Highly Pathogenic Avian Influenza A (H5N1) Caused Mass Death Among Black‐Legged Kittiwakes (Rissa tridactyla) in Norway, 2023

TL;DR

A deadly bird flu outbreak in Norway killed thousands of black-legged kittiwakes, causing a major drop in their population.

Contribution

This study documents the first large-scale HPAI H5N1 outbreak in black-legged kittiwakes, revealing severe neurological and systemic disease.

Findings

More than 15,000 dead kittiwakes were reported, reducing the population by at least 50%.

Necropsies showed severe brain and pancreatic lesions with viral RNA and antigen.

Viral spread may occur through blood vessels due to α2,3-linked sialic acid expression.

Abstract

In 2023, highly pathogenic avian influenza (HPAI) heavily affected gulls in Europe. In July, a mass mortality event was reported in the black‐legged kittiwake (Rissa tridactyla) breeding colony at Ekkerøy in Northern Norway. The cause was confirmed to be infection with the HPAI H5N1 clade 2.3.4.4b virus, genotype EA‐2022‐BB. We describe the outbreak in kittiwakes, including pathological and virological investigations, and discuss the management and zoonotic potential. With more than 15,000 dead birds reported, we estimate that the outbreak caused a reduction in the kittiwake population at Ekkerøy of at least 50%. Diseased birds exhibited neurological signs. Necropsies of 10 birds revealed a peracute fatal systemic disease, with severe lesions in the brain and pancreas co‐localizing with viral RNA and antigen. Vascular expression of α2,3‐linked sialic acids (SAs) and viral RNA/antigen may reflect hematogenous viral spread. Further studies should investigate the long‐term impact of HPAI on kittiwake populations.

Figures

Weekly distribution of HPAIV H5N1 detections and reported mortality in black‐legged kittiwakes (Rissa tridactyla), from May to September 2023, Troms and Finnmark, Norway.

Photos taken during fieldwork at the sampling point at the Storelva River near Ekkerøy, Norway, on July 26, 2023. (A) Kittiwakes suffering from HPAI H5N1, several with squinting eyes and signs of apathy. In the background, dead birds and feathers floating in the water. (B) Kittiwake with opisthotonus (star gazing). (C) Epicardial petechiae and serous fluid in the body cavity and air sacs.

Histopathology and tissue expression of influenza A virus protein and RNA in organs from deceased kittiwakes. Brain (A–I): Hematoxylin and eosin (HE) sections (A, D, and G) demonstrated hemorrhages (A, arrows) and vacuolization of neuropil. Intact blood vessels (A, arrowheads). Necrosis with karyorrhexis and pyknosis (D, arrows) next to viable neurons and neuroglia (D, arrowhead). Single‐cell necrosis with shrunken hyper‐eosinophilic neuron and vacuolized neuropil (G, arrow) and viable neuron (G, arrowhead). Immunohistochemical detection of influenza A nucleoprotein (NP) (B, E, H, brown labeling) was observed multifocally in both the cerebrum (B, arrow) and cerebellum (B, arrowhead). Positive labeling for influenza A NP (E) in the area of necrosis is shown in the serial section (D). Both neurons (H, broad arrows) and neuroglia (H, narrow arrows) were labeled, in both cell nuclei and cytoplasm. Note that the cytoplasm of both the cell body and dendrites was labeled (H, arrowheads). In situ hybridization of influenza A virus RNAscope (C, F, I, red labeling) was detected multifocally in the cerebellum and cerebrum in both neurons, neuroglia, and ependyma (C, arrows), and meninges (F, arrow). Both the cell body (I, arrows) and dendrites (I, arrowheads) of neurons were labeled. Pancreas (J–L): HE‐stained sections of pancreas demonstrated multifocal necroses (J,  ∗). Labeling for influenza A virus NP (K) and influenza A virus RNA (L) was detected adjacent to the necrotic ( ∗) foci in the pancreas. Liver (M–O): HE‐stained sections of liver demonstrated areas of necrosis (M,  ∗) with hypereosinophilic hepatocytes and karyorrhectic and pyknotic nuclei. Immunohistochemical labeling for influenza A NP (K, brown labeling) was observed in both hepatocytes (N, arrowhead) and endothelial cells (N, arrow) of blood vessels. In situ hybridization demonstrated a more widespread labeling of hepatocytes (O, arrowheads), endothelial cells of blood vessels (O, arrow), and endothelial cells of hepatic sinusoids. Scale bars of (A), (C–I) indicate 200 µm, while the scale bar of (B) indicates 1000 µm.

