Seabirds: strategic ornithological compensatory measures: review

7 Practical feasibility reviews

7.1 Selection of actions to include for practical feasibility reviews

As outlined in the methods, the initial list of compensatory measures was, where possible, refined to corresponding conservation actions. Measures that corresponded to clear conservation actions were assessed through a systematic literature review process and evaluated on their ecological feasibility. Where it was not possible to identify corresponding conservation actions directly from the initial compensatory measures, scoping reviews were conducted to identify and evaluate potential conservation actions. Additionally, bycatch mitigation was explored through a targeted review. For those actions where systematic or targeted reviews were conducted could practical feasibility also be assessed.

The actions reviewed through systematic literature reviews, and of which the ecological feasibility was assessed, were presented to the steering group (meeting held 22^nd June 2023) with discussion held on which actions should proceed to WP2 for assessment of practical feasibility. Following the steering group meeting, discussion with the project management group led the agreement to include seven of the eight possible conservation actions. This was based both on outcomes of the ecological feasibility assessment and the opinions of the steering and project management groups. Actions that were considered to build resilience but not necessarily lead to a compensatory effect (i.e. those not forecasted to increase breeding birds but that would potentially reduce vulnerability to short-term impacts) were also included. The final seven actions included are (i.e. 7/8 actions possible, with Disturbance - at sea excluded):

1. Sandeel fishery closure

2. Fishery closure or enhanced management of prey fisheries

3. End of the Gannet harvest at Sula Sgeir

4. Mammalian predator eradication and/or management

5. Avian predator management

6. Reduction of disturbance (at colony)

7. Bycatch mitigation in longline fisheries – automatically included though a targeted review was undertaken

Disturbance at sea was excluded due to there being a consensus that, at this time, there is insufficient evidence to have confidence that this action could lead to a meaningful scale of compensation.

For each action included there is a corresponding practical feasibility chapter, except for the two prey fishery related actions (i.e. Sandeel fishery closure and fishery closure or enhanced management of prey fisheries). These were combined into a single chapter due to many of the principles shared between the two. Given that Mackerel was excluded during WP2, this chapter is termed ‘Closure and management of forage fisheries (Sandeel, Herring, and Sprat)’.

7.2 Practical feasibility: Closure and management of forage fisheries (Sandeel, Herring, and Sprat)

Note: This review was prepared prior to the announcement in early 2024 of the full closure of Scottish Waters to Sandeel fishing. At the time of finalising this report (March 2024), the Scottish Government had recently announced a forthcoming closure of the Sandeel fishery throughout Scottish Waters (Scottish Government, 2024) . The UK Government had also made a similar announcement closing Sandeel fisheries in the UK portion of the southern North Sea (ICES Area 4) (UK Government, 2024) . Neither announcement was made with reference to strategic compensation for offshore wind. With respect to the use of the Sandeel fishery closure as compensation, the Scottish Government has stated^[6]: “Scottish Ministers will consider the suitability of any measures proposed as compensation as part of a case for derogating from the Habitats Regulations at the appropriate stage in the process, if and when such a derogation case may be required. This applies to any closure of the Sandeel fishery in Scottish waters which may be proposed as a compensation measure for offshore wind farm developments.” This report has not been fully updated to reflect these changes as the text was prepared prior to the announcement.

7.2.1 Summary

In this section, we present the practical feasibility of two conservation actions: ‘Sandeel fishery closure’ and ‘fishery closure or enhanced management of prey fisheries’. These two actions are grouped into one chapter due to them sharing the same key principles however, where relevant, we distinguish between the different foraging fish species and their associated fisheries. The focal species include Kittiwake, large gulls, auks, and Gannet.

Forage fish are small, schooling fish that link primary producers and higher trophic levels. ‘Prey fish’ is a more general term for fish consumed by any predator as a primary food source. Here, we specifically focus on the fisheries management of forage fish as opposed to prey fish, given the crucial importance of these highly abundant low trophic level fish as a primary food source for seabirds as well as for other marine predators. Mackerel was excluded from this section as there was less evidence for fisheries affecting their availability to seabirds. Therefore, this section focusses on the three forage fish species of most importance to seabirds in Scotland. Due to the challenges associated with studying seabird diets during the non-breeding season, when birds are at sea, migrating and away from breeding colonies, there is limited knowledge about their diets during this period. Therefore, in this section, we focus primarily on the breeding period.

There has been considerable interest on Sandeel fishery closures over the past several years, with the recent consultations in both England and Scotland both including new assessment on the ecosystem effects of Sandeel fisheries management in UK and Scottish waters, respectively. Internationally, the availability of forage fish, including Sandeel, has been identified as a driver of high breeding failure of surface-feeding seabirds in the North Sea (OSPAR, 2017).

Other prey fisheries management, however, have not received as much attention despite being subject of broader policy commitments. For example, the Scottish Government has fishery policy objectives for restricting or prohibiting the fishing of species that are ‘integral components of the marine food web’, including Sandeel (Scottish Government, 2020a). Similarly, EU and UK fishery delegations recently noted the ecological significance of forage fish species (namely Sandeel, Norway Pout, Herring, and Sprat) as important food sources for seabirds and other marine predators (DEFRA, 2023b). In this section, we provide a high-level summary of the information available on the topic. For more in-depth information, we recommend: the Scottish Sandeel fishing consultation, Engelhard et al. (2013), Heath et al. (2017), and Searle et al. (2023a).

Table 42. Summary of practical feasibility for closure and management of forage fisheries (Sandeel, Herring, and Sprat). Further detail is provided following the table.

This table can be viewed in Excel format.

7.2.2 Background

Sandeel, Herring, and Sprat are important forage fish in Scottish waters, playing a crucial role in the marine ecosystem by representing a large component of the diet of seabirds, marine mammals, and predatory fish. Fluctuations in the abundance (i.e. population size) and availability (i.e. individuals that can be caught by predators) of forage fish can have profound effects on the entire marine food web, influencing both ‘top-down’ (regulation of lower trophic levels) and ‘bottom-up’ (regulation on marine predators) dynamics (Engelhard et al., 2013). Consequently, they play a vital role in maintaining the health and balance of the ecosystem, and fluctuations in their populations can have significant ecological impacts.

Many piscivorous seabird species rely on specific forage fish for their breeding success and population dynamics (Cury et al., 2011; Tasker and Sydeman, 2023). However, due to the concurrent targeting of these forage fish by fisheries, a scenario of potential direct competition arises, where fisheries activities have the potential to reduce prey availability to seabirds. Consequently, the sustainable management, protection, and abundance of forage fish populations is critical for the conservation of seabird populations. However, the relationship between fish abundance and seabird demography is not a linear one and establishing a clear causal relationship between forage fisheries and seabirds is challenging (Figure 28).

Figure 28 illustrates the complex relationship between forage fish abundance, forage fish availability, fisheries, and seabird demography. As seen within the top blue box, the abundance of fish in the water, i.e. population size, is regulated not only by fisheries but also by a range of variables including predation, intra-specific competition, and environmental conditions. Seabirds exhibit diverse foraging strategies. Surface feeding seabirds, such as Kittiwake, can only forage on fish very close to the surface whereas other species such as Guillemot, Razorbill and Puffin can dive to considerable depths. Guillemot can potentially also extract fish from the sediment on the seafloor. Depending on the forage fish, seabirds may only take certain age classes of fish (e.g. older age classes of Herring are generally too large for most species), and often these are different to the age classes primarily targeted by fisheries. For example, most seabirds primarily take the 0-group Sandeel (Wanless et al., 2018), those that are less than a year old, while the fishery targets catches 1+ group Sandeel (Daunt et al., 2008), those that are at least one year. These differences mean that prey availability is not only defined by each seabird species-specific behaviours and fish abundance; but other factors such as fish location within the water column, seasonal and temporal fluctuations (e.g. Sandeel remain hidden in the sand during the night in summer).

While fisheries can significantly influence the total fish abundance within a designated area, the impact of fisheries on fish availability for seabirds is less straightforward. Quantifying the effect of a given fisheries management on seabird demography is therefore extremely challenging, marked by a high degree of uncertainty. Furthermore, lag effects between seabird demographic parameters and environmental conditions further compounds the intricacy and uncertainty of this effect.

7.2.2.1 Forage fish in Scotland

Sandeels (Ammodytes sp., predominantly A. marinus) are the most abundant species group of forage fish in the North Sea and are an important food source to UK seabirds, particularly Kittiwake and Puffin, during the breeding season (Furness and Tasker, 2000; Lewis et al., 2001b; Searle et al., 2023a). There is a seasonal shift in the availability of 0-group and 1+ group Sandeel, which corresponds to a shift from 1+ group Sandeel predominating in seabird diet early in the breeding season then 0-group later (Lewis et al., 2001b). Additionally there has been a long term dietary shift away from Sandeel as a whole and towards more 0-group than 1+ group Sandeel (Wanless et al., 2018).

Lesser Sandeel (A. marinus) are the main target for the largest single species fishery in the North Sea. In west Scotland, they have historically been primarily targeted for their oil and use as animal feed and fertiliser, but currently, no fishery is active nor are stock assessments undertaken. Sandeels have a unique life cycle, relying heavily on suitable sandy substrates. Sandeels lay eggs on sandy substrates in winter, hatch between February and April, and form diurnal pelagic feeding schools between May and June, during which time they bury themselves into the sand at night. They spend winters buried in the sand, emerging briefly for spawning in December and January. This life cycle means that Sandeel have high habitat specificity and are mostly resident, rarely travelling over 30 km from spawning areas. Despite management, most Sandeel stocks have suffered significant declines due to a combination of overfishing and climate change impacts. For more detailed information on their ecology, biology, and status within Scotland, refer to Scottish Government (2023d) and Marine Scotland (2023), and references within.

Atlantic Herring (Clupea harengus) populations rely on benthic habitats for reproduction. In Scotland, Herring reproduce in autumn or spring. Autumn-spawning Herring spawn near offshore banks in the North Sea and to the west of the Hebrides. Spring-spawning Herring, on the other hand, spawn in shallower nearshore areas along the west coast. Herring form dense shoals that migrate between feeding, spawning and wintering grounds, following similar patterns each year. They feed close to the surface at night and remain in deeper water during the day. Juveniles often shoal close inshore, while adults are found more offshore. Fishing for Herring in the North Sea and west of Scotland was banned by the Government in the late 1970s, to allow stocks to recover from collapse. Fishing in the North Sea resumed in 1983 after the stock recovered. After spending their first few years in coastal nurseries, two-year-old Herring move offshore into deeper waters, eventually joining the adult population in the feeding and spawning migrations to the western areas of the North Sea. In Scottish waters, two Herring stocks are assessed by the International Council for the Exploration of the Sea (ICES): the North Sea Autumn Spawning stock, fished in the northeast, and the West of Scotland autumn spawning stock, fished to the west of the Hebrides. The Clyde Herring stock is managed separately since it is only present in UK territorial waters. For further information refer to Frost and Diele (2022), Scottish Government (2022a) and Scottish Herring (2023), and references within.

Less is known about the European Sprat (Sprattus Sprattus). Throughout most of the year they are found at any depth in the water column, spread out over a wide area in local coastal waters, in relatively small shoals, grazing on zooplankton, and are too scattered to be economically targeted by large trawlers. In September and October, they begin to move inshore and assemble into large aggregations in preparation for spawning. All known spawning areas for Sprat are inshore. On the west coast of Scotland, juveniles tend to migrate into sea lochs in July and August. Migration tends to be towards inshore waters during winter though older fish may remain offshore. Sprat shoals move to surface waters at dusk. For further detailed information see ICES (2022) and Froese and Pauly (2023).

Table 43 describes the current understanding of the Scottish stocks and current fisheries management for each forage fish species. Overall, the status and knowledge base for most of the forage fish stocks around Scotland is lacking, especially in the west coast. Monitoring information is important to inform the sustainable management of the fish stocks. Note that the Total Allowable Catch (TAC) is advised by the International Council for the Exploration of the Sea (ICES) and may not necessarily represent the actual catch. The ICES-delimited subareas for management of Sandeel and Herring stocks in British waters and adjacent seas are shown in Figure 29A and Figure 29B, respectively.

The extent to which current ICES fisheries management advice for forage fish considers wider ecosystem considerations, especially of predator-prey interactions, was subject of a recent request to ICES from the EU and UK^[7]. The corresponding advice was recently issued by ICES (2023a). While the headline summary noted that current ICES advice for forage fish (e.g. (ICES, 2023c; b)) does consider ecosystem effects, with these considered via natural predation and qualitative ecosystem considerations, the detailed advice makes clear that ICES advice alone does not ensure there is sufficient biomass of forage fish for marine predators (including seabirds). With relevance to seabirds, the advice noted that advice is at the stock level, so does not consider individual foraging grounds (relevant to breeding seabirds with a restricted foraging range). It went on to note that there is a significant role for national regulations with respect to whether fisheries management is supporting ecosystem functions.

Table 43. Stocks and fisheries management of Sandeel, Herring, and Sprat in Scotland.

Includes assessment of those fisheries where there is potential for seabirds to benefit from changes in fisheries management (NB this is potential only and would require further investigation to understand the likelihood of any benefit).

This table can be viewed in Excel format.

7.2.3 Steps for implementation

The overall implementation of these conservation actions would need to be undertaken by government. The proponent could participate in steps 1 and 5.1.

1. Collation of baseline information on forage fish, fisheries, and seabirds:

To determine whether changes in fisheries management to benefit seabirds is appropriate for a given location, there must first be good baseline knowledge of the dynamics between forage fish, fisheries, and seabirds. A thorough scientific assessment, either through collation of existing information (e.g. ICES reports) or through new data collection, must be conducted to determine the following information:

1.1. Knowledge on the ecology and biomass of forage fish and the role and the impact of fisheries on these species.

• Forage fish data: population structure, seasonal and regional variations in prey distributions, life cycle, population size, and reproduction rates.
• Fisheries data: catch levels, fishery type, and bycatch.
• Analyses and predictions of the impact of fisheries on forage fish stocks, while accounting for other factors affecting these stocks (see Figure 28).
• Assessments of the potential impacts of different fisheries management options and environmental conditions (oceanographic patterns, temperature, etc) on stocks.

1.2. Understanding of the relationship between forage fish stocks and seabirds.

• Seabird diet: proportion of each age/group/species of forage fish consumed by each species and regional, seasonal, and temporal variation on the diets.
• Seabird foraging range and behaviour.
• Analyses of stock size variation and their effects on seabird demography (survival, productivity), and predictions on how environmental changes and fishery pressures will impact them.

1.3. Understanding of the relationship between fisheries and seabirds.

• Assessments of the potential impacts of different fisheries managements on forage fish stocks and availability for seabirds, and, how this translates to impacts on seabird demography.

2. Conduct formal and informal consultations:

2.1. Conduct both formal and informal consultation with key stakeholders, including fishing communities, international/EU fishing organisations (e.g. ICES), conservation organisations, and local authorities, to gather input and discuss the implications of potential management measures.

2.2. Governments could develop and put in place a public consultation to explore one or various management measures aimed at protecting forage fish stocks. The primary objective of the consultation will be to assess the feasibility of implementing different measures, evaluate their economic impacts, and consider public opinion. Examples of relevant recent public consultations are: 1) setting of the TAC for Herring in the Firth of Clyde (Scottish Government, 2023a) and the Scottish Government consultation on the closure of Sandeel fisheries within Scottish waters (Scottish Government, 2023d). Note that these consultations aim to improve fish stock to generate broader environmental and ecosystem benefits rather than solely focussing on seabird protection.

3. Decision on whether to proceed with conservation action (changes in fisheries management):

3.1. Evaluate and incorporate feedback from the consultations, along with any other available evidence. Consult with ICES for additional insight and expertise.

3.2. Government should then decide whether to continue to pursue action after weighing up the ecological benefits with the economic, political, and social costs. This will also include identifying whether fisheries management changes could constitute compensation or be required for other policy obligations.

3.3. Determine the fisheries management to be implemented.

4. Implementation:

4.1. The specific implementation approach will depend on the fishery management measure. All should adhere to standard regulatory and procedural processes followed when changing fishery management.

• Examples of potential considerations include defining geographic boundaries and duration of closures, implementing changes to fishing permits, licenses, or catch limits, and how any management relates to any other area-based conservation management (e.g. Marine Protected Areas (MPAs)).

4.2. Identify and clearly communicate the reasons, objectives, and anticipated outcomes of the proposed fisheries management.

