The 2022 Highly Pathogenic Avian Influenza (HPAI) outbreak that occurred in 2022 in many European countries affected several seabird species. Among them, Northern gannets (Morus bassanus), were particularly impacted. We conducted aerial surveys in waters around the two largest gannet colonies in Ireland (Little Skellig and Bull Rock, together representing 87% of the national population) in September 2022. During surveys, 184 dead gannets were recorded on survey effort, representing 3.74% of the total number of gannets recorded. We estimated the abundance of dead gannets in the surveyed area at 1,526 (95% CIs 1,299–1,793) individuals. The percentage of dead gannets observed was used to estimate a minimum local population mortality of 3,118 individuals across both colonies. Aerial surveys proved a reliable methodology providing key information on gannet mortality derived from HPAI at sea. The study provides the first estimate of gannet mortality in the two largest gannetries in Ireland.

Aerial surveys were conducted along the southwest coast of Ireland encompassing the two largest gannet colonies in the country, Little Skellig (35,294 breeding pairs) and Bull Rock (6,388 breeding pairs), and surrounding waters used for foraging. These two colonies account for 87% of the country’s breeding gannet population (Cummins et al., 2019).

Surveys were conducted in September 2022 over an area of approximately 12,036 km² using a twin-engine fixed-wing aircraft equipped with bubble windows to afford observers an unrestricted view of the sea below the aircraft. Transects were spaced approximately 3.7 km apart (Figure SM1, supplementary material) and flown at a height of 75m above sea level and a speed of approximately 185 km/h using strip-transect methodology (Eberhardt, 1978) in the target sea state of Beaufort ≤3 and >2km visibility. Two observers recorded all seabirds within 200 metres on either side of the aircraft along the transect lines noting species, number of birds, and behaviour, with each sighting date/time linked to location using a Garmin GPS 60.

Gannets could be distinguished from other seabird species due to their larger size and colouration, and dead individuals identified by posture; floating sideways with an extended neck and the wings not folded against their body, while live individuals were in flight or sitting on the water with the wings folded against the body.

Survey effort, and density of gannets were calculated in QGIS 3.22.7 (http://www.qgis.org) using a 4x4km grid across the study area. The total area surveyed per cell was estimated by multiplying the distance flown in each cell by the strip width (0.4 km if both observers were on effort, 0.2 km if one observer was on effort, 0 if both observers were off effort). The number of dead and alive gannets within each cell was counted and divided by the survey effort to obtain the density (individuals/km²) of dead and alive gannets. Estimated total abundance across the survey area was obtained by multiplying density by cell area and summing across all cells (Jessopp et al., 2018). Poisson 95% Confidence intervals (CIs) were calculated following the exact method:

qchiscq(0.025, 2*x)/2 , qchiscq(0.975, 2*(x+1))/2

To assess the representativeness of the survey area in relation to the foraging ranges of gannets from Little Skellig and Bull Rock, tracks from 23 individuals (9 from Little Skellig and 14 from Bull Rock) previously published by Wakefield et al (2013) were used to calculate a 50% Kernel Utilization Distribution (UD), representing the core foraging range, for each colony using the R Package Adehabitat HR (Calenge, 2006). Separate colonies' UDs were merged in QGIS to obtain a combined representative foraging range for both colonies, and the percentage of overlap with the surveyed area was calculated. We made an initial estimate of the population level impact of HPAI in southwest Ireland calculated as:

Estimated population mortality (EPM) = breeding gannets X (mean dead gannets estimate/mean alive gannets estimate)

Maximum and minimum estimates were calculated as:

Min EPM = breeding gannets X (min dead gannet estimate X (max dead gannet estimate/min alive gannet estimate))

Max EPM = breeding gannets X (max dead gannet estimate X (min dead gannet estimate/max alive gannet estimate))

REFERENCES:

• Calenge C. 2006 The package ‘adehabitat’ for the R software: A tool for the analysis of space and habitat use by animals. Ecol Modell 197, 516–519. (http://dx.doi.org/10.1016/j.ecolmodel.2006.03.017)
• Cummins S., Lauder C., Lauder A., Tierney D. 2019 The Status of Ireland’s Breeding Seabirds: Birds Directive Article 12 Reporting 2013 – 2018.
• Eberhardt L.L. 1978 Transect Methods for Population Studies. J Wildl Manage 42, 1-31. (http://dx.doi.org/10.2307/3800685)
• Jessopp, M., Mackey, M., Luck, C., Critchley, E., Bennison, A., Rogan, E., 2018. The seasonal distribution and abundance of seabirds in the western Irish Sea. 
• Wakefield E.D., Bodey, T.W., Bearhop, S., Blackburn, J., Colhoun, K., Davies, R., Dwyer, R.G., Green, J.A., Grémillet, D., Jackson, A.L., Jessopp, M.J., Kane, A., Langston, R.H.W., Lescroël, A., Murray, S., Le Nuz, M., Patrick, S.C., Péron, C., Soanes, L.M., Wanless, S., Votier, S.C., Hamer, K.C. 2013 Space partitioning without territoriality in gannets. Science 341, 68–70. (http://dx.doi.org/10.1126/science.1236077)