NatureScot Scientific Advisory Committee Sub-Group on Avian Influenza Report on the H5N1 outbreak in wild birds 2020-2023

Published: 2023

INTRODUCTION

The avian influenza (AI) outbreak in wild birds from 2020 to date is evidently unprecedented. The geographic scale, range of species of wild birds affected, and severity of impacts are of great concern for the survival of some species or populations.

NatureScot commissioned advice from its Scientific Advisory Committee in order to support Scotland’s Avian Flu Task Force. Specifically the sub-group was asked to consider:

1. An assessment of the current and emerging impact of highly pathogenic avian influenza (HPAI) on wild bird populations in Scotland, noting that the emerging evidence base is developing rapidly (and constrained due to restrictions on ringing and related activities in 2022);
2. An assessment of the current knowledge base regarding impact pathways, the vectors of transmission, species vulnerability, environmental persistence, and epidemiology modelling in relation to H5Nx in wild birds with a view to informing policy and identifying gaps in the knowledge base;
3. Taking account of the policy intention behind the Defra Wild Bird Surveillance Scheme provide advice on complementary surveillance, testing, and carcass collection to expand the evidence base on the extent and spread of HPAI in wild birds. This should be wide ranging advice in terms of what needs to be done, should note ongoing research in this area, consider practical and resource constraints, as well as the current policy intention and advice from SG Animal Health and APHA;
4. Assessment of the impact from various forms of disturbance (e.g., access related to tourism, shooting, general access, generalist land management activities and research on birds involving human–bird contact) in relation to H5Nx in wild birds with a view to informing policy on measures that could alleviate pressures on infection and recovery; 
5. Surveillance and monitoring priorities for passage and wintering populations of waterbirds in autumn-winter 2022-23, and breeding bird populations in spring summer 2023, and beyond. This will include guidance on permissible research activities; and
6. Further research, surveillance, scenario modelling and monitoring needs, and how these could be met most effectively. 

The group was chaired by Dan Haydon with Jane Reid, Anna Meredith, Mark Bolton, Eleanor Watson and Francis Daunt.  Emma Cunningham and Paul Digard provided invaluable input and support. The group was further supported by NatureScot and JNCC staff.

APPROACH TO THE COMMISSION

The primary purpose of this report is to provide a view on the knowledge base surrounding the current HPAI outbreak within wild birds in order that policy decisions can be best informed. Individual panels from the sub group considered one of the questions described above. The consideration of those panels is provided within the Annexes.

The sub group as a whole then reviewed the output from the individual panels and sought to provide a compressive overview that aimed to provide direction for policy makers. This is described in the following section.

OVERVIEW OF THE KNOWLEDGE BASE

Summary

Once HPAI infection is present in a wild bird population there are currently few management tools likely to be effective in controlling or reducing it.  However, relative to the marked natural spread of infection through a population it is highly unlikely that human activities around a site will increase the impact of an outbreak on wild birds.  It follows that our primary focus both now and in the longer-term should be on continued disease surveillance, demographic monitoring, and research to inform the long-term management of these populations in light of HPAI and other pressures.

Key points below are drawn from a series of detailed Annexes addressing the sub-group’s remit.

The epidemiological situation

1. We lack serological data to inform if influenza has previously circulated widely at subclinical levels in UK wild bird populations.  However, the current observed pattern of large-scale mortality in UK wild birds is unprecedented (section 1.1-1.6, Table 1.1)
2. We don’t know what has led to the change in epidemiology.  It could be changing migratory patterns of birds, changes to the virus, enhanced environmental persistence, or simply chance. 
3. However, we should anticipate that influenza will likely continue to be an issue over 22/23 winter (section 1.2), into the 2023 nesting season, and quite possibly well beyond.
4. The virus is currently not thought to be well adapted to mammals (section 2.5.2). There have been cases of foxes, otters, and seals testing positive in the UK but there is no evidence of onward transmission.  There has been only a handful of human cases, and as yet no documented cases of human-to-human transmission.
5. There is some evidence of mammal-to-mammal transmission in seals from the Caspian sea, sea lions from Peru, and mink from Spain.  These are worrying developments, providing a pathway by which the virus could adapt greater zoonotic potential (section 2.5.2).
6. Evidence from Whole Genome Sequencing (WGS) indicates that while the virus may have persisted at some UK sites since 2021, it was very likely to have been introduced into the UK in 2021-2022 by wild birds on more than 80 separate occasions, and to Scotland on at least seven occasions (section 2.3).
7. WGS and other epidemiological evidence indicates that wild birds spread the virus readily between sites within the UK, and between species within the same site.
8. Preliminary analysis of influenza WGS data indicates strong and direct transmission links between some wild Charadriiformes (Suborder Lari, including gulls, terns, skuas and auks) and domestic Galliformes and Anseriformes (section 2.3).
9. WGS data also shows the virus to be genetically very dynamic (with potentially important changes already occurring) and we should be prepared for a change in the manifestation of the outbreak.  This could mean a dampening or worsening of impact on birds and species most impacted.
10. Close attention should be paid to the wider international situation as it relates to the epidemiology of this virus, as there may be important early warnings and lessons to be learned about species vulnerability, virus adaptation, and possible mitigation tactics taking place in other countries.

Ecological impacts

11. Scotland holds important numbers of seabird species, and not just at the national level but also at the wider biogeographic level.  For example, Scotland holds a very high proportion of global populations of great skua, northern gannets and Manx shearwaters (section 1.5).
12. Records to date point to the following being the most seriously affected species:  barnacle geese (section 1.1,1.2), mute swans, northern gannets, great skuas, guillemots, kittiwakes, terns and large gulls (section 1.5), where some breeding sites have recorded large losses (Table 1.1).  However, the number of cases in all species reported to NatureScot are considered to be underestimates of actual incidence of mortality, especially outwith nesting colonies (many of which are well monitored).
13. The ecology of most seabirds, involving  low reproductive rates for long-lived individuals, means that recovery of populations that have experienced heavy mortality will take many years, and possibly decades (Table 1.1). 
14. There are no simple patterns that enable robust predictions regarding which wild bird populations will be impacted. Care should be taken in interpreting species-specific impact data. The available data do not necessarily enable clear distinctions between tolerance to exposure and absence of exposure (section 2.6).  Thus, just because a species escaped substantial impacts in one year does not mean it may do so again in future years.
15. Because influenza is a generalist virus, many species may transmit the virus, and thus rarer species coexisting at low frequency in mixed species communities (e.g. raptors, skuas, terns) are not necessarily anticipated to be at lower risk relative to more common species.
16. In the absence of serological data we cannot estimate what proportion of a population may have acquired natural immunity through exposure (and exposure to previous strains of virus may not confer some immunity on a future strain).
17. There are likely to be substantial indirect ecological effects resulting from the direct demographic impacts of infection and mortality, mediated through intra- and inters-pecific interactions. For example, mortality of one species may open reproductive or foraging opportunities for others, or alter predation rates.  Such effects may be positive or negative; currently we cannot predict with confidence how important and widespread these are likely to be (section 1.8,1.9).

Short-term management options

18. The risk to human health from wild birds is outwith the remit and expertise of this working group, and appropriate guidance should be sought from Public Health Scotland (PSE) particularly prior to conducting fieldwork.
19. There are very limited short-term options to influence the impacts of disease where infection is present (section 2.1, 2.2). 
20. Vaccinating wild birds is not feasible beyond very special cases.  Even placing regulatory issues to one side, there remain:  serious challenges with achieving meaningful coverage; questions about the duration of immunity to homologous challenge; and uncertainty on levels of protection provided by a vaccine to future circulating strains.
21. Benefits of carcass removal for limiting transmission are likely to be small unless a large majority of transmissions arise from carcasses (e.g. through scavengers assembling on carcases), and a large majority of all carcasses can be removed promptly (section 2.4).  This matches the conclusion of a Risk Assessment conducted by EPIC.  However, it would be helpful to examine the potential benefits of carcass removal through a formally designed study, which does not unduly disturb birds (e.g. in colonies).  Any carcass removal will require careful consideration on disposal practices (section 3.4.1).  
22. Reducing the overall infection rate of a highly infectious pathogen requires interventions to be very effective in substantially reducing the transmission rate; many interventions are unlikely to achieve this (for highly infectious pathogens,  theory suggests that even halving transmission might achieve relatively little in terms of overall outbreak size).
23. At sites where the virus is already present, the possible introduction of additional virus passively transported through human visitation can be regarded as a negligible additional contribution to the force of infection already present ($4.2.1).  This view is also consistent with the apparent lack of benefits to restricting human access to sites in 2022 (while recognizing the absence of a comparable counterfactual). 
24. We therefore see little justification for disruption and economic loss caused through suspension of tourism and recreational activities unless the epidemiological case is obviously very strong (section 4.2.1).  This is most likely to be the case when virus is not thought to be on a site, and visitors come in very close proximity to wild birds.
25. In particular, we anticipate no benefits from limiting access to those undertaking monitoring, surveillance and research activities, and indeed a very significant costs to disrupting these activities (section 4.4).  Individual and research team bird surveillance and monitoring is vital to helping us understand the impact of the virus, and we are grateful in particular for the volunteer effort that goes into this.  Some precautions intended to reduce the risk of humans to exposure are advisable, and might include standard biosecurity measures, use of PPE, and vaccination against strains circulating in humans (to minimize any possible recombination between human adapted and non-human adapted strains). 
26. Safe working practices should be adopted, mindful of associated risks and any appropriate local Health & Safety process. The most appropriate safeguards should be identified through consultation as the work is highly specialized, and location specific requiring the input of the relevant field team.  Notwithstanding, where an activity is required to be licensed the statutory nature conservation body, as the licensing authority, has ultimate responsibility for final decisions (section 4.5).
27. More generally, at sites where the virus is not present, biosecurity measures are essential.  A range of precautions should be considered, including imposing fixed time delays between visiting infected and uninfected sites, and careful decontamination of all possible objects that could carry the virus, between site visits (notably footwear, and any clothing which may have been soiled by birds, notably some seabird species). 
28. Effective communication with the public will be important to explain where and why restrictive actions are being taken. 

