Trichocera maculipennis (Diptera)—an invasive species in Maritime Antarctica

Systematics and identification

The diagnostic characters of adult flies, males and females, were given in Volonterio et al. (2013). Their venation and pattern of spots on wings (Figs. 3A, 3B) were used for species identification. Since the larvae in the sewage system appeared abundant, and are the potential target of eradication, their morphology and mode of life are summarised herein.

The larva of T. maculipennis (Rhynehart, 1925; Brindle, 1962) is similar to those in all congeneric species (Figs. 3C–3F). In the genus Trichocera the larva has a strongly sclerotized, dark head capsule, partially retractable into the first (prothoracic) segment. The first four segments behind the head are secondarily divided in two; the next six into three parts. The secondary segmentation is present even in the first instar, although weakly expressed. The last, 11th segment, bears four fleshy lobes, two dorsal and two larger, ventral ones; between them, a pair of large spiracles is situated. Although trichocerid larvae are generally known to be terrestrial, the larvae of this species may thrive in a semi-liquid or liquid substrate because they can raise the anal end with spiracles above the surface; the four anal lobes (Figs. 3E, 3F) are able to adhere to the spiracles to cover them when necessary (Karandikar, 1931).

Distribution

Trichocera maculipennis is widely distributed over the Holarctis. Unlike most congeneric species, it seems to be tolerant to both warm and cold climates. The range of this species in the Northern hemisphere is wide, from the Arctic to the most southern regions of the Mediterranean area (Azores, Canary Islands; Dahl, Krzemińska & Baez, 2002), and in the Far East, to the Kuril Islands (Petrašiūnas & Paramonov, 2014). All records from most northern, Arctic localities, confirm the synanthropic character of the species in these harsh conditions (Iceland, Bear Island, Greenland and Jan Mayen; Dahl, 1957; Dahl & Krzemińska, 2015). In the Southern Hemisphere, T. maculipennis is known from two locations: the Kerguelen Islands (Séguy, 1940; Séguy, 1953) and King George Island (Volonterio et al., 2013). The species was most likely introduced to the Kerguelen Islands by human agency, and dispersed from one of the ports (Dahl, 1970b).

Notably, invasions of two other Holarctic trichocerids to the Southern Hemisphere resulted in established populations which have existed there for c.100 years: Trichocera regelationis L. in South Georgia, first recorded by Bréthes (1925) and Dahl (1970a), and T. annulata Meigen in Australia (Alexander, 1926) and New Zealand (Edwards, 1928). These early introductions were accomplished most likely with soil taken as a ship’s ballast. Both species are closely related with T. maculipennis (Krzemińska, 1999), as discussed in Volonterio et al. (2013).

Observations on the life mode of T. maculipennis at the Polish Antarctic Arctowski Station

The dipterans, discovered in October 2017 in large numbers, were active, flying inside the sewage system. The temperature under the lid of the sewage tank fluctuated from +1 °C to +3 °C, while the outside temperature was −3.5 °C to −2 °C. Only a few adult individuals were noticed outside the facility, at a distance of less than 50 cm. They most likely escaped accidentally when the lid was open. No larvae or pupae were observed outside, whilst adults stayed on the ground or snow, and did not fly (Fig. 4E).

The sample collected contained three females, each at different stages of abdominal maturity (Figs. 4A–4C). The mass presence of the males on the surface of the substrate (Fig. 4F) was unexpected, as males of other trichocerid species swarm in flight, singularly or in groups. The presence of specimens with striped abdomens (Fig. 4D) is notable; this feature appeared variable even in a very small sample (Figs. 4A–4C: A, C with stripes, B, uniformly brown). Similar variation was observed previously in the Uruguayan Artigas Station (Volonterio et al., 2013). This feature seems to occur only sporadically in T. maculipennis, as discussed in Volonterio et al. (2013), therefore this observation seems important when considering a probable source of this species at the Polish Station.

