Invasive herpetofauna pose a threat to biodiversity through disruption of native species interactions (Rice et al., 2011; Richter-Boix et al., 2013; Culbertson and Herrmann, 2019; Hill et al., 2019), displacement of native species through predation (Dove et al., 2011; Piquet and López-Darias, 2021), and the spread of infectious diseases (Rivera et al., 2019; Atkinson and Savage, 2023). While over 780 introductions of non-native herpetofauna were documented globally as of 2015, only 250 species have established as independent populations (Kraus, 2015). The disparity between introduced and established invasive species underscores the importance of life history, environmental variation, propagule pressure and genetic diversity on the ability of an invasive population to establish and thrive. Typically, generalist species are more likely to successfully invade and establish compared to specialists that have more specific habitat requirements (Pitt et al., 2005), and environments more like the invasive species native range tend to be conducive to establishment (Bomford et al., 2009). Introductions of non-native herpetofauna largely result from anthropomorphic activity, such as the pet trade (Willson et al., 2011), farming (Laufer et al., 2008), bio-control efforts (Shanmuganathan et al., 2010), and stowaways in cargo shipments (Richmond et al., 2015). Thus, the numerous factors influencing which species become introduced combined with the complexity of novel ecosystems present a challenge for predicting whether and which non-native amphibian and reptile species will become established.

Invasive populations can arise from novel founder events or expansions of pre-existing invasive ranges, and each of these mechanisms should produce distinct patterns of genetic diversity in the invasive population. Founder events are generally expected to decrease genetic diversity and increase inbreeding depression (Frankham, 1998), which has been observed in the case of the coqui (Eleutherodactylus coqui) invasion in Hawaii (Peacock et al., 2009) and the American bullfrog (Rana catesbeiana) invasion in China (Bai et al., 2012). Contrary to this expectation, invasive populations founded by multiple introduction events or with high propagule pressure (i.e., many individuals introduced in one or more release events) (Rius and Darling, 2014) can maintain genetic diversity at adaptive loci (Selechnik et al., 2019; LaFond et al., 2022) or even show higher overall genetic diversity relative to native populations (Allendorf and Lundquist, 2003; Kolbe et al., 2004). A genetically admixed invasive population has increased adaptive potential and often is capable of faster range expansions (Wagner et al., 2017) and adaptation to novel environmental conditions (Krehenwinkel and Tautz, 2013; Rius and Darling, 2014).

Mitochondrial DNA (mtDNA) is commonly used to establish phylogenetic relationships between populations and trace the origins of invasive populations (Avise et al., 1987; Hamner et al., 2007; Colangelo et al., 2015) because it has a higher mutation rate than nuclear loci, allowing for increased resolution of genetic variation (Gray et al., 1989). Additionally, mtDNA is easy to isolate, has a relatively simple genetic structure (i.e., lacking transposable elements, pseudogenes and introns) and exhibits a straightforward mode of inheritance (Avise et al., 1987). The mitochondrial gene cytochrome b (cyt-b) codes for a central redox catalytic subunit involved in the electron transport chain that is present in nearly all eukaryotic organisms. The cyt-b gene is used to evaluate genetic diversity, trace geographic origins, and distinguish cryptic invasive species across vertebrate groups (Kochzius et al., 2003; Hamner et al., 2007; Colangelo et al., 2015; Jackson et al., 2015). Cyt-b data have been used to investigate the origins of several invasive herpetofauna, including the novel invasion of two subspecies of Boa constrictor in Puerto Rico (Reynolds et al., 2013), the origin and identity of cryptic Xenopus species in Chile (Lobos et al., 2014) and Italy (Lillo et al., 2013), the invasion of the greenhouse frog (Eleutherodactylus planirostris) in Florida (Heinicke et al., 2011) and China (Hong et al., 2022), and the invasive American bullfrog (R. catesbeiana) across the western United States (Bai et al., 2012; Kamath et al., 2016; LaFond et al., 2022).

