The Environmental Protocol is the main legal instrument concerned with the protection of the Antarctic environment within the Antarctic Treaty area. The Protocol has been approved by all 29 Consultative Parties to the Antarctic Treaty but, of the 21 Non-consultative Parties, it has been approved by only six. Further, the Environmental Protocol does not apply to nations that are not signatories to the Antarctic Treaty. The Environmental Protocol has five annexes currently in effect, which set out minimum environmental standards concerning environmental impact assessment (Annex I), conservation of Antarctic fauna and flora (Annex II), waste disposal (Annex III), prevention of marine pollution (Annex IV) and the protected area system (Annex V).

In subsequent sections we highlight text within the Environmental Protocol that is of relevance to non-native species introductions and describe Annex VI—“Liability arising from environmental emergencies,” which has yet to enter into force. We then go on to discuss issues that may deserve more consideration when applying this new international instrument to non-native species introductions.

Crucially, measures for reinstatement of the damaged environment are not included specifically in the annexes currently in force (Annexes I–V). Moreover, control and/or eradication of non-native species is not a specified mandatory requirement under Annex II, although recommended in non-mandatory guidelines contained within the CEP non-native species manual (CEP 2011; adopted by the ATCM through Resolution 6 [2011]). As noted earlier, Annex II does contain a requirement that Parties maintain Antarctic species diversity, habitats and ecological systems (Article 3[3c]), but this is limited in its effectiveness.

To conform with Article 3(3c), Parties should prevent unintentional non-native species introductions and, if necessary, take management action to control or eradicate any species that have been introduced. However, this is not stated explicitly and examination of past examples of response action following the discovery of a non-native species indicates that often little, or no, practical action is taken (Smith 1996; Hughes & Worland 2010; United Kingdom 2013). For example, the non-native grass Poa pratensis has not been removed from the vicinity of the Argentinian Primavera Station at Cierva Point on the Antarctic Peninsula following its introduction in 1954/55, although discussions on eradication have been initiated (Corte 1961; Pertierra et al. 2013), and neither has Poa annua in the vicinity of Arctowski Station, King George Island, South Shetland Islands (Olech 1996; Olech & Chwedorzewska 2011). In contrast, the Liability Annex is much more explicit, with Article 5 stating specifically that response action by an operator that has caused an environmental emergency should be prompt and effective.

The implementation and enforcement of the requirements contained in the annexes currently in force is controlled fully by each individual Party. Despite legal obligations, Parties may, or may not, be inclined to take action to enforce the legislation. For example, Annex II largely does not permit intentional non-native species introductions, but some Parties do not enforce this requirement and non-native decorative plants are visible on some research stations (see figure 5.2 in Hughes, Huiskes et al. 2014). In contrast, should a liability instrument be in force that was considered applicable to the issue of non-native species, then Parties may have an incentive to make non-native species control and eradication a priority issue, particularly if they consider it possible that a second Party might become involved and then attempt to claim back any costs.

The risks and consequences of a non-native species introduction, as an “environmental emergency,” may not have been considered in any depth (if at all) by the legal and policy groups during the greater part of negotiations of Annex VI (Aust & Shears 1996). The issue of the unintentional introduction of macroscopic non-native species became a major priority for the CEP only from 2004 onward (Australia 2004, 2005), and, in particular, after the publication of the influential scientific review paper by Frenot et al. (2005), which clearly set out the scale of the non-native species problem in Antarctica and the sub-Antarctic islands. A workshop entitled Non-native Species in Antarctica, hosted by New Zealand in 2006, further cemented resolve between scientist and policy makers to tackle the issue (New Zealand 2006; Rogan-Finnemore 2008). Therefore, although the issue of non-native species introductions may not have been a focus of those who drafted the Liability Annex, it may also be possible, in principal at least, to use this international instrument for preventing, minimizing and containing the impact of non-native species introductions and to recoup the costs associated with the removal or control of non-native species that have not been eradicated by those responsible for their introduction. However, when the wording of the Liability Annex is read in the context of non-native species introductions, it is clear that the annex presents many potential issues for those operating within the Antarctic Treaty area.