4.3. Establish and conduct monitoring programmes.

• Monitor compliance (see ‘Other – Compliance’ within the Key considerations section): ensure enforcement mechanisms to guarantee compliance with regulations.
• Monitor to check effectiveness of the fisheries management on seabird demography (see monitoring).

5. Establish long-term adaptive management:

5.1. Continuously monitor the impacts of fisheries management on fish stocks, seabirds, and the ecosystem, and make necessary adjustments based on the best available scientific data and stakeholder and international input. Note that studies to monitor the effects of such management interventions will generally take many years/decades to provide clear conclusions and could limit the potential for adaptive management.

5.2. Invest in research and innovation to develop new technologies and practices that enhance sustainability and reduce the environmental impact of fisheries.

5.3. Develop and implement a communication and outreach plan to help build buy-in for actions taken.

7.2.4 Ecological effects of implementing action

There is extensive evidence that show that changes in forage fish abundance, availability, and quality affect the breeding success and survival of seabirds through impacts on foraging efficiency, diet, chick provisioning and parental care to varying extents (Oro and Furness, 2002; Cury et al., 2011; Carroll et al., 2017). There is also evidence that commercial fisheries can deplete forage fish stocks and, as direct competitors, represent an important threat to seabirds (Frederiksen et al., 2004; Jennings et al., 2012; Cook et al., 2014; Carroll et al., 2017; Lindegren et al., 2017; Dias et al., 2019).

The ability of seabirds to prey on forage fish depends on both the absolute numbers of the fish (abundance/stock) and the availability of the fish to seabirds (Figure 28). During the breeding season, seabirds are constrained in both the distance from nest sites that they can forage and the depth in the water column that they can reach, with both foraging range and dive depth varying greatly among species. Additionally, some species have predominantly or exclusively fish diets and differ considerably in their responses to reductions in food abundance compared to seabirds with more generalist diets (e.g. large gull species). Therefore, prey changes and management of forage fish fisheries will have different effects on each seabird species and the ecological context (e.g. presence of other marine predators). The most vulnerable species are, therefore, likely to benefit from measures that allow stocks of forage fish to recover from depletion caused by fishing on those stocks.

As the abundance and quality of Sandeel decreases, there has been marked community level changes in seabird diet composition, especially on auks and Kittiwake over the last three decades towards a higher consumption of Sprat and Herring (Wanless et al., 2018). Therefore, any fisheries management that could secure the stability and/or increase of Sandeel, Herring and Sprat stocks have the potential to benefit seabirds in terms of productivity and survival. Below, a more detailed review on ecological impacts per species is summarised.

7.2.4.1 Kittiwake

Sandeels serve as the primary food source for Kittiwake in the UK during the breeding season. However, Kittiwake also consume Sprat and Herring, particularly in years when Sandeels are less available or of lower quality. Kittiwake are surface feeders, making them particularly susceptible to changes in the vertical distribution and abundance of their prey. Given the ecological characteristics of Sandeels, they are only available to Kittiwake for part of their annual cycle, and their availability depends, to a certain extent, on external factors such as upwelling or predatory fish making Sandeels available. Within Scotland, Kittiwake are the seabird species most vulnerable to the depletion of their food sources by fisheries, especially Sandeel (Furness and Tasker, 2000; Heath et al., 2017). Several studies have demonstrated population-level effects, including decreases in survival and breeding success due to a reduction in Sandeel abundance (Frederiksen et al., 2004; Daunt et al., 2008; Carroll et al., 2017). In Orkney and Shetland, Kittiwake breeding success, and breeding numbers, decreased dramatically after the collapse of the Shetland Sandeel stock (Furness and Tasker, 2000; Wright, 2006).

As a result, increasing the stock of forage fish species through total fishery closures or other fisheries management measures, especially improving the abundance and quality of 0-group Sandeels, has the potential to enhance breeding success and overall survival of Kittiwake, or at least to prevent further declines. However, a recent study by Searle et al. (2023a) examining the long-term seabird demographic effects of the Wee Bankie Sandeel closure shows that while the closure has prevented further declines, and in some colonies led to increases in Kittiwake breeding success, it has not led to a full recovery to pre-fishery levels, despite twenty years of potential recovery time. During this same period there were significant declines in zooplankton (i.e. prey for Sandeel) biomass in the region associated with longer term environmental change (Olin et al., 2022) and ‘scientific fisheries’, which may, in part, explain the lack of a strong recovery in Sandeel. Therefore, the evidence that further Sandeel fishery closures or other fisheries management, will deliver significant population level benefits to Kittiwake is unclear and is expected to highly depend on the ecological context.

7.2.4.2 Puffin

Puffin heavily rely on the focal forage fish, especially 0-group Sandeel. The breeding success of Puffin has been shown to correlate with availability of forage fish (Martin, 1989). In Shetland, during years of low Sandeel abundance, breeding success was low, increasing in years with moderate Sandeel abundance (Furness and Tasker, 2000). Similarly, Fayet et al. (2021) suggests that the poor productivity at north-eastern Atlantic populations is driven by breeding adults being forced to forage far from the colony, presumably because of low prey availability near colonies. Therefore, Puffin would benefit from managements aimed at increasing the stock of forage fish, especially those in proximity to breeding colonies. However, it is worth noting that the Wee Bankie closure did not result in an increase in breeding success. In fact, there was a significant decline in breeding success following the fishery closure (Searle et al., 2023a).

7.2.4.3 Razorbill and Guillemot

Razorbill and Guillemot both exhibit pursuit diving behaviour, but there are distinct differences between them. Razorbill tend to make shallower dives, forage more on Sandeel and less on Sprat, and provision chicks with several smaller fish compared to Guillemot, which provision single fish to chicks (i.e. multi- vs. single prey loaders). Razorbill may be less dependent on fish and able to consume more zooplankton (Mitchell et al., 2004). While both auk species are susceptible to food shortages, Razorbill may be more vulnerable to reduced Sandeel abundance than Guillemot. This is due to the latter’s capacity to access Sandeel within the sand, providing an additional foraging advantage. Nonetheless, both Razorbills and Guillemots are notably more resilient than Kittiwake in the face of food scarcity (Furness and Tasker, 2000). Österblom et al. (2006) found that Sprat quality was a stronger determinant of Guillemot fledgling weight when compared to Sprat abundance. For Guillemot in the North Sea during winter, Sandeel form part of their diet (Sonntag and Hüppop, 2005), presumably foraged though benthic dives to extract burrowed Sandeel. Management measures securing forage fish stocks and fish quality are likely to increase breeding success and adult survival of auks.

7.2.4.4 Gannet

Gannet exhibit a high degree of flexibility in their foraging ranges (Woodward et al., 2019) and diet (in terms of prey species and sizes). Consequently, they are less susceptible to changes in abundance and distribution of individual forage fish species, being able to switch to other forage fish species or prey with no impact on their breeding success (Martin, 1989). However, Guillemette et al. (2018) demonstrated that the breeding success of Gannet begins to decline at approximately 8% of the maximum prey abundance. Similarly, Hamer et al. (2007) suggest that while adults have managed to maintain high reproductive success during years of low prey availability, they may struggle to do so in future years if providing sufficient food for chicks entails further increases in trip duration and/or foraging effort. Both studies demonstrate that even Gannet, a species considered to have low vulnerability to changes in forage fish abundance and distribution, may experience reduced breeding success and adult survival when certain conditions are reached, particularly a reduction of high-energy prey such as Mackerel. This vulnerability may be further exacerbated by factors like the ongoing HPAI outbreak.

7.2.4.5 Large gulls

Large gulls are generalist predators and opportunistic feeders, capable of exploiting a wider range of prey compared to most seabirds, including foraging on both marine and terrestrial habitats. Impacts stemming from the availability of forage fish during the breeding season can be mitigated by shifting to other available prey options (Furness and Tasker, 2000), and as such, there is low confidence that fisheries management will increase breeding success and/or survival of these species (Furness, 2021). Nevertheless, gulls also rely on larger fish species, such as Haddock and Whiting, which, in turn, feed on forage fish. As a result, there is a possibility that gulls benefit indirectly through increases in other prey species.

7.2.4.6 Benefits to other species and the wider ecosystem

Other seabird species with high foraging costs, limited diving ability (i.e., surface feeders) and restricted dietary flexibility, such as terns, Arctic Skua, Red-throated Diver and European Shag, are also likely to benefit in a similar way to the focal seabird species from any increases in forage fish resulting from changes to fisheries management; as has been shown for Common Tern and Sprat (Jennings et al., 2012).

Fisheries management focussed on increasing the biomass of forage fish is likely to yield ecosystem benefits by enhancing ecosystem resilience and reducing the risk of forage fish stock collapse, though may be unlikely to lead to full recovery unless the environmental conditions are suitable (e.g. Essington et al. (2015); (Lindegren et al., 2017)). These benefits extend to increasing the populations of other marine foragers, such as marine mammals (e.g. Harbour Porpoises) and other fish species (Heath et al., 2017), as well as of threatened and vulnerable species that rely on prey fish as a food source. Closures and/or implementation of sustainable fishing practices could lead to reduced mortality by decreasing bycatch of species like Whiting, Mackerel and other seabirds. The extent to which these benefits are realised will depend on factors such as the size of the spatial closure, whether industrial fishing activity is reduced or merely displaced, the time required for stocks to recover, and external factors such as the continued negative impacts of climate change.

7.2.4.7 Time-lags for population level responses

Population-level responses to fisheries management will manifest over the medium and long-term, making it increasingly challenging to disentangle the effects of the management from other variables. There is substantial uncertainty around the period over when benefits to seabirds from changes in managements would be measurable. This uncertainty is tied to the temporal gap between cause and effect, which is influenced by the life cycle of both forage fish and seabirds (age of sexual maturity ~4 to 6 years), combined with the uncertainty associated with the timing of any recovery in forage fish biomass alongside predicted impacts to survival from other pressures, including the operation of wind farms.

The impacts of climate change and other factors introduce additional complexities and uncertainties. These impacts are inherently difficult to predict and consequently, contribute to a reduction in certainty surrounding the expected outcomes of these management measures.

7.2.5 Research and monitoring recommendations

Given the complexity involved in determining the direct effects of fisheries management on seabird demography, several ecological components require monitoring to assess the effectiveness of the action. As changes in fisheries management for conservation purposes are ecosystem-based management approaches, effects will be observed over the long term and need to be understood in the context of any wider ecological changes. Therefore, continuous monitoring is necessary while the action is ongoing. Following Figure 28, several variables should be monitored.

7.2.5.1 Changes in fish abundance and impacts of fisheries

Before and during ongoing management, it is crucial to implement a data collection programme to monitor changes in forage fish abundance and fisheries.

Forage fish populations can be monitored through various methods including statistics on the quantities of each species landed at fishing ports, biological data on the size and age of the fish caught, and ecological and biological data of fish at sea. Such data can be obtained through a variety of techniques such as research trawls (controlled effort destructive sampling), surveys using imaging systems on robotic and autonomous underwater vehicles, hydroacoustic technology, proxies for abundance (e.g. catch per unit of effort), and electronic fish tags and genetic analysis, to understand habitat use, and fish movements.

Where there is not a full closure, the total fish removed from the stock by fisheries should also be monitored. A monitoring programme should be implemented to collect catch data and make this information available for analysis. Monitoring can take various forms, including monitoring of fish landings, logbooks (records from commercial fishers of their location, gear, and catch), electronic monitoring (with cameras), observers (scientists or trained individuals on fishery vessels observing fishing operations to collect data on catch levels, discards and bycatch), and unsystematic surveys (interviews and surveys to fishers).

Stock assessment models, such as surplus production models, virtual population analysis, and statistical catch-at-age models, can then be used to assess the status of exploited fish stocks (Fogarty and Siskey, 2019). These assessments help estimate critical stock characteristics such as biomass, spawning stock biomass (total weight of the reproductively mature individuals in the stock) and maximum sustainable yields (the largest average catch that can be continuously taken from a stock under existing environmental conditions). Currently, Marine Directorate scientists gather data on stocks in the North Sea and the west of Scotland annually. This information is then combined with data from other European nations that fish in these waters and is used by ICES to estimate the current state of the stock and predict possible future changes.

Most data on forage fish populations are collected for fisheries management purposes and, as such, will not necessarily provide detailed information on forage fish availability to seabirds. Therefore, enhanced monitoring that includes data collection aimed at understanding changes in prey availability to seabirds would be important to better link changes in fishery management to seabird foraging. This should include data collection during key periods (e.g. chick-rearing periods for seabirds). New technologies are making this more possible, e.g. by using remote surface and underwater monitoring techniques.

7.2.5.2 Changes in seabird demography and foraging ecology

To assess the impact of fisheries management on seabirds, monitoring should encompass a range of seabird demographic and foraging studies. This includes regular productivity observations, body condition assessments, capture-recapture survival rate studies. Additionally, bird tracking would enhance the understanding of foraging range and behaviour, while diet analysis through observations and sample collection will provide insights into forage fish consumption and dietary variations across species, regions, seasons, and years.

7.2.5.3 Other variables

As illustrated in Figure 28, fish abundance is regulated by several factors, all of which need monitoring. For example, advanced monitoring equipment integrated to traditional sampling gear can collect environmental information. Existing sampling data can also be used (or enhanced), e.g. continuous plankton recorder and Scottish Coastal Observatory. This information can then be used to undertake models to predict the impact of various climate change scenarios on fish stocks. Additionally, it will be essential to conduct simultaneous monitoring of other variables that may affect seabirds, including HPAI and the potential effects of offshore wind farms.

7.2.5.4 Statistical analyses to understand management impact on seabirds

To better understand the relationship between the quality and stock of forage fish and how this translates to seabird breeding success and survival, statistical analyses are crucial. Ecosystem and demographic models can be employed for this purpose. For instance, the use of Before-After-Control-Impact models is a valuable approach in evaluating the impact of management strategies on seabird populations. These models compare demographic data collected before and after the implementation of management measures, provided a baseline dataset exists. This approach enables researchers to isolate the direct effects of management on seabirds while accounting for other environmental variables.

Examples of valuable studies to be replicated include: Heath and Speirs (2011) and Searle et al. (2023a). Heath et al. (2017) highlight essential research priorities that should be addressed. Some ongoing studies may address some knowledge gaps (e.g. the OWEC project PrePARED and the EcoWind projects Pelagio and EcoWings).

7.2.6 Key considerations, potential barriers, and potential solutions

This table can be viewed in Excel format.

7.2.7 Species-specific aspects of implementation

Because this is a fisheries management measure, there are no seabird species-specific aspects regarding its implementation, other than the associated monitoring efforts. All seabird species are likely to benefit from fisheries management measures aimed at increasing forage fish abundance, especially Sandeel, as these are highly attached to particular sand banks and have limited dispersal and movements.

7.2.8 Overall conclusion

Forage fish play a vital role in marine ecosystems, supporting various predators, including seabirds, marine mammals, and larger commercially valuable fish. These fish face multiple pressures, such as fishing, habitat modification, climate change impacts on food resources, and natural predation. Therefore, fisheries management, including closures, though practically feasible, may be best viewed as ecosystem-based measures for wider conservation benefits, so may be less suitable options for compensation focussed on specific feature species impacted by additional pressure from offshore wind (aka. like-for-like compensation). Even with detailed monitoring, it may not be possible to confidently determine the impact of fisheries management on seabird demography, which could compromise the ability to confidently quantify any benefits of changes in fishery management to seabird populations.

Where full fishery closures are not feasible, alternative fisheries management options to improve the sustainability of the fishery such as adjusting fishery quotas (e.g. set-aside policies for forage fish that account for predator needs and areas closed to fishing), imposing limits on time at sea, implementing closures during critical periods (e.g. seabird foraging periods and fish spawning), and restrictions on fishing gear, should be considered.

When designing any changes in fisheries management, particularly as a compensatory measure, then this must be developed in an integrated way. This will require an inter-disciplinary approach between scientists with expertise in benthic processes (e.g. to account for impacts of offshore wind development on fish habitat), plankton and fish ecology, seabird ecology, and wider marine ecosystem changes (from climate change). Of particular importance when seeking to predict future changes is in understanding directional changes in ecosystems, e.g. changes in the plankton communities that support forage fish, to be able to adequately account for both bottom-up (e.g. plankton abundance) and top-down (e.g. natural predation and fisheries) processes. The additional effects (opportunities and risks) from renewable energy development on oceanographic processes and forage fish will also need to be considered. Coordination between different Government departments and directorates will also be critical.