Longer-term management options

29. The paucity of short-term options to control the impact of the outbreak in wild birds makes longer-term planning for the recovery and the future management of these populations crucially important. 
30. Continuity of long-term demographic monitoring, and other related studies including diet, physiology, behaviour, movements and bird ringing (section 3.1, 3.2, 3.3) is imperative.  This work is critical for detection of further impacts of HPAI and its likely interactions with other drivers of population change (for example placement of windfarms, environmental change, effective management of marine protected areas), and status of population recovery.  The continuity of this monitoring is particularly important, and even short interruptions to data acquisition greatly diminish the collective value of these data, especially when we lose key information on the range of drivers beyond HPAI (section 4.4).
31. Continued epidemiological surveillance of HPAI should be regarded as the highest priority - in particular - estimates of disease incidence, mortality rates, and acquisition of samples which might generate additional whole genome sequencing (WGS), and serological data.
32. There is an important need and opportunity for expanded rigorous and integrated demographic and population monitoring of both HPAI impacts and other drivers.
33. While modelling will contribute to our long-term understanding of the epidemiology and ecological impacts of AI in wild birds, it is not likely to be useful to the short-term management of AI over the next year or two.  Modelling is most likely to make a valuable contribution through identifying which species might require,  and would benefit most from,  longer-term conservation management measures (section 2.8).

Priorities going forward

34. These large-scale perturbations to wild bird populations highlight an opportunity to make a step-change in our understanding of their ecology (section 4.3) by bringing improved resourcing and coordination to the diverse and currently somewhat fragmented approaches to their monitoring and surveillance.
35. It is critical that monitoring and research continue during this outbreak in order to investigate interactions between HPAI and other drivers of population changes, such as food supply variation, and e.g. climatic, human disturbance and wind farm effects. 
36. There is an obvious and important opportunity to develop and deploy large-scale Scottish pathogen testing and sequencing capacities as a research asset. World-class facilities and expertise exist across Scotland to conduct this research.  However, these are not currently being deployed to support this vital research, and thus our understanding of the epidemic and the use of this knowledge to inform policy is impeded (section 3.3).  It should be possible to conduct this additional research without in anyway changing the important role of APHA in official reporting (section 3.2).
37. Approved protocols for working in the field and acquiring pathogen samples need to be developed as a top priority (section 3.4).  While working with live virus is undoubtedly complicated, the most important work can be conducted on samples that can be inactivated at the site of collection, which should make the laboratory work much more straightforward. 
38. Considerable resources have been applied to organizing the response so far, and the returns have been good.  The need for this level of resourcing is likely to continue.
39. Linkages and ways of working across agencies and government have improved considerably within Scotland.   Across the UK, there is scope to work more closely to align policy and minimise unnecessary replication across the countries.
40. The sub-group is of the view that there is no silver bullet in terms of managing the short-term impacts of this outbreak.  While we have much to learn about population dynamics from demographic monitoring, transmission from WGS, and background immunodynamics from serological data, this knowledge base does not bear directly on short-term management decision-making.  However, by investing in population studies we will derive greater insights into longer term effects of the disease, and necessary management options.
41. HPAI in wild birds will possibly become an even more pressing issue over the coming few years.  This places Scotland's response to the epidemic in a global spotlight, and generates an internationally important research and policy formulating opportunity.  Careful consideration is needed now as to how these opportunities can be connected with policy development in relation to the conservation and management of already vulnerable and declining species populations (notably, seabirds).
42. Clearly, options for enhanced funding for research and mitigation measures should be pursued within Scotland, and at the UK and international levels.

Annex 1

Remit: An assessment of the current and emerging impact of HPAI on wild bird populations in Scotland, noting the wider geographical and biological context of these impacts, and the emerging evidence base is developing rapidly (and constrained due to restrictions on ringing and related activities in 2022)

Direct effects of HPAI on Scotland’s wild bird populations

1.1 Svalbard barnacle geese

The current outbreak in the Svalbard barnacle goose population in the Solway was first detected in late October 2021, and by the end of the winter, estimates suggest that 13,200 birds - around one third of the flyway population – had been killed by the virus. A total of 31 birds were tested between November 2021 and January 2022 with 29 of those found to be positive. Anecdotal evidence suggests that the breeding population in Svalbard in summer 2022 was down by 30% and that there were still signs of the virus circulating.

Preliminary monitoring data from the 2022/23 winter suggest that the geese had a very good breeding season, perhaps in part due to density dependent effects on the breeding grounds which may assist with recovery. However, an initial peak count of 30,958 geese reported for November 2022 is still considerably lower than the overall season peak of 43,703 reported for the flyway before HPAI in 2020/21. HPAI still seems to be circulation at low levels although there have been very few reports of dead geese on the Solway over the  winter of 2022/23.

1.2 Greenland barnacle geese

The first cases of HPAI H5N1 in the Greenland barnacle goose population were detected in Donegal, Ireland in late January 2022.  By early February 2022, the first Scottish cases were detected on Islay.  No cases were detected in other parts of the Scottish range.  By the spring migration in 2022 it is estimated that 1700 birds were died from the virus in Ireland and 1000 on Islay.

Reports from Iceland over the summer suggest that the virus was still circulating and that c.200 birds from the Icelandic breeding population had died. A co-ordinated count (additional to standard monitoring) took place across the key Scottish wintering sites in the first week of November 2022 which suggested that around 60,000 birds had returned and early estimates of productivity suggest it is around average levels of 10% (Malcolm Ogilvie pers. comm.).

In the 2022-23 winter, the first sick and dead birds on Islay were reported in mid-November at Gruinart Flats.  Swabs were taken from 4 carcasses and 3 of these tested positive. Initially, cases were at a low level, but in mid-December the numbers of sick and dead barnacle geese increased significantly. Over 900 deaths have been recorded since then, but this is a large underestimate due to scavenging and predation by white-tailed eagles, corvids and gulls.  A further 5 birds were tested in January 2023 and all of these were positive. The rate of die-off appears to be stabilising in mid-January 2023, but the locations where dead birds are being found are more widespread across Islay itself, and cases are now being reported on other islands such as Coll, Oronsay and Colonsay. None of the birds from these islands have been tested so far due to access difficulties. On Islay, the average of the November and December counts was 37,501 in 2021/22 and 33,284 in 2022/23; a further count in January 2023 may shed further light on the magnitude of HPAI losses.

1.3 Other wildfowl species

There have been large localised die-offs in Mute Swans (including over 100 at a single site near Glasgow for example). None of these population reductions are large enough to give cause for concern in terms of the overall conservation of the species in Scotland. A total of 32 greylag geese were tested, predominantly from Aberdeenshire, and 28 of these tests were positive for H5N1.  However, no significant concentrations of dead greylags were reported.  A total of 78 pink-footed geese have been tested and 68 of these were positive for H5N1.  These cases were widespread across the country, but there was a concentration of around 200 dead birds reported in the Findhorn area in April 2022. Mortality in pink-footed geese has continued into winter of 2022-23 with concentrations of dead birds in the Findhorn and Firth of Forth areas, but so far the numbers involved do not give cause for concern in terms of population viability.

1.4 Raptors

Up to mid-January 2023, there were 90 positive HPAI cases confirmed in Scottish raptors. The majority of these were buzzards (67), but other species reported were golden eagle (4), white-tailed eagle (5), sparrowhawk (4), red kite (3) hen harrier (1) and kestrel (1). These cases had a wide geographical spread, and are certainly an underestimate of mortality given that testing was limited, and raptors are less likely to be found and reported than more colonial species.

Impacts are unclear at this stage – for eagles, weather was poor at key points in the breeding season which may have been a factor in the low productivity observed, but for white-tailed eagles large chicks also died pre-fledging (which is unusual, with one chick HPAI positive on Mull). Anecdotal evidence also suggests unusually bad breeding seasons (including dead adult birds) for buzzards and red kites in several areas. An analysis of data from the Scottish Raptor Monitoring Scheme for the 2022 breeding season in comparison with previous years has been commissioned from the BTO and will hopefully shed further light on the situation and the reasons for any changes detected. It is due for delivery in March 2023.

1.5 Seabirds

Scotland holds important numbers of seabird species, not just at the national level but also at the wider biogeographic level. For example, Scotland holds 100% of the GB breeding population of great skua (which is around 60% of the world’s population of this species), around 80% of the GB breeding population of gannets (which is around 60% of the biogeographic population) and around a third of the global population of breeding manx shearwaters is found in the Rum Special Protection Area.