Lectin staining of presumptive avian influenza virus receptors. MAL‐II (green) was used to label α‐2,3‐linked sialic acids in the brain (A, B), lung (C), and proventriculus (D) of deceased kittiwakes. Nuclei are labeled by Hoechst 33342 (blue). Epth, epithelium; Mn, meninges; VE, vascular endothelium. Scale bars indicate 50 µm.

Box‐and‐whisker plot showing the quantification cycle (Cq) values from rRT‐PCR for influenza A virus M gene performed on swab and tissue samples collected from deceased kittiwakes (n = 10), Ekkerøy, Norway, July 2023. The central line represents the median, with the whiskers indicating the range of maximum and minimum Cq values. The individual measurements are indicated by diamonds.

Midpoint rooted phylogenetic tree of the hemagglutinin gene from 10 HPAI H5N1 viruses identified in dead black‐legged kittiwakes from Ekkerøy, Norway, in 2023 (red dots), including contemporary genotype BB viruses from Norway (red) and Europe (black). Bootstrap values >50 are shown.

Full text

1. Introduction

From autumn 2020 to 2023, three major highly pathogenic avian influenza (HPAI) epizootics caused by H5Nx clade 2.3.4.4b viruses hit Europe, causing millions of deaths among wild and domestic birds [1]. During this period, H5N1 became the predominant avian influenza virus (AIV) subtype and spread with migratory birds across continents, causing a panzootic [2, 3]. Several genotypes of H5N1 circulated in birds in Europe [4], with the Eurasian BB genotype (H5N1‐A/Herring_gull/France/22P015977/2022‐like), also known as EA‐2022‐BB, becoming predominant from February to October 2023 [5]. This genotype contains three H13 virus‐derived gene segments, likely contributing to adaptation in gull species [6].

In Norway, the BB genotype was first detected in a dead herring gull (Larus argentatus) in April 2023. However, by summer, the black‐legged kittiwakes (Rissa tridactyla) (hereafter referred to as kittiwakes) became the most severely affected species.

Kittiwakes are long‐lived seabirds that breed in colonies on steep cliffs in the boreal and Arctic zones. They migrate along the Eastern Atlantic Ocean, with most birds wintering in the Western Atlantic. This species is listed as vulnerable on the IUCN Red List and as endangered on the Norwegian Red List [7]. The Ekkerøy colony (70°4′ N, 30°7′ E) in Vadsø municipality, Northern Norway, is the largest on mainland Norway, with an estimated 17,000 breeding pairs in 2012 [8].

Despite previous detections of AIVs and antibodies against non‐H5 AIV subtypes in kittiwakes [9–12], limited knowledge exists about the occurrence of HPAI H5 viruses in this species [13], and no studies have described the pathology or avian influenza receptor distribution. The impact of HPAI virus (HPAIV) infection on kittiwake colonies has, up until now, been largely unknown.

Recent spillover of HPAI H5N1 clade 2.3.4.4b viruses to wild and domestic mammals [1], such as farmed fur animals [14, 15], marine mammals [16], ruminants, and cats [17], raises concern for zoonotic transmission. Although human cases remain sporadic [18], increased circulation among domestic mammals elevates the pandemic potential, underscoring the importance of a One Health approach, including as part of an outbreak management plan.

In July 2023, HPAI caused a mass mortality event in the Ekkerøy kittiwake colony. We describe clinical signs, pathology, and virus receptor distribution in affected birds. Further, we characterize the causative H5N1 virus, including screening for mammalian‐adaptive mutations, estimate mortality in the colony, and discuss mitigation efforts and recommendations for future outbreak management.

2. Materials and Methods

Section 2 gives a brief overview of the material and methods used, with further details provided in Supporting Information 1: Appendix 1.

Confirmed HPAIV detections in kittiwakes in Norway from May 1 to July 31, 2023, were extracted from surveillance data from the Norwegian Veterinary Institute (NVI). Kittiwake mortality data were obtained from the County Governor of Troms and Finnmark. To estimate the number of kittiwake pairs breeding at Ekkerøy prior to the outbreak, an ad hoc survey of occupied nests was conducted by the Norwegian Institute of Nature Research and the Norwegian Nature Inspectorate on July 27, 2023.