An “environmental emergency” is defined as “an accidental event that results in, or imminently threatens to result in, any significant and harmful impact on the Antarctic environment” (Annex VI, Article 2[b]). Unlike a catastrophic oil spill, the impacts of which are generally initially great but may dissipate over time, the impacts of non-native species in Antarctica may only become apparent long after the initial, usually minor, introduction event (see Fig. 1). This temporal element is crucial to the question of whether a non-native species introduction constitutes an environmental emergency under the definitions of the annex. If the annex covers only the immediate consequences of sudden accidents or incidents, as suggested by de la Fayette (2010), and not cumulative impact from a continuous situation, then non-native species may fall outside the scope of the annex.

Taking a different perspective, it could be argued also that an introduced non-native species should not be considered an environmental emergency until it starts to become invasive (i.e., spreads and impacts negatively upon indigenous species and ecosystems) as, before then, impacts are likely to be trivial and limited in spatial scale. Non-native species may remain in a persistent state (i.e., surviving and reproducing at a restricted locality, with no expansion of the area colonized over time) for many years before either some genetic change occurs or a physiological threshold is crossed that allows it to become invasive (Olech 2003; Frenot et al. 2005; Hughes & Worland 2010; Olech & Chewedorzewska 2011). For example, climate warming may increase the likelihood of a non-native plant switching to invasive status and spreading more widely, as has already been observed in an analogous fashion with recent population expansion in some indigenous plants on the Antarctic Peninsula (Fowbert & Smith 1994; Convey 2006; Parnikoza et al. 2009). With adequate long-term monitoring, the onset of this switch could be detected and the “imminent threat” of “significant and harmful impact on the Antarctic environment” might make the annex applicable. However, from a practical perspective, rather than initiating a long-term monitoring programme focussed on a newly discovered non-native species, it would likely be less effort to simply eradicate the species as soon after its discovery as possible, as recommended by the CEP (2011).

Consequently, there is a potential case that the presence of even a single individual of a confirmed non-native species should be considered an “environmental emergency in waiting,” and trigger prompt management action, thereby pre-empting the potential application of Annex VI (which itself states that response action by an operator that has caused an environmental emergency should be prompt and effective [Article 5]). If left unmanaged the non-native species may replicate, spread and become invasive before eradication can be funded, planned and attempted (as has already likely occurred with the grass Poa annua near Arctowski Station, King George Island, and the flightless midge Eretmoptera murphyi on Signy Island, South Orkney Islands (Hughes, Worland et al. 2013).

Although the Liability Annex has been adopted as a legally binding measure, it will not enter into force until the Consultative Parties have approved it (Article 2[b]). Approval of each Consultative Party takes place when it deposits a note signifying its approval of the international instrument with the Depositary Government of the Antarctic Treaty. However, the process of approval differs between Parties. For some Parties, such as the UK, an international instrument is only approved once it has been incorporated into domestic law. However, for other Parties approval is the same as initial adoption; for example, Chile and South Africa have adopted and, therefore, approved the Protocol on Environmental Protection to the Antarctic Treaty whilst still not having passed it fully into domestic legislation. In the past the Parties have been slow in making adopted international instruments effective. For example, ATCM Recommendation XVIII-1, which provides guidelines for tourism, is still not effective 20 years after it was adopted initially at Kyoto in 1994. Measure 1 (2005) concerns the Liability Annex, and so far only eight Parties have approved the measure out of the 28 Parties that adopted it in 2005, making it unlikely that the annex will become effective within the next decade.

As the annex cannot be applied retrospectively, all non-native species known already to be present in Antarctica—some of which were almost certainly introduced by, for example, Argentina, Poland, the UK and Uruguay (Corte 1961; Olech 1996; Hughes & Worland 2010; Volonterio et al. 2013)—and any introduced in the years until the annex comes into force will not be subject to this annex.