A total of ca. 20,500 dead seabirds across approximately 160 locations were reported to NatureScot over the period April 4th to September 11th 2022. The most badly affected species recorded were: northern gannets, great skuas, guillemots, kittiwakes, terns and large gulls (see Table 1.1). Some colonies have recorded particularly large losses including Bass Rock, St Kilda, Foula, Fair Isle, Noss, Hermaness, Troup Head and St Abbs. Impacts of HPAI were recorded right across Scotland and were seen in both adults and chicks, with the timing of the HPAI spread differing according to both region and species. For example, great skuas were detected early on as having HPAI, whereas species such as kittiwake and fulmar were not reported as being affected by HPAI until later on in the breeding season. It is likely that there will be further HPAI die-outs in subsequent breeding seasons, although the severity of these will depend on the levels of exposure and immunity built up in the surviving birds. The ecology of most seabirds, which involves slow reproductive rates and delayed age at first breeding, means that recovery of populations that have experienced significant mortality may take a number of years (Table 1.1).  The numbers reported to NatureScot are considered to be a significant underestimate, representing a small proportion of the total birds that have been affected by HPAI. Future seabird colony counts will provide a more reliable estimate of the true extent of population declines.

1.6 Other Species

Mortality due to HPAI is much less well understood for other key bird groups, including waders and passerines.  Very limited quantitative information is available for these species in Scotland, but positive cases have been recorded in passerines in particular, including blackbird, pied wagtail and carrion crow.  Establishing the importance of HPAI in these species will be challenging.

1.7 Changes in Reporting Protocols October 2022 Onwards

In an effort to standardise the way in which HPAI casualties were reported to NatureScot, a new reporting system was put in place in October 2022. This involves inviting professional and known amateur observers (e.g. reserve wardens, RSPB staff, NatureScot staff and local conservation volunteers) to check specific areas such as SPAs, NNRs local nature reserves and beaches adjacent to known areas of high conservation significance for dead birds. A reporting system has been set up using the Epicollect software system (Aanensen et al. 2014) to allow findings to be reported using an app or, where this was not possible, by e-mail. NatureScot can then arrange for testing to be carried out by APHA or collect swabs from the dead birds for testing if needed. This process has the advantage that we get regular coverage of key areas so HPAI impacts on areas of high conservation importance are more likely to be detected early, and the surveys are carried out more regularly so any changes in mortality rates detected over time are more likely to be reliable. The data collected on Epicollect from October 2022 to January 2023 are summarised in Table 1.2, below.

Indirect effects of HPAI on Scotland’s wild bird populations

1.8. Indirect effects of HPAI

Much of the focus on HPAI to date, including in the section above, has been in the context of what can be termed direct effects – the mortality of infected individuals and transmission dynamics to other individuals in the same or different species. Indirect effects have received less attention, yet may be important. These are the effects on other individuals that arise as a consequence of the changes in mortality rates, distribution and behaviour of infected individuals (i.e. where a consequence has another consequence).  Such effects may operate within and between species.

1.8.1 Indirect effects within species

A reduction in density could increase the availability of resources (e.g. food, nest sites) for surviving individuals of the same species. Conversely, in group living species like seabirds, a reduction in density may have negative consequences by, for example, increasing predation risk. Age-related mortality risk from HPAI could lead to a radical change in age structure of the breeding population, with potential consequences on demography.  There is some evidence that chicks have been more affected than breeding adults, but we currently have little information on the relative survival rates of immatures, pre-recruited adults and breeders.  If adult birds have been more strongly affected, age at first breeding could reduce as nest site availability increases, leading to lower average productivity since young breeders tend to be less successful. There may also be a loss of collective memory if a higher proportion of experienced birds die, which may be important, for example, in finding good feeding areas in patchy environments. By contrast, higher mortality among younger age classes may lead to older average age in populations, and senescent effects becoming more apparent. Sex-related mortality risk from HPAI could lead to reductions in effective population size in species with biparental care, the main form of offspring care in Scotland’s seabirds.  Differential survival rates among individuals, whether related to age, sex or other factors, could also lead to changes in social structure among surviving individuals, such as increases in divorce rates linked to the greater availability of alternative mates whose own mate has died, with potential consequences on average breeding success.

Within-species effects may lead not just to changes in mortality risk and resource availability but also distribution. For example, a breeding adult gannet that had been ringed on the Bass Rock was seen at a Norwegian colony during the 2022 breeding season, when under normal conditions breeding adults would not visit other colonies.

1.8.2 Indirect effects between species

Indirect effects between species arise because of the influence species have on each other in the form of community processes such as competition, predation and parasitism.  If a population declines because of an HPAI outbreak, other species occurring in the same locality may benefit from reduced competition for key resources such as nest site or prey availability. There are many examples of competitive interactions among Scottish species that breed sympatrically with similar nesting habitat (e.g. kittiwakes and razorbills) or food/prey preferences (wintering wildfowl; sandeel dependent seabirds).  The converse of this process is facilitation, whereby one species provides opportunities for other species. A good example of this is multispecies foraging assemblages in seabirds, whereby diving species bring fish prey to the surface when they become available to surface feeding species. A reduction in species such as guillemots could have negative consequences for e.g. kittiwakes by reducing prey availability.

A reduction in predation pressure arising from a HPAI-related decline in a predatory bird could have positive benefits for avian prey species e.g. reduction in eagles, great skuas or large gulls could benefit geese, small seabirds etc.  Similarly, a reduction in avian prey hit by HPAI, could have negative consequences for dependent predators e.g. high mortality rates in geese or puffins might adversely affect eagles and great black-backed gulls, respectively. 

As within species, these indirect effects may not just manifest themselves as direct changes in mortality risk or resource acquisition. Species also interact in ways that provide indirect information on key aspects of their ecology – for example, individuals from other species provide public information on e.g. the whereabouts of food or the quality of habitat. A reduction in density of such species will reduce the amount of information available.

Such HPAI-related effects on primary species may most obviously have consequences on demographic rates of species such as productivity and survival probability. Changes in demography could then lead to range shifts via differential population trajectories including local extinctions. Distributional changes could also arise from behavioural shifts, for example if predatory birds seek alternative prey in different localities. Such changes could further effect these species, that themselves may be of conservation or economic importance. Clearly, patterns of change in demography and distribution of associated species could have complex and far-reaching consequences and show considerable temporal and spatial variation.

1.9 Implications of indirect effects

In conclusion, these second order effects may be important but our current knowledge base is poor.  While we intuitively expect these cascading effects (and they’ve been documented in a variety of other systems) we currently can’t predict with any useful level of confidence how important and widespread they are likely to be.  Therefore, it will be important to incorporate robust quantification of these processes in order to gain a fuller insight of the effects of HPAI on Scotland’s birds. This would require synthesising the existing understanding on species interactions, and then devising monitoring and research that investigates changes in these key linkages in this context.

1.10 Concluding key messages

Annex 2

Remit: An assessment of the current knowledge base regarding vectors of transmission and control options (including carcass removal), species vulnerability, environmental persistence, the impact from various forms of disturbance, and epidemiological modelling in relation to H5Nx in wild birds with a view to informing policy and identify gaps in that knowledge base.

The perspective adopted in approaching these topics is entirely pragmatic.  Less a question of what we do or don’t know, or are curious about, but what do we know or what could we realistically find out that would help us identify management options? 

2.1. Vectors of transmission

In general, the detailed process of exactly how diseases are transmitted still remains poorly understood outwith those transmitted by clearly identifiable behaviours (e.g. sexual contact or biting).  Avian influenza is transmitted through direct contact: exposure to virus contained in body fluids, tissues, lesions, oral secretions, or persisting on skin, fur or feathers; and less directly through short range aerosolized droplet and aerosolized transmission.  Some viral strains may be better adapted to some modes of transmission than others .  Because influenza viruses can persist on different types of surface in the environment for up to a few hours (Bean et al. 1982) there is the likelihood of short-term environmentally mediated transmission: exposure to virus excreted by infected individuals into the environment, or persisting on, in and perhaps around carcasses.  This, virus can almost certainly be transmitted through a wide range of behaviours that enhance contact – for example: fighting, provisioning of young, bathing, social preening, carcass scavenging, predation, keptoparasitism etc.  However, we know virtually nothing about the relative importance of these different routes in natural contexts, and it probably wouldn’t help us much if we did as interventions to interfere with these natural behaviours and processes would be impractical or impossible to implement.

2.2. Environmental persistence

Environmental persistence of influenza virus in water has received considerable attention but aside from temperature (low temperatures favouring persistence) there isn’t a clear consensus on the effects of salinity and pH (Dalziel et al. 2016) (Martin, Becker, and Plowright 2018).  Virus is certainly capable of persisting over a time scale of months in fresh and salt water and at pH levels most likely to be encountered in field conditions (Ramey, Reeves, et al. 2022).  There is good evidence to suggest strain specific variation in the ability of virus to persist in different environments (Nazir et al. 2011).  Virus may persist in faeces for at least several weeks (Nazir et al. 2011).  The more likely important influence is whether virus will be maintained in sufficiently concentrated quantities to be infectious (i.e. not in the sea, or intertidal pools, or on rain-swept surfaces, but possibly in smaller persistent waterbodies with low or no flow through, or where guano accumulates).  Such areas have anecdotally been associated with infection(Jones et al. 2015, Camphuysen et al. 2022), and in some cases might provide a management option. There are limited studies of the extent to which virus might remain viable in soil and sediment and evidence it could persist for several months (Nazir et al. 2011).  However, virus was found to be absent from samples of water extracted from mosses lying directly under 4-month old carcasses of Great Skua (Furness, pers comm).