The Cuban treefrog (Osteopilus septentrionalis) is one of the most successful invasive amphibians in the state of Florida (Meshaka, 2001). Invasive populations negatively influence native Florida treefrogs through predation (Wyatt and Forys, 2004), and by influencing pathogen dynamics (Galt et al., 2021; Atkinson and Savage, 2023). The native O. septentrionalis range spans Cuba, the Cayman Islands and the Bahamas, with invasive populations first establishing in the Florida Keys in the late 1800s to early 1900s (Duellman and Schwartz, 1958; Heinicke et al., 2011). Subsequently, O. septentrionalis spread to peninsular Florida by the mid-1900s via multiple anthropogenic introductions (Heinicke et al., 2011) and has now spread throughout much of central and northern Florida and in some areas of the Florida panhandle ( Figure 1 ; Johnson, 2023). Sporadic sightings of O. septentrionalis have been recorded along the southern coastline of Alabama and Mississippi (Morningstar et al., 2024), but no established populations have been reported. Cyt-b from O. septentrionalis has been used to evaluate the ecological drivers of haplotype and species diversity (Rodríguez et al., 2015). Akin to the previous herpetological invasions mentioned above, the origin of invasive O. septentrionalis in Florida was investigated using cyt-b genetic analysis (Heinicke et al., 2011). The resulting mitochondrial phylogeny revealed a deep phylogenetic split among haplotypes of O. septentrionalis, with both clades represented throughout the native range in Cuba and the invasive range in Florida. This finding indicated that current populations are descended from at least two source populations (Heinicke et al., 2011).

In 2017, the first established breeding population of O. septentrionalis in Louisiana, USA, was observed in New Orleans (Glorioso et al., 2016, 2018c). Although O. septentrionalis had been sporadically documented in Louisiana since the 1990s, no breeding colonies had been observed prior to this event (Chatfield and Vance, 2014; Glorioso et al., 2018b, 2018c). In contrast, the discovery of individuals in different age classes, ranging from tadpoles to metamorphs, were found in Audubon Zoo and the adjacent Riverview Park beginning in 2016, leading to further investigation of this invasion event (Glorioso et al., 2018c). Subsequently, a second breeding population was located in St. Charles Parish (Glorioso et al., 2018a). While the novel New Orleans invasion coincided with the arrival of palms that were brought from Lake Placid, Florida for an Audubon Zoo exhibit (Glorioso et al., 2018c), given the occasional observations of O. septentrionalis prior to this event, and that O. septentrionalis are regularly dispersed via anthropogenic activities (Palis et al., 2021), a definitive source for this invasion remains unclear.

Here, we used cyt-b mitochondrial genetics to evaluate genetic variation in the recently established O. septentrionalis population at the Riverview Park and St. Charles Parish sites in Louisiana and assess likely routes of introduction. We sampled and generated cyt-b haplotypes from 31 east Florida individuals and 40 west Florida individuals within central Florida, and 95 individuals in New Orleans, Louisiana, then combined these new data with previously generated sequences from throughout the native and invasive range (Heinicke et al., 2011; Rodríguez et al., 2015). We conducted phylogenetic and haplotype network analyses and compared metrics of genetic diversity in Louisiana to other native and invasive populations. We used these data to assess whether the Louisiana invasion represents a single introduction event or a more complex history of population establishment.

We collected and processed a total of 166 O. septentrionalis between 2017 and 2022. Of those, 31 were from east Florida, 40 from west Florida, and 95 individuals were from Louisiana populations. East Florida samples were collected from one site in Orange County and west Florida samples were collected from one site in Hillsborough County. Louisianan samples originated from two sampling sites; 85 individuals were from Audubon Riverview Park and the adjacent Audubon Zoo in Orleans Parish, while 10 individuals were from 15 kilometers west in St. Charles Parish ( Figure 1 ).

We generated robust cyt-b sequences from all 166 frogs we sampled in west Florida, east Florida, and Louisiana, recovering 14 total haplotypes. Of these, nine are novel (GenBank Accession numbers: PP320231.1 – PP320239.1) and five were previously published (Heinicke et al., 2011; Table 1 ). Novel haplotype se86 was the most common sequence found in our sampling, occurring in 49% of all frogs ( Table 1 ).

In Louisiana we recovered seven haplotypes from 95 individuals, including two previously published haplotypes ( Table 1 ) and five novel haplotypes. Among the five novel haplotypes, se89, se90, and se94 were private alleles recovered exclusively from Louisiana, each from a single individual ( Table 1 ). Novel haplotype se87 was the most common haplotype in Louisiana (N = 37 individuals) and it was also found in one east Florida frog. Novel haplotype se86 was the second most common haplotype in Louisiana (N = 36 individuals), and we also found it at high frequency in west and east Florida ( Figure 2 ). All ten individuals collected from the St. Charles Parish population had this haplotype. Haplotypes se3 (N = 17) and se29 (N = 1) were recovered exclusively in Florida in previous studies. None of the haplotypes present in Louisiana occurred in frogs from Cuba or the Caribbean (electronic supplemental material, Supplementary Table S1 ).