Clinical observations of kittiwakes were mainly conducted at the breeding colony at Ekkerøy and at the river outlet of Storelva, 3.2 km north of the colony, where a high number of birds gathered. To ensure optimal sample quality, necropsy and sample collection were performed from 10 recently deceased kittiwakes. The birds were inspected, measured, and photographed, and their body condition score was assessed. From each bird, tracheal and cloacal swabs and a panel of tissues were collected for virological and pathological analyses, including histopathology, immunohistochemistry (IHC), RNAscope in situ hybridization (RNA‐ISH), Maackia amurensis lectin II (MAL‐II) staining, real‐time reverse transcription polymerase chain reaction (rRT‐PCR), virus whole genome sequencing (WGS), and sequence analyses.

3. Results

3.1. Extent of the Outbreak

The first PCR detection of the HPAI H5N1 clade 2.3.4.4b virus in kittiwakes in Norway was recorded in May 2023, in Harstad (67°0′ N, 16°28′ E), where a smaller outbreak occurred. From May 1 to July 31, 2023, 32 of 40 dead kittiwakes submitted for analysis at the NVI tested positive for HPAI H5N1, confirming that HPAI was present in most breeding colonies in Troms and Finnmark (Figure 1).

Weekly distribution of HPAIV H5N1 detections and reported mortality in black‐legged kittiwakes (Rissa tridactyla), from May to September 2023, Troms and Finnmark, Norway.

Between July 28 and August 28, 2023, 24,594 wild bird carcasses, mainly kittiwakes, were recorded in Troms and Finnmark County in Northern Norway, representing a major HPAI outbreak (Supporting Information 1: Appendix 1 Table S1). Of these, 15,235 carcasses were collected in Vadsø municipality, where the kittiwake colony at Ekkerøy is located. Large numbers of additional dead kittiwakes were observed in areas near the colony at Ekkerøy, including coastal waters, roads, and rooftops, indicating that the true number of dead birds was even higher. For comparison, the total number of kittiwake breeding pairs at Ekkerøy in 2023 was estimated to be 7817 (15,634 kittiwakes) based on colony counts (Supporting Information 1: Appendix 1 Table S2).

3.2. Clinical Characteristics

Clinically diseased kittiwakes exhibited general lethargy and severe neurological signs, including disturbed balance, trembling, falling over, or swimming in circles. Observations included opisthotonus, or head twitching, squinting eyes, and unilateral wing lameness, progressing to paralysis, paresis, recumbency, and death (Figure 2A,B, Supporting Information 2: Video).

Photos taken during fieldwork at the sampling point at the Storelva River near Ekkerøy, Norway, on July 26, 2023. (A) Kittiwakes suffering from HPAI H5N1, several with squinting eyes and signs of apathy. In the background, dead birds and feathers floating in the water. (B) Kittiwake with opisthotonus (star gazing). (C) Epicardial petechiae and serous fluid in the body cavity and air sacs.

3.3. Necropsy

The necropsied kittiwakes, six females and four males, were adults. Two were in good condition, four were in moderate condition, and four were emaciated (Supporting Information 1: Appendix 1 Table S3). Internal inspection revealed ample amounts of serous fluid in the body cavity and air sacs of all birds. Large areas of necrosis and hemorrhages in the pancreas, epicardial petechiae, and splenomegaly were observed (Figure 2C).

3.4. Histopathology

All 10 birds displayed mild to moderate histopathological changes in the brain, affecting both cerebrum and cerebellum, with multifocal vascular damage, hemorrhages, vacuolization of neuropil, and multifocal necroses (Figure 3A–I, Supporting Information 1: Appendix 1 Table S4). This varied from single‐cell degeneration and necrosis of ependymal cells, neurons, and neuroglia to large areas of necrosis. Two birds had multifocal mononuclear meningitis. Apart from this, inflammatory cells were not observed. All the birds had multifocal hemorrhages and moderate‐to‐severe and multifocal‐to‐diffuse necroses in the pancreas (Figure 3J–L, Supporting Information 1: Appendix 1 Table S4). In the liver, there were mild to moderate multifocal necroses in all but one bird (Figure 3M–O). No pathology was observed in the myocardium, except in one bird, which had focal subendocardial hemorrhages. No specific findings were observed in the lung, spleen, proventriculus, or kidney. Intralesional agents such as bacteria, parasites, or fungi were not observed.