In summary, there are many possible routes by which virus might be transmitted within and between species over the short-term and they are almost impossible to differentiate and quantify.  However, this knowledge is not likely to lead to plausible management options.  There is a good deal of evidence that environmental persistence could play a role in the long-term maintenance of viral circulation, although no evidence that it definitively does (Breban et al. 2009).  Again, the management value of filling this knowledge gap is not obvious, so long as it accepted that environmental persistence may play a role in maintaining virus circulation.

2.3. Between species transmission

The transmission routes identified above enable virus to be transmitted within and between susceptible species, and this is likely to be common.  Preliminary analysis of Whole Genome Sequences (WGS) of H5Nx viruses is extremely valuable in shedding some light on what would otherwise remain entirely hidden from us.  WGS analyses indicate that transmission from Charadriiformes (shorebirds and gulls) to domestic Galliformes (poultry) and wild and domestic Anseriformes (waterfowl); and from wild Anseriformes to Charadriiformes and domestic Galliformes, is likely to be common (Lycett per comms).   The same analyses reveal that distinct viral lineages were introduced into the UK on around 80 different occasions in 2021/22 (and at least 7 times to Scotland) – although these are fairly robust lower limits, as upper limits they are sensitive to sample size (Lycett per comms). 

2.4. Carcass removal

Carcass removal is an often-discussed intervention as a means to reduce virus load/exposure risk.  If infections are already well established in the location, then it would seem unlikely that carcass removal would have a significant proportional impact on the force of infection present in the environment.  For example, the number of recoverable carcasses is likely to be small compared with i) the number of alive and infected birds, combined with, ii) the number of carcasses that are not for some reason recoverable, combined with iii) the virus in the local environment.  Even if the force of infection was significantly reduced, the impact might be limited to the ‘shape’ (timing) of the local epidemic, and not the overall final number of dead birds.  If there are benefits to this intervention they are likely to be highly context dependent.  There is so far only one study to our knowledge reporting the impact of carcass removal in terns in the Netherlands, and results are equivocal (Rijks et al. 2022).

2.5. Human transmission

2.5.1 Human to wildlife transmission

Because influenza virus can persist on clothes and shoes for several hours (less on skin) humans could fomite virus into wild bird habitats and breeding colonies if they have recently been exposed to the virus elsewhere.  Likewise, sequential handling of birds in a colony could result in transfer of virus from an infected to a susceptible bird, but while undesirable, if virus is already present in the colony if it arguably unlikely that such transmission will make any long-term difference to the size of the outbreak.

2.5.2 Wildlife to human transmission

H5Nx strains currently circulating in wild birds have to date infected humans only very infrequently indeed, however when they do cross over it can cause severe disease, but to date, no sustained human-to-human transmission has been observed (Ramey et al. 2022) ^{& refs therein}.  A pre-cursor to a more transmissible form of virus between humans might be indicated by within-species transmission within other mammal species. There are positive cases of foxes, otters and seals from the UK, and from other parts of the world (Spain, the United States, Peru and the Caspian Sea) and increasing likelihood of evidence of mammal-to-mammal transmission.  The identification of within species transmission in mammals would be a very concerning development as it could provide a pathway by which the virus could evolve greater zoonotic potential, and high-levels of surveillance are required to identify these events should they arise.

2.6. Species Vulnerability

It is tempting to partition vulnerability into at least two components: i) individual susceptibility to infection by an influenza virus strain; and ii) behavioural and ecological factors that influence individual exposure to infection and individual and population level consequence of infection. 

While there is some modest understanding of the relationship between the virus genome and species tropisms that relates in part to the cell receptor diversity and receptor-binding it is far from sufficient to reliably predict which particular strains are able to infect which host species (here termed intrinsic host susceptibility).  It is entirely possible that different tropisms are explained by a very few specific mutations, and therefore species tropisms could change very rapidly indeed (Verhagen, Fouchier, and Lewis 2021, Yassine et al. 2010).  It is important to stress the highly dynamic nature of the AI genome – importantly different phenotypes can be generated by a very few mutations or reassortments, and these could in principle arise on a time scale of days to weeks.  Indeed, there is already evidence of a number of potentially significant reassortments in the most recent UK genome data (Lycett pers comm).

However, it is often not possible to distinguish between lack of intrinsic susceptibility, and lack of exposure in explaining the absence of infection.  Furthermore, while there is some understanding of the genetic basis of low and high pathogenicity variants (and they can mutate quickly from one to the other (Beerens et al. 2020)) these are almost all studied (and defined) in chickens, and the same strain can change to a different virulence phenotype in another species (thus an HPAI strain in chickens may behave like an LPAI strain in a human or a wild bird).  Thus, the absence of overt disease and mortality cannot be interpreted as absence of susceptibility.    

So in the absence of more detailed data on intrinsic susceptibility  – there are few benefits to trying to distinguish between susceptibility to infection, and susceptibility to exposure. The BTO are in the process of undertaking a species vulnerability assessment accounting for this problem.  Their approach examines the product of scores (on a scale of 1 to 4) for exposure, sensitivity and consequence.  Scores thus range between 1 and 64.

Exposure considers 12 species traits used to assess variation between species in terms of likelihood of exposure to HPAI: migratory status, likelihood of making cold weather movements, gregariousness (separating breeding and migration/winter periods), degree of mixing with other species (separating breeding and migration/winter periods), predatory behaviour, scavenging behaviour, occurrence on farmland, occurence on wetlands, contact risk with humans and contact risk with poultry.

Sensitivity is assessed by scores to two factors: i) the likelihood of a species becoming infected if exposed to the virus, or the prevalence of LPAI/HPAI among the species group from published active surveillance data; and ii) the severity of infection and likelihood of rapid mortality occurring among infected individuals.

Consequence is assessed as the potential population level impact of significant mortality occurring for the species and was assessed using the following eight traits: UK breeding population size, UK wintering population size, UK breeding distribution, UK wintering distribution, UK conservation status, Europe conservation status, global conservation status and breeding strategy. (A potential concern with weighting consequence inversely by population size is that infection can spread very quickly and relatively completely through large and dense populations as the Gannet outcomes demonstrate).

In a trial of this approach the top 10 most vulnerable species were: Pochard, Wigeon, Barnacle Goose, Whooper Swan, Eider, Goldeneye, Gadwall, Bewick’s Swan, White-fronted Goose, Teal, Arctic Skua, Bean Goose, Shoveler, Mallard, Common Scoter, Garganey, Great Skua, Greylag Goose, and Velvet Scoter.  Work is on-going to validate this approach (and some limitations are clear), and to extend it to species populations at particular sites.

2.7. The impact from various forms of disturbance

The Working Group gave careful consideration to how this question might be addressed, concluding the range of disturbances and resulting impacts are likely to be species and context specific, and so well beyond the capacity of the group to fill out a complex matrix of assessments.  However, we make the following points in relation to a possible future study.

2.7.1. General approach recommended

The general approach favoured follows the principles outlined in the NatureScot Research Report 1273 “Development of Marine Bird Sensitivity Assessments for FeAST (2021), where species’ vulnerability to a particular disturbance is assessed through the combination of sensitivity and exposure. Given the wide range of human activities that can potentially cause disturbance that could increase the spread and severity of HPAI in wild birds, this is potentially a huge task, and in order to simplify it to manageable proportions, it is necessary to identify as small a set of categories of “disturbance activities” as possible that adequately captures all the different forms of disturbance that are relevant. Secondly, each disturbance category is then cross-referenced with the bird species, or species groups, which may be impacted by HPAI, in a matrix. Finally, each cell of the matrix (i.e. each disturbance-type x species-group combination) is scored for the extent of the likely impact of the disturbance activity (zero, low, medium, high). This procedure should be conducted for the breeding and non-breeding seasons separately, since disturbance pressures and vulnerabilities will vary by season.  Having identified which disturbance types are likely to cause greatest HPAI impact on which species groups, NatureScot can then consider what policy measures may be appropriate, and proportionate, to alleviate those disturbance impacts.   The Working Group reviewed a wide range of possibly disturbances, the possible ways of grouping species that might be impacted by the same disturbances in similar ways, the mechanisms by which they may increase the spread of HPAI, and approaches to mitigation and the process required to make decisions given the costs and benefits of curtailment/modification of human activities.  However, since there is little available evidence to assess the impact from the very diverse forms of disturbance that may influence the epidemiology of H5Nx in wild birds or to inform policy on measures that could alleviate pressures on infection and recovery, we do not enlarge of these points further in this report.

2.8. Potential contributions of modelling

In the absence of much data at all, it is doubtful that modelling can provide much useful precise guidance regarding the management of AI.  The most useful contribution of modelling may be in identifying which species might require and benefit most from longer-term compensatory management measures.  Thus analysis of population losses, likely recovery times, and the future role of naturally induced immunity may be of the highest priority to initiate.

We can identify three types of questions that might be addressed with modelling  

A) Useful questions that we do have a hope of providing some possibly helpful but general answers for.  This may include some limited scenario exploration. For example estimating:

B) Useful applied questions to which the answers would be valuable but which modelling - informed by available or achievable data - will simply not provide usefully precise answers for management purposes. For example:

i. By what mechanisms could AI remain persistent in wild bird communities where previously it could not?
ii. Will AI become regionally endemic?  In which species?  At what prevalence?  What are the demographic consequences?iii. The effect of possible interventions such as carcass removal         

C) Interesting but less directly useful questions as they do not inform a currently plausible management option.  For example:

i. Interspecies transmission of AI
ii. Spatial spread of AI
iii. Estimating R0 
iv. Estimating the impact of hypothetical interventions such as vaccination

2.9. Concluding key messages

Annex 3

Remit:  Taking account of the policy intention behind the DEFRA Wild Bird Surveillance Scheme, provide advice on complementary surveillance, testing (including of carcasses) to expand the evidence base on the extent and spread of HPAI in wild birds.  This should be wide ranging advice in terms of what needs to be done, should note ongoing research in this area, consider practical and resource constraints, as well as the current policy intention and advice from SG Animal Health and APHA.

