Introduced Prophysaon andersonii (J.G. Cooper, 1872) in Quebec, Canada: first record of Prophysaon (Gastropoda, Eupulmonata, Arionoidea) in eastern North America, confirmed by partial-COI gene sequence

We report for the first time the terrestrial slug Prophysaon andersonii (J.G. Cooper, 1872) from Quebec, Canada. Two specimens were collected in Parc national du Bic. The identification was determined by the external morphology and partial-COI gene sequence data. The genus Prophysaon is endemic to western North America, and the new record indisputably represents an introduction. No species of Prophysaon has, until now, been noticed in North America from outside its native range.


Introduction
The genus Prophysaon Bland & W.G. Binney, 1873 is endemic to western North America, distributed from Attu Island, Aleutian Islands, Alaska (Roth and Lindberg 1981) to California and east to Idaho and Montana (Pilsbry 1948;Smith et al. 2018). This genus has been traditionally placed either in a broadly defined family Arionidae (e.g. Pilsbry 1948) or in the family Anadenidae, when it is raised from its traditional subfamilial rank within the Arionidae (e.g. Wiktor 2000;Bouchet et al. 2017). Pilsbry (1948) divided the genus into two subgenera and recognized eight species. These slugs are known as "taildroppers" on account of their ability to autotomize the end of their tail as a defensive behaviour (Hand and Ingram 1950;Stasek 1967;Deyrup-Olsen et al. 1986). In Canada one species, P. coeruleum Cockerell, 1890, was assessed by COSEWIC (2016) as Endangered; the other three Canadian species-P. andersonii (J.G. Cooper, 1872), P. foliolatum (Gould in A. Binney, 1851), and P. vanattae Pilsbry, 1948-are apparently not at risk. Recently, Proschwitz et al. (2017) documented P. foliolatum (Gould in A. Binney, 1851) in southern Sweden, introduced from North America with the trade of salal (Gautheria shallon Pursh) for the floral industry. One of us (RGF) has observed P. andersonii in association with strongly synanthropic, mostly European, introduced terrestrial slugs and snails in retail plant nurseries in British Columbia. This suggests that at least some members of the genus are likely candidates for introduction elsewhere by the movement of nursery stock and soil. Although the introduced status of at least one Prophysaon species has now been confirmed in Europe (southern Sweden; Proschwitz et al. 2017), there have surprisingly not been any reports of this genus introduced to anywhere in Canada or the eastern USA. Nevertheless, the species was found in southern California, 630 km south of its previously known southernmost limit in central California (Pearce and Richart 2010) We use partial-COI gene sequence data to complement morphological species identification and report P. andersonii from Quebec. This appears to be the first record of any Prophysaon species from eastern North America.

