Invasive species are non-native species that spread in a new region (Blackburn et al. 2011) and/or have negative impacts after establishment (Parker et al. 1999). This management-relevant distinction (Simberloff et al. 2011) can be complicated by some organisms experiencing long time-lags between establishment in a non-native region and subsequent spread or impacts (Crooks 2005; Spear et al. 2021). Spear et al. (2021) termed such invasive species “sleeper populations,” which can experience time-lagged spread or impacts due to a variety of abiotic or biotic factors, including environmental stochasticity, evolutionary change, or patterns of population growth (Sakai et al. 2001). For example, the invasive Spiny Water Flea (Bythotrephes longimanus Leydig, 1860) was present in Lake Mendota, Wisconsin, United States for 14 years before an uncommonly cold summer caused a delayed population boom to a high abundance state (Walsh et al. 2016). The Red Fire Ant (Solenopsis invicta Buren, 1972) was present in the United States for four decades as single-queen colonies before subsequent introductions of multiple-queen colonies initiated rapid spread and impacts of this invasive species (Porter and Savignano 1990; Krieger and Ross 2001). Population declines of well-established invasive species, whether due to human management (Zavaleta et al. 2001) or natural processes (Strayer et al. 2017), can also cause lagged spread or impacts of other invaders as a consequence of reduced predation or competition (Spear et al. 2021).

The Rusty Crayfish (Faxonius rusticus Girard, 1852) has been a prominent invasive species in lakes of northern Wisconsin for half a century (Capelli and Magnuson 1983; Olden et al. 2006). Faxonius rusticus became hyper-abundant relative to other crayfish species in these lakes (Hansen et al. 2013a), and consequently reduced the abundance and/or species richness of aquatic macrophytes, benthic macroinvertebrates, and some fish species (Wilson et al. 2004; McCarthy et al. 2006; Roth et al. 2007). Recently, a control and eradication effort reduced F. rusticus abundance in one lake in this region (Perales et al. 2021), whereas other lakes have experienced substantial F. rusticus population declines without human management intervention (Larson et al. 2019). These population declines might be caused by habitat modification by F. rusticus (Larson et al. 2019), climate factors like frequency or magnitude of drought (Perales et al. 2020), accumulation of pathogens or parasites over time since invasion (Sargent et al. 2014), or interactions with predatory fish communities (Roth et al. 2007). Population declines of F. rusticus provide opportunities for post-invasion community and ecosystem recovery (Hansen et al. 2013b; Szydlowski et al. 2023), but may also provide opportunities for serial or over-invasion (Russell et al. 2014; Karatayev et al. 2023) by other non-native crayfishes due to weakened biotic interactions with this dominant, abundant invader (Simberloff and Gibbons 2004; Strayer et al. 2017).

During routine population monitoring for F. rusticus, we captured three Calico Crayfish (Faxonius immunis Hagen, 1870) across 24 overnight traps in Wild Rice Lake, Vilas County, Wisconsin on August 28^th, 2020 (Figure 1). These crayfish were vouchered as catalog #16809 at the Illinois Natural History Survey Crustacean Collection, Champaign, Illinois, United States. Capelli and Magnuson (1983) documented a similarly low-abundance population of F. immunis in Wild Rice Lake in the 1970s (0.2 male crayfish per trap) but did not detect this species in any of the other 66 lakes they sampled in Vilas County. This species was not detected in Wild Rice Lake or connected lakes in the Manitowish River watershed of Vilas County between Capelli and Magnuson (1983) and Larson et al. (2019). Faxonius immunis is native to southern Wisconsin, but the disjunct Wild Rice Lake population was 260 km from the nearest historical record at the time of Capelli and Magnuson (1983), leading these authors to conclude that F. immunis had likely been introduced to Wild Rice Lake by humans. Faxonius immunis was similarly absent or undetected from the neighboring Upper Peninsula of Michigan, United States in the 1970s, but has been collected at two isolated locations in recent years (Smith et al. 2019). We suspect that F. immunis persisted at very low abundances in Wild Rice Lake over time, but may have been contained from spreading to other lakes due to biotic interactions with formerly hyper-abundant F. rusticus (Figure 1). Faxonius immunis has a large native range in eastern North America but has established non-native populations in both western North America (e.g., Newkirk et al. 2023) and Europe (e.g., Chucholl 2012). Non-native populations of F. immunis have been recently documented to exert strong per-capita predation effects on benthic macroinvertebrates (Chucholl and Chucholl 2021) and reduce the abundance and species richness of these organisms in small lakes (Herrmann et al. 2022), consistent with ecological impacts of other invasive crayfishes like F. rusticus (Wilson et al. 2004; McCarthy et al. 2006). Accordingly, we propose that F. immunis could become a sleeper invader in northern Wisconsin lakes if ongoing F. rusticus population declines facilitate future spread and impacts of this congener.