The 31 frogs we sampled from east Florida had seven haplotypes, including four novel haplotypes (se92, se92, se86, se87) and three previously published haplotypes (se13, se17, se29). Only two individuals had the novel private haplotype se91, and one individual had the novel private haplotype se92. Most east Florida individuals had the novel haplotype se86 (N = 16), which was also found in west Florida and Louisiana. One east Florida individual had the novel haplotype se87, which was the most common haplotype in Louisiana. The three haplotypes se13, se17, and se29 were previously recovered from west Florida and west Cuba, south Florida and the Caribbean, and west Florida and south Florida, respectively (electronic supplemental material, Supplementary Table S1 ). Here, we recovered all three of these haplotypes the first time in east Florida.

In west Florida, we recovered seven haplotypes from 40 individuals: three novel haplotypes (se86, se88, se93) and four previously published haplotypes (se3, se7, se13, se29). A single individual had the private novel haplotype se93 and four individuals had the private novel haplotype se88. The novel haplotype se86 was the most common haplotype in west Florida, similar to its high frequency in Louisiana and east Florida. Se3 was previously recovered only in west Florida, but here we found it in one west Florida frog and in 17 Louisiana frogs. We found se7 in west Florida for the first time, but only in one frog. This haplotype was previously found in a single south Florida individual (electronic supplemental material, Supplementary Table S1 ). We found two other low-frequency haplotypes (se13 and se29) that had previously been detected in Florida and Cuba (electronic supplemental material, Supplementary Table S1 ).

Our haplotype network illustrates that almost all frogs sampled in Louisiana had haplotypes se3, se86, or se87 ( Figure 2 ). These sequences are only found in Louisiana, east Florida (se86 and se87), and west Florida (se3 and se86). Only one Louisiana frog had a haplotype that is recovered more broadly throughout Florida (se29), and this sequence was many mutational steps away from all other haplotypes found in Louisiana. Private Louisiana haplotypes se89 and se90 were respectively three and one mutational steps away from se86. Likewise, private Louisiana haplotype se94 is six mutational steps from se87, a haplotype only rarely found outside of Louisiana. Collectively these patterns suggest the evolution of new haplotypes within the invasive Florida and/or Louisiana populations.

Phylogenetic analysis of O. septentrionalis mitochondrial sequences revealed minimal geographic structure ( Figure 3 ). We recovered a well-supported basal split between two sequences only found in the Caribbean and all other sequences. The next split was also well-supported and separated one sequence private to Cuba with everything else. All other sequences were separated into two large clades with moderate support (posterior probability = 0.8917). Both of these clades contained numerous haplotypes found in multiple localities through the native and invasive range. Haplotypes sampled from Louisiana frogs were dispersed throughout the phylogeny.

Among all known cyt-b sequences, Cuba contained the highest number of haplotypes. Frogs sampled in west Cuba had more haplotypes than any other location, followed by east Cuba and central Cuba ( Table 2 ). The Caribbean, Louisiana, and Florida localities each contained similar numbers of haplotypes ( Table 2 ). Native Cuban localities contained similar percentages of private haplotypes, which were high in comparison to the invasive range. Private haplotypes made up 20–30% of the haplotypes recovered in the invasive range localities of east Florida, west Florida, and south Florida whereas 42.86% of Louisiana haplotypes were private ( Table 2 ). Nucleotide diversity in Louisiana was highest among invasive locations in the United States ( Figure 4 ). Nucleotide diversity in Louisiana (0.039; 95% CI = 0.033–0.056) did not overlap with nucleotide diversity of central Cuba (0.017; 95% CI = 0.0092–0.021) or east Cuba (0.018; 95% CI = 0.010–0.023; Table 2 ; Figure 4 ). Segregating sites were higher in west Cuba and the Caribbean than within invasive localities in the United States ( Table 2 ).

Our study is the first to investigate the origins of the establishment of O. septentrionalis in Louisiana, the first documented establishment of O. septentrionalis in the contiguous U.S. outside of Florida (Glorioso et al., 2018c). We aimed to determine whether this invasion was the result of one or more founding events from remote locations. Our results indicate both Louisiana O. septentrionalis populations are the result of multiple introduction events, likely from central Florida localities. Four of seven haplotypes found in Louisiana individuals were present only in Florida and Louisiana, while the remaining three were rare and only found in Louisiana. This pattern excludes Caribbean locations as direct sources of the novel population and strongly points toward Florida origins. One haplotype lending the most support to our introduction hypothesis is se86, which is novel to our study ( Table 1 ). It was the most abundant haplotype we recovered overall and was at high frequency in our Florida and Louisiana sample sites, consistent with the hypothesis that central Florida is a source of the Louisiana O. septentrionalis.

