The invasive American mink (Neovison vison) preys on native fauna. The pilot phase of the Hebridean Mink Project (HMP) ran from 2001–2006 at a cost of £1.6 million and successfully removed the species from 1100 km2 of the southern islands of the Hebridean Archipelago, the Uists. Mink were also controlled in South Harris to prevent reinvasion. 532 mink were removed, and no further animals were caught or recorded in the eradication area in the last six months of the project. The entire archipelago is now being trapped using techniques developed from the pilot phase. The programme used an adaptive approach, learning as the project proceeded. The lessons learned were also applied to two other scenarios. These included the Isle of Mull, where with limited resources, trapping is carried out by volunteers, and the development of a national management plan in Ireland, where the species is widespread and farmed. The strategies and techniques developed in the Hebrides were modified to fit these differing scenarios. These are discussed together with an exploration of how we can increase our capacity to manage the species over larger landscape scales.
Invasive alien species (IAS) are regarded as one of the greatest threats to global biodiversity (Diamond 1984; Vitousek et al. 1997). They can have large impacts on offshore islands where endemic biotas are particularly vulnerable to extinction (Cronk 1997; Simberloff 2000). Island biotas form most of the world's biodiversity hot spots, accounting for 45% of all bird, plant, and reptile species (Krajick 2005). The protection of these ecosystems yield disproportionately large benefits and their management is recognized as the most cost-effective way of conserving global biodiversity (Myers et al. 2000a, 2000b). Eradication of IAS features highly as part of the conservation tools in these biomes (Atkinson 2001; Donlan et al. 2003; Genovesi 2005)
The complete eradication of IAS from an ecosystem is recognised as the preferred option as it is finite, requiring a single investment. However, eradications can fail due to a number of reasons such as; animals being missed, reinvasion, or lack of resources and time. To reduce these risks, authors have proposed criteria to be fulfilled in order for a planned eradication attempt to work (Bomford and Sinclair 2002; Genovesi 2005). This in turn influences choice of sites for eradication, or long-term control in circumstances where these criteria cannot be fulfilled. It is also a relatively uncommon action, in many cases the view may be taken that eradication is not feasible and no further action is taken. Most mammal eradications have taken place on small islands and their objectives reflect the discrete nature of these locations. Undertaking invasive species control in more open landscapes, too large for a species to be eradicated during the course of a single project, provides particular challenges (Robertson et al. 2017).
Due to the expense of eradication projects, and the inevitable aversion to failure that could result in difficulties in raising future funds, many organizations are reluctant to begin eradication projects in the absence of robust data or experience in similar environments. Many projects are delayed resulting in increased expense (Simberloff 2004; Simberloff et al. 2005) and greater impacts caused by invasive species on species of conservation interest. In the absence of such data, an adaptive approach is re-To quired, learning and gathering data from projects as they are carried out, and using new information and analysis to feedback into project monitoring and decision-making systems in order to fine-tune and improve project operations. This is particularly true as often the data that are required are under recorded in the literature, such as the effort required for species removal in terms of manpower and effort (Roy et al. 2008). In this paper we review three mink control programmes in the UK and Ireland to illustrate these points and the use of interim objectives in planning species removal programmes. In each, the strategy of the programme is outlined together with the techniques used to achieve them. These strategies are in turn described below and are summarized in Table 1; 1) Eradication or as close to eradication as possible from a defined area; 2) Maintaining populations to below acceptable thresholds; and 3) Removal of the species from high-priority sites of rich biodiversity vulnerable to mink predation. Some of the lessons learned and their transferability are then summarized in Tables 2 and 3 and synthesized in a graphic in Fig. 5.
Table 1.
A summary of the differing strategies and resource availability in the three case scenarios
Mink in the UK and Ireland
The American mink (Neovison vison) is a widely distributed invasive carnivore, occurring in 28 European countries having escaped from fur farms. The eradication of invasive Mustelids in general can be difficult due to their elusiveness, neophobia to objects such as traps, and low encounter rates with control mechanisms due to their wide-ranging behaviour (King et al. 2009). Most mink management projects in Europe are long-term control operations or eradications that are quite restricted in range at local river catchment level scales (Bonesi and Palazon 2007) or at a local catchment level in England (Reynolds et al. 2010). The increasing use of volunteers to reduce operating costs has allowed removal over increasingly large areas as seen in Scotland (Bryce et al. 2011; Robertson et al. 2017).