We report here a series of studies on the rediscovered F. immunis population in Vilas County, Wisconsin. First, because the majority of sampling for crayfish in this region has occurred in lakes (Capelli and Magnuson 1983; Larson et al. 2019; but see Olden et al. 2011), we sampled streams and rivers throughout the Manitowish River watershed in summer 2021 to investigate whether F. immunis might be more prevalent than realized in under-sampled lotic ecosystems. This sampling included baited trapping of the outlet flowage between Wild Rice Lake and the downstream Manitowish Chain of Lakes to investigate whether F. immunis was already spreading. Next, we used mitochondrial DNA (mtDNA) barcoding to confirm identifications of F. immunis from Wild Rice Lake, as juvenile or female individuals of this crayfish could be misidentified as a congener like the Virile Crayfish (Faxonius virilis Hagen, 1870). Finally, we compared agonistic interactions between F. rusticus and F. immunis in the laboratory to determine whether differences in aggression between these species may have contributed to historical and contemporary distributions of F. immunis in our study region. Crayfish communities of northern Wisconsin lakes are structured by behavioral dominance and associated competitive outcomes for limited shelter (e.g., rock or cobble substrate) from fish predators (Garvey et al. 1994; Hill and Lodge 1994). Within its native range, F. immunis can be competitively excluded from some habitats by F. virilis (Bovbjerg 1970), and accordingly we suspect that competition with the even more dominant F. rusticus has prevented F. immunis spread out of Wild Rice Lake into adjacent waters (see also Smith et al. 2019). Yet if F. rusticus population declines in this region persist (Figure 1), F. immunis may have opportunities to spread, become more prevalent, and impact freshwater ecosystems in the future.

We did not detect F. immunis from the 24 traps set throughout eight different stream and river sites in the Manitowish River watershed during summer 2021 (Table 1). Similarly, we did not detect F. immunis from nine traps set in the Trout River flowage downstream of Wild Rice Lake (Figure 1). We only collected eight F. immunis from 49 traps set over a total of three nights in Wild Rice Lake (0.12 male CPUE, 0.16 total CPUE), and all F. immunis from this lake were collected in close proximity on the southern shoreline (46.0591, -89.7909). Mitochondrial DNA barcoding confirmed all eight of these crayfish as F. immunis (Figure 2), and two ambiguous individuals collected from elsewhere in the watershed (White Sand Lake at 46.0892, -89.5931) were identified as F. virilis by mtDNA barcoding. Faxonius immunis collected from Wild Rice Lake were most similar to the mtDNACOI sequence of this species from Ontario, Canada (GenBank accession DQ882095.1). Surprisingly, only seven of the 11 crayfish used in the Illinois study were identified as F. immunis by mtDNA, with the remaining four individuals identified as F. virilis. We removed these four individuals and all of their interaction pairs with F. rusticus from linear mixed-models of crayfish behavior prior to analyses. Faxonius immunis collected from their native range at Douglas Creek, Illinois were more diverse than F. immunis from Wild Rice Lake, Wisconsin, with four individuals resembling the previous mtDNACOI sequence from Illinois (Genbank accession AY701220.1) and three other sequences being distinct (Figure 2). European F. immunis were distinct, but monophyletic, to the crayfishes used in our study (GenBank accessions JF438005.1 and JF438006.1). The new F. virilismtDNA sequences generated by our study are excluded from Figure 2 to emphasize F. immunis results.

Faxonius rusticus was significantly more aggressive than F. immunis from our combined location model, as evidenced by an intercept that was significantly lower than zero (Intercept = -1.375, df = 14.202, t = -5.817, P < 0.001). However, we found no significant effect of size difference (Coefficient = 0.050, df = 37.242, t = 1.539, P = 0.132), form difference (Coefficient = 0.230, df = 39.520, t = 0.924, P = 0.361), or location (Coefficient = 0.022, df = 14.273, t = 0.092, P = 0.928) on crayfish behavior. The two location-specific models were highly similar to each other (Figure 3), with F. rusticus significantly more aggressive than F. immunis in both Illinois (Intercept = -1.352, df = 4.966, t = -3.582, P = 0.016) and Wisconsin (Intercept = -1.352, df = 10.580, t = -3.987, P = 0.002). We found no significant effect of size difference on crayfish behavior in either Illinois (Effect = 0.052, df = 18.906, t = 0.993, P = 0.333) or Wisconsin (Effect = 0.041, df = 10.162, t = 1.118, P = 0.289), and no significant effect of form difference on crayfish behavior in Wisconsin (Effect = 0.421, df = 12.179, t = 1.637, P = 0.127). Faxonius rusticus was more aggressive than F. immunis even when size-matched or in most cases where F. immunis was the larger individual in the pair (Figure 3).