7.3 Practical feasibility: End of the Gannet harvest at Sula Sgeir

7.3.1 Summary

Table 44. Summary of practical feasibility for end of the Gannet harvest at Sula Sgeir. Further detail is provided following the table.

This table can be viewed in Excel format.

7.3.2 Background

Sula Sgeir is a small, uninhabited rocky islet located 18 km west of North Rona (59°5’43.44" N, - 6°9’ 22.6188"W), in Northwest Scotland. To date, the Sula Sgeir Gannet population is subject to a licensed annual harvest of 2,000 full-grown chicks, approximately 17% of the colony’s annual chick production (Wanless et al., 2015). This quota is set by the Scottish Government, with advice from NatureScot, and is revised and approved on a yearly basis. The record of the number of harvested full-grown chicks in recent years is shown in Table 45 and, since 2011, has ranged between 1,723 and 2,000, excluding years where harvest did not occur.

The harvest is restricted to accessible areas and occurs over a two-week period between late August and early September, towards the end of the breeding season. During this period, most Gannet chicks have reached an advanced stage of development, making them less susceptible to incidental additional mortality caused by the harvest (i.e. mortality to additional chicks beyond those harvested due to disturbance). Nonetheless, the risk of premature fledging or the inability to return to the nests after the disturbance remains and has not been quantified. Similarly, there is limited evidence regarding the broader impacts of this practice on the overall bird and fauna community within the site (e.g. impacts resulting from disturbance caused during catching, culling, and bird preparation, and due to general human presence).

Table 45. Gannet chicks harvested at Sula Sgeir between 2011 and 2023.

Data from NatureScot (reported by Berwick Bank (2023a)).

This table can be viewed in Excel format.

Of note, in years when Government advice has been issued due to human health concerns (such as in years 2022 and 2023 due to HPAI), the community itself has taken the decision not to proceed with the harvest.

7.3.3 Steps for implementation

1. Pre-decision to proceed:

1.1. Undertake a consultation with members of the Ness community to discuss proposals for ending the harvest.

• The consultation could be targeted to the whole community, to key people of the community, or could be exclusive to those who will be directly affected by the ending of the harvest.

1.2. Initiate negotiations aimed at achieving a mutually beneficial agreement. Given the loss of cultural heritage to the community, the proponent should consider providing compensation through:

• Provision of some kind of community asset (e.g. community centre) or providing improvement to existing community assets (e.g. improvements to harbour infrastructure, improving school facilities, etc). The specific compensation arrangement would require discussion with the community. This would likely require some ongoing commitment for the maintenance of such assets. Alternatively, the proponent could pay into a community benefit fund^[8] to be administered by the community, either on a one-off or annual basis.
• Financially support alternative opportunities that could support other Gaelic community traditions (e.g. promoting and preserving traditional Gaelic arts, music, and language through cultural programmes and events, supporting local Gaelic festivals and exhibitions, providing resources for educational programmes aimed at passing down other traditions to younger generations), thus partially offsetting the loss of cultural heritage from ending the Gannet harvest. Note that each cultural activity has its own value and significance and, therefore, cannot be directly substituted.
• Engage in negotiation and consider a compromise. If discontinuing the harvest is likely to lead to conflict, it may be more effective to consider alternative options such as reducing the harvest quota or adjusting the frequency of the harvest. These alternatives will need community agreement.

2. Decision on whether to proceed:

2.1. Proponent, along with the Scottish Government, should decide whether to continue to pursue action after weighing up the ecological benefits with the financial costs and social impacts.

2.2. If the decision is to permanently stop the harvest, then consider changing the Statutory Order within the Protection of Birds Act that allows this exemption.

3. Implementation:

3.1. A legal agreement would be drafted and signed between the proponent and the Ness community. This would include a summary of the action to be taken (i.e. no further harvesting and thus no licence renewals) together with any conditions agreed (e.g. a compensation scheme for the community financed by the proponent). The agreement (or summary of) should then be shared with relevant authorities (likely to include NatureScot and Scottish Government).

3.2. Put in place any compensation package agreed with the community (step 1.2).

4. Post-implementation:

4.1. Undertake Gannet population monitoring (see monitoring section).

4.2. Proponent to schedule regular meetings with members from the Ness community to:

• Re-evaluate and modify compensation packages, as necessary.
• Present updates on the impact of the harvest cessation on the Gannet population, including findings from ongoing monitoring efforts.

4.3. If stopping the harvest is not yielding positive ecological results, consider alternative compensation options.

7.3.4 Ecological effects of implementing action

7.3.4.1 Gannet

As discussed in detail in WP1, harvesting eggs and/or chicks of any species undoubtedly affects seabird populations. Therefore, ending the annual harvest would increase the annual productivity by ~2,000 chicks per year and would most likely increase the population growth rate and population size of certain Gannet colonies, especially at Sula Sgeir itself but also to some extent surrounding colonies.

Trinder (2016) estimated that the harvest has reduced the Sula Sgeir population growth rate below the level that would be predicted in the absence of the harvest. The author estimated that if the harvest were to end completely, the median annual population growth rate would increase from ~1.023 (i.e. an annual population increase of 2.3% per year) to ~1.035 and to ~1.029 if the harvest quota were reduced to 1,000 chicks. Similarly, results from Wanless et al. (2005b) show that the population growth rate at Sula Sgeir from 1969 to 2004 was significantly lower compared to those from other UK Gannet colonies when accounting for population size, suggesting that an end to the harvest would lead to a population increase.

The annual harvest might also be impacting neighbouring populations that are linked through immigration and emigration, such as those from St Kilda and Sule Stack, Sula Skerry and Flannan Isles. Trinder (2016) estimated that to maintain the current population growth rate, the Sula Sgeir population requires ~270 breeding age recruits each year, likely sourced from colonies in Northwest Scotland. Consequently, the Sula Sgeir population could act as a sink for emigrant Gannet from other colonies, and discontinuing the harvest might lead to increased population growth in those colonies as well. However, it is important to consider density-dependence processes. There is a negative relationship between colony growth rate and colony size, and as such, as the population increases and breeding spaces become limited, the population growth is expected to decrease. Such a relationship between colony size and population growth rates has been demonstrated for Gannet in Britain and Ireland (Lewis et al., 2001a), though the level of density dependence varies depending on environmental conditions (Davies et al., 2013).

Given the increasing population trends observed in the Sula Sgeir Gannet population, along with most other Scottish populations (Figure 30; information prior to the 2021/2022 HPAI outbreak), the harvest itself seems to exert a limited impact on the overall population size and has been considered to be a sustainable practice. Although discontinuing the licensed harvest would undoubtedly increase the productivity and accelerate the population growth rate of the Sula Sgeir colony as well as, to a lesser extent, those of neighbouring populations, its impact at the UK scale is likely minimal. Furthermore, should younger generations within the Ness community decide to cease or reduce their harvesting activities over time on their own accord, the overall degree of benefit derived from this action diminishes accordingly.

7.3.4.2 Other species

Ending the harvest might provide limited benefits to other species present on the islet. The Northern Fulmar, Leach’s Petrel and European Storm Petrel are the only other seabird species remaining at the site during the time of the harvest. As these species are also at late breeding stages, a reduction in disturbance could potentially lead to a slight increase in productivity. Most of the other seabird species have completed breeding at time of harvest (late August/early September) and therefore would not benefit. There is potential for negative impacts on some species were competition for nesting space to increase.

7.3.4.3 Time-lags for population level responses

For Gannet, the effects of this action will become evident within a short to medium timeframe. An increase in annual productivity will be seen effective immediately, as full-grown chicks will no longer be harvested. Survival and population growth, on the other hand, will become apparent as unharvested chicks reach sexual maturity, typically at around five years. As the colonies approach their carrying capacity and reach a ‘saturation’ point where nesting sites are no longer available, the rate of population growth will decline. For other species, the time lag for population level benefits will be longer.

7.3.5 Research and monitoring recommendations

7.3.5.1 Pre-implementation of action

Jeglinski et al. (2023) recently developed a model to understand the metapopulation dynamics and role of density dependent processes in Gannet colonies across the Northeast Atlantic, utilising over a century of census data, including data from Sula Sgeir. The study demonstrated that there were complex regional and broader metapopulation dynamics, including regional and local density-dependent processes, regulating population sizes. The study focussed on the development of a metapopulation model and its general findings with respect to population dynamics in Gannet. As such, it did not, in detail, discuss what the modelling revealed about the strength of different population processes operating for individual colonies or regions. Further development of this model, for instance by incorporating data on actual harvest levels (the current study only considers the presence/absence of harvests), would be a good first approach to understand how individuals from the Sula Sgeir colony are linked to regional and global colonies driven by density-dependence processes and natal dispersal.

Below we provide monitoring recommendations should this action be implemented. Nevertheless, we would advise that such monitoring plans are refined by further development of the model proposed by Jeglinski et al. (2023). Specifically, we suggest conducting a sensitivity analysis to identify which empirical data would best reduce uncertainty in model outputs, thus enhancing our understanding of the effect of changes to harvest practices at Sula Sgeir.

7.3.5.2 Post-implementation of action

Conducting population counts within the Sula Sgeir colony, as well as neighbouring colonies, is a good approach for monitoring population growth and size and assessing the effect of discontinuing the harvest on these colonies. Due to the challenges arising from researching large and remote gannetries, such as those along the NW coast, aerial surveys have emerged as the most common technique for the census of Gannet. This method involves capturing high-resolution photographs from an aircraft and subsequently counting the Apparently Occupied Sites (AOS) (see Wanless et al. (2015) for detailed aerial survey based count methods). However, Sula Sgeir’s uneven terrain poses a challenge for population counts. Given its topography, counts based on aerial surveys are likely to be associated with significant errors. Depending on the expected change in the rate of population increase following the ending of the harvest, smaller population changes may not be detectable with this survey method. Furthermore, to gather more comprehensive data, such as quantifying the number of fledged chicks, juveniles, and immatures across the entire islet or in a representative sample, alternative methods like land, boat, and drone surveys could be more effective, either individually or in combination. These alternative surveys, however, may potentially cause additional disturbances, so the methods used would need careful consideration.

National Gannet surveys are undertaken separately to the main national seabird colony survey programmes (e.g. the recent Seabirds Count or Seabird 2000), and have typically been carried out at ca. ten-year intervals over the past few decades. To effectively monitor Sula Sgeir and ideally nearby colonies, implementing a programme of more frequent surveys during the initial period before and following cessation of the harvest at Sula Sgeir would be appropriate. Conducting the initial surveys, especially within the next few years, is highly important for establishing a robust baseline given likely population level changes from HPAI impacts. The survey frequency should be informed by a sensitivity analysis (see above) to determine what frequency of surveys would be most appropriate, however we would anticipate that relatively frequent surveys would be most useful (e.g. every third year or even annual for the first several years). More frequent surveys will provide thorough detail on both changes in population size and population growth rates over time, while longer survey intervals would only provide coarse data on changes in the rates of population change. Reporting requirements should correspond to the agreed data collection frequency. The initial reporting would focus on documenting the data collection process and, after the first few years, detailed analyses at agreed intervals would be required.

Before-After-Control-Impact (BACI) models are another valuable tool for evaluating the impact of harvests on seabird populations by comparing data collected before and after ending the harvest, provided a baseline exists (i.e. surveys undertaken during harvest years). This approach allows researchers to discern the direct effects of the harvest on seabirds while controlling for other environmental variables.

To enable effective comparisons across colonies, surveys should be conducted during similar time periods. Such comparisons are essential for understanding changes in population sizes and metapopulation processes, such as immigration and emigration rates. Whenever possible, comparisons of population trends and growth rates with those of other Scottish colonies will help understand whether the Sula Sgeir population is exhibiting similar behaviours when accounting for population size and potential impacts of climate change and/or HPAI. While fully disentangling mortality rates caused by climate change, HPAI and harvest can be challenging, the impact of the harvest can be quantified by contrasting the Sula Sgeir population rate with those from neighbouring colonies, where climate change and HPAI are expected to have similar effects. With the recent development of a metapopulation model for Gannet (Jeglinski et al., 2023), and potential for remote monitoring, the effects of these different pressures on individual Gannet colonies should be more possible to disentangle than was formerly the case using single colony closed population models.

Long-term ringing and resighting studies are also an important method of estimating annual survival rates to improve population models. This method involves the systematic ringing and subsequent recapture or resighting of individuals over several years. Annual survival estimates can be calculated using statistical models like the Cormack-Jolly-Seber (CJS) model. Nevertheless, implementing a long-term ringing-resighting scheme on Sula Sgeir may not be possible or advisable, given its challenging terrain and logistical limitations, so effort may be best allocated to more frequent population surveys.

7.3.6 Relevant SPA and SSSI Site Conservation and Management Advice

Sula Sgeir has protection both as an SPA and SSSI (as part of the North Rona and Sula Sgeir SPA/SSSI). The SSSI site management statement notes that the Gannet harvest is a “traditional right that has been exercised by Ness people for at least several hundred years” (SNH, 2010). Under SSSI site management objectives it is noted that the current harvest intensity (2,000 chicks per year) does not appear to be leading to population decline and it is concluded to, therefore, be compatible with maintaining the population in favourable condition (though notes that were the population to subsequently decline then the quota would be reassessed). Full Conservation and Management Advice has not yet been published for the SPA, however, Conservation Objectives are available. The Conservation Objectives are relatively generic (i.e. in common with other seabird breeding colony SPA sites) including e.g. maintaining, in the long term, the ‘Population of the species as a viable component of the site’. As such, the SPA Conservation Objectives do not specifically consider the Gannet harvest.

7.3.7 Key considerations, potential barriers, and potential solutions

This table can be viewed in Excel format.

7.3.8 Species-specific aspects of implementation

Ending the harvest is solely focussed on Gannet, therefore, this section does not apply for this action.

7.3.9 Overall conclusion

Discontinuing the controlled harvest is expected to accelerate the population growth rate of the Sula Sgeir colony and, to a lesser extent, that of neighbouring colonies. The impact at the UK scale is likely to be minimal due to the distance and small levels of population exchange between Sula Sgeir and more distant colonies, however, the metapopulation process is complex and remains unclear. Though practically feasible and likely to provide ecological benefits, this action has several risks. Ending the harvest would impact on local community cultural heritage and traditions that have existed for centuries so could lead to community opposition with associated political risks in pursuing the action. Therefore, the success and implementation of this action relies on the response and engagement of the local community.

There is a lack of evidence regarding the broader impacts of this practice on the overall bird and fauna community within the site so we may have underestimated the benefit of implementing this action to other species. Planning for the implementation of this action, including associated monitoring, would need to account for HPAI, both in terms of how this impacts the Gannet population, and how it impacts the harvest practice (given that no harvest occurred during 2022 nor 2023 due to HPAI).

7.4 Practical feasibility: Mammalian predator eradication and/or management

7.4.1 Summary

This section focusses on predator eradication, mostly focussed on rodents and on islands, and predator control or exclusion which focusses on larger mammals and at mainland colonies and islands. The focal seabird species identified for this conservation action are auks (Common Guillemot, Puffin, Razorbill) and large gulls.

There has been substantial work done in the UK in recent years on predator eradication on islands and associated biosecurity, in this account we provide a high-level summary of key information relevant for using this conservation action as a strategic compensatory measure. We also provide signposting to key sources for further information.

Table 46. Summary of practical feasibility for mammalian predator eradication, control, or exclusion.

Further detail is provided following the table.

This table can be viewed in Excel format.

7.4.2 Background

This action includes different approaches to eliminating or minimising predation by mammalian predators at seabird colonies. The types of management intervention considered are:

• Eradication: Aims to fully remove a predator from an island for permanent reduction in predation (assuming no re-invasion).
• Predator control: In contrast to eradication, control seeks to reduce predation to low levels by reducing predator density (usually for island sites), typically through ongoing predator control (e.g. trapping/poisoning).
• Predator exclusion: Reduced predation by excluding predators from nesting sites, usually by use of predator exclusion fencing around the nesting habitat for ground-nesting species on mainland sites.
• Biosecurity: This has three key components, minimising the risk of (re-)invasion by predators (e.g. rats being carried by boats), regular monitoring (e.g. monthly) for early detection of incursion, and a rapid incursion response to allow prompt and efficient removal before predator establishes, thus avoiding a more costly full-scale eradication attempt. Here, this is considered as an integrative measure to secure the benefits following eradication and/or control measures. Biosecurity can also operate as a stand-alone conservation action to secure existing predator-free status for island seabird colonies, however we do not consider this in detail here^[9].