Methods
The present paper is a result of a general reconnaissance of several Quebec national parks in Sept. 2019 by one of us (AN). The geographic position of the collection site of the new record of Prophysaon andersonii was obtained using a GPS receiver (Garmin eTrex 30x, Southhampton, UK). Specimens were collected by hand and placed directly in 95% ethanol for barcoding and deposited at the Biodiversity Institute of Ontario (BIOUG), Guelph, Ontario, Canada. These specimens were assigned to the project with the project code FTMCA in the Barcode of Life Data Systems (BOLD, http://www.boldsystems.org, Ratnasingham and Hebert 2007).
To map the species' distribution, GBIF (2019) was searched for occurrence data and the results included 535 usable records, from 20 published datasets (https:// doi.org/10.15468/dl.dhkjfe). Added to these were 73 unpublished records held by one of us (RGF, Appendix), as well as one record from Pearce and Richart (2010), 81 records from Smith et al. (2018), and the new record.
For the phylogenetic analysis, Anguispira alternata (Say, 1817) (Discidae) (N = 13) from the FTMCA project was selected as the outgroup. Samples of foot tissue from P. andersonii and A. alternata were used for molecular analysis following protocols in Layton et al. (2019). In this protocol the barcode region of cytochrome c oxidase subunit I (COI) was amplified with the universal "Folmer" primers (LCO1490/HCO2198) (Folmer et al. 1994) and sequenced. After editing and aligning the sequences their quality was checked. The sequence dataset for the genus Prophysaon was completed by additional sequence data from GenBank and from various projects on BOLD. The Barcode Index Number (BIN) algorithm was applied to delineate clusters corresponding to operational taxonomic units at the species level (Ratnasingham and Hebert 2013).
The evolutionary history was inferred using the Neighbor-Joining method (Saitou and Nei 1987). The evolutionary distances were computed using the Maximum Composite Likelihood method (Tamura et al. 2004) and are in the units of the number of base substitutions per site. This analysis involved 341 nucleotide sequences (of which 13 were Anguispira alternata serving as outgroup). Sequences were mined from Genbank and BOLD. In Genbank, the majority of sequences came from western North America (Smith et al. 2018) and one sequence came from Sweden, a specimen of the introduced Prophysaon foliatum (Proschwitz et al. 2017). All ambiguous positions were removed for each sequence pair (pairwise deletion option). There were a total of 655 positions in the final dataset. Evolutionary analyses were conducted in MEGA X (Kumar et al. 2018 Identification (Fig. 2). The external appearance of the slugs conforms to the characters of the genus Prophysaon. Mantle anterior (shell not visible externally) and coarsely granular, with pneumostome in front half of mantle. Tail rounded, not keeled, without mucus gland at tip, and with line of abscission of autotomized tail faintly visible. Sides of body strongly reticulated. In alcohol, body brownish, with darker reticulations, and with one dark lateral band on each side of mantle. Pale dorsal stripe running down length of tail. Length of largest slug ( Fig. 2; contracted, in ethanol): 22 mm. Prophysaon andersonii and P. foliolatum are similar in appearance. Colour of the skin in these species varies considerably, especially in P. foliolatum which typically has the mantle rimmed with bright yellow (Burke 2013). However, the pigmentation of our Quebec specimens, now that they have been in alcohol since their collection, cannot be relied on. No photographs of or notes on the Quebec specimens while alive were taken, but the slugs were immediately recognized as P. andersonii at the time. The preserved slugs, now discoloured, faded, and strongly contracted in alcohol do not readily show the main characters used in distinguishing these species when alive. Prophysaon andersonii is decidedly smaller than P. foliolatum, with the length of 30-60 mm when adult and in a normal, extended position (Burke 2013); P. foliolatum is 50-80 mm (Burke 2013) to over 100 mm (Pilsbry 1948). Assuming the Quebec specimens are fully grown, or nearly so, in their uncontracted state, they would have been within the size range expected for P. andersonii, or at the smallest end of the range for P. foliolatum. The autotomized portion of the tail is somewhat longer and the line of abscission more distinctly marked in P. foliolatum than in P. andersonii (Pilsbry 1948;Burke 2013). In our specimens, the line of abscission is hardly visible (Fig. 2), but perhaps a little more like P. andersonii than P. foliolatum in its position on the body, as compared with the illustrations of preserved (contracted) slugs of these species published by Pilsbry (1948: figs 372, 375). The reticulated pattern on the tail is also somewhat less coarse in P. andersonii than in P. foliolatum. Although it is difficult to tell for certain, our specimens look to have the finer reticulation of P. andersonii-again, in comparison with Pilsbry's (1948) figures 372 and375. Pilsbry (1948) described and illustrated the anatomy of the reproductive system of both species, although the number of specimens studied does not seem to have been great. In the nominotypical subgenus the epiphallus is abruptly enlarged into an oblong, muscular body; Pilsbry (1948) described this structure in P. andersonii as more or less straight, whereas in P. foliolatum, it was found to be slightly curved. He also observed that the slender  part of the epiphallus in P. foliolatum is extremely long, apparently longer than in P. andersonii. Although Roth and Lindberg (1981) and  have since mentioned for other specimens the enlarged, muscular portion of the epiphallus, there has been no comprehensive investigation to evaluate the usefulness of these observed differences in determining species. We did not attempt the dissection of our material because we felt this would not tell us anything. It is of note that Pilsbry (1948) did not use any anatomical characters in distinguishing P. andersonii and P. foliolatum in his key; perhaps he did not place much importance on the data he had.
The BIN AAU1161 was assigned to the COI sequence of the Prophysaon specimen from Quebec which falls within the P. andersonii cluster where some other specimens from British Columbia have the same BIN (Fig. 3). Although the COI genes seems to be highly variable in this genus (Smith et al. 2018), introduced P. andersonii in Quebec and P. foliolatum in Sweden (Proschwitz et al. 2017) can be clearly associated with the clusters of specimens from North America of these species. The node for the P. andersonii cluster has a support value of 65%, while the node of the P. foliolatum cluster has a support value of 98%.

Discussion
In Canada, Prophysaon andersonii is native to British Columbia, where it occurs along both coastal and interior areas (Forsyth 2004(Forsyth , 2005 Fig. 1). Among the species of Prophysaon, this one has the broadest geographic range, from Alaska to California and east to Idaho and Montana (Pilsbry 1948;Roth and Lindberg 1981;Smith et al. 2018), and thus, perhaps the greatest adaptability. In British Columbia, P. andersonii lives in forests mostly, but has been seen in somewhat disturbed habitats (Forsyth 2004), as well as in retail plant nurseries, where it co-occurs with introduced and anthropophilic terrestrial snails and slugs (Forsyth pers. obs.). While the origin and means of transport of P. andersonii to Quebec are unknown, it seems probable to us that this species made its way there via nursery stock.
Parc national du Bic is located on the south shore of the St Lawrence River estuary, ca 15 km south-west of the city of Rimouski. The park has forested, rocky promontories connected by lowlands, which are either forested as well or have old fields. Within the park there are well-developed recreation facilities, including trails and campgrounds. Therefore, the new record of P. andersonii is in a park with quite modified habitats and popular with the public. Aside from the planting of nursery-grown vegetation, a major vector responsible for the introduction of terrestrial molluscs (e.g. Cowie et al. 2008;Bergey et al. 2014), one possible source of introduction could be campers. For example, the introduced slugs Deroceras reticulatum (O.F. Müller, 1774) and several species of the genus Arion A. Férussac, 1819 are common in British Columbia campsites, including some in quite remote areas (Forsyth pers. obs.), and these seem likely to have been transported there by movement of tents (Grimm et al. 2010) and also possibly firewood. It is possible that P. andersonii has been transported to Parc national du Bic in that way, although we are not suggesting that cross-country transport was by this means. Another possible means of transport of either slugs or their eggs is by flowing water.
Wherever the local source of P. andersonii to Parc national du Bic might have been, it seems likely that it was nearby. We suspect that this species might occur elsewhere in the area, and perhaps P. andersonii is more commonly introduced in eastern Canada than currently realised. Investigations of parks, gardens, and plant nurseries in Quebec, and more generally in eastern Canada and the north-eastern USA, might locate additional records of this species.