3.1. Demographic surveillance

Wild bird populations in Scotland are monitored by a number of programmes, co-ordinated by partnerships with Scottish or UK-wide networks comprising a variety of organisations, groups and individuals. These routinely monitor numbers of birds, and to a lesser extent the breeding success of birds and other demographic parameters, at specific locations and times of year. Although disease surveillance is not a specific aim of any of these schemes, the current HPAI outbreak has heightened communication and awareness of signs suspicious or characteristic of HPAI in wild birds to encourage reporting of unusual wild bird mortality/morbidity via the appropriate channels. Relevant programmes include:

• The British Trust for Ornithology Ringing Scheme – (BTO) is the national UK wide ringing scheme that licences all catching and ringing of birds under devolved licences from NatureScot in Scotland and the equivalent government bodies in the devolved nations. This is a major network of fully trained volunteers and researchers organised into regional ringing groups that catch and ring birds across the UK and provides long-term ring recovery data cross all avian species that occur in the UK.  More focused schemes within this framework collect more focused data on survival and productivity of focal species or groups.
• The Retrapping Adults for Survival Scheme (RAS) (BTO/JNCC) is approx. 200 species specific projects used to collect adult survival data including some seabird and wildfowl species (full list www.bto.org/our-science/projects/ringing/surveys/ras
• The Seabird Monitoring Programme (SMP) (BTO/JNCC/RSPB) monitors 25 species of breeding seabirds on an annual basis to provide data for the conservation of their populations.  Scheme participants, both non-professional and professional surveyors, visit sites at both inland and coastal locations to count numbers of breeding seabirds and, where possible, their chicks to monitor breeding success. Additional data on survival, diet and phenology are collected at Key Sites.  
• Volunteer Seabirds at Sea Programme (VSAS) and European Seabirds at Sea Programme (ESAS) (JNCC) collates data from opportunistic surveys such as ferry routes in operation around the UK and shares data – the database only includes data collected by accredited surveyors. In a UK context, these are surveyors who have been assessed by a JNCC accredited ESAS instructor and have met five key standards of data collection. These are: Bird identification, Visual acuity, Application of the methods, Recording stamina, Navigation
• The Goose and Swan Monitoring Programme (GSMP), (BTO/JNCC/NatureScot) monitors 13 of the 14 major geese and migratory swan populations in the UK. Scheme participants, both non-professional and professional surveyors, visit sites in the autumn and winter months to count numbers of migratory geese and swans and, where possible, perform age assessments on the number of young birds present to monitor breeding success.
• Wetland Bird Survey (WeBS) (BTO/JNCC/RSPB) monitors non-breeding waterbirds. Wetland sites are counted once per month, providing data for population and trends in abundance and distribution.
• Scottish Raptor Monitoring Scheme (SRMS) currently focuses primarily on the annual monitoring of the abundance, distribution,movement and breeding success of 19 raptor species native to Scotland. First established in 2002, the Scottish Raptor Monitoring Scheme is now run by the nine-partner Scottish Raptor Study Group.
• Birdtrack (BTO/JNCC/RSPB/WWT/SOC/WOS/CAC) - an unsystematic citizen science project to look at migration movements and distribution of birds throughout the UK and Ireland. The app has a recently (Dec 2022) added facility to record dead/sick birds and whether avian flu is suspected.
• Breeding Bird Survey, and Waterways Breeding Bird Survey (BTO/JNCC/RSPB) annual monitoring carried out by around 3000 public volunteers  to monitor population changes of the UK’s common and widespread breeding birds, including waders, producing population trends for 118 bird and nine mammal species.

Demographic surveillance of specific seabird and raptor populations is additionally undertaken by research and ringing groups/citizen scientists at a number of long-term study sites in Scotland, where more intensive monitoring on annual productivity and breeding success occurs often at an individual bird level.  Key seabird sites include:

• the Isle of May and Inner Forth Islands (Nature Scot, Centre for Ecology and Hydrology, BTO, University of Edinburgh, University of Aberdeen, University of St Andrews and JNCC),
• Bass Rock (RSPB, The Seabird Centre, University of Heriot-Watt, University of Glasgow, and University of Edinburgh),
• Fair Isle (Fair Isle Bird Observatory Trust ,JNCC, NTS),
• Canna (Highland Ringing Group and JNCC),
• Bullers of Buchan (University of Aberdeen),
• Foula (University of Glasgow),
• St Kilda (NTS, RSPB),
• Eynhallow (University of Aberdeen)
• Colonsay and Treshnish Islands and Shiants.

These are representative at time of writing (Jan 2023) but additional organisations are periodically involved at these and other sites e.g. via collaborative research and PhD projects. Specific sites may have involvement of owners, trusts, observatories, researchers at different organisations and universities but ultimately all research work should be licenced through NatureScot scientific licencing or marine planning teams.

Movement studies to document dispersal and migration patterns as well as foraging activity and more local movements around colonies and nest sites are conducted by number of organisations and research groups, both as part of long-term demographic studies and in relation to development requirements for marine renewables. Some of this work is therefore contractually laid out by developers to meet their requirements. There is considerable scope to utilise this type of tracking information in combination with epidemiological information (see below) to inform on disease dynamic and potential routes of transmission. However, there may be restrictions on data access that stems from commercial contracts.

While, in general, demographic data collection is well coordinated through established routes, these surveys were not designed for disease surveillance  and may not operate at a frequency and timescale optimal for disease monitoring. However, they provide a framework for enhanced monitoring schemes going forward.

There has been argument for prioritising monitoring around species of most concern, however understanding variation in response and susceptibility would benefit from broadscale monitoring across species over the same outbreak events and timescales.

3.2 Epidemiological surveillance

General wildlife disease surveillance: At a GB level, the Animal and Plant Health Agency (APHA) Diseases of Wildlife Scheme (APHA DoWS) examines vertebrate wild species that are found dead for all disease and mortality investigations. From 2009, surveillance for vertebrate (apart from cetaceans) wildlife disease in GB is the responsibility of the Great Britain Wildlife Disease Surveillance Partnership, (GBWDSP) under the Chair of the APHA DoWS.  GBWDSP reports to Defra and the World Organization for Animal Health informing policy on wildlife, domestic animal and human health.

3.2.1 Passive HPAI surveillance

Avian Influenza Wild Bird Scheme (AIWBS). As HPAI is a notifiable disease, and in terms of risk to poultry, the APHA also carries out year-round avian influenza surveillance of dead wild birds under the AIWBS, in parallel to the DoWS. The AIWBS was not designed to assess disease risk to wild bird populations, but in some cases APHA may undertake more structured sampling of wild birds, for example in relation to infected commercial premises (IPs). Reports of dead wild birds are submitted via the public, BTO volunteers and warden patrols across GB on behalf of Defra, Welsh Government and Scottish Government, and the sensitivity/intensity of surveillance varies according to current disease risk, as determined by the Animal Disease Policy Group (ADPG). Defra’s approach to AI surveillance in wild birds is to:

i. align with biodiversity conservation targets and species recovery programmes;
ii. help understand the risk posed to and from poultry and other captive birds;
iii. improve knowledge on what virus strains are circulating and how they are evolving, and why some species are more resistant;
iv. develop models to predict future evolution and spread.

In Scotland, SRUC Veterinary Services (VS) are the APHA partner organisation for AIWBS (and DoWS), and have a remit from Scottish Government (Agriculture and Rural Economy Directorate) to undertake wild bird disease surveillance through its network of veterinary centres. The SRUC VS centres that can accept birds for post mortem examination under the AIWBS are Aberdeen, Dumfries and St Boswells. SRUC VS in Pentlands Science Park, Edinburgh, only takes carcases submitted for forensic examination by enforcement bodies (e.g. police) and not birds collected under the direction of Defra for AI testing. All wild bird surveillance is incorporated under a defined budget from Scottish Government to SRUC for both domestic and wild animal disease surveillance. Increased HPAI surveillance in wildlife therefore has both financial and personnel resource constraints.

As a notifiable disease, all HPAI sample testing and confirmation is undertaken under SAPO4/CL3 by APHA Weybridge (National Reference Laboratory). For H5N1 APHA Weybridge is currently only one of two locations in the UK that can handle diagnostic samples that are suspected to contain active virus (along with The Pirbright Institute). As wild birds are not defined as ‘kept’ under law, reporting is not a legal requirement, as it is for domestic/captive birds, but is encouraged through the AIWBS. 

SRUC do not undertake investigations where there is a suspicion or confirmation of any notifiable disease, including HPAI. However, as wild birds found dead, even during an HPAI outbreak, could have died from multiple unknown causes, post mortem examination can be legitimately undertaken by SRUC. Strict biosecurity precautions are always taken due to the possibility of HPAI or other notifiable/zoonotic infection. However, in the face of the current HPAI outbreak, samples are submitted to APHA to rule in/out HPAI and further investigations would only then proceed on negative birds, and is not generally advised. Once submitted, APHA/SRUC cannot release carcases to third parties.

