First record of Chelicorophium curvispinum ( G . O . Sars , 1895 ) from Lake Mälaren , SE Sweden

The Ponto-Caspian amphipod Chelicorophium curvispinum (G.O. Sars, 1895) was first recorded in the Baltic Sea catchment area in the early 20 century. Since then the species has considerably extended its distribution. Here, we present the first record of the invasive amphipod in Sweden (Drottingholm, Lake Mälaren, SE Sweden). It is plausible that C. curvispinum was initially introduced via shipping to the brackish waters of the Stockholm Archipelago. National ferry traffic may have provided a subsequent means of transport for this species from Stockholm centrum to Lake Mälaren.


Introduction
Chelicorophium curvisipinum (G.O. Sars, 1895) is a tube dwelling amphipod native to fresh and brackish waterbodies of the Ponto-Caspian region. However, this species has exploited European river systems and man-made canals to expand its distribution, which currently includes Belarus, Belgium, Bosnia-Herzegovina, Croatia, Estonia, France, Germany, Hungary, Ireland, Lithuania, Netherlands, Poland, Romania, Russia, Serbia, UK and Ukraine (CABI 2017). Chelicorophium curvispinum had reached the southern parts of the Baltic Sea drainage area and brackish lagoons by the beginning of the 20 th century (e.g. Bij de Vaate et al. 2002). First records of C. curvispinum in the northern Baltic Sea were reported in the 1920's from Curonian Lagoon, Lithuania (Szidat 1926), in 2005 from Narva Bay, Estonia (Herkül and Kotta 2007), from the easternmost parts of the Gulf of Finland, Luga Bay, Russia in 2006 (Malyavin et al. 2008) and from Lake Ladoga, Russia in 2009 (Kurashov et al. 2010). Here, we report the first record of C. curvispinum in Sweden and discuss the possible route of introduction.

Material and methods
In May 2017, during a marine archaeological survey conducted in Drottingholm Bay, Lake Mälaren, Sweden (59º19.456′N; 17º53.656′E WGS84), seven individuals of an unknown Corophiidae species were collected from wood samples retrieved from a wooden structure at a depth of ~13 meters, (Figure 1). Tube structures typical to tube dwelling species of corophiids were clearly visible on the wood samples, and animals were crawling outside the tubes when the wood samples were brought to the surface ( Figure 2). The specimens were stored in 70% alcohol and transported to the Department of Environmental Sciences, University of Helsinki, Finland for identification. The unknown Corophiidae was initially inspected and body length measured under a binocular microscope. One specimen was also dissected, mounted on temporary slides with glycerol and further inspected under a compound microscope for species determination and illustration. The original descriptions and later taxonomic study by Cărăusu (1943) and the latest revisionary work by Bousfield and Hoover (1997) were used in identification.

Results
The short rostrum, separate urosomal somites, laterally inserted uropod 1 and spinose rami of uropods 1 and 2, small antennal cone, not elongated maxilliped palp segment 2, and other characters clearly attributed the species to the genus Chelicorophium Bousfield and Hoover, 1997. Presence of both median and small distal processes on antenna 2 segment 5, four posterior claw-like spines on dactylus of gnathopod 2, nearly bare anterior margin of pereopods 3 and 4 basis, spinose outer margin of uropod 1 peduncle and both margins of outer ramus, and bare inner margin of uropod 2 inner ramus indicated that the specimens belong to C. curvispinum ( Figure 3). However, there are some differences from the original description and figures of G.O. Sars (1895). He did not mention the presence of distal process on antenna 2 segment 5, nor that the uropod 2 peduncle was devoid of outer spines. Study of the type and other material from the Caspian Sea is necessary to resolve the importance of these differences.  Chelicorophium curvispinum were 3.9−5.9 mm in length (average 5.2 mm, SD 0.7 mm). The sample contained both adult females and males. Based on photographs, these non-indigenous amphipods covered the topmost areas of the wooden structures at densities of 500−2000 individuals m -2 . However, no specific abundance assessments were made during the archaeological survey.

Discussion
Here we report the first record of C. curvispinum in Scandinavia. This finding is not surprising as this novel Corophiidae is ranked among the most successful invaders in European fresh and brackish waters (CABI 2017). Moreover, C. curvispinum is listed in the watchlist of Swedish invasive species (Frammande arter 2012) so its arrival was expected. Currently, the nearest known locations of C. curvispinum populations are in Estonia (Herkül and Kotta 2007) and Lithuania (Olenin and Leppäkoski 1999), at an approximate distance of about 500 km from Lake Mälaren.
It is difficult to postulate the actual transport vector for this introduction, but some remarks can be made. Even though C. curvispinum has successfully used channels and rivers as invasion corridors (e.g. Bij de Vaate et al. 2002), this amphipod probably arrived in the Stockholm Archipelago area in ship ballast water (Bij de Vaate et al. 2002) and/or as a fouling organism on vessel hulls (Lucy et al. 2004). The Baltic Sea is heavily navigated by fast vessels and ship traffic is regarded as the main vector of dispersal for non-indigenous species in the area (Leppäkoski and Olenin 2000). Both proximate locations of C. curvispinum have big ports that are well connected to the Stockholm Archipelago.
Lake Mälaren is connected to Stockholm centrum, which is a brackish water (< 0.33%; Karlsson et al. 2010) seafront in the inner archipelago of Stockholm. The distance between Drottingholm Bay and Stockholm centrum is about 10 km. Stockholm centrum is a busy hub of multiple ports and an extensive network of ferry routes. In addition to leisure crafts, scheduled tourist ferries operate between Stockholm centrum and Drottingholm castle every hour during the summer. Drottingholm pier is located ~100 meters from the site where C. curvispinum was retrieved (Figure 1). This clear connection between Drottningholm bay and Stockholm centrum suggests that C. curvispinum arrived first to one of numerous international ports in Stockholm and was subsequently transported to Lake Mälaren via small craft.
The environmental characteristics of the sample site accord well with the habitat preference of C. curvispinum. The strait often has periods of very high current velocity that improves food availability for C. curvispinum (van der Brink et al. 1993). The wooden constructions are very suitable substrate for C. curvispinum as the species is known to inhabit submerged wood in rivers and streams (Den Hartog et al. 1992;Hoffmann and Hering 2000). Usually, C. curvispinum is recorded in relatively shallow waters (< 5 m; Van der Brink et al. 1993;Herkül and Kotta 2007) but our findings suggest that the species can have much wider depth range.
As the population range and density of C. curvispinum in Lake Mälaren is currently unknown, it is difficult to predict the environmental consequences of this introduction. Chelicorophium curvispinum may potentially outcompete other macroinvertebrate species in the River Rhine ( Van der Velde et al. 2000) but as yet there are no data regarding the invertebrate community in our sample site. In addition, large and dense populations (up to 750 000 specimen m -2 ; Van der Brink et al. 1993) may result in clogging and fouling problems. On the other hand, C. curvispinum is a suitable food item for many fish species (Van der Brink et al. 1993) and may thus prove to be beneficial for the native fish community.
Although this study provides only a brief glimpse of C. curvispinum in Sweden, our finding is yet another example of the high dispersal potential of this species. Hopefully, our paper will trigger more comprehensive research on the dispersal of C. curvispinum in Swedish waters. We believe that this novel amphipod species is currently relatively common in Sweden, but determination of its extent is hampered by lack of taxonomic knowledge and poor representation of monitoring sites in potential C. curvispinum habitats. In addition, the morphological discrepancies highlighted in this study should be investigated.