In WP1, we concluded that there was strong evidence supporting predator eradication as an effective conservation tool for island breeding seabird populations. However, the size of benefit and the appropriateness of predator eradication is influenced by many factors (inter alia type predator present, nesting strategy of target seabird species, distance to mainland/other islands). This review found that eradication schemes were generally most successful for small mammals (e.g. rodent species such as Brown Rat) while control could be more appropriate for medium mammals (e.g. American Mink). Effective schemes for eradication and control of invasive species is an ongoing and complex process that demands continuous monitoring, preventive measures, and sustained resource investment (Holmes et al., 2023).

Predator eradication or control and associated biosecurity are non-targeted site-based measures which typically benefit multiple bird and non-bird species for a given site. Eradications take considerable planning and will not always be practically feasible for a given location or may be unsuccessful. Eradications typically take several years, once the planning phase is included, requiring significant funding (millions £s per site) and ongoing biosecurity. As such, if using as a compensatory measure there are benefits to strategic delivery, which could deliver a programme of eradication and associated biosecurity across multiple sites. While control approaches are most suitable for sites where it is impossible to fully remove (e.g. mainland sites) or reduce re-invasion probability to low levels (e.g. island sites separated from mainland by small sea distances). In the short-term control measures will often be cheaper to implement than full eradication, but the long-term costs will typically be higher than eradication (and associated long-term biosecurity).

The history of eradications of invasive mammalian predators on islands in the UK is summarised by Thomas et al. (2017a). The key focus of island eradication schemes on seabird breeding islands has been for removal of Brown Rat and, to a lesser extent, other species (e.g. Black Rat and Stoat). The key seabird species that have been the target beneficiary species are burrow-nesting seabirds (i.e. Manx Shearwater, European Storm Petrel, and Atlantic Puffin). Thomas et al. (2017a) identified that there had been at least 12 eradication attempts for rodents on islands in the UK between the 1960s and 2017, with many successful though not all (and some sites required multiple attempts). Predator exclusion through anti-predator fencing is a standard measure for protection from predation for ground-nesting species (especially waders, terns, and gulls) on mainland sites (White and Hirons, 2019; Babcock and Booth, 2020). Its benefit has been demonstrated for terns (Babcock and Booth, 2020) and gulls (Dalrymple, 2023). The method has been used especially for wader species, including in Scotland^[10].

7.4.3 Steps for implementation

The following steps focus primarily on predator eradication on islands; the steps for predator control and predator exclusion fencing will be similar, with key differences noted at the end of the section.

1. Pre-implementation:

1.1. Long-list sites: Identify potential sites and the species likely to benefit for each site.

• Undertake a desk-based study to identify suitable island sites for predator eradication or control and use a scoring approach to prioritise amongst these (for detail see Pre-implementation monitoring: predator control or eradication on islands).
• For predator exclusion, a more targeted long-listing approach would be appropriate. This would focus on mainland and some island sites where predation by medium-sized mammals (especially Fox) have been identified as a threat to ground nesting species (principally terns and gulls).

1.2. Determine coordination and delivery approach.

• There are various options for how a strategic island eradication programme could be coordinated and delivered. It could potentially build on existing initiatives (e.g. the Seabird Island Biosecurity Programme (SIBP), the Scottish Biodiversity Strategy and Delivery Plan, Biosecurity for Scotland’s Seabird Islands) and minimally would at least need to coordinate with these initiatives. However, most existing initiatives are focussed on biosecurity rather than eradication.

1.3. For long-listed sites, undertake a detailed feasibility study (predator eradication).

• Below are set out key factors to consider when assessing the feasibility of predator eradication for long-listed sites. A detailed guide on undertaking site-specific feasibility studies (focussed on eradication of rodents but broadly relevant to all mammalian predator eradications) is available in Section 3 of the UK Rodent Eradication Best Practice Toolkit (Thomas et al., 2017b). Table 1 therein sets out seven key criteria to consider which are focussed on practical aspects:
  • Technical feasibility – the logistical feasibility of trapping, catching, or killing all of the target species at a site;
  • Sustainability – whether the risk of re-invasion can be reduced to an acceptable level;
  • Socially acceptable – there is full support from landowners and managers, the local community, and other island users;
  • Politically and legally acceptable – that all required permits and consents can be obtained (e.g. for use of second generation anticoagulant rodenticides);
  • Environmentally acceptable – any impacts on the environment can be reduced to a low and acceptable level;
  • Capacity – that all required resources (including trained personnel, and equipment) can be sourced within the project timeframes;
  • Affordable – that the total project costs in the long-term (i.e. including both the initial eradication and ongoing biosecurity) with contingency (for e.g. if the eradication takes longer than anticipated/re-invasion occurs) can be funded.
• There must also be an assessment of whether a proposed eradication is likely to lead to the desired outcome (i.e. recovery/increase in the target beneficiary seabird species).
  • Gain a good understanding of the potential for the seabird species of interest to benefit from eradication of predators at the target site/s. This should include understanding the vulnerability of the seabird species to predation by the mammalian predator/s present, including any site-specific factors (e.g. accessibility to the predators of occupied and unoccupied potential nesting habitat). To assess the potential scale of benefit, a detailed habitat survey should be conducted, using this to map and quantify the extent of currently occupied and unoccupied habitat. The current population status of the seabird species and any historical knowledge on this (especially where there are data from pre-invasion) are important context. For a significant population level response to a successful eradication attempt, the population must not be significantly limited by other factors (e.g. food availability), therefore having a good understanding of the species ecology is important.
  • Undertake predator surveys at the site (see Research and monitoring recommendations), and any nearby areas (e.g. neighbouring islands) from where re-invasion could occur. Using these surveys to confirm the presence of predators and potentially the presence of previously unrecorded predator species at a site.
  • Eradication feasibility: For an eradication programme to be followed it must be realistically possible to remove the whole population of the invasive predator, if this is not possible then eradication is not appropriate (in such cases predator control may be a more appropriate alternative). A key consideration will be whether there are multiple invasive mammalian predators or a single predator species. Where multiple predator species are present it will often be advisable to eradicate multiple species at the same time to avoid the risk of the remaining predator/s producing greater impacts post-eradication of a competitor predator (i.e. predatory release/ or reduced interference competition).
  • Biosecurity feasibility: This must be undertaken pre-decision to undertake an eradication attempt at a site. While the acceptable level of risk of predator re-establishment will vary (e.g. in proportion to the size of potential benefit and resources available to commit in event of predator incursion), it is crucial to ensure that this risk can be reduced to an acceptable level. The pathways by which invasive mammals could return to an island should be determined, including by vessels (e.g. recreational vessels, cargo ships, and ferries) and via predators swimming considering predator specific swimming abilities (e.g. based on models of predator distribution, for Brown Rat see Tabak et al. (2015)).

1.4. Short-list sites: Prioritise sites based on results from feasibility study.

• Produce short-list of sites, initially removing those not deemed to be feasible then re-prioritising using updated information gained from the above preparatory work. Where the feasibility of eradication is assessed to be low, but the benefit of reduced predation assessed to be significant then it may be appropriate to have an alternative objective of predator control (i.e. seeking to minimise the abundance of the predator).

2. Implementation of management intervention:

2.1. For selected site/s produce detailed eradication design and biosecurity plans (see Annex 4: Biosecurity and Incursion Response in Thomas et al. (2017b)), assemble resources required for eradication (staff and equipment), and obtain all permits and licences required.

2.2. Initiate pre-eradication monitoring (see Research and monitoring recommendations).

2.3. Initiate eradication scheme.

2.4. Initiate biosecurity. This will include producing detailed biosecurity plans which should be developed and implemented working with the land manager/s and community. Ongoing surveillance for mammalian predators should be established, including training personnel and potentially training a specialised biosecurity dog. The users of a site (e.g. any inhabitant community and local businesses) should be trained in biosecurity to understand how they can minimise the risk of introducing mammalian predators. An emergency response plan must be developed so that this can be deployed in the event of the mammalian predator returning, or when a high risk of this occurring is identified (e.g. a ship wreck). Preventative measures should be established along key incursion pathways. For full details of planning biosecurity, see Annex 4 of Thomas et al. (2017b) which provides a guide to planning and mounting an incursion response (this is developed for rodents, so while broadly relevant to other invasive mammals there will be additional considerations to those included).

3. Post-implementation:

3.1. Undertake ongoing monitoring to confirm absence of eradicated predator (see Post-eradication monitoring).

3.2. Maintain ongoing biosecurity with routine surveillance. Review and update biosecurity plans annually and ensure that equipment is maintained and replaced as required both for monitoring and of incursion equipment stored at any associated rapid response hub. This will be most sustainable if part of a wider regional or national programme (e.g. building on the current Biosecurity for Scotland’s Seabird Islands project).

3.3. Monitor seabird response (by e.g. population counts and productivity monitoring) post eradication (see Post-eradication monitoring).

7.4.3.1 For predator control and predator exclusion:

The steps to take for predator control on islands will follow the same principles as above which is primarily focussed on predator eradication. A key difference for predator control is that there is a less clear distinction between the implementation and post-implementation phases, as control of predators must continue indefinitely. However, control would typically start with a higher effort period to reduce predator abundance, then subsequently lower the effort to maintain the low predator abundance.

Predator exclusion is generally a lower scale initiative, this would still require similar preparation in terms of understanding which sites may most benefit and in establishing the likely level of ecological benefit. Once these sites have been identified, the action would likely be most effectively managed on a site-by-site basis (i.e. rather than as part of a wider strategic programme). Guidance on establishing predator exclusion fencing is available, which would need to be considered if implementing (White and Hirons, 2019; Babcock and Booth, 2020). Guidance is included on the preparatory work required to decide whether a predator exclusion fence may be appropriate for a site, on the design of the fencing, and on planning for their installation (White and Hirons, 2019). This should be considered in conjunction with guidance written specifically on predator exclusion fencing for ground-nesting seabirds, which is available for tern species (Babcock and Booth, 2020), and case-studies, such as (Dalrymple, 2023) which reports on the use of predator exclusion fencing for the benefit of ground-nesting gulls (Herring Gull, Lesser Black-backed Gull, and Great Black-backed Gull) in North-west England.

7.4.4 Ecological effects of implementing management interventions

Mammalian predators impact seabirds by predating on eggs or chicks and in some cases adult birds; therefore, management to reduce predation can lead to increased productivity and survival. These predators most impact species nesting on more accessible terrain, thus ground and burrow nesters are most sensitive to predation. Species that nest among boulders or on low-lying accessible rocky shores (e.g. Razorbill at some sites) can also be sensitive. Species that primarily breed on steep rocky cliffs will rarely have significant benefit (including Common Guillemot and Kittiwake). In WP1, we produced an overview of the evidence for different seabird species benefitting from control or eradication of various mammalian predator species. Below we draw on that review and some additional sources to summarise the ecological effects of these measures for the focal seabird species and briefly other seabird species.

7.4.4.1 Ground-nesting colonial species

Large gulls (including Herring Gull, Lesser Black-backed Gull, and Great Black-backed Gull): As ground nesting species often breeding on mainland coastal sites (including adjacent to/or on more accessible parts of coastal cliffs) these species are vulnerable to mammalian land predators. On island sites these species should generally benefit from mammalian predator eradication and/or control. On mainland sites, depending on the topography and terrain, predator exclusion fencing may be suitable, which has been shown to reduce predation by Foxes increasing productivity in all three large gull species for a breeding site in England (Dalrymple, 2023).

Terns are vulnerable to predation by ground predators, so can benefit from any measures to reduce ground predators including eradication (for island nesting sites) and predator exclusion fencing (reviewed in Babcock and Booth (2020).

7.4.4.2 Auks

Guillemot and Razorbill: Both species nest on rocky cliffs, however Guillemot are largely confined to steep rocky cliffs but do nest among boulders or on flat open ground in some areas (e.g. on tops of sea stacks), while Razorbill generally nest in slightly more accessible areas including scree around the top of cliffs and boulder-fields. As such both species are relatively protected at most sites to predation by most mammalian ground predators. However, Razorbill will typically be more vulnerable than Guillemot, so stand to potentially benefit from island eradications at some sites.

Puffin, as a burrow nesting species, are highly vulnerable to ground predators, particularly rodents. On Lundy Island in the Bristol Channel both Brown Rat and Black Rat were eradicated in the early 2000s, subsequently Puffin returned to breed on the island (Lock, 2006) with 1335 Puffin reported breeding in 2023 (InsideEcology, 2023). However, at Ailsa Craig in the Firth of Clyde, where an eradication programme was completed in the earlier 1990s, Puffin only slowly established with 186 individuals recorded in 2015 (Thomas et al., 2017a). While at Cardigan island, Ceredigion in Wales, where eradication occurred in the 1960s, Puffin have never re-established (Thomas et al., 2017a). This demonstrates the significant potential for Puffin to benefit from eradication of mammalian predators but the scale or such benefit is likely to be highly site-specific.

7.4.4.3 Other species

The most cited species for benefits from island predator eradications are burrow nesting seabirds, including Manx Shearwater and European Storm Petrel. In addition to seabirds, wader species are often the focal beneficial species for predator eradication or control programmes. As a broad site-based measure it is expected that for many sites where implemented additional species will benefit beyond the target species, however which species and to what extent will be highly site-specific. There may also be wider ecosystem benefits.

7.4.4.4 Time-lags for population level responses

In the case where predation targets seabird chicks and the applied management successfully reduces disturbance and chick predation rates, an immediate increase in annual productivity can be anticipated. If predation targets adults and the management successfully reduces adult predation rates, an increase in adult survival rates is expected within a short period. Population recovery will be dependent on the wider ecological context, including whether there is a source of new recruits (relying on the health of the species across the wider region), and availability of prey. Reestablishment is most uncertain, and it can take decades for a species to return once absent, if at all.

7.4.5 Research and monitoring recommendations

7.4.5.1 Pre-implementation monitoring: predator control or eradication on islands

Produce long-list of candidate sites for predator eradication/control using desk-based study. Several previous studies have sought to identify and prioritise islands for predator eradication in the UK (Ratcliffe et al., 2009; Stanbury et al., 2017; Mitchell et al., 2018) which can be used as a starting point.

The study of Stanbury et al. (2017) is the most complete and comprehensive study covering nearly all offshore islands in the UK and considering the presence of mammalian predators and the potential conservation benefits of removing these. For taking a strategic approach to the selection of sites for predator eradication (or control), a useful approach would be to build on the original analysis and associated islands database included as supplementary material within the study. Key components to update or revise could include: 1) updating population numbers (Seabird 2000 was used for the original study) using the Seabirds Count census results (published November 2023) and the recent additional colony counts undertaken during 2023 in response to the HPAI outbreaks. 2) including information on site protections (SPA/SSSI and associated qualifying features), useful around understanding current site management context (relevant e.g. to additionality issues). 3) potential to refine the prioritisation criteria and formulas used (see next point).

Develop and agree prioritisation criteria. For strategic compensation, it may be relevant to produce separate prioritisations for key species and/or develop an aggregate score including multiple species. Stanbury et al. (2017) and Ratcliffe et al. (2009) developed similar prioritisation scoring approaches which could be further developed for the purpose of strategic compensation.

Produce initial long-list of candidate sites based on desk-based assessment of feasibility, level of benefit (including to the target species) prioritised according to the agreed prioritisation criteria.

7.4.5.2 Pre-eradication monitoring

For the seabird species of interest, suitable indicators should be identified (see section 5.4 in Thomas et al. (2017b)). Baseline data should be collected pre-eradication to provide a point of comparison. For demographic rates (e.g. productivity and annual survival rates), that can display high inter-annual variability, data should ideally be collected for multiple years pre-eradication. For most sites there will currently be recent population counts available following the Seabirds Count census, however as this data gets older (or where significant changes are anticipated – e.g. for species or sites impacted by the HPAI outbreak) repeating colony counts to establish a firm baseline will be required.

The abundance and distribution of the target predators should be established pre-eradication to provide a baseline against which progress can be tracked for the adopted predator management action. The general approach here will be as for post-eradication predator monitoring.

7.4.5.3 Post-eradication monitoring

Undertake ongoing monitoring to confirm the absence of the eradicated predators including surveillance focussed especially on high-risk areas (e.g. vicinity of settlements and entry points). The methods used will vary depending on the predators being monitored for. Methods for rodent detection are set out by Thomas et al. (2017b) (see Annex 3 therein), with many of these methods suitable for multiple predator species. Standard practice is to undertake an intensive monitoring check two years post-eradication to confirm the absence of predators, this would include monitoring throughout a site including in inaccessible areas (e.g. where rope access is required). Monitoring for the presence of predators then becomes part of the routine surveillance as part of the biosecurity plan.