Across Great Britain, a current (from February 2023) reporting threshold of a single dead bird of prey, dead gulls or wild waterfowl (swans, geese or ducks), or five or more dead wild birds of any other species at the same place at the same time, is in place.  Members of the public and warden patrols use this system, and may also report sick/moribund birds, although no intervention will take place until after death.  Reports are made using a Defra online/telephone helpline; reports are triaged and some birds will be collected for HPAI testing by APHA with criteria adjusted to vary the sensitivity of surveillance; – carcasses will most likely go to nearest veterinary centre but all samples for influenza testing go to APHA Weybridge. Note ringed dead birds should be reported via http://www.ring.ac/.

The surveillance programme will not collect further wild bird carcasses from the same location (defined as a 3km radius of where the birds were found) and once carcasses have been collected from a given location, no further carcasses of the same species will be collected for at least 14 days. A maximum of 5 birds will be collected from a particular location for testing when a mass die-off is reported. HPAI testing becomes unreliable as carcasses decompose, so if, after four days from the report, there has been no collection or no contact can be made with the person reporting the whereabouts of the carcasses, the carcasses will not be collected and, if necessary, disposed of appropriately. If disposal is required, the responsibility for this lies with the landowner on privately-owned land and the local authority on public land.

Collection of dead birds for diagnostic testing (wild and domestic) is performed by the Defra-contracted UK Farmcare Ltd, at Defra request, using appropriate biosecurity and PPE. In Scotland, Defra liaise with SRUC Veterinary Services (SRUC VS) who submit samples to APHA Weybridge for HPAI testing. In some circumstances, dead wild bird carcases are submitted directly to APHA on advice from Defra. NatureScot staff may also swab dead birds and submit carcases to SRUC VS under a SOP. Data on reported birds and test outcomes are held by APHA and are published on the Defra website. This information is copied to SRUC and Scottish Government (Chief Veterinary Officer (CVO) office).

Advice on collection of dead birds for research purposes is still to be clarified. Samples from dead birds for research purposes may be feasible if collected in a way that any virus is inactivated at source: approved methods to do this have yet to be verified and approved.

3.2.2 Other passive surveillance: 

The RSPB beached bird survey is conducted once a year (in February) by volunteers. It is systematic, but temporally and spatially limited and was initially designed to monitor effects of oil pollution and not HPAI surveillance. As such, it is unlikely to help with rapid identification and reporting of mass mortality events. A review of this survey is ongoing.

As part of the APHA DoWS, garden bird disease and mortality can be reported voluntary and investigated through the Garden Wildlife Health (GWH) programme and Garden BirdWatch scheme, and avian carcases submitted with prior veterinary approval. This is generally undertaken at the Institute of Zoology, London but Scottish garden birds might on occasion be examined by SRUC at the request of GWH.

For Scottish raptors, a single point of contact to coordinate dead raptor submission and undertake post-mortem investigation and tissue archiving was established at the University of Edinburgh (Royal (Dick) School of Veterinary Studies), working in conjunction with the Predatory Bird Monitoring Scheme which maintains a tissue and egg archive used for monitoring and research and receives 3-400 carcasses per year, and the Scottish Raptor Study Groups and SRMS. From 2016 this scheme undertook post-mortem raptor examinations, sampling and archiving, and co-ordinates with the Science and Advice for Scottish Agriculture Wildlife Incident Investigation Scheme, and with APHA if there is suspicion of HPAI. However, funding for the dedicated personnel for this scheme ceased in 2021.

Wildlife rescue organisations, such as the SSPCA National Wildlife Rescue Centre, and private veterinary practices may be presented with sick wildlife which may be infected with HPAI, with or without clinical signs. Once in captivity, wild birds fall under the legal definition of ‘kept’ and  reporting of suspicion of notifiable disease is required, with subsequent actions if AI is detected, e.g culling of all birds within the facility. Such birds cannot be submitted to SRUC VS. Under active Defra/SG HPAI restrictions (prevention zones), wildlife rehabilitation centres are not permitted to admit wild birds.

3.3 Active and opportunistic disease surveillance

Research projects licensed by NatureScot and the Home Office (Animals (Scientific Procedures) Act allow active capture and sampling of some wild bird or other wild species for research purposes.

There are four categories of projects relevant to active and opportunistic surveillance:

1. Projects already monitoring for pathogens including AI. This includes work that has specifically collected blood samples to test for current and previous exposure to a range of pathogens, including antibodies to AI in European shags on the Isle of May - all samples prior to cessation of work have been negative for H5N1. The Avian Flu Research Consortium (FluMap) funded by Defra and BBSRC launched in May 2022 to deliver research into how AI viruses are emerging in wild populations and the risk to both domestic and wild birds, including why some bird strains/species are more resistant to AI strains.
2. Projects that have subsequently been given permission to collect opportunistic samples for AI. Revision of some licences to include opportunistic sampling relevant to HPAI (e.g. choanal/cloacal swabbing) has been made in 2022 e.g for gannets on Bass Rock, though a limited number of samples were collected (n=19).
3. Projects, with and without licenses, with other primary purposes that involve collection of bird or environmental samples that could subsequently be used for opportunistic/retrospective analysis regarding HPAI status

There are other Scottish research groups that are routinely collecting blood and swab samples, and environmental/water sampling for other types of pathogen/parasite detection and/or other ecological research, that could also be utilised for HPAI screening. However, there is currently no simple way of identifying all potentially relevant activities that are being undertaken in this respect. One route may be to identify active researchers through licencing activities submitted to NatureScot and scope exists to reach out to the research community to enable researchers to identify where they may be able to offer data, samples or expertise e.g. faecal sample collection from geese in Orkney (Moredun).

4. Remote monitoring of populations – Remote monitoring of populations by drones is being trialled at several locations around the Scottish colonies by different groups. There is scope to develop these techniques to train AI protocols to e.g. recognise dead/live birds and pilot studies are being conducted to look at the feasibility of reading colour rings offering opportunities to expand demographic monitoring remotely.

Post sample collection there is considerable expertise to utilise genomic, serological, demographic and spatial data using different bioinformatic, statistical and modelling approaches to inform on potential routes of disease transmission and progression with notable concentrations at The Roslin Institute, School of Biology and School of Geosciences at the University of Edinburgh, the School of Biodiversity, One Health and Veterinary Medicine at the University of Glasgow, Biomathematics and Statistics Scotland, UK Centre for Ecology and Hydrology Disease Team, University of Aberdeen, University of Highlands and Islands, and EPIC (Centre of Expertise in Animal Disease Outbreaks).

3.4 Biosecurity and biosafety issues

3.4.1. Carcass collection

UK Farmcare Ltd are contracted by Defra to collect wild bird carcasses and submit for HPAI investigation and testing. Due to the potential zoonotic risk such carcasses should only be collected by trained personnel with appropriate biosecurity and PPE measures in place. The general public are advised not to touch a dead wild bird. Wild bird carcasses suspected or confirmed as having a notifiable disease must be disposed of as a category 1 Animal By-Product (ABP) via an approved disposal route.   If removal is needed, it is the responsibility of the landowner, or local authority in the case of public areas, to safely arrange disposal and to cover any costs associated, including disposal in compliance with relevant ABP rules: Dead or sick wild birds: what to do - Avian influenza (bird flu): how to spot and report the disease - gov.scot (www.gov.scot).  Removal of dead wild birds, for example in a breeding colony situation, to decrease risk to other birds via indirect or direct contact, e.g. through scavenging, or environmental contamination, is controversial. The recent (August 2022) Mitigation Strategy for Avian Influenza in Wild Birds in England and Wales, does not generally recommend removal of wild bird carcasses. This is due to the limited evidence to indicate whether this reduces transmission risk within colonies given the significant levels of environmental contamination that may remain in the area, and the close contact between birds in these colonies.  However, the Mitigation Strategy notes that there is emerging evidence from seabird colonies in Continental Europe that carcass removal may be effective in reducing incidence in some species when the risk of movement of the virus around the colonies by carcass collectors can be mitigated, together with the welfare impacts of entering the colony areas. Case-by case assessment is recommended, taking into consideration species behaviour, site accessibility and characteristics and timing within the breeding cycle, with a view to how disturbance and removal might alter outcomes in relation to disease transmission (e.g. entering a colony early or late in the season may lead to higher levels of disturbance and mixing of birds than for example the incubation and non-mobile chick phases), as well as ability to dispose of carcasses in line with ABP disposal rules and the health protection of personnel involved in carcass removal.

For collection and disposal of dead birds (and collection of swabs for additional surveillance activities), appropriate personal protective PPE must be worn, including FFP3 respirator, a coverall, goggles, rubber or polyurethane boots and disposable gloves. The respirator must be fitted for the individual (following Guidance on respiratory protective equipment (RPE) fit testing).  Footwear should be cleansed and disinfected using Defra approved disinfectants at the correct concentration should be used and disposed of appropriately to minimise environmental impact (see also associated guidance) and coveralls should either be disposed of or washed. Staff should also be trained in safe methods to get PPR on and off without contamination e.g. removing single-use gloves without contaminating your hands.

Advice on collection of dead birds for research purposes is still to be clarified. EPIC (Scottish Government’s Centre of Expertise on Animal Disease Outbreaks) has produced a scoping document to conduct a risk assessment (under review) for the movement of wild bird carcases (suspected or confirmed) that will also inform final protocols.