Larvae were not buried, staying on the wet wooden surface inside the sewage cover box. One larva was most likely third instar, whilst all others were fourth instars (Fig. 4G). The dark brown color of the largest and possibly oldest larvae is noteworthy. It most likely comes from a gradual, passive absorption of coprosterols (present in human excrements; the breakdown product of cholesterol) which make the larvae inconspicuous on the substrate of the same color. Nowhere in the literature has such a dark color been recorded in larvae of this or other species of Trichocera; the trichocerid larvae are usually whitish or grey in colour.

Biology

For the successful eradication of this species, knowledge on the biology of its life stages (egg, larva, adult) is essential; particularly their resistance or vulnerability to external conditions and the habits which enable this particular species to spread outside its genuine distribution.

All stages of Trichocera species are vulnerable to desiccation and freezing or warming; according to Dahl (1969), temperatures lower than −10 °C and higher than 32 °C kill adults and larvae. Steady temperature oscillating around 0 °C (−2 °C to +5 °C) seems especially important for development of the eggs and larvae of this species, as may be deduced from its preference of synanthropic life in habitats such as mines (even to 400 m deep and in complete darkness; Grimshaw, 1906), cellars, and storage houses, but also in natural caves where temperature conditions are more or less stable (Plachter, 1983; records since XIX century are reviewed by Petrašiūnas & Weber, 2013). These habitats also provide rich sources of food for the larvae: rotten vegetables, animal droppings (especially guano in the caves), compost and sewage waste. According to Karandikar (1931), the duration of life cycle strongly depends on the environmental conditions (Table 1); in the laboratory, the cycle may be completed in c. 40 days, while outside, in a synanthropic habitat, it may last up to one year. This discrepancy is difficult to explain. Extended appearance of adults in anthropogenous habitats can be a result of repeated, fast developing generations, or by gradual emergence of females of one generation.

Other information on the life mode of T. maculipennis is restricted to casual observations of the adults’ appearance. In the Arctic region, the adults appeared in July to August (Dahl, 1967a); in the subarctic region in June to August, and in Sweden and Finland from early spring to September (Dahl, 1967b). In the Southern Hemisphere, in Kerguelen, the adults appear during the Antarctic summer (January to March), winter (July to August) and spring (September), which is nearly all year round. Some adults were found in caves, and others near human settlements (Port Jeanne d’Arc, Port aux Français).

Invasion of King George Island

Until now, in most cases, the presence of flying, potentially invasive, invertebrates in Antarctica corresponded with ships’ or aircrafts’ operations during supply periods for scientific stations (Hughes et al., 2011). The Chilean Frei Station’s airport on Fildes Peninsula, King George Island, may be considered as a potential transport hub for non-native species between South America and other regions of King George Island, other islands in the South Shetlands archipelago, and Antarctic Peninsula region. Other sources of spread must also be considered, such as ships with tourists or personnel of stations, and above all, the cargo of multiple operators from their home countries. Many studies (e.g., Lee & Chown, 2009; Tsujimoto & Imura, 2012; Chwedorzewska et al., 2013) suggest that almost all cargo for scientific stations in South Shetlands Islands can be a potential vector for non-native organisms.

The most likely way of introduction to Antarctica would be by the cargo of root vegetables stored in the ships going directly from Europe or Asia, as there is no evidence of this species in South America. Potatoes, swedes, and other tubers are known to be the food source for the larvae of T. maculipennis (Dahl & Krzemińska, 2015). However, females do not lay eggs in fresh bulbs or tubers, neither can any larvae be found within them because the tissue of such vegetables is too hard. The attack is possible only after vegetables have been softened by previous infection by molds and bacteria, therefore a significant threat is posed if the vegetables show traces of rotting. Deep freezing of vegetables would be lethal to trichocerids in all life stages, and would eliminate future introduction of non-native insects to Antarctica.