Histopathology and tissue expression of influenza A virus protein and RNA in organs from deceased kittiwakes. Brain (A–I): Hematoxylin and eosin (HE) sections (A, D, and G) demonstrated hemorrhages (A, arrows) and vacuolization of neuropil. Intact blood vessels (A, arrowheads). Necrosis with karyorrhexis and pyknosis (D, arrows) next to viable neurons and neuroglia (D, arrowhead). Single‐cell necrosis with shrunken hyper‐eosinophilic neuron and vacuolized neuropil (G, arrow) and viable neuron (G, arrowhead). Immunohistochemical detection of influenza A nucleoprotein (NP) (B, E, H, brown labeling) was observed multifocally in both the cerebrum (B, arrow) and cerebellum (B, arrowhead). Positive labeling for influenza A NP (E) in the area of necrosis is shown in the serial section (D). Both neurons (H, broad arrows) and neuroglia (H, narrow arrows) were labeled, in both cell nuclei and cytoplasm. Note that the cytoplasm of both the cell body and dendrites was labeled (H, arrowheads). In situ hybridization of influenza A virus RNAscope (C, F, I, red labeling) was detected multifocally in the cerebellum and cerebrum in both neurons, neuroglia, and ependyma (C, arrows), and meninges (F, arrow). Both the cell body (I, arrows) and dendrites (I, arrowheads) of neurons were labeled. Pancreas (J–L): HE‐stained sections of pancreas demonstrated multifocal necroses (J,  ∗). Labeling for influenza A virus NP (K) and influenza A virus RNA (L) was detected adjacent to the necrotic ( ∗) foci in the pancreas. Liver (M–O): HE‐stained sections of liver demonstrated areas of necrosis (M,  ∗) with hypereosinophilic hepatocytes and karyorrhectic and pyknotic nuclei. Immunohistochemical labeling for influenza A NP (K, brown labeling) was observed in both hepatocytes (N, arrowhead) and endothelial cells (N, arrow) of blood vessels. In situ hybridization demonstrated a more widespread labeling of hepatocytes (O, arrowheads), endothelial cells of blood vessels (O, arrow), and endothelial cells of hepatic sinusoids. Scale bars of (A), (C–I) indicate 200 µm, while the scale bar of (B) indicates 1000 µm.

3.5. Immunohistochemical Detection of Influenza A Virus Nucleoprotein (NP)

IHC staining for influenza A virus NP in brain tissues revealed positive labeling multifocally in and adjacent to necrotic areas, in a large number of individual neurons, glial cells, and ependymal cells, and in a lower number of meningeal and vascular endothelial cells (Figure 3B,E,H). Evaluation of the pancreas revealed a positive signal multifocally, and commonly adjacent to the necrotic foci (Figure 3K). In the liver, a positive signal was present in hepatocytes and endothelial cells of sinusoidal capillaries and larger blood vessels (Figure 3N). No signal was detected in the lung, spleen, or heart muscle.

3.6. In Situ Detection of Influenza A Viral RNA

RNA‐ISH of brain, pancreas, liver, and spleen tissues from selected birds revealed signals of various intensities, indicating the presence of viral RNA (Figure 3C,F,I,L,O). The labeling was strong in the cerebrum, cerebellum, and pancreas, aligning with the IHC, albeit stronger and more widespread in the vascular endothelium. Labeling of the liver varied across individuals, from sparse pinpoint signal to strong, multifocal labeling of hepatocytes and endothelial cells throughout the tissue. Spleens from two birds were examined for viral RNA, one displaying a weak positive signal multifocally and the other being negative.

3.7. Distribution of Virus Receptors

MAL‐II labeled the surface of vascular endothelial cells in all examined kittiwake organs, suggesting the presence of α‐2,3‐linked sialic acids (SAs) that function as AIV receptors. Labeling in the brain was sparse and confined to vascular, ependymal, and meningeal structures, but no convincing signal was observed in neurons or glial cells (Figure 4A,B). Labeling was present in the mucus‐secreting epithelium of the proventriculus and the lung epithelium (Figure 4C,D).