Given an assumption that the Liability Annex might be applicable to non-native species introductions at some stage, the annex states that any preventative measures and response action should be “reasonable,” taking into consideration the “rate of natural recovery” of the Antarctic environment, and also be technologically and economically feasible (Article 2[e]). Once a species is introduced, natural recovery of the Antarctic environment to its pre-introduction state may occur if the species dies out, most likely due to physical factors; however, should a species become invasive then ongoing habitat alteration, rather than natural recovery, may occur over time. The longer a non-native species is allowed to persist in Antarctica, the more likely it is that its distribution range will increase both through natural processes and by further anthropogenic transfer. Therefore, if no action is taken by either the responsible operators or another operator at some point following the introduction, eradication of a non-native species may become neither technologically nor economically feasible.

Annex II requires that “Each Party shall require that precautions, including those listed in Appendix C to this Annex, be taken to prevent the introduction of micro-organisms (e.g., viruses, bacteria, parasites, yeasts, fungi) not present in the native fauna and flora.” While it is important that measures are taken to prevent microbial introductions (including pathogenic species that may cause disease in wildlife), once introduced, eradication would likely be unfeasible or impossible, in contrast to an introduced macro-organism such as a grass or invertebrate, which may realistically be controlled by eradication attempts.

Parties are required to ensure their operators take measures to reduce the risk of environmental emergencies and their potential adverse impacts (Annex VI, Article 3). Listed measures include the following: (1) “specialised structures or equipment incorporated into … facilities and means of transport,” which in the context of non-native species could include designated quarantine areas for the inspection or storage of imported material, or equipment for removing propagules from cargo and equipment; (2) “specialised procedures incorporated into the operation and maintenance of facilities and means of transport,” which could include appropriate biosecurity procedures; and (3) “specialised training of personnel,” which could include education on the risk associated with non-native species and steps that can be taken to reduce this risk (e.g., cleaning of cargo, vehicles and equipment, boot washing and careful management of fresh foods [see COMNAP & SCAR 2010]).

Article 4 sets out the need for contingency plans to help respond to environmental emergencies but, to date, little discussion of practical contingency planning has occurred within the CEP or COMNAP with regard to non-native species and it is doubtful that many, if any, such contingency plans exist within the Treaty Parties. However, the CEP non-native species manual (CEP 2011), which provides information and non-mandatory guidelines on non-native species and biosecurity, does mention the preparation of a “rapid response guideline” as an area of potential future work (see Annex I, point 17).

The voluntary CEP non-native species manual (CEP 2011) states that “responses to introductions should be undertaken as a priority, to prevent an increase in the species’ distribution range and to make eradication simpler, cost effective and more likely to succeed.” However, assuming an introduction can be attributed to the actions of an operator, and that eradication is feasible and would not itself cause excessive environmental damage, it is unlikely that the responsible operator could be compelled to eradicate the non-native species under Annex II alone.

Articles 6 and 7 of Annex VI concern the allocation of liability following an environmental emergency and legal actions to reclaim costs. If non-native species are eradicated at the earliest stages of introduction, costs are likely to be trivial in comparison with the larger-scale actions required inevitably once a species has become established over a larger area. However, if a non-state operator does not take these initial eradication steps, it is feasible that a different Party could suggest that they intend to undertake a response action (i.e., eradication) and reclaim costs through the domestic courts using national legislation that incorporates the Liability Annex into national law, thereby giving the non-state operator a financial incentive to take prompt action to remove the non-native species. As discussed earlier, whether the small scale presence of non-native species consistent with an initial introduction event would constitute an environmental emergency as defined in Article 2 may be questioned. If a state operator was responsible for the introduction, its Party may be reluctant to engage in a lengthy and potentially time-consuming engagement with the legal system surrounding Annex VI, or to subject themselves to reputational damage, and rather take the potentially simpler and easier option of ensuring eradication at an early stage, as would be required under Annex II, Article 3(3c).