Monitor the target seabird species. This should include regular population counts to track the changes in population over time post-eradication. Depending on site accessibility, practicality, and species ecology the intervals between population counts would vary (e.g. species that are slower to establish/recruit to a site could be monitored at reduced frequency). For larger sites where whole colony counts are impractical to undertake frequently a stratified sampling/indicator area approach could be taken. Monitoring could also include measuring key demographic rates expected to change post-eradication (e.g. adult survival and productivity) using standard approaches depending on the species of focus.

7.4.5.4 Predator control and predator exclusion monitoring

For predator control the monitoring requirements will be largely the same as for predator eradication. As predators are not eliminated, but rather maintained at low densities, the monitoring for predators will have a different purpose, but the same monitoring methods could be used. For predator control, monitoring would be used to assess the abundance of predators and reactively adjust control methods (e.g. trapping) to avoid predator populations increasing.

For predator exclusion the monitoring approach for the target benefitting species (i.e. the seabird species) would take the same general approach. There will also be some specific requirements around checking that the exclusion fencing is intact and fully functioning (and repairing/adjusting as required), thus monitoring of the fencing would be integrated with maintenance (see chapter 10 in: White and Hirons (2019)).

7.4.6 Relevant SPA-specific information

7.4.6.1 Biosecurity

For the UK Marine Strategy Assessment 2018 (Mitchell et al., 2018) it was assessed that most island SPAs with seabird features lacked adequate biosecurity measures (only 20% assessed to be sufficient). However, following the Biosecurity for Life project and other initiatives it is expected that when re-assed (due in 2024), over 90% of SPAs will be deemed to have adequate biosecurity. The Biosecurity for LIFE project was a 5-year, EU LIFE funded initiative, started in 2018 to put robust and sustainable biosecurity measures in place for 42 island SPAs across the UK with breeding seabirds as designated feature species. This was a partnership project, which with respect to Scotland, included RSPB and the National Trust for Scotland. Its main outputs are:

• Biosecurity plans: working with island land managers and communities to develop and implement biosecurity plans.
• Biosecurity surveillance: supporting biosecurity surveillance for invasive predators by training personnel, deploying surveillance equipment, and training a specialised biosecurity dog (to detect invasive Brown Rats).
• Rapid response hubs: developing a network of regional hubs equipped to rapidly respond in the event of an incursion on any of the 42 island SPAs.
• Industry training: providing training to marine industries and businesses on how to undertake and implement effective biosecurity when operating in areas of risk.

In Scotland, Biosecurity for LIFE has been continued by the Biosecurity for Scotland’s Seabird Islands project funded by the Scottish Government’s Nature Restoration Fund up until 2026. Rapid response hubs with associated volunteer networks have been established in four locations (Edinburgh, Stornoway, Orkney and Shetland) providing coverage for all of Scotland.

7.4.6.2 Sites for predator eradication

Stanbury et al. (2017) produced an assessment for all identified UK islands (including crown dependencies) on the presence of invasive mammals and reptiles together with the presence of species of conservation interest. This was a desk-based review, so relied on various sources, so for any given site the information should be ground-truthed. They ranked islands based on the feasibility of eradication and the level of conservation benefit (this was not restricted to seabird species). Many of the top islands identified are SPAs for breeding seabirds located in Scotland (see Table 3 therein). This is a useful starting point for considering sites that could benefit, but some of the information used is now out-of-date (e.g. on seabird abundance) and the prioritisation approach used was not designed with strategic compensation in mind (see Pre-implementation monitoring: predator control or eradication on islands).

7.4.6.3 Information on specific locations (selected cases only)

Rum SPA: There is concern about the potential impacts of Brown Rat on the Manx Shearwater population on Rum. The current management plan for Rum SPA, which is a National Nature Reserve managed by NatureScot, includes a project under objective NH4.3 to: “Undertake surveillance of impacts and activity of brown rats where they come in contact with priority species and habitats” (NatureScot, 2023h). Previous studies on Rum have suggested that Brown Rat are impacting Shearwaters but the level of impact may be low due to the location of the Shearwater colonies (inland on higher altitude parts of the island) (Lambert et al., 2015). Also present on Rum are Wood Mice, which a recent study suggests may be impacting the Shearwaters (Lambert et al., 2021). For Rum there is a need to better understand the impacts of both the Brown Rat and Wood Mice on the Manx Shearwater and the interactions between these two mammals to inform any potential future control or eradication programme.

Shiant Isles SPA: The Shiant Islands Seabird Recovery Project, started in late 2014, successfully removed Black Rat with the island declared rat free in March 2018 (Main et al., 2019).

Handa SPA: In the late 1990s, Brown Rat were eradicated from Handa Island (Stoneman and Zonfrillo, 2005). This led to an initial recovery in seabird populations. However, subsequently Brown Rat re-established, potentially naturally from the mainland (which is ca. 300 m distant). Berwick Bank (2023b) examined the feasibility of re-eradicating rat from the site.

Canna and Sanday SPA: Between 2005 and 2006 Brown Rats were eradicated from Canna and Sanday which are two neighbouring islands that connect at low tide. This has led to a recovering in many of the seabirds nesting on the islands, including Puffin breeding on the main islands again (previously being restricted to offshore stacks), however some other species did not show obvious benefits including Common Guillemot and Kittiwake (Luxmoore et al., 2019).

Orkney Islands: Stoats established on the Orkney Islands relatively recently, being first recorded in 2010, since then becoming widespread on Mainland Orkney and connected islands and were considered a threat to the bird life (especially of ground nesting species) (Fraser et al., 2015). There is currently an eradication scheme ongoing through the Orkney Native Wildlife Project, though at the time of writing this had identified a significant funding shortfall, so its continuation is not guaranteed.

Further locations are mentioned in relation to specific mammalian predators below (see Predator-specific aspects).

7.4.7 Key considerations, potential barriers, and potential solutions

This table can be viewed in Excel format.

7.4.8 Predator-specific aspects of implementation

For offshore islands in the UK (and its Crown Dependencies), the known or suspected predator status is reviewed in Stanbury et al. (2017). This was completed based on published and reported information and included some simplifying assumptions (e.g. assuming certain invasive species were present on islands with larger resident human populations) rather than field visits and the status may have changed since that study was completed. Therefore, this information should be confirmed when considering potential sites for eradications. On the mainland there are both native (e.g. Fox) and non-native (e.g. Mink) mammalian predators present with the approaches used for control or management of these consequently differing (e.g. exclusion versus control). Below we provide some background on predator management each focal predator species.

7.4.8.1 Rodents

Rodents, including rats and mice species, are the most widespread invasive non-native species present on UK islands and have been the main targets of island eradication schemes in the UK for the benefit of seabirds (Thomas et al., 2017a) and globally (Howald et al., 2007), with rats considered the predators most threating seabird species worldwide (Dias et al., 2019). Therefore, there is extensive experience of eradicating rodents with best practice methods developed (see Thomas et al. (2017b)).

Brown Rat (also known as Norwegian Rat) are the most common target of mammalian predator eradication schemes for benefit of seabirds in the UK (Thomas et al., 2017a). They are present on many islands around the UK including SPA sites (e.g. Rum, see Relevant SPA-specific information above).

Black Rat, where present, have potential to be a greater threat to seabirds (in terms of the range of species impacted) than Brown Rat due to their greater agility, potentially being able to access some cliff nesting habitat. On UK islands they have a highly restricted distribution. They were present previously on Lundy and the Shiant Islands but have been successfully eradicated from those sites. For seabird breeding island sites in Scotland, they are now only understood to be present on Inchcolm (Firth of Forth) where their presence is confirmed^[11].

Mice species are considered threats to seabirds globally, particularly to albatross, petrel, and shearwater species (Spatz et al., 2023a). Mice have not been the focus of eradications in the UK^[12], however their impacts may have been previously underestimated with recent studies demonstrating potential for impacts on St Kilda (Bicknell et al., 2009; Bicknell et al., 2020) and Rum (Lambert et al., 2021). It will not always be appropriate or desirable to remove these species where long established (e.g. on St Kilda there is the endemic St Kilda Field Mouse, Apodemus sylvaticus hirtensis, thought to have been introduced by the Vikings centuries ago) but the presence of mice and how this could affect the success or otherwise of removal of other species should be considered (see e.g. Lambert et al. (2021)).

7.4.8.2 Medium-sized mammals

American Mink is a medium sized mammal which, though more easily detectable than rodent species, have strong swimming abilities meaning that they can cross relatively wide expanses of water (>2 km), given the difficulty in preventing reestablishment, control measures rather than full eradication are often used. Two large scale mink control or eradication projects have been conducted in Scotland in recent years, the Hebridean Mink Project and the Mink Control Project. The Hebridean Mink Project aimed to reduce Mink numbers to low numbers (and locally eradicate) amongst the Outer Hebrides running from 2001 in a phased project, with by 2018 only a few (low single figures) individuals caught that year. The Mink Control Project targets Mink on the Scottish mainland across northern Scotland, this has been running from 2018 trapping Mink across a wide area. This is a long-term project, part of the wider The Scottish Invasive Species Initiative, which is currently funded for 2023-2026 by a grant from the Scottish Government’s Nature Restoration Fund and in-kind funding by project partners and volunteers.

Fox are a native predator, however they can negatively impact ground nesting birds on mainland sites, particularly waders, terns, and gulls. Predator exclusion fencing has been used to exclude Foxes from nesting areas for gulls and terns (Babcock and Booth, 2020; Dalrymple, 2023).

Stoats are present on the Orkney Islands and subject to an eradication programme (see Information on specific locations (selected cases only)).

Other mammalian predator species are present on some islands in Scotland (see Stanbury et al. (2017)) including feral cats and hedgehog. The latter species is subject of a removal programme on the Uists, primarily for the benefit of wader species.

Non-invasive native mammals can also predate on seabirds, for example Otter have been recorded predating on Black Guillemot and European Storm Petrel (Bolton et al., 2017; Johnston et al., 2020). To fully understand predation impacts it is important to consider both invasive and non-invasive predators, however it would rarely ever be appropriate to control native non-invasive predators.

7.4.9 Key references

This account has given a high-level summary of key issues to consider when developing a mammalian predator eradication, control, or exclusion programme for the purposes of strategic compensation. As well as the references cited within the text, we recommend considering the resources summarised in Table 47, which provide greater detail on various aspects of this action.

Table 47. Key resources for further information on mammalian predator eradication or control and associated biosecurity.

This table can be viewed in Excel format.

7.4.10 Overall conclusion

Predator eradication on islands would benefit from strategic delivery through a coordinated programme and as such has potential as strategic compensation. Strategic delivery would better allow for sites for eradication to be prioritised appropriately. Predator eradication is only a viable measure where biosecurity can be put in place in the long term to secure predator-free status. Predator control may be a suitable alternative in some circumstances, where eradication is not realistic, or where there is a high probability of re-establishment (e.g. islands close to mainland/other islands where predators are present). However, predator control can be financially prohibitive in the long term due to the need for ongoing control (though eradication also has ongoing costs associated with biosecurity).

Predator exclusion, particularly on mainland sites will generally be most suitable as a site-specific measure for the benefit of ground nesting species, predominantly gulls and terns. The number of sites where this may be suitable for seabirds in Scotland is unclear but is likely to be limited. As such this measure has more promise for project-specific compensation rather than for strategic level delivery.

Key issues remaining include the status of the measure with respect to additionality, particularly for protected sites (SPAs and potentially also SSSIs), and its potential to deliver for the species most impacted by offshore wind (aka. like-for-like compensation) with cliff nesting species least likely to benefit. Given that the measure is broad-based (particularly predator eradication), with wider conservation benefits for sites where implemented, there may be potential to develop this measure as a wider strategic conservation programme, with benefits apportioned between compensation and wider biodiversity net gain or nature restoration targets, depending on the species benefitting for a given site.

7.5 Practical feasibility: Avian predator management

7.5.1 Summary

This conservation action, the management of avian predators (i.e. raptors, gulls, and crows), focusses on a set of different types of management interventions, each of which will be described separately. The management interventions are diversionary feeding (DF), removal techniques (REM), and deterrence (DET). The focal seabird species identified for this conservation action are auks and large gulls.

Table 48. Summary of practical feasibility for avian predator management.

Further detail is provided following the table.

This table can be viewed in Excel format.

7.5.2 Background

Avian predators can exert significant top-down effects on seabird populations (Perkins et al., 2018; Anker‐Nilssen et al., 2023), either directly, by taking adults, eggs and chicks, or indirectly, through disturbance or other means such as kleptoparasitism, when an individual steals food or prey from other individuals (Finney et al., 2001; Sanz-Aguilar et al., 2009; Perkins et al., 2018). Among these predators, raptors, corvids, and large seabirds like gulls and skuas, have been observed preying on seabirds at Scottish colonies (Votier et al., 2004; Perkins et al., 2018; Langlois Lopez et al., 2023). In WP1, we explored the effect of various avian predator management interventions on seabirds and waterbirds to identify the specific predation reduction actions that had most potential as strategic compensation for benefitting auks and large gulls (the focal species).

The most commonly employed management interventions included: a) diversionary feeding, a non-lethal, temporary method aimed at providing predators with alternative prey items to divert them from predating on seabirds (DF), b) the targeted or non-targeted removal of predators’ nests, eggs, chicks and/or breeding individuals (REM), and c) the use of deterrent objects, including physical and acoustic means (DET). It was clear throughout WP1, that the effectiveness of each management intervention is contingent upon multiple factors, such as the avian predator involved, the nesting ecology of the target seabird species, site-specific conditions, and the chosen management intervention. Additionally, the review in WP1 highlighted a notable scarcity of evidence concerning the applications of these management interventions on the focal species, and very little evidence testing such management in Scotland or the wider UK.

7.5.3 Steps for implementation

Each management intervention possesses inherent advantages and disadvantages that require careful consideration during the planning and implementation stages and should always be tailored to the specific seabird species, avian predator in question and overall context. Therefore, the implementation steps are here outlined at a high level only (and may not apply to all situations). Further detail on the monitoring related elements is discussed in the monitoring section.

1. Collation of baseline information and field studies including field trials:

1.1. Review knowledge on impacts of avian predators on Scotland’s seabird populations generally.

1.2. Identify which predators have the most significant impact on which seabird species and at which breeding colonies.

1.3. Identify which predator management intervention/s may be possible to apply in order to reduce impacts from predators and identify candidate field trial locations.

1.4. Gather comprehensive ecological, breeding, and behavioural data from both predator and seabird species.

1.5. For all management interventions, except potentially predator removal, it would not be possible to proceed directly to delivering as a compensatory measure, with research and trials first required (see monitoring section for specific recommendations).

• Design and conduct preliminary field trials to test methods and assess the efficacy of the different management interventions. Note that the duration of the trials may vary depending on the management intervention, potentially requiring several years to determine suitability in some cases.
  • (DF) Trials focussing on DF should conduct experiments to compare DF delivery methods (e.g. feeding on ground, feeding platforms, nest feeding) to determine the most effective approach.
  • (DET) Trials researching the impact of deterrence should aim to identify which deterrent mean or group of deterrent means would be most efficient for each specific predator while also noting their effects on the focal seabird species (i.e. avoiding disturbance).

2. Decision on suitability as a compensatory measure and, if suitable, determining which management intervention to implement:

2.1. Based on the information collated in step 1, determine the management intervention with the most potential for the specific species and site.

3. Detailed design of management:

3.1. Undertake site visits to explore terrain and overall site and colony characteristics to identify whether the management could be implemented at the site and towards those species.

• Determine the land ownership status by identifying whether sites are privately-owned, designated as a reserve, public land, etc.
• Use of maps or field visits to comprehend the terrain’s characteristics.
• Gain detailed knowledge on the target avian predator, including its breeding ecology, behaviour, feeding habits, and diet preferences. (DET) For deterrence, also obtain information on previously used deterrent methods and whether these were efficient and to what degree.
• Research the breeding ecology of focal seabird species within the site, including information on laying, hatching, peak chick provisioning period, and fledgling dates.
• Understand the ecological dynamics between seabird and predator by gathering data on e.g. current predation rates, peak predation timing (in terms of date and time of day), behaviour, and seabird colony responses to the predator.
• Research the conservation status and relevant protections of target species.

