The Lesser Antilles are plagued by the continuing presence and spread of non-native I. iguana, which is a major threat to the native insular Iguana populations. Here we report on the presence of non-native iguanas on two of the last remaining islands previously believed to lack such invasive populations, Saba and Montserrat. Although no evidence of incursions was found for Montserrat, we did identify a recent non-native incursion on Saba as flagged by the integrated analysis of multiple genetic and morphological data. In addition to scale and coloration patterns, we highlight how morphometrics can aid the in-situ identification of non-native and hybrid iguanas within native populations, especially in absence of rapid genetic analytic techniques.
Passenger- and cargo transport to Saba is mainly channelled through St. Maarten, though private vessels can have other origins. A recently reinstalled ferry service also first passes St. Eustatius, which is home to a native I. delicatissima population. Additionally, a biweekly service delivers cargo from Martin County (Florida, USA), where non-native I. iguana is present (Meshaka et al. 2004; iNaturalist, 2022). On St. Maarten I. delicatissima has already gone extinct and currently a large non-native I. iguana population is present. Based on microsatellite, ND4 and MLH3 data, our results identify that non-native iguanas on Saba have their ultimate genetic origin in Central America, as well as the ABC islands. However, given the high inter-island connectivity with St. Maarten and since iguanas from the same genetic backgrounds have been identified there (van den Burg et al. 2018b), their more immediate origin is presumed to be the large non-native iguana population on St. Maarten. Likewise did two other non-native reptiles that arrived in recent years from St. Maarten (van den Burg et al. 2021d).
Among the Lesser Antillean islands, individual non-native Iguana iguana have been traced back to several native range origins, e.g., mainland South America (Brazil), Central America (e.g., Honduras and El Salvador), the ABC islands, as well as Saba and St. Lucia (Vuillaume et al. 2015; van den Burg et al. 2018b; Breuil et al. 2019; Pounder et al. 2020). In addition, our work as well as that by others show that Lesser Antillean native Iguana populations have at some point or other, also made to other (neighboring) islands (Vuillaume et al. 2015; van den Burg et al. 2018b), be it either by natural or human-mediated dispersal. The demonstrated occurrence of Saba/Montserrat haplotypes on St. Maarten (van den Burg et al. 2018b) particularly shows the need to utilize multiple genetic markers when trying to identify non-native iguanas on Saba, as introgressed Saban iguanas could theoretically re-invade the original native population. Considering mtDNA data across non-native populations present in the Greater (De Jesús Villanueva et al. 2021) and Lesser Antilles (the current study and references above), non-native populations on Anguilla and St. Maarten stand out given their high genetic diversity given the presence of three out of the four major mtDNA clades of the I. iguana complex.
Although 26 novel alleles were identified across 15 loci for Saba and Montserrat, these can present either undiscovered native variation or the presence of a non-native gene pool. As relatively few samples were previously analysed from Saba and Montserrat it is not surprising that we identified novel alleles, also for Montserrat from where our data highlights an absence of non-native iguanas. Considering all loci, L20 stands out as novel alleles were identified on both islands, though we argue that their significance strongly differs. Namely, the 182 allele identified on Montserrat coincide closely with the previously known diversity suggesting this constitutes a rare haplotype. In contrast, the novel alleles from Saba suggest non-native presence given the large gap with known native diversity (but see Estoup et al. 2002). Given the large, combined sample size from Saba (35) and Montserrat (53), especially fixed loci can aid identification of non-native presence. As a total of four variable loci (L3, L8, L16, L17; Fig. 3) are fixed across both island populations (except for several aberrant samples from Saba) this provides additional evidence that non-native iguanas are present on Saba but absent on Montserrat. This was corroborated by our nDNA and mtDNA sequence data.
Even though a complete overview is still lacking, scalation and coloration patterns have already been found to be highly variable throughout the I. iguana complex. We assessed presence of enlarged nasal scales and a melanistic facial patch in both the Saba and Montserrat population, given previous morphological assessments of these populations (Breuil et al. 2020; van den Burg et al. 2022). Among both populations only three Saba animals had nasal horns and were additionally assigned as non-native by microsatellite and sequence data. Similarly, these three iguanas, including one other individual from Saba, genetically identified as non-native, lacked a clear melanistic black patch (though two had some faint partial black scales). On Montserrat several adult iguanas genetically assigned as native had a patch with roughly 50% melanistic scales, while two individuals only had 10–20% melanistic scales. In other words, the melanistic patch of native Montserrat animals was less pronounced than in native animals of Saba. As we found no genetic evidence of non-native presence on Montserrat, we believe that the less evident melanistic patch is a native characteristic of the Montserrat population. This is corroborated by Breuil et al. (2020), who indicated that the Montserrat population is less melanistic.