It is likely that only very clear-cut cases of environmental negligence will be pursued under the annex. It may be difficult to prove—to an appropriate legal standard—responsibility for a non-native species introduction. This may be true particularly at locations under the control of one operator, but visited regularly by ships and/or aircraft belonging to other operators, including those operated by the tourism industry and national operators. Examples include Fildes Peninsula on King George Island and, in the same archipelago, Deception Island, as well as Rothera Research Station, Adelaide Island. In addition, the potential for colonization via natural pathways may complicate attribution to the operator in question (Hughes & Convey 2012).

The annex does not apply to fishing vessels and ships transiting through the Antarctic Treaty area; however, these vessels could be sources of non-native species introductions to the marine environment (Article 1). Furthermore, the application of the “sovereign immunity rule” whereby ships or aircraft operated by a Party for governmental non-commercial activities are beyond the jurisdiction of the annex (Article 6[5]), is relevant as the majority of ships and aircraft operating in Antarctica are owned or operated by states. Nevertheless, state operators are still obliged to take prompt and effective response action and are not exempt from the planning and contingency arrangements in the annex.

Article 9(1b), sets out the limit of liability for an environmental emergency arising from an event which does not involve a ship at three million Special Drawing Rights (currently ca. 4.5 million USD). This should be more than sufficient to undertake a simple one off eradication programme over a limited spatial scale, but may not be adequate for a larger-scale multi-year eradication of a non-native species belonging to a more mobile biological group (e.g., some invertebrates) or one which is able to disperse progeny rapidly. In the marine environment, eradications of marine species are almost always unsuccessful (see IUCN 2013), and if attempted in Antarctic waters are likely to be extremely expensive.

In a situation where one Party considers it is appropriate to initiate action under the Liability Annex, the claim will quickly fail if support is not obtained from all other Consultative Parties, as Article 12 states that reimbursements for expenses arising from the response to an environmental emergency shall be approved by way of a Decision, which requires consensus agreement by all Consultative Parties. Achieving this level of agreement may be extremely difficult, given the interactions between Parties both at and away from the meeting table at the ATCMs. Furthermore, payment of costs from the fund held by the Antarctic Treaty Secretariat relies on the Party whose operators caused the environmental emergency also approving the Decision to release funds to the Party that undertook the response action. In other words, it may be possible for the Party found liable to block payment from the fund by not approving the Decision.

Currently, there is a strong financial incentive for Parties to implement strict procedures to prevent a shipping incident (where the ship may be put at risk) or an oil spill (as fuel has a high value in Antarctica), as the cost of an incident may be high, even before clean-up costs are taken into consideration. Policy makers have highlighted the importance of non-native species issues and have worked with COMNAP to generate guidelines to reduce the risk of non-native species introductions (COMNAP & SCAR 2010; CEP 2011). However, the introduction of a non-native species incurs no immediate financial cost to a Party, while ongoing biosecurity measures may be considered expensive, inconvenient and time-consuming. As a result biosecurity implementation may not be prioritized, or an introduction even recognized as an environmental emergency with the associated requirements for contingency planning, as laid out in the Liability Annex. This may leave operators open to later potential claims.

There are well-considered arguments for rapid action following the identification of an anthropogenically introduced non-native species within the Treaty area, specifically to prevent further spread (see Fig. 1). However, over-zealous removal of “supposed” non-native species without appropriate research may be problematic and in some instances may prevent natural colonization when mistakes are made (Hughes & Convey 2012).