Lectin staining of presumptive avian influenza virus receptors. MAL‐II (green) was used to label α‐2,3‐linked sialic acids in the brain (A, B), lung (C), and proventriculus (D) of deceased kittiwakes. Nuclei are labeled by Hoechst 33342 (blue). Epth, epithelium; Mn, meninges; VE, vascular endothelium. Scale bars indicate 50 µm.

3.8. Virological Investigation

High virus levels were detected in swab and tissue samples from all individuals by rRT‐PCR. The highest mean level, indicated by the quantification cycle (Cq) values, was detected in the brain (Cq 15.2), followed by tracheal swabs (Cq 21.8), cloacal swabs (Cq 23.6), liver (Cq 24.8), and heart (Cq 26.0) (Figure 5).

Box‐and‐whisker plot showing the quantification cycle (Cq) values from rRT‐PCR for influenza A virus M gene performed on swab and tissue samples collected from deceased kittiwakes (n = 10), Ekkerøy, Norway, July 2023. The central line represents the median, with the whiskers indicating the range of maximum and minimum Cq values. The individual measurements are indicated by diamonds.

HPAI H5N1 clade 2.3.4.4b viruses with the hemagglutinin (HA) cleavage site PLREKRRKR/GLF were confirmed in all 10 birds, using rRT‐PCR and WGS. No evidence of co‐infection with other AIV subtypes was detected by WGS. Similarity searches in GISAID BLASTn showed that the viruses from kittiwakes at Ekkerøy shared 99%–100% identity with contemporaneous genotype BB viruses from Europe. The phylogenetic analysis showed that the viruses clustered together and were highly similar to other viruses in gulls circulating in Norway, as well as Europe, during the spring and summer of 2023 (Figure 6, Supporting Information 1: Appendix 1 Figure S2 A–G, Supporting Information 3: Appendix 2).

Midpoint rooted phylogenetic tree of the hemagglutinin gene from 10 HPAI H5N1 viruses identified in dead black‐legged kittiwakes from Ekkerøy, Norway, in 2023 (red dots), including contemporary genotype BB viruses from Norway (red) and Europe (black). Bootstrap values >50 are shown.

Mutation analyses showed that the viruses detected in Ekkerøy kittiwakes remained mainly bird‐adapted and had a typical profile of the BB genotype, including NP‐Y52N and M2‐A30S (Supporting Information 4: Appendix 3). Key substitutions in the HA or polymerase proteins previously associated with increased mammalian adaptation were not identified. No neuraminidase mutations associated with oseltamivir resistance were present.

3.9. Management and Outbreak Response

During the outbreak, the municipality organized the removal and registration of carcasses. People handling the birds used disposable gloves, coveralls, goggles, and FFP3 masks. A sealed container with top‐loading access was installed in the area to enable hygienic and secure disposal of bird carcasses. Following necropsy and sample collection, the remains were deposited in the container. Once full, it was transported to a certified incineration facility in compliance with environmental and biosecurity regulations.

Compared to Ekkerøy, the observed mortality in the neighboring kittiwake breeding colony at Hornøya in Vardø municipality was low (observation, Tone Reiertsen). To prevent disturbance of vulnerable birds and further viral spread, movement restrictions to Hornøya and two other nature reserves were prioritized and implemented on July 27, 2023 [19]. At Ekkerøy, signs providing information about the outbreak and discouraging traffic in the area were posted. No restrictions were imposed on livestock grazing in the area.

4. Discussion

This study documents mass mortality in kittiwakes caused by HPAI H5N1 and enhances the understanding of the disease, pathology, and AIV receptor distribution in this species. Our estimate of the Ekkerøy breeding population size prior to the outbreak is associated with uncertainty, as it does not take into account immature individuals and “floaters” (i.e., sexually mature birds that do not breed). This implies that Ekkerøy’s total population in 2023 might have included as many as 30,000 kittiwakes [20]. Hence, with 15,235 reported dead birds and an assumption of a significant number of unreported cases, HPAI is likely to have caused a colony population reduction of at least 50%.

Seabird populations grow slowly due to their low reproduction rates and the years they spend at sea as immatures before breeding. Acute declines can severely affect future population status [21]. The recovery potential depends on long‐term trends in the population. Declining populations rarely recover, and the time to extinction is usually shortened [22]. The HPAI outbreak may therefore have dramatic long‐term impacts on the kittiwake population in affected areas.