The main aim of this paper is to compare three different mink control programmes, their objectives, and the strategies deployed to achieve these. Firstly we explore the strategies and approach used in a pilot eradication scheme in the Outer Hebrides. We describe the adaptive approach used to develop the work and lessons learned. We then describe how this approach was adapted to develop a mink control strategy for Mull in the Inner Hebrides off the west coast of Scotland. Finally we look at how a strategy was developed for Ireland, and how this approached the challenges of managing the species at a nation-wide scale. The paper then concludes with conceptual developments that could make larger scale control more feasible, especially when combined with advances in technology.
Case study one;
The Hebridean Mink Project 2002–2006
Having escaped from the fur farms on the Isle of Lewis in the 1950s (Fig. 1; Cuthbert 1973; Angus 1993), mink spread throughout the 2800 km2 island archipelago through to the southern tip of South Uist within 40 years. Within this period unsuccessful attempts were made with the objective to stop them spreading across the main water barrier between the islands of Harris to North Uist (Angus 1993).
Fig. 1.
A map of the Outer Hebrides showing Carloway, where the fur farms originally were, the buffer zone of Harris (1) and the removal zones of the Uists (2), taken from Roy et al. (2015).
Mink threaten a number of important natural resources, including important populations of ground nesting birds at critically important breeding sites. These sites or Special Protection Areas (SPAs) have been designated by the European Commission (EC) under the Birds Directive (Council Directive 2009/147/EC on the conservation of wild birds). In addition, mink are important predators of internationally important salmon populations at sites identified as Special Areas of Conservation (SACs) under the European Habitats Directive (Council Directive 92/43/EEC on the Conservation of natural habitats and of wild fauna and flora). Both the birds and habitats directives have resulted in a legally binding obligation for European Union member states to identify and protect key sites of conservation interest for species listed in the annexes in these pieces of legislation.
As well as threatening biodiversity, mink are an economic pest, damaging stock at fish farms and poultry in small holdings (Clode and MacDonald 2002; Areal and Roy 2006). The archipelago was considered too large, and the uncertainties too great, to aim for complete eradication as the project objective from the onset. Instead, an initial pilot phase was designed with the more limited objective of removing mink from North Uist, Benbecula and South Uist and to reduce mink density from neighbouring South Harris to minimise recolonisation of the Uists (Moore et al. 2003; Roy 2006, 2012). Following the success of the first phase and the information gained on the costs and methods required for success, the work has continued with the objective of the complete eradication of mink from the entire archipelago (Lambin et al. 2014).
The project used an adaptive approach from the outset; initially by collecting data from the literature on life history parameters and developing simple population models based on best estimates (Dunstone and Birks 1983; N. P. Moore et al. unpublished report). This formed an important component in a subsequent EU LIFE funded project.
Live trapping in riparian and coastal areas was the main technique used by the project. A total of 2545 live capture cage traps was dug into the ground during the first three months of the project, although only 10% was open at any one time, with the remainder left locked shut to prevent captures. All set traps were monitored daily. This was later supplemented by the use of den locating dogs. In the Uists this resulted in 100 824 trap nights over four years. Overall a total of 228 mink was caught in the Uists, with the last capture in March 2005. After this date, despite a further seven months of intensive trapping and searching effort, no further signs of mink were found and they were considered likely to have been removed from this region. After a cessation of trapping to secure further funding and continue with an island-wide eradication campaign, mink were subsequently re-discovered in the Uists, although it is not clear if these were recolonizers or relict populations (Lambin et al. 2014). In the buffer area of South Harris, 41 674 trap nights over four years resulted in 240 captures with few animals being caught by the end of the project. This effort greatly reduced the risk of recolonisation from this region, although there was still a possibility of extant isolated populations remaining within the region, particularly on offshore islets. Roy et al. (2015) provide a full description of the project.
Adaptations and lessons learned
As the project progressed, a number of logistic changes were made to maximise effectiveness and reduce costs. These included:
The ease with which traps could be checked was improved by fitting them with solid metal, non-mesh doors, allowing the trapping staff to easily find traps and to determine from a distance if the trap had been triggered. It has been estimated that this has saved the project approximately £14 000 annually (Roy 2012).
As mink numbers were reduced there was a need for cost effective methods of monitoring. By setting traps along the road networks and in other sites where they could be seen easily from vehicles and boats. This allowed a low density of traps to be maintained over large areas at reduced cost to help monitor mink presence or absence.
It was found that some staff were more successful at setting traps to catch mink than others (Roy 2012). Staff were rotated between different areas to ensure these successful trappers viewed all of the different trap rounds on the islands and improved the quality of setting. They were also used to train others in their methods to spread best practice.
