Pomacea Spp.) in Florida Stephanie A

Nova Southeastern University NSUWorks

HCNSO Student Theses and Dissertations HCNSO Student Work

12-7-2017 Forecasting the Spread and Invasive Potential of Apple Snails (Pomacea spp.) in Florida Stephanie A. Reilly Nova Southeastern University, [email protected]

Follow this and additional works at: https://nsuworks.nova.edu/occ_stuetd Part of the Aquaculture and Fisheries Commons, Oceanography and Atmospheric Sciences and Meteorology Commons, Other Animal Sciences Commons, and the Terrestrial and Aquatic Ecology Commons Share Feedback About This Item

NSUWorks Citation Stephanie A. Reilly. 2017. Forecasting the Spread and Invasive Potential of Apple Snails (Pomacea spp.) in Florida. Master's thesis. Nova Southeastern University. Retrieved from NSUWorks, . (460) https://nsuworks.nova.edu/occ_stuetd/460.

This Thesis is brought to you by the HCNSO Student Work at NSUWorks. It has been accepted for inclusion in HCNSO Student Theses and Dissertations by an authorized administrator of NSUWorks. For more information, please contact [email protected]. Thesis of Stephanie A. Reilly

Submitted in Partial Fulfillment of the Requirements for the Degree of

Master of Science

M.S. Marine Biology

Nova Southeastern University Halmos College of Natural Sciences and Oceanography

December 2017

Approved: Thesis Committee

Major Professor: Matthew Johnston

Committee Member: Bernhard Riegl

Committee Member: Kenneth A. Hayes

This thesis is available at NSUWorks: https://nsuworks.nova.edu/occ_stuetd/460 HALMOS COLLEGE OF NATURAL SCIENCES AND OCEANOGRAPHY

FORECASTING THE SPREAD AND INVASIVE POTENTIAL OF APPLE SNAILS (POMACEA SPP.) IN FLORIDA

By: Stephanie A. Reilly

Submitted to the Faculty of Halmos College of Natural Sciences and Oceanography In partial fulfillment of the requirements for the degree of Master of Science with a specialty in:

Marine Biology

Nova Southeastern University

December 2017

Thesis of

Stephanie A. Reilly

Submitted in Partial Fulfillment of the Requirements for the Degree of

Masters of Science:

Marine Biology

Nova Southeastern University Halmos College of Natural Sciences and Oceanography

Approved: 12/5/2017

Thesis Committee:

Major Professor: ______Dr. Matthew Johnston, Ph.D.

Committee Member: ______Dr. Bernhard Riegl, Ph.D.

Committee Member: ______Dr. Kenneth A. Hayes, Ph.D. List of Tables Table 1: Physiological tolerances of select species of apple snails ...... 12 Table 2: List of studies using MaxEnt including the purpose, extent, and organisms studied ...... 15 Table 3: Summary of environmental layers used in each model run ...... 16 Table 4: Risk analysis- conversion of output model into a prediction of potential invasion risk...... 18 Table 5: AUC values of base SDM's ...... 20 Table 6: AUC values of extended SDM's ...... 20 Table 7: Percent of forecasted habitat suitability in Florida for selected invasive Pomacean snails ...... 21 Table 8: Strongest indicators of invasion success based on jackknife and marginal response curves ...... 21 Table 9: Base SDM variable importance for P. canaliculata ...... 25 Table 10: Base SDM variable importance for P. diffusa ...... 25 Table 11: Base SDM variable importance for P. maculata ...... 25 Table 12: Extended SDM variable importance for P. canaliculata ...... 30 Table 13: Extended SDM variable importance for P. diffusa ...... 30 Table 14: Extended SDM variable importance for P. maculata ...... 30 Table 15: Forecasted percent cover of Florida at risk of Pomacea spp. invasion ...... 31

List of Figures Figure 1: Example record from EDDMaps database ...... 13 Figure 2: Florida sub-basins (HUC 8) ...... 19 Figure 3: Base SDM of P. canaliculata ...... 22 Figure 4: Base SDM of P. diffusa ...... 23 Figure 5: Base SDM of P. maculata...... 24 Figure 6: Extended SDM of P. canaliculata ...... 27 Figure 7: Extended SDM of P. diffusa...... 28 Figure 8: Extended SDM of P. maculata ...... 29 Figure 9: P. canaliculata risk analysis map using FEMA one-hundred year flood data ... 32 Figure 10: P. diffusa risk analysis map using FEMA one-hundred year flood data ...... 33 Figure 11: P. maculata risk analysis map using FEMA one-hundred year flood data ..... 34