Though we did not sample any localities in the Florida panhandle, the likelihood that frogs spread naturally from Florida to Louisiana in a range expansion is minimal. The high nucleotide diversity of haplotypes in Louisiana is inconsistent with the expectation for a range expansion. Range expansions of invasive species typically result in decreased genetic diversity as the leading edge expands (Wilson et al., 2005; Davis et al., 2011). Instead, we see higher nucleotide diversity in Louisiana compared to all locations except for west Cuba and the Caribbean ( Figure 4 ), which is where the species originated (Heinicke et al., 2011; Johnson, 2023). Notably, the nucleotide diversity in Louisiana is nearly equivalent to values found in west Cuba. Furthermore, there is no evidence for a leading edge, as established populations of O. septentrionalis between the Florida panhandle and the Louisiana population have not yet been documented (Johnson, 2023). Future work that samples O. septentrionalis from the Florida panhandle and the Gulf Coast of Alabama and Mississippi, and analyzes their genetic composition could enable assessment of potential new population establishments and identification of their source populations.

High genetic diversity metrics in the Louisiana population suggests a high propagule pressure invasion and/or multiple introduction events (Kolbe et al., 2008; Simberloff, 2009; Wang et al., 2019). Given that human transport is the most plausible route of invasion, multiple founding events is the most likely scenario explaining the patterns we recovered in Louisiana. For example, repeated transport of one or a few individuals traveling on ornamental plants purchased from locations within Florida and planted in New Orleans is a plausible source of the novel invasion. Anecdotal accounts of O. septentrionalis being found within many purchased plants from Florida (Johnson, 2023), the scale and economic importance of the nursery industry in Florida (Khachatryan et al., 2022), and low efficacy of state-level biosecurity measures in preventing the spread of established invasive species (Paini et al., 2010) lend support to this hypothesis. Therefore, it is plausible that the known planting of ornamental palms from Lake Placid, Florida in the Audubon Zoo contributed to this invasion (Glorioso et al., 2018c). Sampling in or around the nursery in Lake Placid—which lies farther south than our sampling sites—to verify the presence of haplotypes shared with Louisiana could strengthen support for this hypothesis. Additionally, increased sampling at or near ornamental horticultural nurseries and retailers in top economic contributing regions, such as central and south Florida (Hodges et al., 2016), could give greater insight into the impacts of this industry on the spread of O. septentrionalis.

Human traffic near the invasion site has the potential to facilitate introductions from distant source populations. New Orleans is a large city that is visited by nearly 20 million tourists annually, with many visitors arriving via vehicular transportation or large cruise ships that depart from docks approximately 5 miles from the invasion site (Adams, 2024). Florida is a frequent point of origin, ranking as the 3^rd most common origin of interstate travel entering Louisiana in 2023 (Adams, 2024). As vehicular transport is the primary mode of invasion of O. septentrionalis (Johnson, 2023), it is plausible that multiple instances of individual frogs hitch-hiking within vehicles departing from central Florida have contributed to the New Orleans invasion. Additionally, Interstate 10 (I-10), a major thoroughfare linking Florida to Louisiana, provides direct access to New Orleans and is heavily utilized by Floridians during hurricane evacuations (Ghorbanzadeh et al., 2021). While hurricane evacuees likely are minor contributors in comparison to tourists, increasing hurricane severity and resulting evacuation efforts may increase the frequency of these individual translocation events. Additionally, cruise ships departing and arriving at the Port of New Orleans frequent many Caribbean islands and various Floridian ports. Caribbean sampling of O. septentrionalis is minimal, leaving unclear resolution on what haplotypes are present in this region. Increased Caribbean sampling may reveal haplotypes from this study found exclusively in Louisiana may also occur in the Caribbean and may be derived from there. Overall increased sampling efforts throughout Florida, as well as the Caribbean could provide increased clarity on the contribution of translocation events to the founding of the Louisianan population.

Heinicke et al. (2011) previously reported two divergent clades of O. septentrionalis to be phylogeographically grouped based on relation to the Guanahacabibes peninsula in remote west Cuba. They additionally reported Bahamian haplotypes form a basal monophyletic group. With the additional samples of Rodríguez et al. (2015) as well as our novel sequences, our phylogenetic analysis groups together west Cuban haplotypes within and outside of the Guanahacabibes peninsula together, indicating this peninsular boundary does not impede gene flow to other localities in Cuba. Our study does not expand on the Heinicke et al., 2011 finding of Bahamian samples forming a monophyly, as we did not include additional Bahamian sampling outside of Heinicke et al. (2011) providing another reason for additional sampling in the Caribbean and specifically the Bahamas.