Clinical presentation and pathological lesions of HPAI vary by species and viral strain [23]. The clinical signs of the infected kittiwakes indicate a disease prominently affecting the central nervous system (CNS), aligning with the neurotropic nature of the virus [24–32].

A pathological finding commonly described in gulls infected with HPAIV is a subacute disease with lymphoplasmacytic encephalitis and areas of gliosis [26, 27, 33]. This differs from our findings, where deceased kittiwakes demonstrated widespread necrosis in the brain, pancreas, and liver, with no or limited signs of inflammation. This could suggest hematogenous spread of the HPAI H5N1 genotype BB infection with limited time for an inflammatory reaction to take place prior to death, defining it to be a peracute, highly pathogenic, systemic, and fatal disease in kittiwakes. One study detected considerable species variation in the severity of CNS inflammation among wild birds naturally infected with the HPAI H5N1 virus, ranging from severe in swans and Canada geese to undetectable in a herring gull [27]. Experimental inoculation of ducks and wild‐caught laughing gulls (Larus atricilla) with the HPAI H5N1 virus produced necrosis in the brain and pancreas of birds that died or were euthanized due to severe illness, corresponding to our findings in kittiwakes [25]. In birds that recovered, mild lesions such as lymphoplasmacytic perivascular encephalitis and heterophilic pancreatitis were demonstrated. As our study only included deceased birds, we do not know if any infected birds recovered from the infection and whether these then would present similar subacute inflammatory lesions.

Aligning with hallmark features of HPAI, the kittiwakes showed signs of widespread systemic infection, with detection of HPAI H5N1 viral RNA in the brain, liver, and heart, in addition to the trachea and cloaca. By the time of death, the highest viral levels were detected in the brain, supported by low Cq values and extensive labeling by IHC and RNA‐ISH. This corresponded with the histopathological lesions being most evident in this organ. Widespread histopathological lesions, in addition to extensive IHC and RNAscope labeling, indicate high virus levels also in the pancreas. We did not perform rRT‐PCR on pancreatic samples. However, based on the extensive pathology in this organ, we recommend that rRT‐PCR of pancreatic samples be included in future studies.

Virus attachment to cells is a key factor for determining viral tropism, with AIV HA preferentially binding α2,3‐linked SAs [34]. HPAIV replication in endothelial cells is common in chickens and swans, but less so in ducks and mammals [35]. Interestingly, α2,3‐linked SAs were prominently detected on vascular endothelial cells in kittiwakes. Vascular damage with hemorrhages and endothelial expression of viral RNA and antigen were observed in the brain and pancreas. This indicates that kittiwake endothelial cells are susceptible to HPAIV, aligning with the hypothesis of hematogenous spread.

Despite the strong expression of influenza A viral RNA and NP in the brain, neurons were not labeled by MAL‐II, suggesting that H5N1 targets structures in neuronal cell types that do not overlap with MAL‐II binding sites (Sia‐α2,3‐Galβ1‐GalNAc). Gangliosides are the main carriers of both α2,3‐ and α2,6‐linked SAs in the CNS and support the attachment of different influenza A virus strains [36, 37]. While comparative studies of noninfected kittiwake tissues would be interesting, samples from healthy kittiwakes were not feasible due to their endangered status in Norway. The entry point of the virus could not be determined in our study. Like previous observations in herring gulls, laughing gulls, and ring‐billed gulls (Larus delawarensis) [38], the mucus‐secreting epithelium of kittiwake proventriculus and lung epithelium expressed MAL‐II targeted SAs.

The genetic similarity between the HPAI H5N1 viruses detected in our study and genotype BB viruses circulating in Europe during spring 2023 suggests introduction to breeding colonies in Northern Norway by migratory birds from Europe. The viruses in kittiwakes were highly similar to each other and clustered with the BB genotype viruses detected in other gull species in Norway, indicating that the outbreak at Ekkerøy was part of a larger epizootic. The temporal succession of positive cases within Northern Norway supports a gradual move of the virus from southwest to northeast. We cannot exclude virus introduction to kittiwakes during their pelagic phase in the Atlantic Ocean when they mix with other seabirds. However, we found no data to support this, and such an introduction would have required a longer presymptomatic phase of infection. Two HPAIV detections in kittiwakes in Finland in July 2023 clustered among the viruses from gulls in Northern Norway, indicating the same origin. Another virus from Southern Norway in July 2023 (A/Black‐legged_kittiwake/Norway/2023‐07‐1645‐1k/2023) clustered with HPAI H5N1 viruses detected in gulls in Denmark, England, and Scotland in the summer of 2023 and probably represents a separate introduction from the North Sea region.