iii

Table of Contents List of Tables ...... ii List of Figures ...... iii Abstract ...... 1 1.0 Introduction ...... 2 1.1 Invasive Apple Snails ...... 3 1.2 Invasive Apple Snail Impacts ...... 4 1.3 Introduction and Spread in Florida ...... 6 1.4 Removal and Eradication Efforts ...... 7 1.5 Species Distribution Models...... 8 1.6 Study Motivation and Purpose ...... 9 2.0 Methods...... 10 2.1 Data Collection ...... 10 2.2 The Species Distribution Models...... 14 2.3 Base Species Distribution Models ...... 15 2.4 Expanded Species Distribution Models...... 16 2.5 Area Under the Receiver Operating Curve ...... 17 2.6 Risk Analysis ...... 17 3.0 Results ...... 20 3.1 Base SDMs ...... 20 3.2 Extended SDMs ...... 25 3.3 Risk Analysis Maps...... 31 4.0 Discussion and Conclusion ...... 35 Acknowledgements ...... 40 References ...... 41 Appendices ...... 55

Abstract Forecasting the potential range of invasive species is a critical component for risk assessment, monitoring, and management. However, many of these invasive species are not yet at equilibrium which can be problematic for many modelling approaches. Using the climate matching method, MaxEnt, a series of species distribution models (SDMs) and risk analysis maps were created for select apple snail species in Florida: Pomacea canaliculata, P. diffusa, and P. maculata. Apple snails, freshwater gastropods in the family Ampullariidae, are native to South America and were introduced to the United States via the pet trade approximately 40 years ago. These highly invasive species have already been introduced in ten states and established in at least seven. The models and risk analysis in this study show the majority of Florida was at least moderately suitable for all apple snails modeled, with P. maculata posing the greatest threat.

Keywords: Apple Snail, Species Distribution Model, Invasive Species

1.0 Introduction An invasives species is defined as a non-native species whose introduction does or is likely to cause economic or environmental harm or is detrimental to human health (US Federal Executive Order 13112 1999). Invasive species have been recognized as one of the greatest threats to global biodiversity for decades (Herbold and Moyle 1986, Usher 1986, Pimm 1989, Huxel 1999, MacDougal and Turkington 2005). Any species introduced to a novel habitat has the potential to become an invasive species due to a lack of natural controls (i.e., predators, diseases, parasites, competition) that keep its population in check. Species that tend to become successful invaders are abundant, have a generalist feeding strategy, are broadly tolerant of environmental conditions, have short generation times, and have high genetic variability (Ray 2005). Once established in new environments, invasives alter population and community structure (Sandlund et al. 2001). While many species spread via natural dispersal mechanisms (i.e., wind, water currents, and continental shift), the ranges of native fauna are generally limited by geographic barriers such as open oceans or mountain ranges. Rapid global homogenization of biodiversity, as a result of expanding global economy, occurs when human-mediated mechanisms rapidly transport large numbers of exotic species across natural barriers facilitating massive invasions (Carlton and Geller 1993). The cost of biological invasions to the global economy lies in the trillions of dollars (Millennium Ecosystem Assessment 2005) and monetary impacts in the United States (US) alone are estimated to be $120 billion United States Dollars (USD) annually (Pimentel et al. 2005). Additionally, profound, negative ecological changes have been instigated by numerous invasive species such as lionfish (Pterois volitans and Pterois miles) in the Atlantic basin (Johnston and Purkis 2011), brown tree snakes (Boiga irregularis) in Guam (Savidge 1987), and water hyacinth (Eichhornia crassipes and Eichhornia natans) in Nigerian waters (Akinyemiju 1987).The growing problem of biological invasions has received increased attention from resource managers over the past few decades due to the burgeoning number of species introduced into non-native environments and the subsequent negative impacts wrought by the exotics in these systems (Lockwood et al. 2013). As a major port of entry for exotic species, subtropical Florida is particularly susceptible to the introduction, spread, and establishment of non-native species, with

2 greater than 500 observed cases. Some prominent examples include the Burmese python (Python bivittatus), lionfish (Pterois volitans), Brazilian pepper tree (Schinus terebinthifolius), Australian pine (Casuarina equisetifolia), and Argentine black and white tegus (Tupinambis merianae) with additional arrivals every year (Florida Fish and Wildlife Conservation Commission 2017). Several of these exotic species have caused widespread changes to native flora and fauna, however, many others have not yet received adequate study to understand their impacts. This study will focus on a set of non-native mollusks from the genus Pomacea Perry, 1810. Apple snails (Ampullariidae) are recent US inductees and studies have shown them as an emerging threat to aquatic ecosystems (Howells et al. 2006, Rawlings et al. 2007). One of the greatest challenges when forecasting the spread of Pomacea species stems from the difficulty in identification (Hayes et al. 2012). It was assumed for many years that P. canaliculata (Lamarck, 1822) was the main invader in the Florida and Southeast Asia. However, recent taxonomic revisions that included anatomical, phylogenetic, and biogeographic studies have demonstrated that multiple Pomacea species are present in Florida, with P. maculata Perry, 1810 being the predominant invader, and P. canaliculata restricted to northeast regions of the state (Cowie 2002, Rawlings et al. 2007, Hayes et al. 2008).