Our study confirmed an F. immunis population in Wild Rice Lake, Vilas County, Wisconsin, which had not been documented since the 1970s (Capelli and Magnuson 1983) and could potentially spread to become an invasive species in the future (Spear et al. 2021). Faxonius immunis currently occurs at very low abundances in Wild Rice Lake and has not been documented elsewhere in the Manitowish River watershed, whether by regular crayfish sampling in lakes (Larson et al. 2019) or our stream and river sampling in summer 2021. We believe that this F. immunis population has been prevented from spreading out of Wild Rice Lake due to competition with invasive, hyper-abundant F. rusticus in neighboring lakes over the past several decades, particularly as F. immunis was previously documented as a poor competitor relative to congeners (Bovbjerg 1970; Smith et al. 2019). Laboratory behavioral assays supported F. rusticus as consistently dominant over F. immunis across source population, body size, and sex or form differences. Accordingly, F. immunis may be unlikely to emerge as a sleeper invader in this region per Spear et al. (2021). However, ongoing population declines of F. rusticus (Larson et al. 2019) increase uncertainty about species composition and potential ecological interactions of crayfish communities in this region in the future.

We found F. rusticus to be consistently dominant over F. immunis despite using individuals from different regions (Illinois or Wisconsin), invasion histories (native or non-native populations), and methods of collection (trapped or seined). This contradicts some recent research that has found region of origin and invasion history of populations to affect behavioral outcomes between crayfish species pairs (Glon et al. 2018; Reisinger et al. 2020), but supports past work indicating that F. immunis is a poor competitor relative to congeners (Bovbjerg 1970; but see Chucholl et al. 2008). For example, F. rusticus was dominant over F. immunis even when F. immunis had a moderate size advantage and despite some mismatches by sex or male reproductive form that affect crayfish behavioral interactions (Bovbjerg 1956; Figler et al. 2005). Although F. immunis is an emerging invasive species in Europe (Chucholl and Chucholl 2021; Herrmann et al. 2022), its spread and ecological effects in northern Wisconsin have likely been limited by biotic interactions with F. rusticus and other organisms (e.g., competitive dominance by F. rusticus may expose F. immunis to increased fish predation due to limited shelter access; Garvey et al. 1994; Hill and Lodge 1994). To reduce the intensity of these biotic interactions (Creed 2006), F. immunis often uses habitats like ephemeral wetlands or intermittent streams rather than permanent waterbodies (Bovbjerg 1970; Smith et al. 2019; Newkirk et al. 2023). As such, the distribution of F. immunis in our study region could be under-estimated due to a monitoring emphasis for crayfish on large, permanent lakes with both crayfish competitors and fish predators (Larson et al. 2019; Perales et al. 2021).

We sampled streams and rivers of the Manitowish River watershed in summer of 2021 to survey for F. immunis populations outside of lake ecosystems commonly monitored in Vilas County, Wisconsin. We did not detect F. immunis from any of these stream or river sites, including relatively intensive trapping of the Trout River flowage downstream of Wild Rice Lake. This suggests that F. immunis is not spreading out of Wild Rice Lake into downstream ecosystems at present, although more widespread, intensive sampling focused on habitats neglected for crayfish monitoring in Wisconsin may reveal other F. immunis populations (Olden et al. 2011; Newkirk et al. 2023). Like Capelli and Magnuson (1983), we interpret F. immunis as an isolated, non-native population in Wild Rice Lake given its remote location relative to F. immunis populations in southern Wisconsin (Creaser 1932; Hobbs and Jass 1988). Faxonius immunis was similarly absent or undetected by systematic crayfish sampling in the nearby Upper Peninsula of Michigan in the 1970s, but has been documented from two locations within the past decade (Smith et al. 2019). Phylogeographic or landscape genomics studies might be used to investigate whether F. immunis is a non-native species in northern Wisconsin and the Upper Peninsula of Michigan or instead a rare, under-sampled native species (Larson et al. 2012; Schmidt et al. 2023). This seems unlikely, as F. immunis is not difficult to detect by common methods, like baited trapping, relative to congeners within its native range (O’Shaughnessey et al. 2021; Smith et al. 2019; Newkirk et al. 2023). Regardless, future molecular studies could provide insights into F. immunis distributions and ecological interactions, including the possibility that strong effects of invasive F. immunis in Europe may be unique to this highly diverged lineage or possible cryptic species (Filipová et al. 2011).