Due to the consistent clinical signs, pathological findings, and detection of high levels of HPAI H5N1 infection in all examined birds, together with the absence of intralesional infectious agents (such as bacteria) and inflammation, an extensive differential diagnostic workup was not performed in this outbreak. Notably, WGS did not identify any evidence of mixed AIV infection.

Kittiwake mortality varied considerably between colonies, for instance, between Ekkerøy and Hornøya. A similar pattern was seen in tern colonies in the German Wadden Sea area in the summer of 2022, with some sites experiencing up to 40% mortality and others mostly spared [39]. Factors influencing mortality numbers include host susceptibility, colony size, species diversity, behavioral changes, bird movement patterns, immune status, timing, magnitude of virus introduction, environmental factors, and varying recording [40–42]. Understanding transmission within and between colonies requires wildlife community surveillance focusing on all interacting species.

Several factors likely contributed to the severity of the HPAI outbreak in kittiwakes. Due to the pelagic lifestyle of kittiwakes, preexisting immunity to H5 is assumed to have been low prior to the epizootic in 2023 [9, 12]. Virus properties, specifically the genetic constellation of the BB genotype, also likely enhanced virus replication in gulls [43]. Kittiwakes nest closely together, facilitating virus transmission. The outbreak at Ekkerøy coincided with a high number of susceptible chicks and adults with a reduced body condition toward the end of the breeding period that could have lowered their resistance and resilience to infection. At the Storelva River near Ekkerøy, numerous infected carcasses, feathers, and feces combined with low water levels likely promoted virus spread through freshwater. Environmental samples could have provided more information and should be obtained in future outbreaks.

The viruses detected in kittiwakes at Ekkerøy were avian‐adapted with mutation profiles similar to other BB genotype viruses, thus not indicating elevated risk for mammal adaptation. However, a few mutations that could increase the zoonotic risk were present. NP‐Y52N associated with evasion of human BTN3A3 restriction was identified [44].

Mitigation measures can be implemented to reduce HPAI spread and protect vulnerable wild bird populations. One measure is removal of carcasses, but this must be balanced against the risk of disturbing the remaining birds in the colony. The effect of carcass removal is difficult to measure. A study of HPAI‐affected sandwich tern (Thalasseus sandvicensis) colonies found that carcass removal reduced adult mortality by an average of 15% [45]. To keep the infection pressure in the environment low and reduce spillover to other birds and mammals, carcass removal should start early in an outbreak and be performed by trained personnel using personal protective equipment. Vadsø has only 5800 inhabitants, a long coastline, and minimal livestock farming. The high number of bird carcasses at Ekkerøy was challenging to handle due to limited staff in this remote area, coinciding with summer vacation, underlining the need for joint effort in extreme situations.

Vaccination of endangered bird species may be considered as a preventive measure for conservation in extreme cases [46]. However, for kittiwakes, oral or bait‐based delivery is not feasible due to their feeding behavior, and intramuscular administration poses logistical challenges in large seabird colonies. Additionally, vaccine efficacy may vary according to species and circulating viral strain.

Early detection of viruses with increased ability to infect mammals is crucial for pandemic preparedness. Management of HPAI mass mortalities in wildlife, therefore, requires a multisectoral One Health approach with clearly defined organizational roles. Active cross‐species surveillance to detect potential spillover was not performed during the outbreak in 2023 but is recommended during future wildlife outbreaks, with a primary focus on scavengers, predators, and marine mammals. Passive surveillance for avian influenza was ongoing but with limited samples from the area, and some cases in wildlife may have gone undetected. White‐tailed eagles (Haliaeetus albicilla) were observed at Ekkerøy during the outbreak. However, no disease or mortality was noted, suggesting that they were less susceptible to the BB genotype [47].

After it became known from the USA in March 2024 that ruminants can be infected with HPAIV, we performed post‐outbreak surveillance in sheep that had grazed among sick and dead kittiwakes. H5 antibodies were detected in one sheep 11 months after the outbreak in kittiwakes, indicating that spillover to mammals had occurred rarely [48].