1.1 Invasive Apple Snails The Pomacea canaliculata species complex comprised of more than 13 apple snails, some of which are among the most invasive freshwater snails in the world: P. canaliculata (i.e., the most widely studied species of the complex), P. maculata, P. lineata (Spix, in Wagner, 1827), P. dolioides (Reeve, 1856), P. megastoma (Sowerby, 1825), P. figulina (Spix, in Wagner, 1827), P. poeyana (Pilsbry, 1927), P. paludosa (Say, 1829), and six undescribed species (Hayes et al. 2009, 2012). In just the last few decades, members of this group of non-native apple snails, predominantly P. maculata and P. canaliculata, have been documented widely throughout tropical and subtropical regions of the world outside of their native ranges (Cowie and Hayes 2012, Horgan et al. 2014). Still, the snail complex has only just recently been the subject of intense study that is

3 beginning to bring taxonomic clarity to the native and non-native species of apple snails (Cowie 2002, Hayes et al. 2008, 2009, 2012, 2015). Ampullariids are naturally distributed in tropical and subtropical ecosystems globally, but the genus Pomacea is endemic to South and Central America, with one exception P. paludosa, which is native to the Southeastern US (Martin et al. 2001, Cowie 2002). In their natural habitat, apple snails can remain active throughout the year if environmental conditions are favorable (i.e., warm and relatively humid), but usually aestivate (i.e., a state of torpor or dormancy that is maintained throughout a hot or dry period) during parts of the year to prevent drying out. Temperatures and rainfall are the two main factors that determine their activity levels (Coelho et al. 2012). High humidity allows the snails to remain out of water for some time without desiccation. It has been reported that the snails can disperse across land during these high humidity periods (Kappes and Haase 2012). Such land dispersal activities account for localized spread, and do not make up the majority of apple snail dispersal; downstream dispersion via moving water and during flooding events are the likely major natural dispersal pathways. In combination with other transport vectors, such as intentional and unintentional human- mediated introductions and predators that transport and drop captured snails, apple snails can rapidly spread between sites. Subsequently several members of Pomacea have been widely introduced and established in many natural and artificial aquatic ecosystems globally (Cowie 2002). One species, P. canaliculata is listed as one of the world’s top 100 worst invaders and has had deleterious ecological and economic impacts in regions of Southeast Asia (Naylor 1996, Lowe et al. 2000).

1.2 Invasive Apple Snail Impacts P. canaliculata was intentionally brought to Taiwan in 1979 to be farmed as a possible high-protein food source and potential export (Naylor 1996). Their rapid reproduction and growth combined with low investment costs made apple snails ideal for aquaculture and an export commodity, earning it the nickname “the golden miracle snail” (Carlsson and Lacousière 2005). Shortly after being introduced, they quickly spread to the rest of Asia, becoming well established throughout the region by 1990. Another species, P. maculata, may have been inadvertently being introduced during this period

(Hayes et al. 2008). However, low market values and negligence, consequences of new stringent health regulations and the snails’ perceived poor taste by locals, resulted in the release and escape of Pomacea spp. into Asian waterways, where it eventually established as a major agricultural and natural wetland pest (Cowie 2002). In Asia, introduced snails cause damages during rice-planting season, when the snails emerge from aestivation and build up energy reserves by consuming rice seedlings. The damage done to rice fields is positively correlated with the density and size of the snails (Naylor 1996, Ito 2002, Sin 2003, Carlsson and Lacousière 2005, Howell et al. 2006, Horgan et al. 2014, Pan et al. 2014). While the most noticeable impacts are to food crops, snails have instigated complete collapses of aquatic plant communities in surrounding wetlands in locations where they are abundant and also displace native species (Carlsson et al. 2004, Karraker and Dudgeon 2014, Monette et al. 2016). In many cases oligotrophic, macrophyte dominated ecosystems are transformed into turbid plankton- and algae-dominated ecosystems (Peh 2010). Efforts to restore damaged wetlands have been hindered by re-invasions of P. canaliculata and the subsequent deterioration of water quality in the surrounding area (Wang and Pei 2012). In addition to the negative impacts to native flora and fauna, apple snails are known vectors of multiple diseases (Hollingsworth et al. 2006). Most notably, the snails host the parasite Angiostrongylus cantonensis, the rat lungworm, which can cause angiostrongyliasis manifested as eosinophilic meningoencephalitis (Teem et al. 2013, Kim et al. 2014). Consumption (intentionally or accidentally) of raw or undercooked infected snails can result in infection by the parasite (Lv et al. 2013). While eosinophilic meningoencephalitis associated with apple snails has not been reported in the US, there are populations of P. maculata hosting A. cantonensis in New Orleans, but no reported cases of human disease associated to P. maculata (Howell et al. 2006, Teem and Gutierrez 2008, Teem et al. 2013). Apple snails are also potential vectors for cercarial dermatitis and echinostomiasis (Horgan et al. 2014). In addition, several reports note that apple snails have the potential to accumulate high concentrations of heavy metals within the viscera (Frakes et al. 2008, Hoang et al. 2008). These parasites and pollutants have the potential to impact higher trophic levels via trophic cascades and consumption, including human health (Hayes et al. 2015).