Species identification challenges could also account for some uncertainty around F. immunis distributions in northern Wisconsin and elsewhere, as evidenced by our misidentification of four F. virilis individuals from Douglas Creek, Illinois as F. immunis prior to mtDNA barcoding. Crayfishes of the family Cambaridae should be identified using male form I gonopods (e.g., Taylor et al. 2015). In the seasonal absence of reproductively active male crayfish, biologists often use life colors or other morphology to make species identifications. Faxonius immunis generally differs from F. virilis by a pale, continuous dorsal stripe extending the length of the carapace and abdomen as well as a deep incision at the base of the moveable finger of the claw (Taylor et al. 2015). Other identification sources (e.g., Chucholl et al. 2008), including a crayfish identification guide from the Wisconsin Department of Natural Resources (Roesler 2019), propose tufts of hairs or setae at the base of the moveable finger of the claw as diagnostic of F. immunis. However, two crayfish we sequence-confirmed as F. virilis from outside of Wild Rice Lake in Vilas County, Wisconsin also had hairs at the base of the moveable finger of the claw, which had motivated our mtDNA barcoding of these individuals. These crayfish otherwise resembled F. virilis by life color and morphology. Conversations with several crayfish taxonomists suggested that hairs on the claws should not be used for F. immunis identifications, as F. virilis can routinely share this morphological trait (Christopher A. Taylor, Illinois Natural History Survey and Bronwyn W. Williams, North Carolina Museum of Natural Sciences; personal communications). Molecular sequencing may assist identification and monitoring of F. immunis and other crayfish populations, although we caution that mtDNA barcoding cannot detect hybrids that may be common among some crayfishes of the genus Faxonius (Rozansky et al. 2021; Merovich et al. 2022). We do not at present suspect F. immunis of hybridization with either F. rusticus or F. virilis given their phylogenetic distance (Figure 2), whereas the more closely related F. rusticus and F. propinquus are well-documented to hybridize in lakes of northern Wisconsin (Perry et al. 2001).

Our study could be interpreted to support F. immunis as an innocuous non-native species, rather than a harmful invasive species, in Wild Rice Lake. Invasive species spread in new regions (Blackburn et al. 2011); by contrast, F. immunis has not apparently spread over the half century since its discovery in Wild Rice Lake by Capelli and Magnuson (1983). Invasive species often achieve hyper-abundance resulting in negative ecological impacts (Parker et al. 1999; Hansen et al. 2013a); by contrast, F. immunis has remained rare in Wild Rice Lake over time (Larson et al. 2019; Figure 1). One explanation for the lack of success of F. immunis in the Manitowish River watershed may be poor competitive ability against the widespread, abundant F. rusticus. Even if F. rusticus population declines in this region continue, F. immunis may be unlikely to become an invasive species. For example, F. rusticus has been present but rare in Wild Rice Lake for decades (Larson et al. 2019), and this lake evidences minimal community or ecosystem impacts of F. rusticus (Szydlowski et al. 2023). Faxonius immunis has been unable to take advantage of these low abundances and weak ecological effects of F. rusticus in Wild Rice Lake by increasing its population size or spreading. Factors that have limited more robust population growth of F. immunis within Wild Rice Lake over the past half century are unknown. Further, populations of the native crayfish F. virilis may recover as F. rusticus declines (Perales et al. 2021), and F. virilis is also competitively dominant over F. immunis (Bovbjerg 1970). Lastly, the still-uncertain causes of population declines for F. rusticus, like pathogens or fish predation, may also be rendering these lakes less suitable for subsequent crayfish invaders (Sargent et al. 2014; Roth et al. 2007). A minority of non-native species are able to spread or cause negative impacts as invasive species (Williamson 1996; Jeschke 2014), and managers and policy-makers use this distinction to prioritize which organisms require prevention, control, or eradication efforts (Lodge et al. 2016). In our study region, managers and policy-makers may be well-justified in identifying F. immunis as a species that is unlikely to become invasive in the future.