3.2. Confirm the feasibility of implementing the management at the given location/s and target species.

4. Pre-implementation of management:

4.1. Prepare for management implementation.

• Obtain necessary permits for land use and species-related actions.
  • Where the target site is an SPA colony, then an HRA will be required.
• Offer training to personnel who will be responsible for implementing management.
• Install necessary infrastructure and secure required materials and equipment:
  • (DF) Erect feeding poles/stations (locations will be determined based on baseline knowledge), construct, mount and/or identify hides or observational points, install surveillance cameras and identify where and how food for diversionary feeding is going to be stored and disposed of.
  • (DET) If applicable, install deterrent objects.
• Identify individual predators with higher predation rates and/or locate predator nests/pairs.
  • Conduct searches to locate predator nests near or around the seabird colonies by monitoring displays, nest-building, prey transportation, and food-passing interactions behaviours.
  • The identification of individual predators with higher predation rates on seabirds could be done by conducting behavioural observations and quantifying successful predation rates and/or by inspecting the food items present in pellets located within/in the vicinity of the predators’ nests.
• Management-specific considerations:
  • (DF) Identify when and what food type and quantities are going to be fed to predators. Secure the source and quality of diversionary food.
  • (DF) Decide whether all predators or just a subset of them should be targeted for DF.
  • (DET) The specific implementation approach will depend on the deterrent used. If necessary, install deterrent objects, such as speakers, canes, scarecrows, etc prior to the start of the seabird breeding season, so as to avoid unnecessary disturbance.

5. Implementation of management:

5.1. (DF) For detailed information on methods, see Redpath et al. (2001), Ludwig et al. (2018), Smart and Amar (2018) and Mason et al. (2021).

• If predators are breeding, provide DF from post-hatching until predator fledgelings have left area or until all seabirds have fledged. Note that the frequency of which the DF will be provided will depend on the predator’s dietary requirements. For example, smaller items within DF may need replacing at least daily while larger items may likely be available for longer.
• If predators are not breeding, provide DF during the peak of seabirds’ breeding season.
• Ensure that predators are feeding on diversionary food.
• Throughout the season, ensure that amount of food is accordant to brood sizes/ages and/or number of predator individuals using the DF.
• If applicable, dispose of uneaten food daily.

5.2. (REM) For examples of potential removal techniques see Finney et al. (2001) and Sanz-Aguilar et al. (2009).

• Search potential nests and constantly monitor predators throughout the seabirds’ breeding season.
• Destroy, remove and/or replace, eggs, chicks and/or adults, as required. Potential removal techniques are egg pricking and substitution, destruction of empty active nests, trapping, shooting or poisoning adults.

5.3. (DET) Example of study using bamboo canes as a deterrence: Boothby et al. (2019)

• Some deterrents will require active intervention, such as operating speakers, lasers, ensuring that canes are set correctly, etc. Others, once in place will not require any actions until removal, although regular checks will still be necessary.

5.4. Undertake prey and predator observations to monitor the success of management, e.g. whether predator’s rates reduce, and if prey productivity, survival, abundance and recruitment increase (see monitoring).

6. Post-implementation of management:

6.1. Evaluate and reflect on what worked and/or did not work to enhance efficiency in subsequent years.

• Note whether individuals got habituated to management (e.g. deterrent object) over time. If so, reassess management programme.

6.2. Measure effectiveness of management.

• Perform appropriate data analyses to understand the impacts of the management on the productivity of both prey and predator, as well as its effect on other wildlife.

7.5.4 Ecological effects of implementing management interventions

The management interventions considered have rarely been trialled at seabird colonies. Consequently, there is a high degree of uncertainty regarding the potential and degree of ecological benefits of this conservation action as a whole, especially on the focal species of this review. Any effects, if present, are likely to be highly site- and context-specific and we are, therefore, unable to provide detailed information in this section.

7.5.4.1 Ground-nesting colonial species

Ground-nesting colonial species, such as terns and waders, are the species for which there is strongest evidence for benefits from general avian predator management. Studies by Smart and Amar (2018) and Mason et al. (2021) demonstrated the efficacy of DF of raptors in reducing predation rates and doubling the productivity of Little Terns and Northern Lapwings, respectively. DF of certain raptors, such as Kestrels and Red Kites, can be a highly effective predation management tool that enhances productivity of ground-nesting colonial species that could consequently lead to increased population growth rates. As ground-nesters, large gull species may potentially benefit from DF through reduced nest predation, though the extent to which gull colonies could benefit in Scotland is unclear. Ground-nesters are also highly likely to benefit from predator removal and deterrent techniques. Deterrents like lasers and bamboo canes, are most likely to work for the protection of these species.

7.5.4.2 Auks

While no direct evidence of the use of the management interventions on focal species was found, if raptors were to prey on an auk colony and the management interventions were successful in reducing disturbance, predation and/or kleptoparasitism, avian predation management could immediately increase productivity and/or adult survival rates.

7.5.4.3 Other seabird species

Given that the benefit to seabirds strongly depends on the successful implementation and response (e.g. disturbance and predation rate reduction) of the avian predator management interventions, seabirds regulated by top-down effects from avian predators, such as Black-legged Kittiwake (Oro and Furness, 2002; Anker‐Nilssen et al., 2023) and Northern Fulmar (Swann, 2002; Reid et al., 2023), should theoretically show some degree of benefit.

7.5.4.4 Predators

Target predators are likely to experience adverse effects from removal techniques, particularly those involving lethal methods, as well as from deterrent efforts. These management interventions have the potential to reduce their survival rates and population size, and/or force individuals to relocate to other sites. On the other hand, non-target predators may benefit from compensatory responses. Additionally, DF may attract other predator individuals, either from the same species or different ones, owing to the increased food availability and ease of access to it. While this benefit could lead to higher population density and improved survival rates, regulation may be necessary if it leads to an increase in seabird predation, which would undermine the purpose of the management.

7.5.4.5 Time-lags for population level responses

In the case where predation targets seabird chicks and the applied management successfully reduces disturbance and chick predation rates, an immediate increase in annual productivity can be anticipated. If predation targets adults and the management successfully reduces adult predation rates, an increase in adult survival rates is expected within a short period. Distribution expansion and population growth are expected in the short to medium term, depending on the continued success and implementation of the management. The extent of these effects, however, will depend on the effectiveness of the management intervention.

7.5.5 Research and monitoring recommendations

7.5.5.1 Baseline information required to decide which management intervention is suitable

Establishing a baseline understanding of the predator-prey dynamic before implementing any type of avian predator management is crucial. The initial step involves identifying and quantifying the impact of predators, both in terms of predation and disturbance, on key seabird species and colonies as well as understanding their life history and survival traits. In some cases, this information may also be available in existing literature. In most cases, however, this information would need to be collected through field studies, which can be obtained through traditional field observations, dietary studies, survival analysis and/or installing cameras at nests. Subsequently, this information could be employed to calculate overall predation levels, such as the number of predated adult Puffins per year or the number of predated Guillemot eggs and/or chicks per year. Predation levels life history and survival data can then be used to conduct population viability analyses, following a similar approach to Langlois Lopez et al. (2023), to model the population-level impacts of avian predators on seabirds.

Experimental trials will be essential for most management interventions. These trials serve a dual purpose: first, to measure the management’s effectiveness in terms of predation reduction and seabird benefit, and second, to refine and test methods. The latter will involve determining correct quantities (e.g. the density of bamboo canes at nests or how many nests should be removed and which ones), the timing (i.e. when management is most effective), the frequency (i.e. how often a certain action needs to be conducted), and, in the case of deterrence, the time of predator response to deterrent and effectiveness of combining different approaches. All experiments should adhere to well-established experimental protocols; therefore, experiments should be well-designed, carefully controlled and appropriately analysed, with the aim of providing reliable and meaningful insights into the effectiveness of the management.

7.5.5.2 Ecological information required prior to implementing management intervention

Once the population-levels impacts are understood, and the decision has been made to proceed with the management intervention at a specific site, detailed biological data becomes crucial. This data encompasses population size, productivity rates, predation rates and predation behaviour of both predator and seabird. This information can be gathered through fieldwork in the years preceding the initiation of the management. This may involve regular observations from hides and/or vantage points throughout the breeding season, through cameras, or, if available, through existing long-term datasets. For certain management interventions it will be necessary to locate nests or potential nesting sites of predators and, whenever possible, identify individuals with higher predation rates prior to implementing the management. Colour-ringing individuals could facilitate the identification of specific individuals for this purpose. If available, it is important to obtain information on previously used management interventions at the site and their efficiency concerning both predators and seabirds, either by reviewing existing literature, reaching out to potential sites that might have implemented such actions, or conducting experiments before the official management implementation.

7.5.5.3 During implementation of the management interventions

Monitoring should take place while the management is ongoing. While there will be specific details to evaluate for each management interventions, the primary objective is to monitor predation attempts (the frequency of predator attacks) and success rates (the proportion of successful attempts) through in-situ observations at various points during the seabirds’ breeding season, with particular focus during the peak of breeding or when chicks and adults are most vulnerable.

For diversionary feeding: Observations of the predator's food intake, provisioning rates, predation rates, and productivity are crucial aspects to understanding the effectiveness of the management. These observations should account for potential variations in diurnal hunting patterns and seasonality and can be conducted through direct observations (from hides or vantage points), cameras (either miniature cameras at nest or at feeding platform), by collecting and analysing regurgitated pellets, or by weighing the food that was not consumed. The key aspects at this stage are to confirm whether birds are indeed consuming the provided food and to quantify their intake of the focal seabird species. Some studies, for example, have dyed the diversionary food with a distinct colour to aid in distinguishing it from natural prey items during observations.

For removal: Conduct constant monitoring efforts around the seabird colony to gather evidence of predation. This includes monitoring direct predation rates, localising active nests and analysing the content of pellets found within nests. Additionally, monitor overall predator population trends to provide a clear understanding of the likely impact of removal on the predator population.

For deterrence: To detect habituation, the responsiveness of predators to the deterrent object should be monitored over time. Observations should quantify differences in reaction time, proportion of times individuals are deterred, intervals between predation attempts, and successful predation attempts.

Overall, to understand the impact of the management intervention on the focal seabird species monitoring should encompass aspects like productivity, behaviour, population counts, and predation rates and attempts. Surveys can be conducted in person, throughout the breeding season, and variables such as the number and growth stage of chicks should be monitored to calculate the annual productivity. Subsequently, employ statistical analyses to quantify the effects of the management intervention on predation rates and seabird productivity either by comparing years and/or sites with versus without management and, if applicable, differences between treatments. Observation should also focus on recording any unforeseen impacts on other wildlife, such as the presence of other non-target species and how these are impacted.

7.5.6 Key considerations, potential issues, and potential solutions

The parenthesis before each consideration indicates the management intervention it is applicable to: DF = diversionary feeding, REM = removal, and DET = deterrence. If not stated, then the point is applicable to all management interventions.

This table can be viewed in Excel format.

7.5.7 Species-specific aspects of implementation

Overall, how successful DF would be for the focal seabird species remains uncertain. However, targeting nests with DF is more likely to reduce raptor predation, as it minimises scavenging and non-target predator consumption. Deploying DF more broadly to target various predators decreases its effectiveness, as extensive scavenging by non-target predators and gulls at gull colonies could deplete the food source too rapidly to influence the target predators. This, however, would require field trials to confirm.

7.5.7.1 Large gulls

Including Herring Gull, Lesser Black-backed Gull, and Great Black-backed Gull. Note: gulls to benefit through predator management.

DF, REM and DET could benefit gulls in certain circumstances, but has not been proven. Results will depend on the response of the avian predators. Notably, gulls may be attracted by the DF which could prove beneficial in terms of food availability (though this could reflect an incorrect DF programme). REM should be conducted in locations where it would have minimal impact on the targeted gull species. Physical and acoustic DET are unlikely to be suitable, as gulls are likely to react to them. However, they may be effective if they do not provoke a response.

7.5.7.2 Puffin

DF requires a stable terrain and may prove difficult or impractical to implement close to burrows. Puffins, known for their sensitivity to disturbance, require DF to be conducted at a distance from the colony, which diminishes its effectiveness as gulls often engage in kleptoparasitism directly in front of burrows. Therefore, burrow-nesting auks are less likely to benefit from DF. Puffin could, however, benefit from REM and DET efforts, if performed away from the burrows and they do not respond negatively to them.

7.5.7.3 Guillemot and Razorbill

Guillemots and Razorbills are cliff-nesters within large colonies. This means that, for DF to prove successful, it would need to occur on the outskirts of the colony, potentially reducing its effectiveness depending on the predator’s location. If effective, REM could decrease disturbance and predation rates at Guillemot and Razorbill colonies. DET objects could be easily placed at a stable terrain close to the colonies. However, it may be challenging to protect individuals from sea-based predation.

7.5.8 Overall conclusion

While all the management interventions (diversionary feeding, removal, and deterrence) included within this conservation action (avian predator management) show potential, further research and trials are necessary to determine their suitability for use as strategic compensation. Given the limited evidence and knowledge available on the topic, removal techniques appear to be the most ecologically effective, followed by diversionary feeding and deterrence. However, potential conflicts may arise during implementation, particularly if the predators themselves are birds of conservation concern (e.g. gulls, raptors), making predator removal inappropriate in many cases despite their ecological feasibility. Therefore, the practical feasibility of management interventions, especially those involving lethal measures for avian predators, requires careful consideration, rendering avian predator management particularly challenging as a strategic compensatory measure.

We have sought to outline the necessary steps that would be required to implement this conservation action as compensation, adding important information for each management intervention. However, given the significant level of uncertainty surrounding their efficacy and specific implementation, primarily due to the lack of specific research, these steps will require refinement as evidence is collected and field trials are conducted.

7.6 Practical feasibility: Reduction of disturbance (at colony)

7.6.1 Summary

During the breeding season, seabirds, especially those located and/or nesting in accessible locations, are exposed to human disturbance occurring from water, land, and air (through Unmanned Aerial Vehicles; hereafter UAVs). Therefore, management measures to reduce disturbance at seabird colonies should be tailored to the source. When applicable, each source is described separately within this document. The focal seabird species identified for this conservation action are auks and large gulls.

Note: This section focusses primarily on incidental disturbance resulting from recreational activities. There are other sources of potential disturbance in the vicinity of seabird colonies, including e.g. seaweed harvesting (Goodship and Furness, 2019) , inshore fishing activity (e.g. creeling), agricultural activity adjacent to nesting sites, and scientific research activities.

Table 49. Summary of practical feasibility for the reduction of disturbance (at colony).

Further detail is provided following the table.

This table can be viewed in Excel format.

7.6.2 Background

Seabirds in colonies are highly vulnerable to human disturbance, which can originate from various sources including water, land, and air. This type of disturbance is mostly linked to tourism, recreation, photography, research, pedestrians, pets, and UAVs (for recreational and research purposes). The impact of the disturbance can vary in intensity and effect, contingent on colony-specific characteristics and individual differences, but can range from changes in behavioural responses and physiology to effects on demographic parameters and at a population level, including changes in population size and permanent colony abandonment (Carney and Sydeman, 1999; Blanc et al., 2006; Ellenberg et al., 2006).

In WP1, we explored the effects of various human disturbance management measures in and around seabird colonies, mostly for recreational purposes, to identify those with the most potential for strategic compensation for our focal species in Scotland. The most commonly tested management measures included visitor measures (e.g. set-back distances, cap number of visitors and use of signs), marine vessel measures (e.g. set-back distances and speed limits), and UAV regulations. Despite a limited number of studies identified throughout WP1, it was clear that measures to reduce disturbance had a beneficial impact on seabirds, but the degree of the benefit varied between studies and was, at times, unclear. Overall, the effectiveness of these management measures greatly depends on location characteristics, the extent of the impacts of the disturbance, and the type and quantity of visitors. It is important to note that many of these management measures are already implemented in Scotland, particularly in reserves and protected areas, but the efficacy of these measures is not commonly studied, published, or publicly available.

7.6.3 Steps for implementation

1. Collation of baseline information on disturbance levels and their impact on seabirds including field trials:

1.1. Review evidence on what type and to what degree human disturbance occurs at Scottish seabird colonies.

1.2. Identify at what locations human disturbance management measures are already in place. Existing management measures are likely to be in place for many SPAs or reserves.

2. Decision on feasibility of implementing management measures at given colonies and focal seabird species:

2.1. Based on the information collated in step 1, undertake site visits to explore the site and colony characteristics to determine the management measure with the most potential for each specific species and colony.