1.3 Introduction and Spread in Florida Non-native apples snails were first imported into to the US for the pet and food trade in the 1950’s. In the late 1970’s snails were released or escaped into the wild and have since been recorded in at least ten states with increasing frequency (Carlsson et al. 2004, Howells et al. 2006, Burlakova et al. 2009, Cowie and Hayes 2012, Horgan et al. 2014). The earliest report of the non-native apple snail P. canaliculata in Florida was in 1978, although recent studies indicate these specimens were likely the morphologically similar species P. maculata (Howells et al. 2006, Rawlings et al. 2007, Bernatis 2008, Hayes et al. 2012). However, P. canaliculata was also introduced into Florida at some point in time, and an established population reportedly still occurs in north Florida, but has yet to be confirmed (Rawlings et al. 2007, Hayes personal communication). Pomacea diffusa Blume, 1957, Marisa cornuarietis (Linnaeus, 1758) another member of family Ampullariidae, and an additional unidentified Pomacea species have also been reported as established in localized regions of central and south Florida since the early 1980’s, Palm Beach County since 1989, and south Florida since the 1970’s, respectively, but have not yet spread beyond these regions and may in fact already have been extirpated (Rawling et al. 2007, Capiner and White 2011, Hayes personal communication). Ongoing introductions of the snails are attributed to poor aquaculture and aquarium practices that allow individuals to escape (Rawlings et al. 2007), deliberately release as a biocontrol agent for invasive algae (Cowie 2002), and storms (i.e., hurricanes and tropical storms), which may distribute apple snails over long distances as demonstrated in Texas during Tropical Storm Allison (Howells and Smith 2002). In their native South American wetlands, apple snails are spread via downstream flows in interconnected waterways, but also experience more expansive spread during the wet season’s annual floods (Kappes and Haase 2012). It is logical therefore that the snails could be spreading in the same way in Florida. In total, five species of non-native apple snails have been recorded in Florida: P. canaliculata, P. maculata (previously reported as P. insularum (d’Orbigny, 1835), P. sp. (not yet named, but was previously reported as P. haustrum (Reeve, 1856)), P. diffusa (previously reported as P. bridgesii Reeve, (1956)), and Marisa cornuarietis. Of these five species, all but P. diffusa are considered invasive (Cowie and Hayes 2012, Hayes et al. 2015). The presence of P. maculata and P.

6 canaliculata in Florida, in particular, is cause for concern as numerous studies have shown that these snails can have devastating impacts where they have been introduced including ecosystem trophic shifts, changes in biodiversity, and disease transmission (Naylor 1996, Byers et al. 2013, Horgan et al. 2014). The life history characteristics (i.e., amphibious life style, high fecundity, trophic generalists, and adaptability) of non-native apple snails make them successful invaders to Florida’s wetlands, where environmental conditions are similar to their native habitats in South America (Hayes et al. 2008). Apple snails are trophic generalists, typically feeding on a variety of items comprised of periphyton, biofilms, macrophytes (floating, submerged, and emergent), living animals, and carrion (Cowie 2002, Burlakova et al. 2009, Hayes et al. 2008, 2015). Macrophytes and periphyton are both integral parts of Florida’s wetlands that provide a variety of ecosystem services such as purifying water, controlling erosion, retaining pollutants, and maintaining biological diversity (Horgan et al. 2014). Loss of such integral structures would cause ecosystem shifts which reduces the quality of habitat (Karunaratne et al. 2006). Non-native apple snails also have greater feeding rates and higher conversion efficiencies, which is attributed to their feeding behaviors (Kwong et al. 2009, Horgan et al. 2014); both P. canaliculata and P. maculata prefer macrophytes, and exhibit higher growth rates compared to the native P. paludosa that primarily feeds on periphyton (Sharfstein and Steinman 2001, Morrison and Hay 2011). Thus non-native Pomacea can impact native snails by competing for resources and cause ecosystem shifts (Carlsson et al. 2004, Burlakova et al. 2009, Morrison and Hay 2011). Furthermore, Pomacea species are highly fecund and can lay from 20 (P. paludosa) up to approximately 2000 (P. maculata) eggs per clutch, with up to five clutches per month depending on the species. This high rate of reproduction and subsequent population increase can compound the deleterious effects of the snails on native flora and fauna (Turner 1996, Cowie 2002, Teo 2004, Conner et al. 2008, Robertson 2012).

