New records of the non-native virile crayfish Faxonius virilis (Hagen, 1870) from the upper Snake River drainage and northern Bonneville Basin of the western United States

The virile crayfish Faxonius virilis (Hagen, 1870) is a widespread non-native species across North America, and may be an emerging invasive species globally. We report new records for F. virilis from the upper Snake River drainage and northern Bonneville Basin of the western United States. We sampled 162 sites for crayfish during the summers of 2016 and 2017, detecting F. virilis at 22 of these sites, which ranged from small streams to wadeable rivers, as well as artificial reservoirs and a large natural lake (Bear Lake, Idaho and Utah). We also report the first records for F. virilis from the large, mainstem Snake River. However, we did not find F. virilis in the Snake River headwaters of western Wyoming, and detections of this species were few in the Snake River drainage below Shoshone Falls in south central Idaho. We also found no F. virilis in eastern Oregon. The widespread distribution of F. virilis in the upper Snake River drainage and northern Bonneville Basin may be a conservation concern for native crayfishes in this region, specifically the pilose crayfish Pacifastacus gambelii (Girard, 1852) and the Snake River pilose crayfish Pacifastacus connectens (Faxon, 1914). We recommend that research is undertaken to address ecological interactions of non-native F. virilis with native species, communities, and ecosystems in the western United States, and that education, outreach, and regulatory options be pursued to further limit the introduction and spread of this species.


Introduction
The introduction of non-native crayfishes is a leading contributor to population declines and increased risk of extinction for native crayfishes globally (Lodge et al. 2000;Richman et al. 2015).Non-native crayfishes are regularly introduced via pathways such as fishing bait, stocking for ecosystem management or harvest, and releases of aquarium pets or laboratory organisms (Svärdson 1995;Larson and Olden 2008;DiStefano et al. 2009;Peters and Lodge 2009).Non-native crayfishes can manifest unwanted impacts as invasive species by displacing native crayfishes through mechanisms including competition and disease transmission (e.g., Bohman et al. 2006;Twardochleb et al. 2013).Further, these invasive crayfishes can transform the structure and function of entire freshwater ecosystems due to their high population densities, generalist feeding habits across trophic levels, and burrowing behaviors that mobilize sediment and occasionally damage infrastructure (e.g., Harvey et al. 2011;Hansen et al. 2013;Kreps et al. 2016).
The virile crayfish Faxonius virilis (Hagen, 1870; Figure 1), formerly Orconectes virilis (Crandall and De Grave 2017), is under-studied in proportion to its large non-native range.Native throughout much of the north central United States and south central Canada east of the continental divide, F. virilis has been introduced to California (Light et al. 1995), the desert Southwest (Moody and Sabo 2013), the Pacific Northwest (Larson et al. 2010), and throughout the Appalachian region of the eastern United States (e.g., Kilian et al. 2010).Faxonius virilis has also occasionally been found established outside of North America, including the United Kingdom (Ahern et al. 2008) and the Netherlands (Souty-Grosset et al. 2006;Filipová et al. 2010).Given its expansive nonnative range, studies evaluating impacts of, and developing management options for, F. virilis are surprisingly scarce.However, F. virilis has been found to compete with rare native fish for food in the southwestern United States (Carpenter 2005), and affect ecosystem processes like rates of leaf litter breakdown in desert streams that historically lacked crayfish (Moody and Sabo 2013).In response to these impacts of invasive F. virilis, control and eradication methods including electrofishing removal and treatment with rotenone have been explored (Rogowski et al. 2013;Recsetar and Bonar 2015).
During the summers of 2016 and 2017, we investigated the current distributions and habitat associations of two poorly known native crayfishes of the western United States: the pilose crayfish Pacifastacus gambelii (Girard, 1852) and the Snake River pilose crayfish Pacifastacus connectens (Faxon, 1914) (unpublished data).We sampled 162 sites across Idaho, Nevada, Oregon, Utah, and Wyoming coinciding with the proposed native ranges of these two species (Larson and Olden 2011).Our sampling focused on the upper Snake River drainage and adjacent endorheic desert basins like the Bonneville Basin, which includes the Great Salt Lake (Larson and Williams 2016).During our sampling, we found non-native F. virilis widespread in many waterbodies from which the species had not previously been documented.In particular, we expanded the known distribution of F. virilis in this region relative to past distributional records of the species from Idaho (Clark and Lester 2005), Utah (Johnson 1986), and Wyoming (Hubert 2010), previously summarized by Larson and Olden (2011) and thus updated distributions for F. virilis from the upper Snake River drainage and northern Bonneville Basin of the western United States.