It was found that mink had distinct seasonal patterns of behaviour depending on their breeding cycle, which influenced their vulnerability to traps (Dunstone 1993; Roy 2012; Roy et al. 2015). Some parts of the year they were highly mobile and easier to trap, while in others they were restricted in movement particularly when they were rearing young in their dens. We changed the seasonal profile of trapper effort to increase trapping effort during the periods when the animals were most vulnerable to this method, and instigated the use of scenting dogs during the denning season to identify den sites for targeted trapping.
When initially placed, there were large between trapper differences in the density of traps used. Based on experience, we standardised the distance at which traps were set along waterways (400 m) to maximise the length of linear habitat covered by a given number of traps.
Mink redistributed themselves and moved towards coastal habitats as their numbers were reduced. Stable isotope analysis carried out on carcasses showed that, as populations were culled, untrapped mink moved to secured, vacated, and nutrient rich coastal territories, increasingly relying on richer marine food as the population diminished (Bodey et al. 2010). This is despite constant trapping pressure on coastal and inland riparian habitats. To further corroborate the fact that coastal habitats provided more resources, it was seen that animals from coastal habitats had smaller home ranges (travelling less to secure more food) and were in better condition (Sandell 1989; Helyar 2005). We further increased trapping effort in coastal habitats as the work progressed as a consequence.
Mink, like many mustelids, use scent to communicate. They have well developed scent glands, and early on in the project trials were carried out to test the efficacy of using the scent glands to bait traps. It was found that scent gland bait traps caught mink at a significantly faster rate than traps baited with fish alone (Roy et al. 2006). As a result, the use of scent glands became standard practice.
These lessons and techniques form the basis of a comparison table (Table 2), to show which of them could be transferred to the other scenarios explored in this paper.
Case study two;
mink control on the Isle of Mull
A mink population was confirmed on the Isle of Mull (Fig. 2) in 2006, and at the request of Scottish Natural Heritage, a small-scale pilot study was carried out to explore the different control options available on this island (Roy 2008). Based on the density seen on the Outer Hebrides, the population was estimated at a carrying capacity of approximately 400 individuals estimated (Sugoto Roy 2007, unpublished report). However, the strategies that could be applied to this scenario are very different to those in case study one. Firstly, the reinvasion risk was considered very high as the Isle of Mull is within swimming distance of mink found on the mainland of Scotland, which is five kilometres away, without accounting for offshore islands in between the two land masses. Secondly, resources were very limited. Trapping was carried out by a handful of volunteers, with limited traps and money. Given the reinvasion risk, eradication was not an option, and the priority was to reduce the impacts of mink on species of conservation concern, in particular the important seabird colonies found off the West Coast of Mull. The objective was therefore to reduce predation on these colonies during the breeding season by reducing mink populations to zero at least around the periods of seabird breeding.
Fig. 2.
Showing species richness of seabird species across Mull and adjacent islands (darker, redder spots show a greater number of seabird species). Areas in green are Special Protected Areas designated by the EU. The large red circles are areas where any mink presence at all would have very negative impacts on biodiversity, as even a few individuals in a low-density population could wipe out entire bird colonies. (Taken from Roy 2008).
Table 2.
The transferability of techniques, in the three differing scenarios
Transferrable techniques
A number of the techniques developed on the Outer Hebrides could be transferred to support the trapping campaign on Mull. Table 2 outlines the techniques from the Hebrides that could be transferred to this scenario.
The Mull programme cannot be described as an eradication, but if well executed it could keep the critical areas of the island mink free, at least during the seabird breeding season. The complete removal of a species from an area, albeit with the continued risk of reinvasion, is still different from ongoing control, while providing many of the benefits of true eradication at a local scale.
Case study three;
managing mink on protected areas across Ireland
In the third case study, the different strategies that could be applied to the management of mink in Ireland were explored. This included a review of the possible techniques that could be utilized in the event of any nationwide control effort (Roy et al. 2009).
In this scenario, eradication was not an option as mink farms were still in operation in the Irish Republic. Animals were still escaping from these facilities, and sightings of feral mink often occurred in their vicinity (Fig. 3). As a result, there was a continuous source population from which re-invasion could occur. Also, the available resources were insufficient to carry out an eradication over an area of 840 000 km2 (Robertson et al. 2017).
Fig. 3.
Location of Irish mink fur farms 1900-1960 (left) and mink sightings (right) from 1961–2008. The figure is taken from Roy et al. (2009).