1.4 Removal and Eradication Efforts Efforts to reduce or eliminate invasive apple snail populations globally have been minimally successful in part due to high costs and lack of efficiency of the control efforts

(Pias et al. 2012, Martin et al. 2012). Molluscicides, such as endosulfan and niclosamide, and pesticides are the most widely used. These methods, however, are non-selective and often cause unintended negative impacts (i.e., death and deformities) to other organisms. Another method often considered as a potential solution to control aggressive invasive species is bio-control, which is the intentional use of a predatory species to control a pest population (Hoddle 2004, Prabhakaran et al. 2017). Bio-control methods, such as introducing the common (Cyprinus carpio) and black (Mylopharyngodon piceus) carp, green pufferfish (Tetraodon nigroviridis), ducks, leech (Whitmania pigra) and the Chinese soft-shelled turtle (Pelodiscus sinensis) have been attempted in many Asian countries to control the spread of apple snails (Teo 2001, Sin 2006, Dong et al. 2011, Guo et al. 2016, Guo et al. 2017). The bio-controls that have been implemented to control the snails, however, have generally not been effective at reducing or eliminating invasive populations (Naylor 1996, Plant Health Australia 2009). Some studies suggest that baiting and trapping, either using food or pheromones, may be potential methods to capture snails and have minimal impact on the surrounding environment (Cowie 2002, Howell et al. 2006, Plant Health Australia 2009). The most successful apple snail control technique to date is hand picking individual snails at all life stages (Bernatis et al. 2014). When combined with careful water control (i.e., draining excessive standing water), populations of invasive apple snails can be reduced by preventing the snails from exploiting high water levels that are needed for range expansion (Naylor 1996). Despite the relative success of manual removal, other methods will need to be utilized to fully control apple snail populations as hand picking is labor intensive and not a feasible long-term solution.

1.5 Species Distribution Models One way to identify and forecast the spread of invasive species such as apple snails is by using species distributions models (SDM). SDMs are widely implemented tools that aim to forecast a species ecological niche in the context of its ‘fundamental’ and ‘realized’ niche, using species observation records and the environmental, spatial characteristics of a study area (Elith et al. 2011, Elith 2013). Defined, a fundamental niche is the ‘multi-dimensional hyper-volume’ where environmental conditions are

8 suitable for the survival of the species and the realized niche is a subset of the fundamental niche in which the species actually inhabits (Hutchinson 1957). SDMs typically do not represent the actual, realized niche of a species as most models do not include limitations imposed by biotic interactions (i.e. competition, predation) due to a lack of study or available data (Elith 2013). When presence and absence data are available (e.g., from a standardized survey), general-purpose statistical methods, such as generalized linear models (GLM), generalized additive models (GAM), artificial neural network (ANN) and genetic algorithm for rule-set production (GARP), are preferred (Guisan and Zimmerman 2000,Elith 2002, Fielding and Bell 2007). However, a vast amount of occurrence data are presence-only and even when absence data is available, their value may be questionable in many situations – for example, absences may be recorded when the species may be present in extremely low quantities (Anderson et al. 2002, Anderson et al. 2003). In such cases, methods using Bioclim, Domain, and MaxEnt models – all which solely require presence-only records – may be more appropriate. Such models have been utilized to identify the distribution of a species (Peterson and Nakazawa 2008, Phillips and Dudik 2008, Endries 2011), model sensitive ecosystems (Sarell et al. 2003), analyze economic impacts (Emerton and Howard 2008), investigate evolutionary processes (Kozak et al. 2008), and of interest for this study, as a tool to predict species invasions (Peterson 2003, Thuiller et al. 2005, Ward 2007).

1.6 Study Motivation and Purpose It is evident that a few apple snail species belonging to the genus Pomacea have become well established in the state of Florida, continue to spread in the southeastern US, threaten agriculture, and endanger native species (Naylor 1996, Lowe et al. 2000, Carlsson et al. 2004, Savrick et al. in review). The current and potential impacts of non- native apple snails and the likelihood of continued spread makes understanding their range limits critical for assessing and managing these invasives. As such, and given the uncertainty surrounding the present population status of non-native apple snails in Florida, there is a need to understand where, when, and what species may spread in the future. These data are important to quantify how much area may be negatively impacted and to direct prevention, removal, and possible eradication efforts in Florida. Given this