In order to help distinguish non-native from native iguanas in invaded populations, we here explored the use of novel tools, in addition to molecular data (Stephen et al. 2013; Martin et al. 2015; Vuillaume et al. 2015; van den Burg et al. 2018, 2021; Miller et al. 2019; Pounder et al. 2020; Breuil et al. 2022; Mitchell et al. in prep) and scale and color pattern characters (Breuil 2013). Size-dependent comparisons across 16 variables indicated that on Saba non-native iguanas have larger subtympanic plate scales than native iguanas (Figs. 5 and 6). The diagnostic power and usage of these characteristics in distinguishing between non-native and native iguanas will depend not only on sample size but on the geographic and taxonomic origin and admixture of non-native individuals. In short, we hypothesize that if four iguanas with different genetic origins and admixture were to arrive on Saba, other SVL-dependent variable(s) could be used to identify non-native animals. Therefore, additional analyses on intraspecific variation among the measured variables in I. iguana should be useful (and are ongoing; van den Burg et al. unpublished data). This approach is likely most useful to apply in small insular populations as opposed to widely-connected mainland populations given the former tend to show less variation and given their vulnerability to non-native incursions. However, ex-situ, within the (illegal) pet trade (Noseworthy 2017; van den Burg and Weissgold 2020), these variables will remain of low value until a range-wide and high-sampled dataset is present. Therefore, we suggest that size-dependent variables will be of greatest value in distinguishing between more distantly related species and that they should especially be evaluated for use in conservation management for I. delicatissima.
While the initial presence of non-native species can be expected to be highest around the incursion point, an initial presence at locations more distant from the incursion point can occur when biosecurity regulations are bypassed (Brisbane et al. 2020). Saba knows two principal incursion points to the island. These are the Fort Bay harbor and the Juancho E. Yrausquin airport, situated on the southwest and northeast sides of the island, respectively. Non-native iguanas were mainly located at a site 0.4 km east from the harbor, although one animal was found north of the settlement of Windwardside (Fig. 2). How this latter individual reached the center of the island is impossible to determine precisely. We speculate that it might concern an intentionally and illegally released pet (Jesse et al. 2016) arriving either via the airport or the habour, or that it might have arrived at the harbour as an intra-island stowaway that originally arrived at the harbor. Occurrence of three non-native iguanas near the harbor strongly suggest their arrival via transport by boat or ship, which in the case of the neighboring island of St. Eustatius, also appears to be the principal way of incursion of non-native iguanas (Debrot et al. 2022).
Introgressive displacement has been identified as the main threat to I. delicatissima (Vuillaume et al. 2015). Hybrid and non-native iguanas are believed to be more territorially aggressive, and have higher fitness thanks to larger clutch sizes compared to I. delicatissima (van Wagensveld and van den Burg 2018). Although data on clutch size from the Montserrat and Saba populations are practically unavailable (Blankenship 1990; van den Burg et al. 2022), insular Iguana populations have generally been found to have smaller clutch sizes than continental populations (e.g., Fitch and Henderson 1977; van Marken Lichtenbelt and Albers 1993; Bock et al. 2018), even though this is assumed to be partially climate dependent (e.g., Novosolov et al. 2013). The available data suggest that the Montserrat and Saba population produce clutch size within the lower range recorded for I. iguana, potentially allowing reproductive and genetic swamping when non-native or hybrid iguanas would reproduce within these island populations.
Whether hybridization has occurred on Saba remains unresolved given our limited data, though preliminary data could suggest it is taking place. The presence of non-native iguanas on St. Maarten with a (partial) I. i. iguana origin limits our ability to assess whether the I. i. iguana nDNA from the three admixed iguanas on Saba represents native or non-native DNA. Although SAB58 could be the maternal parent of SAB08 and 57 given mtDNA, nDNA (MLH3) and body size, our microsatellite allele data rejects this hypothesis. All data combined, suggests that none of the non-native iguanas had an identical pair of parents. The relatively low number of non-native iguanas observed on Saba implies that on-island hybridization might still be absent or rare. However, the presence of two iguanas with faint melanism with the facial patch could further suggests that these animals are hybrids. Additional fieldwork and genetic assessment is needed to determine whether hybridization is still truly absent.
Our results in any case show that there is a pressing need for rapid action to completely remove non-native iguanas on Saba, and that prospects for this are relatively good in light of apparently lacking or only limited hybridization. However, only immediate action can result in the removal of non-native iguanas before they start interbreeding (Debrot et al. 2022). Failure to act at an early stage of the invasion process elsewhere has led to strong growth of invasive iguana populations, and serious existential threats to or even extinction of the original native iguana populations e.g., on Grand Cayman (Rivera-Milán and Haakonsson 2020) and in the French West Indies (Angin 2017). More extensive size-dependent morphometric data collection from the native population will additionally be essential to further help discern differences with non-native iguanas, which can help to rapidly identify and cull non-native and or hybrid iguanas.
We conclude by pointing out that our findings illustrate the need for continuous monitoring for non-native iguanas within the remaining native Iguana populations in the Lesser Antilles. Although genetic data from across the native Iguana range currently allows identification of non-native or hybrid presence, this field-to-lab workflow is often excessively time consuming. At present genetic assessment requires the shipment of samples and involves delays due the required permitting processes (e.g. CITES permits). Improved in-situ identification screening to rapidly distinguish non-native from native individuals is urgently needed. Here, in addition to scale and coloration patterns, we show size-dependent morphological variables can also be utilized for rapid screening. We urge our regional partners to build morphometric databases for native Iguana populations as soon as possible that can help to rapidly distinguish non-native iguanas for culling because even if biosecurity is eventually improved, incursions will from time to time continue to take place.