Good biosecurity practices are also important for on-going demographic surveillance.  These should cover both risk of transmission from birds to bird nd site to site nd the potential risk to individuals handling birds., and active and opportunistic disease surveillance. Individual organisations carry out their own risk assessments for their own employees following the most up to date HSE/UKHSA guidance and the BTO is updating guidance information for volunteer fieldworkers.  UK HSA has published a preliminary technical risk assessment and is currently putting together a technical team to explore potential risks to human health more fully.

The Advisory Committee on Dangerous Pathogens (ACDP) has issued guidance for laboratory workers handling flu viruses, which includes highly pathogenic avian influenza viruses. This covers both intentional and diagnostic work with viruses that are highly pathogenic for birds and details on SAPO containment (SAPO4) and laboratory requirements.

3.5 Data management

There is considerable discussion across various committees and platforms on how best to collect and collate and share data on HPAI in wild birds, particularly real-time event tracking, with a large number of bodies and channels currently being used. In certain circumstances, generation of sequencing data will be required, for understanding virus transmission events and for the identification of newly circulating strains. Workflows which enable appropriate sequence data to be generated must be considered, as well as capacity to generate, handle and make such data accessible to researchers. Other data management considerations include:

• Privacy and IP requirements e.g. GDPR, genome sequencing data
• Validation, consistency of methodology (e.g. through accreditation) and quality assurance of test results and data outputs

3.6 Concluding key messages

Annex 4

Remit: Surveillance and monitoring priorities for passage and wintering populations of waterbirds in autumn-winter 2022-23, and breeding bird populations in spring and summer 2023, and beyond. This will include guidance on permissible research activities. 

4.1. Background

Scotland currently fosters an impressive range of ornithological research activities, generating data that have been, and will be, highly valuable in providing necessary evidence to underpin key policy decisions surrounding wildlife conservation and conflicts. This includes the forthcoming 25-year Scottish Biodiversity Strategy; agricultural policy and associated agri-environment schemes; upland management; offshore renewables and MPA designations; wildlife crime; impacts of predators and invasive species; etc. The work is undertaken by professionals from combinations of academic, governmental, non-governmental, charitable and private sectors, alongside an invaluable army of highly skilled and experienced ‘citizen scientist’ volunteers.

We already have well-established frameworks for planning, authorising and undertaking ornithological fieldwork, through heuristic risk-reward (or cost-benefit) evaluations. We recognise that any form of ecological / ornithological research comes with potential negative impacts, for example due to short-term disturbance of animals and habitats, and risk of accidental harm (including potential for inducing mortality and/or reproductive failure). This is why many activities are licensed. Yet we consider that such impacts are acceptable, within reason, given the valuable data, insights and evidence that research activities generate.

Designated procedures are evaluated and approved through statutory agencies and Home Office licensing and ethical reviews. Other key activities, such as bird ringing, are licensed separately (e.g. through British Trust for Ornithology and in some cases NatureScot). Fieldworkers are trained to evaluate risks (e.g. from inclement weather conditions and catching methods), and to adjust activities accordingly. Further evaluation and approval of potentially delicate risk-reward balances is provided by Schedule 1 and Special Methods licensing. Hence, the need to appropriately evaluate risk-reward balances in ornithological research is nothing new. Now, we need to identify and incorporate new potential risks and rewards stemming from HPAI into our evaluation systems.

At the current time (Jan 2023) there are no HPAI-related suspensions (with the exception of barnacle geese) on ringing or research activities on any wild birds in any country within the UK. However, it is possible, and perhaps likely, that some taxa- and/or area-specific suspensions will be introduced if there are major HPAI outbreaks during 2023. Research activities involving affected species and areas will then need to be granted exemptions in order to proceed.

We suggest a framework for evaluating risks and rewards of proposed work, concerning HPAI specifically, and also encompassing other research activities. The ambition is to provide pragmatic guidance in the face of considerable uncertainty and the need for rapid decision-making and actions. NatureScot can take a positive lead in nurturing ecological/ornithological research and resulting knowledge gains in the context of current risks.

The framework conceptualises key ‘risks’ and ‘rewards’ of research activities, which can then be (qualitatively) balanced. Here, it is imperative to recognise that the rewards (and risks) are much wider than just HPAI. Many of Scotland’s bird populations were already facing major threats before HPAI arrived, including from climate change, habitat and prey loss including due to fisheries and offshore renewables, disturbance, invasive species, wildlife crime, etc. These other threats have not gone away. For some or many species, they currently pose bigger threats to population persistence than HPAI, and hence should remain a focus of attention.

Hence, while decisions to cease activities due to HPAI-associated risks could be viewed as ‘precautionary’ with respect to HPAI, they could be damaging with respect to other threats. Ceasing all activity, or all activity for particular sets of species, is unlikely to be justifiable as a blanket strategy, given that we know that there are other critical issues and evidence needs at hand.

As with (almost) all scientific endeavours, and certainly all field ornithology, we cannot quantitatively evaluate risk-reward balances. We cannot know the exact negative impact of any proposed action, and we cannot predict the current and future positive impacts of resulting knowledge. But, we need to make sure that such uncertainty does not paralyse all activity, and prevent collection of key data. We must then learn from errors that are discernible. This is in line with current practice, for example for cannon-netting, where data on mortalities are collected and causes noted to help shape best practice and identify any instances of clear malpractice which can then be addressed.

4.2. Risks of research in the era of HPAI.

There are three broad categories of risk relating to research activities conducted in an environment with HPAI:

1. Risks to wild birds
2. Risks to people (researchers / fieldworkers)
3. Risks to commercial / domestic birds (primarily poultry etc)

This paper primarily considers point 1, with some comments on points 2 & 3.

4.2.1. Risks to wild birds

Research activities undertaken in environments with HPAI could have non-zero risks to wild birds. This could include risk of spreading infection among individuals that are the target of research activities (e.g. through sequential handling), and risk of collateral impacts on non-target individuals and species. But, a position that we can only accept zero direct or collateral risk would be prohibitive, and poorly aligned with existing research and licensing practices where some risks are accepted as inevitable and justified.

Our ambition should be to minimise the risk that research activities will cause substantial excess impacts of HPAI.  Here, substantial excess impact is defined as large numbers of birds dying, suffering, or failing to reproduce that would not have done so had research activity not been undertaken.  This ambition captures the key point, that activities that may directly cause infection, or more serious infection in some extra individuals may have very little impact on the overall timecourse, endpoint, or hence overall magnitude of an outbreak.

Direct excess impact will be high if our activities expose or infect many vulnerable individuals that would not otherwise be exposed or infected; or greatly increase the probability of mortality, suffering or reproductive failure of infected individuals (e.g. due to disturbance, when sick individuals just need to rest and recover). However, direct excess impact will be low in species that are substantially HPAI-resistant (i.e. not easily infected, or where infections do not typically result in mortality / suffering / reproductive failure); at locations where focal individuals are likely to get infected whether or not they are impacted by some activity (e.g. if they live somewhere where infection is already present); and in species where the probability of dying once infected is already high.

Overall, the biggest risk of substantial excess impact on wild birds is likely to come if our activities introduce HPAI into a population of a highly sensitive species where HPAI is not already present. This is the primary situation that we should aim to avoid (i.e. potential to instigate new exponential growth in infection). In other circumstances, ceasing research activities may have little or no benefit in terms of reducing HPAI impacts. However, the degree of impact could also be viewed in the context of population size, and wider threat status or degree of vulnerability (as developed in the BTO’s risk assessment). This could identify situations where individual birds are of particular conservation value (e.g. very small populations of rare species).

4.2.2.   Risks to people (researchers / fieldworkers)

This is largely outwith the remit of this working group.  At time of writing there is no evidence that current widely circulating strains of HPAI pose a major risk to humans, however it is essential that this risk is kept under constant review.

Guidance on safe fieldwork practices in the context of HPAI, including PPE and methods for sanitising equipment, has been provided (for example advice to bird ringers from British Trust for Ornithology).  The US CDC advise fieldworkers to get a standard human flu vaccine a few weeks before fieldwork, stating that this may help reduce severity of symptoms in the event on infection, and minimize the chances of recombination with different human adapted influenza viruses.

Any further guidance on PPE that may be attached to exemptions from fieldwork suspensions should bear in mind that HPAI-related PPE may cause major risks regarding other dimensions of fieldwork health and safety, or simply be unworkable. This may be the case, for example, where fieldworkers need to climb cliffs or trees. Here, items such as full-face coverings that restrict visibility, suits that restrict movement, and rubber boots that restrict dexterity and grip, may be dangerous.

Blanket instructions on mandatory PPE in the context of HPAI should therefore be avoided. Rather, guidance should be provided, but appropriate safe working practices should be considered with or by fieldworkers given their knowledge of sites, activities and associated wider risks. As is standard procedure, health and safety practices should then be applied in proportion to all relevant risks.

4.2.3. Risks to commercial / domestic birds (primarily poultry etc)

Some guidance on minimising risks of infecting commercial / domestic birds following contact with wild birds has been provided, for example in the British Trust for Ornithology’s guidance for bird ringers.