9 need, the aim of this study was to produce SDMs based on current apple snail distributions in Florida to identify suitable habitats and locations with the greatest risk of invasion under present and future environmental conditions. The SDMs were parameterized by presence records and literature-derived environmental conditions (Table 1) that are known to facilitate the survival and spread of apple snails outside of their native range. Accordingly, two SDM types were produced for each study species by: 1) including only biological variables collected from WorldClim (Appendix 1, Hijmans et al. 2005) that were deemed relevant to the spread of the snails according to the literature (hereafter the ‘base SDMs’), and 2) by including the same biologically- relevant variables in the base SDMs in addition to an expanded set of environmental data that are thought to influence snail distribution (hereafter the ‘expanded SDMs'). Next, risk analysis maps were created for each species, informed by the potential environmental envelope derived from the SDMs and given contemporary habitat types in Florida combined with one-hundred year flood data (FEMA 2017). These ‘risk analysis’ maps identify additional areas of potential spread, given an extreme flooding event in the future. Unfortunately, and though five species of invasive apple snails are present in Florida, only data from three of the five species were sufficiently well represented to allow analysis in this study. Going forward, study data will be useful to identify where the impacts of invasive apple snails on Florida’s ecosystem may be the greatest and can be used to help formulate and direct mitigation and eradication efforts.

2.0 Methods 2.1 Data Collection All training data – i.e., data used to develop the model and discover potential relationships – was collected from the Early Detection and Distribution Mapping System (EDDMaps), a database developed by the University of Georgia for early detection and rapid response programs and invasives management (Bargeron and Moorhead 2007). This robust system combines observational data stored in different formats (i.e., herbaria records, research projects, regional survey programs) into a single overall distribution map, with each record containing all known information (see also Figure 1) about the observation (Bradley and Marvin 2011, Rawlins et al. 2011, Wallace et al. 2014, Klug et

10 al. 2015). Observation records spanning the date of their first record in Florida to the end of 2015 were kept as training data for three selected invasive Pomacea species: P. canaliculata, P. diffusa, and P. maculata. Marisa cornuarietis and Pomacea sp. (formerly identified as P. haustrum) were excluded from the model due the lack of observation records in EDDMaps. Additional geo-referenced observation data spanning the date of first record to the end of 2016 were collected from the Florida Fish and Wildlife Conservation Commission (FWC), United States Geological Survey (USGS), and the Florida Museum of Natural History and used as test data - i.e., data used to evaluate model performance subsequent to the analysis runs. Taxonomic experts Jennifer Bernatis and John Slapcinsky were consulted to authenticate all records. The dataset was assessed for quality (i.e., duplicate records for each species were removed across all data sets), and a preliminary literature review was conducted to determine the physiological range of tolerances for each apple snail species found in Florida (Table 1). Bioclimatic (BioClim) variables (Appendix 1) were initially collected from WorldClim with a spatial resolution of 30 seconds (~1 km2) (Hijmans et al. 2005). Since collinearity of predictor variables in MaxEnt can produce spurious results, only two temperature variables- maximum temperature of the warmest month (BIO5) and minimum temperature of the coldest month (BIO6) - and three precipitation variables - annual precipitation (BIO12), precipitation of the driest quarter (BIO17), and precipitation of the warmest quarter (BIO18) - were selected a priori for the base model founded on biological importance following Byers et al. 2013. Additional environmental variables - land cover from the National Land Cover Dataset (NLCD) (Appendix 2, Homer et al. 2015), soil type from the USGS Geologic Map Database (USGS2 2017), and pH measurements from the EPA STORET database (United States Environmental Protection Agency 2017) - were collected for use in the expanded models to enhance the forecast of suitable habitats. To identify potential areas of future invasion risk in extreme flooding events, one-hundred-year flood data from the Federal Emergency Management Agency (FEMA) (FEMA 2017) and hydrological sub-basin units were collected from the National Hydrography Dataset (NHD) and Watershed Boundary Dataset (WBD) (USGS3 2017). All environmental data layers were clipped to the study region, the state of

Florida, and divided into 1 km2 spatial cells within ESRI ArcMap 10.3 and converted into ASCII files for use in MaxEnt (version 3.3.3k).

Table 1: Physiological tolerances of select species of apple snails Temperature (°C) Salinity (psu) pH P. paludosa >13 (Stevens et al. 0-8* (Glass and Darby >6.5 (optimal 7.0-8.0) 2002) 2009) (Glass and Darby 2009)

P. maculata >0 (Yoshida 2014) 0-8.0 (Bernatis 2014, 5.5-9.5 (optimal 7.5- >6 (Byers et al. McAskill 2015) 9.5) (Bernatis 2014) 2013) 0-15.0 (optimal at 0-5) 4.0-10.5 (optimal 7.0- 15.2-36.6 (Martin and Valentine 9.0) (Ramakrishnan (Ramakrishnan 2014) 2007) 2007) 0-13.6 (optimal at 0- 6.8) (Ramakrishnan 2007)