Material and methods
We sampled for crayfish at 78 sites between July 16 th and August 10 th 2016, and 84 sites between July 2 nd and August 3 rd 2017.At each site, we either conducted an hour long timed search with two observers (105 sites), or set baited crayfish traps overnight (57 sites).We used timed searches at 88 lotic (wadeable streams to large rivers) and 17 lentic sites (lakes and reservoirs), and baited trapping at 42 lotic and 15 lentic sites.In most cases, choice of timed search or baited trapping was ad hoc in response to our schedule that day (e.g., we often trapped at sites near where we camped for the night), although baited trapping was sometimes required at sites where timed searches were not feasible (below).
Timed searches used hand nets, D-frame nets, or seines depending on habitat size or other attributes.We used hand nets in the smallest streams where larger nets were difficult to use, and to search the wadeable littoral zones of reservoirs and lakes by overturning potential crayfish shelter like cobble and large woody debris.We used D-frame nets and seines in larger wadeable streams and rivers, following an approach approximating quantitative kick seining for crayfishes (e.g., Engelbert et al. 2016).Timed searches generally covered 100-200 m of linear habitat in either lotic or lentic environments.
At some sites-including those too deep, too steep, or with too limited public access to sample by our timed search methods-we set crayfish traps (0.42 m long by 0.21 m diameter with two 60-mm openings) overnight that were baited with dry dog food.When trapping, we set four to six traps per site for approximately 16 hours at depths ranging from a half meter to several meters deep.All baited traps were set from the banks of streams or rivers or the shores of lakes or reservoirs; we did not use a boat to sample offshore or deeper locations.
We identified all crayfish collected to species, and crayfish identifications were verified by a recognized expert (B.W. Williams).Voucher specimens were preserved in 90% ethanol and deposited at the North Carolina Museum of Natural Sciences (NCSM) Non-Molluscan Invertebrate Collection.We provide NCSM catalog numbers for F. virilis vouchers in the Supplementary material Table S1.

Results
We detected F. virilis at 22 of 162 sites sampled for crayfish during the summers of 2016 and 2017 (Figure 2, Table S1).Faxonius virilis almost always occurred allopatrically, but was sympatric with the native crayfish P. gambelii at one site near the town of Murtaugh on the mainstem Snake River.Twelve sites where we detected allopatric F. virilis were historically occupied by the native crayfish species P. connectens or P. gambelii (Larson and Olden 2011;Larson and Williams 2016).
Ten of the 22 sites where we found F. virilis were in the Bonneville Basin of Utah and Wyoming, including the upper Bear River and one of its mainstem reservoirs (Woodruff Narrows) and one of its tributary reservoirs (Sulfur Creek), multiple sites in the middle Bear River of southeastern Idaho, the northern and southern shorelines of Bear Lake in both Idaho and Utah, and a small reservoir in northeastern Utah.These represent the first published records of F. virilis from Bear Lake.
Within the Snake River drainage, the majority (8 of 12) of F. virilis detections were above Shoshone Falls.These included four sites in the mainstem of the Snake River, ranging from near the town of Roberts downstream to near the town of Murtaugh, with intervening locations at Massacre Rocks State Park and the town of Burley.These represent the first records of F. virilis from the Snake River.We also found F. virilis in the Portneuf River near the town of Pocatello, a reservoir in a headwater tributary to the Snake River above Shoshone Falls, a very small tributary stream close to the Snake River near Massacre Rocks State Park, and Dierkes Lake near the town of Twin Falls adjacent to the Snake River.
Below Shoshone Falls, no F. virilis were found in any mainstem Snake River sites.We did collect F. virilis from two reservoirs in the upper Salmon Falls Creek drainage, the Little Wood River near the town of Richfield, and a small reservoir southeast of Boise.Cumulatively, we found F. virilis in habitats including extremely small tributary streams, midsized wadeable rivers, reservoirs, a large natural lake, and the large, mainstem Snake River (Figure 3).
Sites where we found F. virilis were generally in lower elevation, warmer valley bottoms, whereas the species was generally absent from higher gradient, colder mountain streams and rivers, such as those draining the Teton or Uinta mountains.We do not know if this distributional pattern reflects biological tolerances or habitat preferences of F. virilis, or instead is attributable to differing introduction efforts by humans between these sparsely vs. more densely populated locations.Faxonius virilis was abundant at most sites where it was detected by hour-long timed searches (Table S1).We collected a mean of 79.9 F. virilis (range of seven to 312) per hour at sites sampled by timed search, with high F. virilis catches recorded in both lotic and lentic environments (Table S1).Conversely, F. virilis catch per crayfish trap was relatively low regardless of habitat type.We collected a mean of 0.60 F. virilis (range of 0.17 to 1.83) per trap at these sites.Given the apparent superior detection probability of hourlong timed searches for F. virilis, it is possible that we failed to detect this crayfish from some of our trap-sampled sites.