4.3. Rewards of research in the era of HPAI.

Evaluations of applications for exemptions to undertake fieldwork in the context of HPAI should consider the rewards of the work alongside the risks. It is imperative that evaluations consider a breadth of rewards, and not solely focus on HPAI.  Overall, there are three broad categories of valuable research that should be strongly supported in the era of HPAI:

i) Work that is directly HPAI-related (e.g. infection surveillance, strain monitoring, environmental detection, estimation of survival/mortality rates, etc).

ii)  Work that is not directly HPAI-related, but is enabled by the impacts of HPAI. For example, opportunities to evaluate density-dependence in key life-history and movement parameters, and impacts of predator-prey interactions, as population sizes of single or multiple populations rapidly decrease. Density-dependence is often a key parameter in population models and adaptive management strategies (e.g. relationship between goose density and crop damage). It is often extremely difficult to estimate, and the impacts of HPAI may provide some opportunities to do so, as this could help with developing future population management models, including ones that have nothing directly to do with HPAI.

iii) Work that is neither directly HPAI related nor enabled, but is extremely valuable or urgently required to advance knowledge in pure or applied science and the resulting evidence-base for policy. This includes work that is valued by policy makers and advisors, and/or by private, non-governmental, charitable and/or academic sectors. High value can be (partly) evidenced as work that is funded, which typically means that it has been through some competitive assessment, and defined as of high/timely strategic or economic value and/or internationally excellent and ground-breaking science (depending on the sector). This encompasses activities that are primarily enacted by volunteer citizen scientists, but are actively supported and funded by organisations like NatureScot and BTO in partnership with JNCC – for example seabird, wildfowl and raptor monitoring programmes and structured ringing activities. Topics are likely to include, but are not limited to, evaluating ecological and evolutionary impacts of climate change; offshore and onshore renewables; agri-environment schemes; movements; evidence of wildlife crime; invasive species or other diseases; providing information to support MPA/HPMA designations etc.

4.4.  The risks of suspending surveillance, monitoring and research activities

The resilience of proposed work to short term disruption or, conversely, the severity of damage done by relatively short-term cessations, should also be considered. For some projects, it may not matter hugely whether fieldwork is done this year, next year or the year after (although it may still be frustrating to researchers/volunteers who would like to do it uninterrupted). For other projects, it could be highly damaging or even catastrophic to the planned knowledge generation. This will particularly apply where data are needed for imminent policy decisions, and where continuous runs of multi-year and/or multi-generation data are required and such ‘high investment’ work is already in mid-flow. In Table 4.1 we define five broad categories for risk-reward consideration:

It should be noted that such evaluations can change over time. For example, there may be relatively little damage to knowledge by ceasing some unstructured ringing activities for some short period. However, a longer hiatus could substantively damage capacity to monitor key demographic rates in wild bird populations, and hence capacity to understand full impacts of HPAI and other threats.

Evaluations of applications for exemptions should also consider the essential human element. Many key fieldwork projects are undertaken partly, substantially or entirely by ‘citizen science’ volunteers, whose work needs to be pro-actively valued and supported to ensure that they remain engaged and willing to contribute.  There can also be value in allowing fieldwork activities whose rewards come in the form of skills training and student development rather than immediate scientific value. Ceasing such activities could severely impact future capacity for knowledge generation.

Hence, the overall ambition should be to positively support fieldwork activities except in cases where HPAI or other risks are deemed unacceptably high.

4.5. Decision making

Permissible research and fieldwork exemptions in the context of future HPAI suspensions should be considered through a heuristic balance between apparent risk and reward-damage. Ultimately where an activity is required to be licensed it will be for the statutory nature conservation body as the licensing authority to adjudicate.

This should be done through dialogue between the licensing authority, BTO, key researchers and fieldworkers who would lead fieldwork activities. Such dialogues should proceed with the aim of permitting fieldwork activities whenever it is reasonably safe and valuable to proceed, with the recognition that some restrictions may be necessary.

Evaluations could be initiated through fieldworker self-assessment, recognising that the best people to judge risk-reward for any project on any species at any particular location will likely be expert fieldworkers on the ground. This could highlight difficult cases that would need to be discussed in more detail with NatureScot (and/or BTO or others). This is effectively analogous to current practices of special licensing. An initial example of what this could look like is in Table 5.2.  Such self-assessment schemes will need to be devised and explained to fieldwork leaders in advance of any further HPAI outbreaks, to ensure that exemptions can be considered and supported in a timely and transparent fashion.  It will be essential to ensure that approved time-sensitive fieldwork activities can proceed with minimal disruption.  This will require training of decision-making staff, to ensure that the risk-reward mindset is established with sufficiently broad and balanced appreciation of the range of potential risks and rewards. For sites where NatureScot is the land-owner or manager (e.g. NNRs), this will need to include site staff, to ensure that local and central evaluations of exemptions are congruent rather than conflicting.

Table 4.2 shows some examples of some fieldwork decisions that could be taken for work that is typically undertaken in Aberdeenshire (Bullers of Buchan, Whinnyfold and Forvie NNR), under the broad umbrella of Grampian Ringing Group. This sketch is based on scenarios that arose in 2022. Decisions would be sensibly re-evaluated in 2023 and subsequently, as knowledge of HPAI dynamics and impacts improves. These are intended as illustrative examples of decisions that could have been taken for particular activities at one particular time. They should not be taken as a prescriptive list of types activities that should or should not be carried out in coming years.

*Disturbance caused by these activities would likely be less than other forms of human disturbance that were observed in 2022, even after ringing activity was stopped (e.g. sea kayakers, wild swimmers, drone flyers, bait diggers etc…).

4.6. Concluding key messages

References

Aanensen, David M., Derek M. Huntley, Mirko Menegazzo, Chris I. Powell, and Brian G. Spratt. 2014. ‘EpiCollect+: Linking Smartphones to Web Applications for Complex Data Collection Projects’. F1000Research 3 (August): 199. https://doi.org/10.12688/f1000research.4702.1.

Bean, B., B. M. Moore, B. Sterner, L. R. Peterson, D. N. Gerding, and H. H. Balfour. 1982. ‘Survival of Influenza Viruses on Environmental Surfaces’. The Journal of Infectious Diseases 146 (1): 47–51. https://doi.org/10.1093/infdis/146.1.47.

Beerens, Nancy, Rene Heutink, Frank Harders, Alex Bossers, Guus Koch, and Ben Peeters. 2020. ‘Emergence and Selection of a Highly Pathogenic Avian Influenza H7N3 Virus’. Journal of Virology 94 (8): e01818-19. https://doi.org/10.1128/JVI.01818-19.

Breban, Romulus, John M. Drake, David E. Stallknecht, and Pejman Rohani. 2009. ‘The Role of Environmental Transmission in Recurrent Avian Influenza Epidemics’. PLoS Computational Biology 5 (4): e1000346. https://doi.org/10.1371/journal.pcbi.1000346.

Camphuysen, C.J., Gear, S.C,, Furness, R.W. n.d. ‘Avian Influenza Leads to Mass Mortality of Adult Great Skuas in Foula in Summer 2022’. Scottish Birds 42: 312–23.

Dalziel, Antonia E., Steven Delean, Sarah Heinrich, and Phillip Cassey. 2016. ‘Persistence of Low Pathogenic Influenza A Virus in Water: A Systematic Review and Quantitative Meta-Analysis’. PLoS ONE 11 (10): e0161929. https://doi.org/10.1371/journal.pone.0161929.

Jones, Jeremy C., Stephanie Sonnberg, Richard J. Webby, and Robert G. Webster. 2015. ‘Influenza A(H7N9) Virus Transmission between Finches and Poultry’. Emerging Infectious Diseases 21 (4): 619–28. https://doi.org/10.3201/eid2104.141703.

Martin, Gerardo, Daniel J. Becker, and Raina K. Plowright. 2018. ‘Environmental Persistence of Influenza H5N1 Is Driven by Temperature and Salinity: Insights From a Bayesian Meta-Analysis’. Frontiers in Ecology and Evolution 6. https://www.frontiersin.org/articles/10.3389/fevo.2018.00131.

Nazir, Jawad, Renate Haumacher, Anthony C. Ike, and Rachel E. Marschang. 2011. ‘Persistence of Avian Influenza Viruses in Lake Sediment, Duck Feces, and Duck Meat ▿’. Applied and Environmental Microbiology 77 (14): 4981–85. https://doi.org/10.1128/AEM.00415-11.

Ramey, Andrew M., Nichola J. Hill, Thomas J. DeLiberto, Samantha E. J. Gibbs, M. Camille Hopkins, Andrew S. Lang, Rebecca L. Poulson, et al. 2022. ‘Highly Pathogenic Avian Influenza Is an Emerging Disease Threat to Wild Birds in North America’. The Journal of Wildlife Management 86 (2): e22171. https://doi.org/10.1002/jwmg.22171.

Rijks, Jolianne M., Mardik F. Leopold, Susanne Kühn, Ronald in ‘t Veld, Fred Schenk, Allix Brenninkmeijer, Sander J. Lilipaly, et al. n.d. ‘Mass Mortality Caused by Highly Pathogenic Influenza A(H5N1) Virus in Sandwich Terns, the Netherlands, 2022 - Volume 28, Number 12—December 2022 - Emerging Infectious Diseases Journal - CDC’. Accessed 24 November 2022. https://doi.org/10.3201/eid2812.221292.

Verhagen, Josanne H., Ron A. M. Fouchier, and Nicola Lewis. 2021. ‘Highly Pathogenic Avian Influenza Viruses at the Wild-Domestic Bird Interface in Europe: Future Directions for Research and Surveillance’. Viruses 13 (2): 212. https://doi.org/10.3390/v13020212.

Yassine, Hadi M., Chang-Won Lee, Renukaradhya Gourapura, and Yehia M. Saif. 2010. ‘Interspecies and Intraspecies Transmission of Influenza A Viruses: Viral, Host and Environmental Factors’. Animal Health Research Reviews 11 (1): 53–72. https://doi.org/10.1017/S1466252310000137.