The objective of mink control in Ireland is to protect globally and regionally important species of migratory birds, such as ground nesting farmland birds and seabirds, together with freshwater species of fish and invertebrates that are protected under the annexes of the EU Habitats Directive. Any strategy for Mink control would need to be developed with a focus on protecting species of conservation concern.
Again in terms of techniques, many of those developed in the previous projects could be applied here and these are outlined in Table 2.
In addition to these approaches, a further method developed for use on slow flowing lowland waterways — the mink raft (Reynolds et al. 2010), — was also considered. This combines detection (through footprint tracking media) followed by subsequent trap operation to ensure the traps are only open when and where animals are detected, making large-scale trapping operations more economic.
The most feasible possible option available in this current scenario was to map out all protected areas in Ireland, and categorize them in terms of the protected species of conservation concern that they are designated for under the EU Habitats Directive. These species were then classified on the basis of their vulnerability to mink predation based on size. These sites were then prioritized accordingly; i.e., protected areas with a large number of vulnerable species were given high priority and vice versa. The process is described in detail in Roy et al. (2009). Maps summarizing high-priority areas are given in Fig. 4. This programme is best described as on-going control using the terminology of Bomford and Sinclair (2002).
Fig. 4.
A map of the SPAs in the Republic of Ireland, colour coded according to an index incorporating the number of Annex one species present on them and how vulnerable these species are to mink predation taken from Roy et al. (2009).
Programme objectives
The three case studies illustrate different objectives, reflecting the costs, feasibility, and scale of the different scenarios. These can be contrasted with the criteria for eradication proposed by Bomford and O'Brien (1995) to illustrate some of the choices faced by managers.
In the first case of the Hebrides, the long-term objective was to achieve true eradication from the entire archipelago, but this was considered too expensive and uncertain to be funded. Consequently, the objectives were reduced to the complete removal of mink from part of the island chain to a natural barrier. This reduced the costs and risks, and on the projects successful completion the work was extended to achieve eradication throughout the island chain. When dealing with large scale eradications such a step-wise approach is pragmatic, even though the individual elements to not meet all of the proposed criteria to achieve eradication.
The second case study, from Mull, had the objective of limiting damage to a resource, the seabird colonies, and true eradication was not considered feasible given the chances of reinvasion. However, it was considered possible to achieve complete removal of mink from the island with only the need to deal with incursions as they occurred. This is also a common situation on many islands, and a feature of larger programmes where the risk of reinvasion cannot be discounted. This strategy of maintaining an area as animal free is not eradications, but also does not fit comfortably in the definition of ‘ongoing control’ proposed by Bomford and O'Brien (1995).
The last case study of control in Ireland falls more clearly into the category of ongoing control, with a continued need to remove animals at a relatively small scale to protect local resources.
All three case studies demonstrate successful approaches to reduce the impact of mink, but with different objectives and costs. Maintaining a flexible approach to the choice of objectives while maintaining clarity in different cases is key to the successful management of IAS impacts. This need for flexibility is not uncommon in very large scale control programmes (Robertson et al. 2017).
Technological advancements
The choice of objectives is often determined by the availability and costs of available methods, and these constraints may change as technology develops. For mink control, the need to check traps daily for humaneness reasons imposes limitations on how far and wide an area staff can cover. The different case studies have been described using a schematic diagram in Fig. 5. Efficiency measures such as those described below, may be useful in changing the parameters and limitations dictated by resource availability. As a result, larger, more difficult, and more ambitious eradication can be undertaken. Countering these limitations is key and may be achieved through the development of new techniques as outlined below:
Developing methods to detect animals that do not require their capture or daily checking. This is one of the fundamental aspects of the use of mink rafts (Reynolds et al. 2010). Here tracking systems associated with traps first detect the presence or absence of mink, and the traps are only opened once an animal is known to be frequenting the area. This can greatly reduce the need for daily checking.
In addition, the use of alternative sources of data, such as data from public sightings, can be aggregated and used to identify areas with a high likelihood of presence or absence of a species during removal. This was carried out post-hoc for the Hebridean project, and analysis of sighting data suggest that they are a useful addition to other detection techniques (Faulkner et al. 2017).
Remote checking of traps. Systems exist which can send a signal, for example over the mobile phone network, to give notification that a trap has been sprung (e.g., www.minkpolice.com). At present these systems are limited by cost and the availability of communications coverage which can limit their use. However, the cost of the equipment is likely to be outweighed by the savings in manpower, particularly when the rate of capture is expected to be low.