P. 13-37 (Byers et al. 8.0-12.0 (Howell et al. >8 (Glass and Darby canaliculata 2013) 2006) 2009) >18 (Hayes et al. >5.5 (Byers et al. 2015) 2013) Aestivate at 10-17.5 6.29-7.46 (Ito 2002) (Estebenet and Martin 2002)

P. diffusa 18-27 (Howell et al. 7.0 (Howell et al. 7-8 (Howells et al. 2006) 2006) 2006) <44.9 (Mu et al. 0-6.8 (Jordan and 2015) Deaton 1999) To our knowledge no salinity tolerance studies for P. paludosa. Its ecology has been noted to be most similar to P. maculata and P. canaliculata so the widest temperature range has been used for this species (Rawlings et al. 2007, Stevens et al. 2002).

Figure 1: Example record from EDDMaps database

2.2 The Species Distribution Models The spatial distribution of invasive Pomacea spp. in Florida was forecasted using MaxEnt and visualized as predictive habitat suitability maps using ArcMap (ESRI 2011). MaxEnt is a machine-learning method for characterizing the probable distributions from incomplete information – i.e., the maximum entropy (closest to uniform) of a species distribution (Phillips et al. 2006, Pearson 2010). In principle, the estimated distribution must agree with everything known, or inferred, from environmental conditions where species have been observed and avoid assumptions not supported by the data (Phillips et al. 2006, Elith et al. 2011). MaxEnt models have achieved high predictive accuracy (Phillips and Dudik 2008) and have been utilized across many ecological, evolutionary, conservation, and biosecurity applications (Table 2). Unlike other SDM techniques, and important in the context of this study, MaxEnt does not require species absence data. Instead, it uses presence data and background environmental variables (continuous or categorical), the dataset that describes pseudo-absences- a random sample of non- occurrences from the study region, across the study area to evaluate a continuous probability of suitable habitat (Phillips et al. 2006). This is advantageous as absence data are lacking for most Pomacea species (Byers et al. 2013).

Table 2: List of studies using MaxEnt including the purpose, extent, and organisms studied Primary purpose Extent Organisms Reference Predicting current Colombia Spiny pocket mice Shcheglovitova and distributions (for Anderson 2013 conservation, risk South Bradypus assessment, or new America variegatus and Phillips et al. 2006 surveys) Microryzomys North minutus Endries 2011 Carolina Aquatic species

Predicting the Iran Heteropterans Sohjouy-Fard et al. 2013 potential distribution Global Rana catesbeiana Ficetola et al. 2007 of invasive or pest Global Plant species Thuiller et al. 2005 species

Predict species Amazon Amazonian trees Saatchi et al. 2008 richness or diversity Basin

Predict distribution California Odocoileus Pease et al. 2009 for morphological or hemionus genetic studies (phylogeography or phyloclimatic studies)

Understand patterns Chile Endemic birds Moreno et al. 2011 of endemisim

Forecast distribution Bavaria Abies alba Falk and Mellert 2011 in relation to climate Australia Eulamprus Dubey et al. 2013 change or land leuraensis transformation

Test model Global Many species Hijmans 2012 performance against California Phytophthora Václavík and Meentemeyer other methods ramorum 2009 Global New Eriocheir sinensis Herborg et al. 2007 Zealand Ant species Ward 2007

2.3 Base Species Distribution Models Of MaxEnt’s output options - raw, cumulative, and logistic - the logistic output was selected for ease of interpreting the probability of presence, measured as a

15 continuous habitat suitability value between 0 (unsuitable) and 1 (the most suitable) (Phillips and Dudık 2008). Each run was set to execute a cross validation, a resampling technique that divides the data into two sets of data – i.e., training and test datasets. The model function was first fit to the training data set only, then validated using the test dataset. A total of 100 replicates were run with 10,000 randomly selected points for background data, and a maximum iteration of 10,000 to allow adequate time for model convergence. The initial environmental layers consisted of the BioClim variables selected a priori (also see Table 3) that were deemed as important to the spread of the snails as derived from the literature (Byers et al. 2013). Variables that did not contribute significantly - i.e., those that contributed less than five percent - were then removed as model parameters and a more parsimonious model was re-run as the final base model.