Humans were followed up according to the existing routines recommended by the Norwegian Institute of Public Health at the time. These guidelines stated that the municipal doctor should consider offering postexposure prophylaxis with oseltamivir and recommend PCR testing if clinical symptoms developed within 14 days of exposure. No such cases with clinical symptoms were reported in connection with the outbreak. Serological testing was not an option at the time.

Our study does not determine to what extent kittiwakes survived infection and may have gained immunity. Studies that can provide such data may be of great value for projecting the vulnerability of this and other kittiwake colonies to future HPAI H5Nx outbreaks.

5. Conclusions

We recommend further monitoring of the long‐term impact of HPAI on the kittiwake populations. HPAI outbreaks in seabird colonies need awareness and must be mitigated using a multisectoral One Health approach to protect endangered wildlife species, mammals, and public health.

Author Contributions

Conceptualization, funding acquisition: Grim Rømo, Caroline Piercey Åkesson, Lars Austbø, Johanna Hol Fosse, and Ragnhild Tønnessen. Methodology, investigation: Grim Rømo (field work, epidemiology), Caroline Piercey Åkesson (pathology), Lars Austbø (PCR investigations), Johanna Hol Fosse (receptor studies), Johan Åkerstedt (epidemiology, geographic mapping), Maryam Saghafian (staining techniques), Cathrine Arnason Bøe (whole genome sequencing, genotyping), Olav Hungnes and Ragnhild Tønnessen (phylogeny, mutation screening), and Tone Kristin Reiertsen (mortality/population estimation). Writing – original draft, writing – review and editing: All authors. Project administration: Grim Rømo and Ragnhild Tønnessen. Supervision: Ragnhild Tønnessen and Caroline Piercey Åkesson.

Funding

The work presented in this paper was funded by the Norwegian Veterinary Institute.

Disclosure

The work presented in this article has been published as a preprint [49].

Ethics Statement

Carcass handling and disposal procedures followed national regulations and guidelines, and no specific permits were required for this study.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting Information

Additional supporting information can be found online in the Supporting Information section.

Supporting information

Supporting Information 1 Appendix 1: Detailed methods and additional results. Table S1: Number of dead birds reported by the municipalities in Troms and Finnmark County, Norway, between July 28 and August 28, 2023, due to the HPAI H5N1 outbreak in kittiwakes (Rissa tridactyla) in Northern Norway. Table S2: Counts of apparently occupied nests during the outbreak, used for the estimation of the number of breeding pairs of black‐legged kittiwakes at the Ekkerøy colony, Norway, in 2023. Table S3: Individual data and pectoral muscle condition score (PMCS) of 10 black‐legged kittiwakes that died from HPAI H5N1 and were necropsied at Storelva, Norway, in July 2023. Table S4: Score of histopathological examination: Hematoxylin and eosin (HE)‐stained slides of organs from 10 black‐legged kittiwakes that died from HPAI H5N1 at Ekkerøy, Norway, in 2023, were examined and scored semiquantitatively based on histopathological observations. Figure S1: A photo segment of the kittiwake cliff on Ekkerøy from July 2023, showcasing the image resolution. Apparently, occupied nests were counted from several such pictures to estimate the size of the breeding population in 2023. Photo: Knut Sverre Horn. Figure S2: Midpoint rooted phylogenetic trees showing the genetic relationship between 10 HPAI H5N1 viruses identified in dead black‐legged kittiwakes from Ekkerøy, Norway, in 2023 (red dots), and contemporary viruses from Norway (red) and Europe (black), including the coding part of the gene segments (A) PB2, (B) PB1, (C) PA, (D) NP, (E) NA, (F) M, and (G) NS.

Supporting Information 2 Video clip. Video footage from a highly pathogenic avian influenza (HPAI) H5N1 outbreak among black‐legged kittiwakes (Rissa tridactyla) in Vadsø, Norway, in July 2023. The video shows mass mortality and birds exhibiting clinical signs of neurological dysfunction.

Supporting Information 3 Appendix 2: Data included in the phylogenetic analyses.

Supporting Information 4 Appendix 3: Table showing markers which might indicate mammalian adaptation identified using FluMut in whole genome sequenced HPAI H5N1 clade 2.3.4.4b genotype BB viruses from 10 black‐legged kittiwakes at Ekkerøy in Norway, 2023.