• Identify the source/s of human disturbance that have significant impacts on seabirds at each colony. Note seasonal changes.
• Research the breeding ecology of focal seabird species within the site, including information on laying, hatching, peak chick provisioning period, and fledgling dates and how each parameter is affected by disturbance.
• Determine the land ownership status by identifying whether sites are privately-owned, designated as a reserve, public land, etc. Identify whether there are already management measures in place.

2.2. Confirm the feasibility of implementing management measures at the given location/s and target species.

2.3. For some management measures, such as fixed set-back distances, speed limits and UAV-related regulations, it would not be possible to proceed directly to delivering as compensation, with research and trials first required.

• Design and conduct preliminary field trials to test methods and assess the efficacy of the different proposed management measures. Note that the duration of the trials may vary depending on the measure.

3. Pre-implementation of management measure:

3.1. Review the Scottish Outdoor Access Code, the Scottish Marine Wildlife Watching Code, the Guide to Best Practice for Watching Marine Wildlife, and the Scottish Drone Regulations, to understand what can and cannot be done, and what should be already in place, noting that these documents provide guidance on access rights and responsibilities as a wildlife watcher and UAV operator and are not legally enforceable (except Drone Regulations).

3.2. Prepare for implementation of management measures.

• Obtain necessary permits for land use and vessel management measures.
  • Where the target site is an SPA colony, then an HRA may be required depending on the type of measure.
• Offer training to personnel who will be responsible for implementing any management measure (this is mainly focussed on wardens).
• Offer environmental education, workshops, meetings, or signage to inform residents and site users, particularly marine vessel owners, about the new management measures and the motives behind them.
• Create designs and prototypes of paths and signs.
• Secure familiarisation of each measures’ best practice for reducing seabird disturbance (e.g. Edney et al. (2023) provides a set of best practices for using UAVs in seabird monitoring and research and NatureScot will soon publish a guidance for use of drones at seabird colonies).

4. Implementation of management measures:

4.1. Once the measure/s have been established, the next step is implementation, which will vary depending on their nature. Some examples include:

• Install informative signs to educate the public about the potential impact of human disturbance on seabirds, promote responsible behaviour, and clearly outline the newly implemented management measure.
• Hire a warden and/or personnel to monitor visitor numbers, increase compliance with new management measures and offer guidance and information to incoming visitors about the origin and importance of following measures.
• Construct pathways and or set up infrastructure to define designated visitor areas, clearly indicating areas that should be avoided.

4.2. Undertake observations to monitor the success of the measures (see Research and monitoring recommendations).

4.3. Given that most seabirds can be found at the colonies exclusively during the breeding season, some management measures would only need to be implemented during the seabird breeding season (though this will be species-specific).

5. Post-implementation of management measures:

5.1. Measure effectiveness of management measures in terms of:

• Human disturbance reduction: evaluate whether the new measures were followed effectively by visitors.
• Benefit to seabirds: perform relevant data analyses to understand the impacts of the measures on the productivity of seabird species.

5.2. Assess whether additional measures or adjustments to existing ones are needed to enhance effectiveness.

7.6.4 Ecological effects of implementing management interventions

7.6.4.1 Large gulls

Management measures aimed at reducing human disturbance have rarely been tested for the benefit of large gulls. Consequently, there is a considerable degree of uncertainty regarding the potential ecological benefits for these species. Due to their size and aggressive responses to perceived threats, large gulls may experience limited benefits from reduced human disturbance. Nevertheless, these species exhibit strong intra-species competition and frequently predate on neighbouring nests. Therefore, any management measures that effectively minimise disturbances that lead to gulls leaving their nests unattended could potentially reduce egg and/or chick predation, subsequently increasing productivity. While management measures relating to UAVs may reduce stress, attacks, and flushing responses, it is unlikely that they would significantly enhance productivity and survival.

7.6.4.2 Auks

Overall, cliff-nesting auks are likely to benefit from management measures and reduction of human disturbance due to stress and predation reduction, and increased parental care, productivity, and survival. It could, potentially, also lead to colony expansion or recolonisation if habitat is available, with effects likely to be strongest at the top of the cliffs. Coastal cliff-nesters, such as Guillemot and Razorbill, are the species expected to benefit most from measures to minimise disturbance originating from sea, as they are the most exposed to this type of disturbance.

Guillemot productivity is influenced both by physical nest characteristics (e.g. nest height, proximity to the sea, gradient of cliff slope, number of neighbours) and human disturbance. At St Abbs Head, for example, halving the number of visitors was predicted to result in an increase of nesting success from 70.1% to 87.2% (Beale and Monaghan, 2004). The benefit, however, lies in the reduction of flushing behaviour and stress, which has the potential to lead to increased productivity, although the degree of it is uncertain. Razorbill tend to occupy crevices in cliffs, so have more protection. As no studies have been conducted on testing such measures to reduce disturbance for Razorbill, their effectiveness on the species is highly uncertain.

7.6.4.3 Puffin

Puffin are less likely to be threatened by human disturbance through visitors, marine vessels, and UAVs. This is because during the day, they are typically away foraging at sea and at the colony are hidden within burrows, minimising proximity to marine vessels and reducing their impact. Additionally, UAVs are unlikely to cause significant disturbance to burrowing individuals. Puffin, however, may benefit from a reduction in burrow trampling, if visitors are indeed causing such an impact.

7.6.4.4 Other seabird species

There is strong evidence, particularly for Kittiwake and terns, supporting the potential benefits of reducing human disturbance (Beale and Monaghan, 2004; Medeiros et al., 2007; Vogrin, 2013; Bishop et al., 2022). Human presence has been strongly linked to poor nesting success in these species, with the impact attributed to factors such as people load and proximity to the nests. Kittiwake, for example, were found to be more sensitive to human disturbance than Guillemot at St Abbs Head, possibly due to their closer proximity to viewpoints (Beale and Monaghan, 2004). Similarly, petrels, cormorants, and Fulmar are highly affected by human disturbance at their colonies (Dias et al., 2019). Even with larger species such as Gannet, there is evidence that installing signs can increase reproductive success (Allbrook and Quinn, 2020).

7.6.4.5 Time-lags for population level responses

If management measures successfully reduce human disturbance at a colony, one can anticipate several immediate positive effects. These include an immediate reduction in stress and negative physiological responses, a decrease in nest destruction and predation rates resulting from human disturbance, and a more efficient allocation of energy expenditure. As a direct consequence, an improvement in overall body condition and parental care can be expected, potentially resulting in an immediate increase in annual productivity. In the short to long term, assuming no other threats are impeding progress, distribution expansion, an increase in adult survival rate, recolonisation, and population growth could be anticipated.

7.6.5 Research and monitoring recommendations

7.6.5.1 Pre-implementation of management intervention

Establishing a baseline understanding of the effects from human disturbance on a site and determining the primary source of disturbance affecting seabirds are crucial for understanding the effectiveness of potential management measures. The initial step involves identifying and quantifying the impact of human disturbance on key seabird species. In some cases, this information may be available in existing literature, especially from reserves where this has been monitored previously, or it could be inferred from literature on similar sites or those managed by similar organisations (e.g. RSPB, National Trust for Scotland).

In most cases, however, this information would need to be collected through field studies, utilising traditional field observations to quantify changes in behaviour, parental care, productivity, and/or survival analysis. Experiments monitoring physiology responses can also be conducted. Subsequently, this data can be used to calculate overall human disturbance levels caused by different sources. In some instances, inferring information from similar sites or by comparing plots with different degrees of disturbance may be employed. Alternatively, the optimal approach is to conduct experimental trials to quantify and refine the effectiveness of management measures in terms of reducing human disturbance and benefitting seabirds (e.g. determining the maximum distance a visitor can approach individuals without causing major effects).

7.6.5.2 Implementation of management intervention

Monitoring should take place while the management measures are in place. While there will be specific details to evaluate for each disturbance type and source, the primary objective is to monitor human disturbance (effectively, whether humans comply to management measures), and to determine whether such compliance influences seabird’s demography. This requires in-situ observations at various points during the breeding season, with a particular focus during the peak of breeding or when chicks and adults are most vulnerable.

To understand the impact of management measures on the focal seabird species, monitoring should encompass aspects like productivity, behavioural responses, and population counts. Due to the unclear and challenging nature of measuring the impact of human disturbance, which may manifest as physiological changes, and the influence of other threats on seabird demographics, detecting and quantifying the direct effects of reducing human disturbance is challenging. Nevertheless, surveys and counts should be conducted throughout the breeding season, with variables such as the number and growth stage of chicks monitored to estimate annual productivity. Subsequently, statistical analyses should be employed to quantify the effects of the regulations or management measures on seabird productivity and behaviour either by comparing years and/or sites with versus without measures (or those less impacted by human disturbance) and, if applicable, assessing differences between treatments (see e.g. Watson et al. (2014)). Before-After-Control-Impact (BACI) models, for example, are a valuable tool for evaluating the impact of management on seabird populations by comparing data collected before and during management, provided a baseline exists.

Eventually, management measures should be reassessed periodically. For example, if a colony expands towards newly added paths, a decision should be made on whether the paths need to change after a few years or whether the action has already met its objectives, and no further management measures apart from those already there, are needed.

7.6.6 Key considerations, potential barriers, and potential solutions

This table can be viewed in Excel format.

7.6.7 Species-specific aspects of implementation

The implementation of disturbance reduction management measures is mostly site-specific and will generally not vary between the focal species.

7.6.8 Overall conclusion

Different management measures would bring varying benefits to focal species. For instance, large gulls are likely to benefit most from UAV management, assuming these are causing significant disturbance. Guillemot and Razorbill, on the other hand, stand to gain the most from any form of human disturbance reduction, whether on land, at sea, or by UAVs. Puffin are expected to benefit primarily from visitor management measures that prevent burrow trampling and are unlikely to experience significant benefits from at-sea and UAV management.

An important consideration in this conservation action is additionality. Existing guidelines, such as the Scottish Outdoor Access Code, Scottish Marine Wildlife Watching Code, Guide to Best Practice for Watching Marine Wildlife, and Scottish Drone Regulations, provide recommendations for behaviour around wildlife in Scotland to minimise human disturbance and should be applied by all visitors across Scotland. Moreover, many of the proposed management measures are already implemented at reserves, including SPA colonies. Hence, most management measures are either already in place or should be. This conservation action might be one of the cheapest to undertake and would require less planning compared to most of the other conservation actions considered.

Since the effect of human disturbance at colonies on seabird demographic parameters is unclear and challenging to quantify, there is considerable uncertainty about whether implementing this action would effectively mitigate the impacts of offshore windfarms. It is difficult to ascertain whether this conservation action would work effectively as a strategic compensatory measure.

7.7 Practical feasibility: Bycatch mitigation in longline fisheries

7.7.1 Summary

This section focusses on bycatch mitigation in floated demersal longline fisheries. The focal species is Gannet with Fulmar listed as a secondary species.

Table 50. Summary of practical feasibility for bycatch mitigation in longline fisheries.

Further detail is provided following the table.

This table can be viewed in Excel format.

7.7.2 Background

The incidental bycatch of seabirds by fisheries has been assessed as one of the top three threats to seabird species globally, and the threat with the greatest average level of impact (Dias et al., 2019). Longline fisheries, in particular, have a high risk of seabird bycatch (Anderson et al., 2011) which has been most well studied in albatross and petrel species in the southern hemisphere with significant knowledge gaps for longline fisheries operating in the North-East Atlantic. Seabirds are attracted to fishing vessels for offal, bait, discards, and for the catch itself, depending on the fishery and the seabird species. Concerns of the potential for bycatch to impact UK breeding seabirds have prompted a series of studies. Initially, vulnerability assessments were conducted using fishery gear parameters and seabird species sensitivity (Bradbury et al., 2017). Subsequently, fishery observer programmes were established, along with assessments to quantify levels of bycatch mortality (Northridge et al., 2020; Kingston et al., 2023).

Miles et al. (2020) used the bycatch estimates of Northridge et al. (2020) to assess the potential for population level impacts. They found significant impacts from longline fisheries on Fulmar and Gannet, where it was estimated that the population could be up to 7% and ca. 1% larger in the absence of bycatch, respectively. This led to a further analysis on bycatch hotspots in UK fisheries (Northridge et al., 2023), which identified longline fisheries operating along the shelf edge west and north of Scotland as leading to significant mortality of Fulmar. Following this work, the Scottish Government commissioned a detailed study on bycatch in longline fisheries in Scotland, focussed on Fulmar (Kingston et al., 2023). This study confirmed significant levels of bycatch, although it produced lower estimates compared to Northridge et al. (2020).

Alongside research and monitoring, policies regarding seabird (and other wildlife) bycatch have been evolving over the past decade. This to support existing commitments around monitoring and to minimise bycatch included in the UK Marine Strategy (DEFRA, 2019), the Conservation of Habitats and Species Regulations (UK Government, 2017), and under the OSPAR Convention (OSPAR, 1992). This was brought together at a UK level under the Marine wildlife bycatch mitigation initiative (DEFRA, 2022), which outlined five policy objectives:

1. “Improve our understanding of bycatch and entanglement of sensitive marine species through monitoring and scientific research.

2. Identify “hotspot” or high-risk areas, gear types and/or fisheries for bycatch and entanglement in the UK in which to focus monitoring and mitigation.

3. Develop, adopt, and implement effective measures to minimise and, where possible, eliminate bycatch and entanglement of sensitive marine species.

4. Identify and adopt effective incentives for fisheries to implement bycatch and entanglement mitigation measures.

5. Work with the international community to share best practice and lessons learned to contribute to the understanding, reduction and elimination of bycatch and entanglement globally.”

Seabird tracking data, when combined with fishery vessel tracking information (either from vessel monitoring system or automatic identification system data), provides insights into the seabird ecology that influences bycatch risk. Gannet are attracted to fishing vessels, which changes their behaviour when in the vicinity of vessels (Votier et al., 2010). Similarly, foraging distributions of Fulmar are associated with fishing vessels, with over half of individuals showing this association, although this was predominantly for trawl-type vessels (Darby et al., 2021). Studies using light-level sensors on Fulmar demonstrate that they also interact with fishing vessels at night (demonstrated by the detection of artificial light at night) and over the winter, including higher concentrations of interactions to the north and west of Scotland (Dupuis et al., 2021). The frequency of vessel interactions (as detected by artificial light at night) has increased over the past two decades, which could increase bycatch risk (Darby et al., 2023).

The primary longline fishery operating in Scottish Waters uses floated-demersal longlines (also referred to as ‘piedra bola’) to target European Hake (Merluccius merluccius). This gear is composed of longlines with a series of baited hooks interspersed with weights and floats that act to hold the hooks up above the seabed (Figure 31). Recently, the sink rates of the gear in Scottish longline fisheries has been investigated (Rouxel et al., 2022). Authors showed relatively slow sink rates (mean <0.1 ms^-1 at the surface from 0-2 m depth for the centre point between the floats and hooks), which is below international guidance (>0.3 ms^-1) (ACAP, 2023).

Given the concerns around seabird bycatch, various mitigation options have been developed. For UK operating fisheries, including longlines, bycatch mitigation options were reviewed by Anderson et al. (2022). For the floated demersal longline fisheries specifically, mitigation options were reviewed by Kingston et al. (2023). International best practice recommendations have been developed by the Agreement on the Conservation of Albatrosses and Petrels for pelagic and demersal longline fisheries (ACAP, 2023). Though many bycatch mitigation options have been identified (for further detail see Bycatch mitigation options below), only a few have been trialled in the floated demersal longline fishery operating in Scotland, which leads to uncertainty on which may be most effective.

The floated demersal longline fishery primarily targets European Hake and, to a lesser extent, other demersal fish (principally Ling, Molva molva). Hake is a shared stock between the UK and neighbouring states. For the stock shared between the UK and EU (representing most of the Scottish Waters for the fishery), in the North Sea (UK/EU Waters of ICES areas 2a and 4) 46% is allocated to UK vessels, while to the northwest (UK/EU Waters of ICES areas 6 and 7 and international waters of areas 5b, 12, and 14), 20% is allocated to UK vessels (DEFRA, 2023b). The UK longline fleet typically is comprised of 10 to 15 vessels each year (though the total number of vessels since 2010 is ca. 40, but this number is thought to be misleading given changes in the ownership and names of vessels and operators) (Kingston et al., 2023). As such, the UK registered fleet represents a significant, but not majority, component of vessels operating in Scottish Waters. This is important to consider both in terms of bycatch monitoring and mitigation, where this focusses only on UK/Scottish registered vessels.