We expect that O. septentrionalis will continue to negatively affect native species through resource use and influences on local pathogen dynamics, but changes in the genetic composition of O. septentrionalis populations are a potential cause for concern. Anecdotal sightings of O. septentrionalis outside their established invasive territory are plentiful and range from Georgia to as far north as Vermont along the east coast of the United States (Morningstar et al., 2024). These sightings are likely a result of anthropogenic transport either via horticulture or other items transported from their invasive range to wider regions of the U.S. and are not yet representative of established breeding colonies. Cold temperatures play a large role in limiting the northern distribution of O. septentrionalis, which have been observed dead at temperatures below 0 °C (Haggerty and Crisman, 2015), but shifting global temperatures threaten this climatic barrier to expansion. With fewer freeze events occurring in regions north of the current range of O. septentrionalis, these populations may persist and establish farther north than previously expected (Osland et al., 2021).

Admixed populations at range limits may also benefit from an increased evolutionary potential resulting from heterosis, thereby increasing the extent of dispersal and probability of permanent establishment (Forsman, 2013; Wagner et al., 2017). Novel genotypes resulting from heterosis (Verhoeven et al., 2011) may allow for adaptation to colder temperatures. Evidence shows O. septentrionalis from the northern latitudinal limits of their Florida range had critical thermal minima trending lower than their southern counterparts and that in general O. septentrionalis has a higher tolerance for cold weather than previously believed (Simpson, 2013). Though cold tolerance tends to result from phenotypic plasticity (McCann et al., 2014; Mittan and Zamudio, 2019), it is possible that exposure to freeze events and the existence of novel genotypes within northern populations result in adapted cold tolerance. Though we observed patterns suggesting local adaptation of the cyt-b mitochondrial gene in Florida and Louisiana, more work could help to assess adaptation in functionally relevant genes. Future experimental studies on mitochondrial gene transcription associated with thermal regulation across invasive populations could provide deeper insight into thermal adaptation (Hong et al., 2024).

Disparities in size between northern and southern populations of O. septentrionalis have also been recorded (McGarrity and Johnson, 2009). Studies do not yet indicate if this shift to smaller sizes in colder latitudes is genetically based or if these changes result from environmental effects such as shifts in food availability or metabolic constraints resulting from cold temperatures. It is possible that smaller body size and increasing urban development may aid in their spread, as O. septentrionalis can exploit urban refugia to escape or minimize exposure to thermal temperature limits (Haggerty and Crisman, 2015; Meshaka, 2001).

Invasive species present a unique opportunity to study evolution in action by assessing novel selection pressures in situ. These systems can serve as natural experiments, with introduced populations functioning as treatment groups and native populations acting as controls. Invasive O. septentrionalis have the potential to serve as a powerful model for studying rapid adaptation to novel environmental conditions, especially in the recently established Louisiana population. While mitochondrial genes markers have historically served as a valuable tool in determining the origin of invasive species, next-generation sequencing (NGS) techniques offer better resolution to fine scale population structure and more robust findings on this recent population divergence (North et al., 2021). Whole mitochondrial genome analysis can expand on findings derived from single genes for more comprehensive insights as demonstrated with Rhinella marina, the invasive cane toad (Cheung et al., 2024). Like many amphibian species, O. septrentrionalis has not yet had its nuclear or mitochondrial genome sequenced, due in part to the difficulty of assemblage (Kosch et al., 2024; O’Connell et al., 2024) inhibiting the depth of evolutionary questioning utilizing this species.

Our study concludes that the newly established Louisiana population of O. septentrionalis is derived from multiple introductions, primarily from central Florida. These introductions likely occurred via anthropogenic means (Johnson, 2023), though we cannot conclude with certainty whether the transport of flora from Florida to Louisiana contributed to this establishment. To fully resolve the origin of Louisiana’s population, future work could prioritize increased sampling across the Florida panhandle, Caribbean, and key Floridian localities. By identifying the source populations and the degree of admixture within the Louisiana population, we gain insight into where management efforts could be prioritized and understanding of the genetic variation available for selection to act upon. In the case of O. septentrionalis, this knowledge can be used not only for reconstructing invasion history, but could also anticipate the species’ capacity to adapt to colder climates, urban environments, and other novel pressures.