al. 2022). However, as most samples were from faecal matter it is possible that remains from larger prey (i.e. otoliths) that would be expelled in pellet form are underrepresented here.

The fish otoliths that were identifiable within this study were mostly from gadoid species, and Saithe especially (Table 2). While Lesser Sandeels are the more typical prey in the North Sea (Howells et al. 2018), Shag diet is known to be more varied in other areas of Scotland (Swann et al. 2008) and neighbouring seas (Hillersøy & Lorentsen 2012). Notably, Saithe have been recorded on cameras mounted to tidal turbines within Bluemull Sound (Hutchison et al. 2020). This is likely to be linked to the predominant habitat type in the area, in turn, influenced by the strong tidal currents unique to this geographic region.

It is important to acknowledge the small sample size (N = 41) and bias of the samples towards faecal matter, which likely mean that the full diet of Shags at this site was not captured in this study. As the Shag roost is active year-round and accessible, we recommend conducting further studies. Collecting year-round information on Shag diet could help assess vulnerability to extreme weather events and marine renewable energy development. As the channel contains tidal stream turbines, determining the extent to which Shag diet is composed of species that may be affected by current energy extraction and/or underwater infrastructure is highly relevant. Tagging Shags there especially would allow for important habitats to be identified and for foraging behaviour to be characterized, which, combined with information on diet, could greatly aid in conservation efforts directed towards this vulnerable diving seabird.

Acknowledgements

The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Research Ethics Committee of University of the Highlands and Islands (OL - ETH SHE – 1230, approved on 21/02/2019). Thanks to Martha Devine and Glen Tyler for assistance with roost selection, Chevonne Angus, David Riley, and Saro Saravanan for assistance during laboratory work and Derek Jamieson for granting permission to use the site. Thanks to Ben Wilson and Maria Bogdanova for extensive comments on the PhD thesis chapter draft of this study and to Nina Dehnhard and editors for review of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was funded by the Bryden Centre project, supported by the European Union’s INTERREG VA Programme, managed by the Special EU Programmes Body (SEUPB), and by the Seabird Group through their Research Grant. The views and opinions expressed in this paper do not necessarily reflect those of the European Commission or the Special EU Programmes Body (SEUPB).