Alternatively, time lags between the introduction and establishment of a non-native species and subsequent spread or negative impacts as an invasive species can complicate management decisions and policy priorities (Sakai et al. 2001; Crooks 2005). Long-present non-native species may become invasive following major abiotic or biotic changes (Spear et al. 2021). We propose that population declines of F. rusticus may create opportunities for serial or over-invasion by subsequent crayfish species (Russell et al. 2014; Karatayev et al. 2023), including F. immunis as a possible sleeper invader in the future (Spear et al. 2021). For example, invasive status in another region is often an accurate predictor of invasion risk in a focal region (Gordon et al. 2008). Faxonius immunis is invasive in Europe, where it competes effectively against other invasive crayfish of the genus Faxonius (Chucholl et al. 2008) and has community and ecosystem effects on benthic macroinvertebrates comparable to F. rusticus (Chucholl and Chucholl 2021; Herrmann et al. 2022). We also suggest that future crayfish invaders in northern Wisconsin may not necessarily resemble F. rusticus by habitat preference, life history, or ecological function. Manoomin or Wild Rice (Zizania palustris L.) is an example of a species that could be vulnerable to a novel crayfish invader in lakes of Vilas County. Zizania palustris is a culturally significant resource to Native American tribes in the Great Lakes region but has experienced recent declines (Matson et al. 2021). Faxonius rusticus has been investigated as a risk to Z. palustris through direct consumption or habitat modification (Wenner 2017), but this crayfish generally requires firm substrates distinct from the softer substrates where Z. palustris grows (Hill and Lodge 1994). By comparison, F. immunis tolerates soft substrates and low dissolved oxygen concentrations typical of some wetlands (Bovbjerg 1970; Smith et al. 2019), and was collected in our study from Wild Rice Lake adjacent to Z. palustris habitat. Failure to compete effectively with F. rusticus may not exclude a subsequent crayfish invader from having novel impacts on other ecosystem processes or aquatic community members like Z. palustris. A risk-averse management strategy might choose to control or eradicate F. immunis from Wild Rice Lake while its population remains small and localized (Vander Zanden et al. 2010; Lodge et al. 2016).

Lakes of northern Wisconsin have experienced a sequence of natural, post-glacial colonization and human-assisted invasion by crayfishes that will likely continue following declines of F. rusticus populations. Creaser (1932) found only F. virilis and the semi-terrestrial burrowing crayfish Lacunicambarus nebrascensis (Girard, 1852) from our study region in the early 20^th century. By Capelli and Magnuson (1983), F. propinquus and F. rusticus had sequentially invaded these lakes due to human introductions, with F. rusticus ultimately displacing F. propinquus by a combination of competition, differential vulnerability to fish predation, and hybridization (Garvey et al. 1994; Hill and Lodge 1994; Perry et al. 2001). As F. rusticus experiences ongoing population declines (Larson et al. 2019; Szydlowski et al. 2023), other crayfish species may arrive by tracking their preferred environmental conditions under climate change (Hellmann et al. 2008) or as a consequence of continuing human introductions (Seebens et al. 2017). Faxonius immunis offers managers and policy makers a thought exercise for how to meet these future crayfish arrivals to north temperate lakes. For example, the resist-accept-direct (RAD) framework offers alternative approaches for how to address both climate-tracking species (Feiner et al. 2022; Rahel 2022) and biological invasions (Dunham et al. 2022). Faxonius immunis, for example, could be resisted by an eradication effort (Hansen et al. 2013b), accepted with no management intervention, or directed away from sensitive habitat for Z. palustris by barrier design or maintenance (Cowart et al. 2018). Managers and policy makers should apply the well-documented lessons of F. rusticus invasions in northern Wisconsin lakes in preparing responses to future crayfish invasions, while also considering that future crayfish invaders may not necessarily resemble F. rusticus given the considerable ecological diversity of this taxonomic group (Taylor et al. 2019).

Funding was provided by a University of Illinois ASPIRE fellowship to EKS and United States Department of Agriculture Hatch Project ILLU-875-984 to ERL.

Crayfish were collected under fishing licenses consistent with Illinois and Wisconsin regulations.

EKS: Funding Acquisition, Conceptualization, Investigation, Data Curation, Formal Analysis, Writing - Original Draft, Writing - Review and Editing; JHH: Methodology, Data Curation, Formal Analysis, Writing - Review and Editing; DKS: Conceptualization, Investigation, Writing - Review and Editing; ERL: Funding Acquisition, Conceptualization, Investigation, Writing - Review and Editing, Supervision