Discussion
Our study demonstrates that the non-native crayfish F. virilis is widespread in the upper Snake River drainage and Bonneville Basin of southeastern Idaho, western Wyoming, and northern Utah in the United States.We expand on past, more isolated records for this species from Idaho (Clark and Lester 2005), Utah (Johnson 1986), and Wyoming (Hubert 2010).We report the first records of F. virilis from the mainstem Snake River, as well as the first published records of F. virilis from Bear Lake, where this nonnative crayfish was found on both northern and southern shorelines.Our failure to find the native crayfishes P. connectens and P. gambelii at 21 of the 22 sites occupied by F. virilis-despite 12 of these being historic locations for native crayfishes (Larson and Olden 2011;Larson and Williams 2016)-suggests that F. virilis may represent a conservation threat to native crayfishes in this region.Similarly, its poorly known community and ecosystem effects would benefit from further study (Moody and Sabo 2013).
Faxonius virilis was first reported from this region by Johnson (1986), who documented the species from several reservoirs in the Provo and Weber River drainages of Utah.Some early F. virilis introductions within Utah were initiated by management agencies to provide forage for non-native, warm water sportfish (Johnson 1986).Our record for F. virilis from the southern shoreline of Bear Lake in Utah is particularly notable in this state.In Idaho, Clark and Lester (2005) reported F. virilis from China Creek, a tributary to Salmon Falls Creek.We greatly expand on published records for F. virilis in southern Idaho, including multiple sites in the mainstem Snake River.Within Wyoming, F. virilis was previously found to be widespread in the Bear River and adjacent reservoirs in the vicinity of the town of Evanston (Hubert 2010).Our collections support this finding, but replicate it in a more broadly accessible publication.Our sampling confirms the proposal of Larson and Olden (2011) that F. virilis is more widespread in parts of the western United States than previously recognized.In contrast, we did not find F. virilis in eastern Oregon, and the species was rare in the Snake River and associated tributaries below Shoshone Falls.Similarly, the species was not found in the Snake River and its tributaries within western Wyoming.These results demonstrate that although relatively common, opportunity remains to contain further spread, or prevent new introductions, of F. virilis in this region.
Given its widespread distribution within the upper Snake River drainage and northern Bonneville Basin, F. virilis may pose a conservation threat to the native crayfishes P. connectens and P. gambelii.Our work on conservation status, current distributions, and habitat associations for these native crayfishes is ongoing.However, like Hubert (2010), we almost never found these native crayfishes occurring in sympatry with F. virilis, and F. virilis had seemingly displaced native crayfishes from their historic range at 12 study sites from which they were previously reported.Both P. connectens and P. gambelii are extremely data poor, lacking effectively any published studies on their ecology or conservation status (Larson and Olden 2011;Larson and Williams 2016).We recommend immediate study on potential interactions of P. connectens and P. gambelii with F. virilis that could guide conservation of native crayfishes in this region.For example, we do not know if apparent displacement of native crayfishes by F. virilis is driven by factors like competition or disease transmission (Bohman et al. 2006;Twardochleb et al. 2013), which may warrant different management responses.Similarly, the effects of non-native warm water sportfish like smallmouth bass Micropterus dolomieu (Lacepède, 1802) on both native and non-native crayfishes in the western United States are unknown (Carey et al. 2011), but differential susceptibility to fish predators has been found to affect native crayfish displacement by invasive crayfish elsewhere (DiDonato and Lodge 1993).
Finally, we suggest that education and outreach to anglers, aquarium hobbyists, the biological supply trade, or users of live seafood could prevent further introductions and spread of F. virilis, as could regulation or policy changes related to the transport and possession of live crayfish (DiStefano et al. 2009;Peters and Lodge 2009;Larson and Olden 2011).We know little about past or present pathways of introduction of this crayfish to this region, aside from some early introductions for sport fish forage in Utah (Johnson 1986).Given the lack of study of crayfish in this region, we also cannot quantify rates of secondary spread or natural dispersal of F. virilis populations from existing records (e.g., Wilson et al. 2004).Regardless, much of the western United States, such as Oregon, is not yet invaded by F. virilis, and there are many areas where this non-native crayfish cannot spread on its own due to dispersal barriers like drainage divides, dams, and waterfalls (Fausch et al. 2009).Urgent priority should be given to minimizing the introduction, establishment, and spread of F. virilis in this region through educational outreach and regulatory mechanisms focused on human users of live crayfish.Such an effort would also benefit from better understanding the habitat or environmental requirements that can support F. virilis populations in order to prioritize where to apply education or management intervention most efficiently (Vander Zanden and Olden 2008).We are skeptical that sufficient presence records exist for F. virilis in our study region to accurately model the habitat requirements of an invader that is still spreading and not yet at equilibrium with its environment (Václavík and Meentemeyer 2012).However, a larger-scale habitat modeling effort for this crayfish using available data layers such as Domisch et al. (2015) and its known native and non-native occurrence records would benefit many management jurisdictions concerned with this invader (e.g., Morehouse and Tobler 2013).

Figure 2 .
Figure2.Sites where virile crayfish Faxonius virilis (Hagen, 1870) were and were not detected from the upper Snake River drainage and northern Bonneville Basin of the United States during the summers of 2016 and 2017, as well as where this species was reported within or immediately adjacent to the student extent byJohnson (1986),Clark and Lester (2005), and Hubert (2010).

Figure 3 .
Figure 3. Representative sites where virile crayfish Faxonius virilis (Hagen, 1870) were collected from the upper Snake River drainage and northern Bonneville Basin of the United States, including wadeable rivers like the Bear River north of Evanston, Wyoming (A), the southern shoreline of large, natural Bear Lake in northeastern Utah (B), reservoirs like Lost Creek Reservoir in northeastern Utah (C), and the large, mainstem Snake River at Massacre Rocks State Park, Idaho (D).All images from E.R. Larson.