Migration between the stations and possibility of establishment in natural habitats

The Polish Antarctic Arctowski Station is the fourth station on King George Island invaded by T. maculipennis (Fig. 1). The flies were observed first in the Fildes Peninsula region at the Uruguayan Artigas Station in 2006, then in 2009/2010 at the Frei Chilean Station, and in 2013/2014 in the Korean King Sejong Station on the Barton Peninsula (Peter et al., 2013; Korea United Kingdom Chile Uruguay, 2016). In January 2015 adult specimens were caught also inside the Escudero Station (P Convey & T Contador, pers. comm., 2015). Separate introduction events to these stations seem less probable than spreading of one original population.

Methods of spread are currently unknown. The species can fly in temperatures around 0 °C, thus flight seems an obvious possibility. However, Trichoceridae are poor flyers, and flight cannot take place in windy conditions (Dahl, 1965). Nevertheless, a passive dispersal on wind currents should be considered. Observations of several other congeneric species suggest that the flies may spread by walking combined with short episodes of flight over the surface - either ice-free or snow covered surfaces (e.g., Hågvar & Krzemińska, 2008). A recommended method of checking dispersal ability would be to monitor areas between stations, combined with pitfall traps which are known to attract the trichocerid females.

Recolonisations of sewage systems at the Artigas Uruguayan Base in Maxwell Bay, after unsuccessful eradication attempts, indicate this species’ ability to survive in natural environments (Volonterio et al., 2013). A breeding place and substrate to lay eggs in, alternative to those offered by human activity, may be the guano of birds or pinnipeds, with ornithogenic soil providing a temperature higher than freezing point (Zdanowski, Zmuda & Zwolska, 2005). Some shelter may also be provided by the low vegetation, which includes mosses and vascular plants: native hair grass Deschampsia antarctica and pearlwort Colobanthus quitensis, and an invasive bluegrass, Poa annua L. (Olech & Chwedorzewska, 2011; eradication of the grass is now in progress; Galera et al., 2017).

Human traffic between stations and the transfer of eggs or larvae on shoes or personal equipment could be another alternative form of invasive species dispersal. During the season in 2016/2017, the Polish Antarctic Station was visited by 52 scientists from different countries, including Brazil, Chile and Argentina. Most of them travelled to King George Island from South America via Chilean Frei Station’s airport in Maxwell Bay. The station’s personnel would also visit other bases on occasion during holidays, many times taking fresh food as gifts, a potential source of larvae. There is no data showing the intensity of this kind of human traffic between scientific stations. The Arctowski Station was also visited by tourist ships coming from South American ports; in the 2016/2017 season, 649 people landed on the shore of Admiralty Bay (Annual Report Polish Antarctic Arctowski Station 41. Antarctic Expedition 2016/2017; B Matuszczak, 2016/2017, unpublished data). Tour operators in Antarctica have strict biosecurity measures including boot washing and cleaning personal equipment, and this way of introduction is least likely. Furthermore, the eggs of all trichocerids are very delicate, short–living and immediately desiccate outside wet substrate. However, some cruise ships send fresh food as gifts to the stations they visit, so that may be a risk of repeated introduction of alien species too.

On the South Shetland Islands, where environmental conditions are similar to those experienced by T. maculipennis in the Northern Hemisphere (Dahl, 1970a), the species has the physiological capacity to colonise suitable habitats in the Antarctic Peninsula region, as it had done before in the Kerguelen Islands where it was observed within and outside human settlements (Dahl, 1970b). However, Kerguelen Island is 13° further north (49°S, compared to 62°S at the Admiralty Bay). The vegetation on that island, although low, is present all year round, and climatic conditions, although harsh and windy, are less severe than those in Admiralty Bay (Convey et al., 2018). The temperature on Kerguelen Island is higher than on King George Island (+1 °C in August and +6° C in January on Kerguelen Island, and +0.8 °C in January and −6 °C in August on King George Island; Petlicki et al., 2017). Presence and further spread of this species on King George Island indicate its adaptation to colder temperatures.