In another scenario, if a non-native species eradication is likely to cause substantial damage to the indigenous biota in the affected area, it may be difficult to justify approving the introduced species’ removal. For instance, following years of inaction regarding the establishment of Poa annua on King George Island (Olech & Chwedorzewska 2011) and the chironomid midge Eretmoptera murphyi and the enchytraeid worm Christensenidrilus blocki on Signy Island (Hughes & Worland 2010), it could be argued that the scale of damage that would now inevitably be caused to local ecosystems by any eradication attempt cannot be justified. This illustrates a situation in which different interpretations of Annex II come into conflict. For example, attempting the eradication of a non-native species using methods normally used elsewhere (such as digging up or use of herbicides or pesticides) may damage habitats which, within the area concerned, would be contrary to the Annex II requirement to maintain the species diversity, habitat and balance of Antarctic ecological systems (Article 3[3c]). Crucially, however, should the invasive species not be eradicated and, rather, be spread to other Antarctic locations, far greater impacts might result over a considerably larger spatial scale. There are no guidelines on how to deal with the conflicts arising from the implementation of the Environmental Protocol, although the issue has been identified as an area for further work by the CEP (2011).

Other legal issues may need consideration before eradications may proceed. Pesticides, which may be essential for successful eradications, are not permitted within the Treaty area other than for “scientific, medical or hygiene purposes” (Annex III, Article 7). Furthermore, permits can be given by appropriate national authorities to undertake harmful interference with native flora and fauna, but only (1) to provide specimens for scientific study, (2) to supply specimens for museums and other educational establishments, or (3) to “provide for the unavoidable consequence of scientific activities … or of the construction and operation of scientific support facilities.” Can we consider the introduction of a non-native species an “unavoidable” consequence of our activities in Antarctica? Parties may see environmental impacts as an inevitable consequence of undertaking science in Antarctica and justified by the benefits this research brings. It may be more difficult to justify any introductions that come about through tourist activities, from this perspective. Such complications may make Parties reluctant to resort to potentially costly enquiry procedures or arbitration, as described in Annex VI, Article 7(5a).

The non-native black fungus midge Lycoriella sp. was first reported to have colonized the sewage system at Casey Station (Australia) in 1998, but several subsequent attempts to eradicate it have failed (Hughes et al. 2005). In the case of some other species, the likelihood of a successful eradication might also be low. For example, the fly Trichocera maculipennis found recently both in the sewage system and outside the Uruguayan General Artigas Station on Fildes Peninsula (Volonterio et al. 2013) may be virtually impossible to eradicate now, given that it is a flying insect that has probably established in the environment, and is likely to be highly pre-adapted to maritime Antarctic conditions given its natural distribution, including high-latitude boreal locations. It may be that a Party could reclaim costs from an operator responsible for an introduction, even if the Party's response action was ultimately unsuccessful, as it could be claimed that full eradication is impossible in practical terms, but that management action to control or limit spread of the species was taken instead. However, longer term control and management will not be chargeable as liability of the operator ends 15 years after the initiation of the response action (Article 7[1]).

Currently, the importation of foods (excluding livestock) for human consumption within Antarctica is not subject to the measures to prevent intentional non-native species introductions set out in Annex II, Article 4, of the Protocol (“Introduction of non-native species, parasites and diseases”). However, it should be noted that the exemption of food products from Article 4 is only granted provided that foods are kept under carefully controlled conditions and are disposed of in accordance with Annex III (“Waste disposal and waste management”) and Appendix C to Annex II (“Precautions to prevent introductions of microorganisms”). It is unclear if Parties that fail to conform with these requirements, with the result that a non-native species becomes introduced to the Antarctic environment, could be held liable under the Liability Annex. Furthermore, hydroponic facilities present at several Antarctic research stations have in the past been closed down due to infestations by non-native invertebrates, and it is possible that these species may be released outside the hydroponic facility, as has happened within the sub-Antarctic islands (Frenot et al. 2005; Greenslade 2006; Hughes & Convey 2012; COMNAP 2013b). Such environmental incidents highlight the need for adequate contingency planning, which is also a requirement of the Liability Annex.