We now have an understanding on the threats that bycatch in longline fisheries could pose to seabirds in the UK, policy commitments to reduce and eliminate bycatch where possible and potential bycatch mitigation options identified. However, there is still substantial uncertainty in bycatch estimates due to low monitoring effort, meaning that we cannot currently confidently quantify the population level impacts. There has also been some work on developing bycatch mitigation for Scottish longline vessels, so while there are a number of promising options, further work is required to develop best practice approaches. These points are developed further in the rest of this section.

7.7.3 Steps for implementation

The following steps focus on UK registered vessels. For non-UK registered vessels there would need to be coordination with the countries from which those boats are registered, though the broad approach would be similar.

1. Pre-implementation:

1.1. Establish baseline: Refine estimates of bycatch rates and understanding of current bycatch mitigation practice.

• Refine estimates of bycatch rates by enhanced monitoring (see Research and monitoring recommendations) with a high proportion of trips including observers and deploying observers using a stratified sampling approach (i.e. a less opportunistic sampling approach). Also, introduce remote electronic monitoring for all vessels. Note, initial work on this could commence immediately but this would continue through implementation and establishment of baseline should not delay trials.
• Determine current bycatch mitigation practice. This is important context both for bycatch rates (i.e. to determine to what extent these rates are with some mitigation applied), and to understand what mitigation could be considered over and above existing practice (i.e. additional).

1.2. Agree initial bycatch mitigation measures to apply.

• Some bycatch mitigation measures may have sufficient evidence to apply immediately, e.g. use of bird-scaring lines (see Bycatch mitigation options), although some of these mitigations may already be in place for some (potentially all) vessels.

1.3. Determine mechanism for establishing enhanced bycatch mitigation programme for floated demersal longline fisheries.

• An enhanced bycatch mitigation programme should be developed to allow for continuous improvement. International best practice recommendations should be followed (see Good et al. (2020)).
• Consider options to secure enhanced bycatch mitigation programmes, potential options include:
  • Voluntary approach – through a collaborative approach with the fleet, with uptake of mitigation measures being voluntary. This could foster a more collaborative environment, but mitigation measures may not be taken up by the whole fleet, leading to reduced overall mitigation and potential to undermine initiative due to lack of a level playing-field between vessel operators. There would be a significant risk that the scheme would not ensure the long-term implementation of bycatch mitigation. This approach would require an auditing mechanism (underpinned by monitoring) to have confidence that all operators are following whatever approach is agreed to.
  • Regulatory approach – this could be via licence conditions or other approaches (e.g. spatial management measures). This would provide a compliance mechanism. It would also help ensure all UK vessels/ and or vessels operating in Scottish/UK waters (depending on how it was implemented) are held to a common standard (level playing-field). However, this could lead to reduced goodwill from vessel operators if they feel it is imposed, reducing sense of agency. This could reduce scope for bycatch mitigation trials which are essential for continuous improvement. The mechanism would determine which vessels are subject to the bycatch mitigation requirements, e.g. licence conditions would only apply to UK registered vessels (though there may be potential for these conditions to be mirrored by other fishery licensing authorities), while area-based measures could apply to all vessels operating in an area but may take longer to develop, and what is possible may be limited outwith designated MPAs.
  • In both cases an oversight group should be established including SNCBs, relevant stakeholders, industry representatives, and those with relevant scientific expertise. The oversight group would be responsible for making key decisions within the initially established framework for the enhanced bycatch mitigation programme.

1.4. Agree objectives for the bycatch mitigation programme.

• Specific and measurable objectives should be set for minimising bycatch (see Good et al. (2020) and Sharp (2016)).

1.5. Initiate bycatch mitigation trials.

• Trials could be undertaken both pre- and post-implementation of establishment of an enhanced bycatch mitigation programme. There are benefits in initiating trials at an early stage as this could allow more rapid reduction in bycatch once the programme is put in place. For details on trials see below.

2. Implementation of enhanced bycatch mitigation programme for longline fisheries:

2.1. Put in place the enhanced bycatch mitigation programme designed in the pre-implementation phase. The steps required for this will vary depending on the type of programme to be adopted, e.g. a voluntary or regulatory based approach.

2.2. Put in place enhanced vessel monitoring (see monitoring).

2.3. Consider enhanced colony monitoring for target seabird species (see monitoring).

2.4. Put in place initially agreed bycatch mitigation.

2.5. Undertake bycatch mitigation trials (see monitoring).

2.6. Establish compliance arrangements.

3. Post-implementation:

3.1. Continue monitoring put in place in earlier steps.

3.2. Continue bycatch mitigation trials.

• Informed by agreed objectives, continue programme of bycatch mitigation trials until bycatch is reduced to agreed level.

3.3. Update bycatch mitigation requirements as required.

• Based on monitoring and bycatch mitigation trials, update bycatch mitigation requirements.

3.4. Have ongoing compliance arrangements in place, with incentives (positive or negative) for full implementation of agreed mitigation.

7.7.4 Ecological effects of implementing action

The ecological background is covered further in the background section above and in the underpinning targeted review (see Bycatch mitigation in longline fisheries). These demonstrate that current bycatch rates are substantial for some species and that there are well proven successes in reducing bycatch rates of seabirds, so beneficial ecological effects are anticipated. Bycatch is also an animal welfare concern, both through mortality and injury (hooked birds that are released alive). The level of sub-lethal impacts for surviving birds (e.g. physical injury and stress) is not known with monitoring currently only recording dead birds. This could be substantial so could increase the population level benefit from bycatch mitigation compared to that anticipated if solely considering direct mortality levels.

7.7.4.1 Fulmar

The biggest benefits from bycatch mitigation in the longline fishery would accrue to Fulmar, where bycatch mortality is high (>1000 individuals/year for UK registered vessels) (Kingston et al., 2023). As bycatch mortality primarily affects adult and sub-adults at sea, its impact on the population is more significant to if it were predominantly young birds (Miles et al., 2020).

The fishery primarily operates along the shelf break along the west and north-west of Scotland and around the Shetland Isles. Fulmar have very extensive foraging ranges, suggesting that any UK breeding Fulmar could be bycaught in this fishery. However, given proximity and core forage range we would anticipate benefits to be greatest for Fulmar populations breeding along the west coast of Scotland and in the northern Isles with less benefit to colonies along the east coast of Scotland. However, as the fishery operate year-round, during the non-breeding season when Fulmar are less tied to their breeding colonies populations will be more mixed.

7.7.4.2 Gannet

Bycatch levels of Gannet in the longline fishery are substantially lower than for Fulmar but still at potentially significant levels in population terms (ca. 100 individuals/year in UK registered vessels) (Kingston et al., 2023).

In common with Fulmar, Gannet have wide foraging ranges, but given the distribution of the fishery, Gannet populations along the west coast of Scotland and in the north of Scotland would stand to benefit most with much lesser benefit to east coast populations.

7.7.4.3 Other species

Smaller numbers of bycatch have been recorded for Great Skua and Great Shearwater so there would be some benefit to these species from bycatch mitigation. Given recent significant mortality in Great Skua following the HPAI outbreak this source of additional mortality could be more significant, so bycatch mitigation could have population level benefits. Great Shearwater breed in the South Atlantic, so birds present in Scottish Waters will either be birds on migration or non-breeding sub-adults.

7.7.4.4 Time-lags for population level responses

As bycatch mortality acts directly on the adult populations of Gannet and Fulmar, reducing bycatch mortality can have near immediate population level benefits. However, as noted above, bycatch mitigation methods have only been trialled to a limited extent in the floated demersal longline fishery. Therefore, it could take a few years of trials and monitoring to arrive at the most effective bycatch methods. As such while reducing seabird mortality from bycatch would have immediate population level benefits, the effectiveness of bycatch mitigation should improve over the first few years post-implementation of a bycatch mitigation programme.

7.7.5 Research and monitoring recommendations

7.7.5.1 Bycatch monitoring (throughout)

Bycatch monitoring should follow standard practices (e.g. Dietrich et al. (2007)). This would best be done by building on the existing monitoring under the UK Bycatch Monitoring Programme⁵, but with significantly increased coverage (coverage should be >20%, and ideally higher, to allow accurate quantification of bycatch rates, Babcock et al. (2003)). Incorporating remote electronic monitoring with cameras (see Kindt-Larsen et al. (2012); (WWF-UK, 2017)) could provide more complete coverage and allow more efficient use of observer time.

Remote electronic monitoring should be rolled out as standard on vessels using this gear and will be important for monitoring the efficacy of mitigation measures at minimising bycatch and improving understanding of bycatch rates. This will be best achieved with some human observer support to verify certain activities in the startup process. Given the relatively small number of vessels involved in this fishery all vessels should be covered. Acceptance of observers should be mandatory with mandating of REM use on vessels using this gear in Scottish Waters, this would ensure full coverage and reduce potential for any biases in monitoring (i.e. systematic biases where coverage is incomplete/only on subset of fleet).

With improved monitoring data, more detailed analyses will be possible to refine understanding on risk factors, e.g. locations, season, time of day; these can inform on which mitigation options may be most effective.

7.7.5.2 Bycatch mitigation trials

Bycatch mitigation trials should test mitigation methods under a variety of conditions. The data to collect in trials will mostly be the same as for regular bycatch monitoring, although some additional data may be collected depending on the method (e.g. sink-rates when using weighting). Bycatch mitigation trials should be carried out in a scientifically robust way (e.g. use of paired-trials). The ACAP review process is useful to consider here (see pages 8-9 in Petrels (2021)).

7.7.5.3 Seabird population monitoring

The key monitoring with respect to seabirds is the assessment of bycatch rates. However, to help understand population level impacts and any population response post bycatch mitigation, colony monitoring would be useful to inform on any changes to the population trajectories of colonies identified as likely to benefit. It may be possible to detect differences between pre- and post-mitigation for colonies in areas likely to experience higher bycatch than other areas (i.e. a BACI type design), though unless there is a very strong response it may not be possible to separate this out from other factors.

To better understand the demographic effects of bycatch and any changes in this pre/post application of mitigation, measuring annual survival rates for different age classes and sexes would be useful. Previous studies have shown that in many species bycatch rates are higher for adult males than other population components (Gianuca et al., 2017). This would follow standard recapture/resighting adults for survival (RAS) methods.

7.7.6 Key considerations, potential barriers, and potential solutions

This table can be viewed in Excel format.

7.7.7 Species-specific aspects of implementation

Bycatch mitigation methods will generally benefit all seabird species bycaught within the demersal longline fishery. However, it should be noted that the effectiveness of the methods will vary depending on the species.

7.7.8 Bycatch mitigation options

To reduce bycatch risk, mitigation options must counter the risk factors influencing bycatch rates. Key risk factors for longlines include:

• Time of day – bycatch risk is generally reduced when gear are set at night, however this may not be the case for Fulmar which are active also at night (Dupuis et al., 2021; Darby et al., 2023).
• Location – bycatch risk will be highest where fisheries operate in core foraging areas of seabirds. Gannet and Fulmar have long foraging ranges meaning that most of Scottish Waters will be within foraging range during the breeding period, though bycatch risk may be higher when vessels operate in core foraging areas.
• Sink rate – baited hooks are only accessible to seabirds when close to the sea surface, therefore if gear sinks slowly it will be accessible to seabirds for a longer period increasing bycatch risk (Rouxel et al., 2022).
• Hook type – birds may be more likely to become caught on some designs of hooks than other.
• Bait type – certain bait will be more attractive to seabirds and the risk of being hooked will also vary by bait type.
• Offal handling practices – how offal is handled and discarded by vessels will affect when and how seabirds are attracted to vessels (though they are also attracted for the bait itself). Changes in handling practice may reduce bycatch risk.

Bycatch mitigation options fall into three main types:

1. Gear adaptations – include any modifications to the fishing gear itself, e.g. change to weights, but also associated physical mitigation, e.g. bird-scaring lines.

2. Operational adaptations – these include changes to when and how gear are deployed, e.g. setting gear at night to reduce risk to diurnally active seabirds.

3. Management options – these include restrictions on when and where fisheries can be active, including seasonal spatial closures.

Bycatch mitigation options for UK operating fisheries, including longlines, were reviewed by Anderson et al. (2022). Specifically for the floated demersal longline fisheries, that are the subject of this account, mitigation options were reviewed by Kingston et al. (2023). International best practice recommendations have been developed by the Agreement on the Conservation of Albatrosses and Petrels for pelagic and demersal longline fisheries (ACAP, 2023). National and international^[13] databases of bycatch mitigation methods also exist. In the below table (Table 51) we provide a summary of key options (for further details and options see the cited references). It is important to note that many bycatch mitigation options are possible to use in parallel, and bycatch will generally be best reduced through using a combination of mitigation options. Some mitigation options may also enhance the effect of others, for example increasing the sink-rate will also increase the effectiveness of bird-scaring lines by increasing the proportion of baited hooks at accessible depths (i.e. prior to sinking) that are covered by the scaring lines (Rouxel et al., 2022).

Table 51. Summary of bycatch mitigation options for floated demersal longline fisheries (selected measures only).

These are examples of potential measures rather than a list of recommended options.

This table can be viewed in Excel format.

7.7.9 Overall conclusion

Bycatch rates have been assessed to be significant for some species (notably Fulmar) in the floated demersal longline fishery in Scotland. Bycatch mitigation measures have dramatically reduced bycatch rates in many longline fisheries worldwide, so it is likely that bycatch rates could be significantly reduced with consequent population level benefits for impacted species. As such, an enhanced bycatch mitigation programme could be suitable as a compensatory measure. However, there are some issues that need to be considered in its use as compensation.

Bycatch rates can only be quantified with wide confidence intervals currently due to low monitoring effort. This limits our capacity to quantify, with confidence, the potential population level benefits from a bycatch mitigation programme. Improving monitoring would be a priority as part of a coordinated enhanced bycatch mitigation programme. Based on currently available evidence, the benefits to Gannet could be relatively low, however, benefits for Fulmar are likely to be significant. Similarly, although many potential bycatch mitigation options have been identified, few have been trialled in floated demersal longline fisheries operating in Scotland. There is therefore uncertainty on which measures would be most effective and of how much each would reduce bycatch rates. However, the benefits of enhanced bycatch mitigation are likely to be significant.

There may be issues in implementing as compensation given existing bycatch reduction policies and commitments, i.e. its status with respect to additionality. However, there are potential benefits of implementing as compensation, as this could give a source of finance for monitoring, bycatch mitigation trials, and for any resulting costs in changes to gear or operating procedures.

The UK registered component of the floated demersal longline fleet operating in Scottish Waters is in the minority, so bycatch levels could not be fully reduced without international coordination (or potentially area-based restrictions) to ensure all vessels operating use effective mitigation methods.

Implementing bycatch mitigation through a programme approach (i.e. an enhanced bycatch mitigation programme) would deliver the greatest benefit. Through this approach, monitoring and bycatch mitigation trials would reduce current uncertainties. Over several years it is likely that significant bycatch reduction could be achieved. The monitoring elements would be particularly critical if implemented as compensation to give confidence on the level of compensatory effect (i.e. how much bycatch had been reduced).

7.8 Results summary for practical feasibility reviews

We evaluated seven actions (though the two forage fisheries-related actions were combined into a single chapter for this section), with the key findings summarised in Table 52. For each component considered in the table, we provide some general summary of results across the actions and highlight any key differences.

It is not possible to meaningfully provide an overall rank for the actions as there are many different criteria to consider and how these are weighted would vary on how these are prioritised. Furthermore, scoring will also depend on the species of interest. In this section, we provide a high-level summary of key results across actions assessed in WP2.

Table 52. Summary of practical feasibility for the seven assessed actions. For further detail see the action chapters. Type of measure – general refers to actions that act over a wide area and site to actions applied at one or more sites; direct to where the action benefits the species directly (e.g. reduced predation) and indirect to actions that are some steps removed from seabird population response (e.g. reducing fishing may increase prey abundance which may then lead to population level responses in seabird populations). Costs of implementing are high-level relative estimates (low-medium-high) of direct costs of implementing measures only and have high uncertainty (i.e. these do not include any indirect costs, e.g. to loss of earnings resulting from the implementation of a measure) and will depend on how a measure is implemented. The overall feasibility score provides a relative assessment from Low–High. Note this is in terms of considering these measures specifically in the context of compensatory measures and with current knowledge. We have not evaluated the measures in terms of their use for wider conservation purposes which would require consideration of different criteria with consequent changes in scoring.

This table can be viewed in Excel format.