A case of colonisation similar to that of T. maculipennis was observed previously by Hughes et al. (2013) of an alien chironomid Eretmoptera murphyi Schaeffer, introduced accidentally from South Georgia to Signy Island, during a soil transplantation study in the 1960s (Edwards & Greene, 1973; Block, Burn & Richard, 1984). This species became invasive and acquired a biomass exceeding more than twice that of the entire native microarthropod species at the location where E. murphyi occurs at the Signy Island, and is thought to have a significant impact on nutrient cycling in the environment (Hughes et al., 2013).

Eradication in human settlements

Presence of flies in one of four sewage systems at the Polish Antarctic Station (Fig. 2) indicates an early stage of colonisation in the Admiralty Bay region, and immediate management actions are necessary to stop further spread. Previous unsuccessful eradication attempts of T. maculipennis at Uruguayan and Korean Station on Fildes Peninsula (Volonterio et al., 2013; Korea United Kingdom Chile Uruguay, 2016) have shown that one-off action would not cause total extermination.

Protocols implemented against an invasive sciarid fly (Lycoriella sp.) at the Australian Casey Station also appeared insufficient (Hughes et al., 2005), whilst the same species was effectively eradicated at the Rothera Station. The difference between living conditions of populations in both stations were significant. At Rothera, the flies were inside the building (in the alcohol bond store), while at Casey Station the pupae were hidden in nooks and crannies in the sewage system, and their eradication was hindered, so much so that they appeared resistant to the chemicals used. Synanthropic species are difficult to eradicate (Hughes et al., 2005), especially in such delicate ecosystems as terrestrial Antarctica, where very strict norms of using insecticides are to be obeyed.

The first steps of a comprehensive eradication protocol were implemented at the end of the season in 2017/2018 at the Arctowski Station. The protocol included several different methods targeting different life stages of invasive flies. Firstly, the affected sewage system was cleaned and treated with the solution of hydrochloric acid. The treatment appeared effective; the sewages are now free of insects, but reappearance can be expected as it had happened at the Uruguayan and Korean stations. Monitoring will be continued and the protocol repeated if needed.

As a next step (season 2018/2019), the locations of insect occurrence will be fumigated with insecticides containing lambda-cyhalothrin. This substance is a synthetic pyrethroid which disrupts insects’ nervous system. Insecticides containing proposed chemicals belong to the long-acting substances (up to six to eight weeks), and are used for controlling adults, larvae and eggs of insects. At the Arctowski Station they will be used only inside the sewage system, without contact with natural environment. The substance breaks down into harmless compounds after six to eight weeks. UV lamps and electric flying insect killer lamps will be used inside the sewage’s cover box and in the vicinity of the sewage system to control flying adult specimens.

It is necessary also to control the natural environment: trichocerids on flight, especially just above the ground, and walking on snow (Fig. 4E), and to implement and check the pitfalls. Flying trichocerids are easily collected with a net; the pitfalls will not eliminate the species, but may confirm its presence in the field. As there is no molecular data for T. maculipennis, either from the Northern hemisphere or Maritime Antarctica, the barcoding studies may establish the origin of its population on King George Island. Molecular studies are necessary to reveal the sources of fly introduction and to determine whether it was a one-off incident or possibly the result of multiple occasions. Further research on the potential influence of T. maculipennis on native fauna and their possible reproduction outside the research bases must be considered.

Subsequent complex measures, taken jointly by all parties, should be implemented as soon as possible. First steps to establish an action plan to manage the non-native flies in King George Island were taken jointly by Korea Uruguay Chile United Kingdom (2017) (ATCM XL–WP 26). This action should be extended and supplemented.

National operators of scientific stations on King George Island are aware of the problem of non-native flies, and together, are planning further works, starting with systematic monitoring. A proposed extermination protocol will be shared between operators and discussed.