Self-resetting traps. The use of lethal traps which kill rather than capture the animal may have different checking requirements and these can vary between countries. However, traps still need to be visited to be reset. A variety of novel designs offer the potential for self-resetting traps (e.g., www.goodnature.co.nz) (Carter et al. 2016). Although lethal trapping options were not discussed here, if the risk for non-targets species is low, such traps would enhance the efficiency of trapping campaigns in terms of area, time, and staff resources. Traps could be left unattended for long periods of time and only checked periodically.
Fig. 5.
A schematic diagram demonstrating how area and resource availability determines the overall strategy and approach to the management of mink. Where campaigns are well resourced and areas are not too large, such as in zone 1, eradication is possible. If areas are very large, a lot of resources will be needed to achieve eradication. In zone 3, where areas are relatively small, even if poorly resourced, mink populations can be maintained to low levels, especially if labour saving techniques are used. In zone 4, where areas are large and resources are limited, campaigns need to focus on interventions at high priority sites. Zone 1 can be increased by increasing the areas we can cover (denoted by A), e.g., through use of roads, checking traps from a distance etc. It can also be increased by making better use of existing limited resources (denoted by B), e.g., by using self-resetting or self-reporting traps, thus reducing manpower needs.
Strategic advancements
Fostering collaboration: In large landscape scale management operations, collaborative efforts between organizations and individuals can greatly enhance our ability to manage invasive species more effectively. An economic study carried out after the end of the project, explored different scenarios under which fish farms in the Scottish islands could more effectively manage mink (Areal and Roy 2006). It found that by acting together, the cost of mink control, in particularly the overhead costs and the initial outlay in equipment costs, could greatly reduce the financial burden on any one organization, and could increase the area over which operations could be undertaken. Joining together to develop management teams and equipment pools would greatly improve mink control at the landscape scale. This has also been found on the mainland Scotland where different landowners and volunteer forces have greatly improved our ability to manage the species (Bryce et al. 2011).
Better adherence to the precautionary principle (Wittenberg and Cock 2001): the precautionary principle advocates a hierarchical approach to invasive species management, starting from prevention all the way through to early detection and rapid response, eradication, and long term control. Improving systems at an earlier stage of invasion would greatly improve the cost-effectiveness of regional invasive species management strategies. A number of species, some quite novel, are now being traded on the Internet (Parrott and Roy 2009), and species are being moved around from between the mainland of the UK and offshore islands deliberately. These include recent populations of pine martens on Mull and Skye (Solow et al. 2013), and fox sightings on the Outer Hebrides (Cramb 2006). The development of formal surveillance programs, the recording and reporting of early detection and sightings, and having in place rapid response teams and equipment with greatly enhance our ability to manage invasion.
Table 3.
Technological and strategic developments that could greatly improve removal efficacy in the different scenarios
A more useful way of developing regional strategies for invasive species management would be to apply all aspects of the precautionary principle to all situations. For example, any future planned eradication on Mull would need to incorporate preventative measures to stop animals coming in from the mainland, such as mink management on the adjoining Scottish mainland, early detection and rapid response of any invaders, formal eradication of defined areas of conservation importance, and longer term, but efficient, control strategies for the remaining areas. Similarly, any eradication of mink on the Outer Hebrides still requires surveillance of any remaining animals, or new invaders brought in through human agency. Using novel technologies, self-setting lethal traps or selfreporting detecting mechanisms could be used to make long-term surveillance more sustainable.
All the technological and strategic developments outlined above are summarized in Table 3, comparing their applicability to the different scenarios. These include techniques and ideas that have developed after the completion of the Hebridean project through post-hoc analysis, for example where sighting data was found to be a useful addition to field data. Ultimately, the objective of any species management campaign will be dependent on the area that needs to be managed and the resources available for management. Eradication is only possible where resources are sufficient and areas are not prohibitively large, as summarized in Fig. 5. If insufficiently resourced, alternative objectives, such as long term management or site specific control need to be developed, using labour saving techniques to improve efficacy with limited resources.
As a final note, engaging with policymakers to strengthen operational biosecurity measures on offshore islands, would help to maintain their biological status and uniqueness.
Acknowledgments:
The author wishes to thank the staff and trappers of the Hebridean Mink project, Scottish Natural Heritage, landowners on the Isle of Mull, and on the Outer Hebrides, National Parks and Wildlife Service of Ireland for the numerous projects that have led to this discussion paper.