Table 3: Summary of environmental layers used in each model run Initial Environmental Layers in Model Species Environmental Parsimonious Model Layers P. BIO5, BIO12 canaliculata BIO5, BIO6, BIO12, Base P. diffusa BIO17, BIO18 BIO6, BIO12 SDM P. maculata BIO5, BIO6, BIO17, BIO18 P. BIO5, NLCD, pH, GEO_SOIL canaliculata BIO5, BIO6, BIO12,

BIO17, BIO18, BIO6, BIO12, NLCD, Extended P. diffusa NLCD, pH, GEO_SOIL SDM GEO_SOIL BIO5, BIO17, NLCD, pH, P. maculata GEO_SOIL P. BIO5, BIO6, BIO12, Extended Model and FLOOD canaliculata BIO17, BIO18, Risk P. diffusa NLCD, pH, Extended Model and FLOOD Analysis GEO_SOIL, P. maculata Extended Model and FLOOD FLOOD* *FLOOD refers to flood risk under one-hundred year flood condition data from FEMA

2.4 Expanded Species Distribution Models To enhance the forecasted distribution of the base SDMs, an expanded SDM was created for each snail species using additional variables chosen from a literature review and that were not available in the BioClim dataset. Land cover, soil type, and pH were

16 selected due to their effects on snail growth and reproduction. PH was chosen as it may be a limiting factor for growth, shell maintenance, and hatchling success (Ramakrishnan 2007). Land cover was selected as it is a primary factor that influences structure and composition of aquatic communities. Apple snails are able to inhabit a wide range of ecosystems (i.e. swamps, ditches, ponds, lakes, and rivers), but prefers lentic water. Furthermore, the snail’s amphibious life style allows it to inhabit areas with oxygen poor waters when emergent vegetation is present for respiration (Allan and Johnson 1997, Mantyka-Pringle et al. 2014, Ferreira and Capítulo 2017). Soil types were chosen due to their effect on water chemistry and snail persistence (Harman 1972, Hanelt et al. 2001, Sawasdee and Kohler 2009, Darby et al. 2004). The expanded SDMs followed the same procedure as the base SDM, but their environmental layers were comprised of the above expanded set of environmental data, in addition to the BioClim variables selected a priori in the base model (also see Table 3).

2.5 Area Under the Receiver Operating Curve Predictive performance and model accuracy were tested by performing an area under the receiver operating curve (AUC) analysis for each model run (Elith et al.2006; Phillips et al. 2006, Phillips 2010). AUC values ranged from 0.00, a model with no predictive ability, to 1.00, a model with perfect predictability (Peterson 2003, Pearson et al. 2006, Raes and Steege 2007). Model AUC values greater than 0.75 were considered acceptable following Pearce and Ferrier 2000. Each model was visualized using ESRI’s ArcGIS 10.3 as a raster map to show habitat suitability across a range of 0 (low suitability) to 1 (high suitability). A jack-knife test - a resampling technique that works by leaving one observation out at a time from the sample set and estimating variance and bias of the statistic - was used to evaluate the relative importance of each environmental variable for each model.

2.6 Risk Analysis To forecast apple snail distribution given the prospect of future flooding, one- hundred year flood data was incorporated into the expanded SDM. Flood data was chosen as hydrological cycles (flood and droughts) are common in the snails’ natural habitats of

South America and Florida undergoes similar hydrological cycles. An observation assessment was created by linking apple snail observations to watershed sub-basins (HUC 8) in ArcMap. Areas were flagged as high risk if snails were observed within a watershed sub-basin (HUC 8) (Figure 2) that had the necessary environmental parameters for snail survival as identified by the SDM under normal and flood conditions, or as defined by a suitability index of 0.75 or higher. Counties with missing flood data were conservatively given a suitability score of 0.5 (i.e., random). Areas were labeled at a moderate invasion risk if environmental conditions were sub-optimal (between 0.25 and 0.75), or were likely to become invaded only during one-hundred-year flood conditions. Areas at low risk were all other locations where high or medium risk conditions were not met (Table 4).

Table 4: Risk analysis- conversion of output model into a prediction of potential invasion risk Risk Observation SDM and/or one-hundred-year flood conditions* assessment High Present Both Moderate Present or Absent Either Low Absent None *Counties with missing flood data were conservatively considered a moderate risk.

Figure 2: Florida sub-basins (HUC 8) 0)Lower Conecuh, 1)Florida Bay- Florida Keys ,2)Pea, 3)Lower Chattahoochee, 4)Apalachicola, 5)Apalachicola Bay, 6)Sarasota Bay, 7)Northern Okeechobee Inflow, 8)New, 9)Withlacoochee, 10)Crystal-Pithlachascotee, 11)Oklawaha, 12)Kissimmee, 13)Upper Suwannee, 14)Lower St. Johns, 15)Yellow, 16)Myakka, 17)Withlacoochee, 18)Caloosahatchee, 19)Perdido, 20)Escambia, 21)Hillsborough, 22)Alafia, 23)Big Cypress Swamp, 24)Alapaha, 25)Lower Choctawhatchee, 26)St. Andrew- St. Joseph Bays, 27)Aucilla, 28)Western Okeechobee Inflow, 29)Everglades, 30)Lower Suwannee, 31)St. Marys, 32)Daytona- St. Augustine, 33)Pensacola Bay, 34)Lower Ochlockonee, 35)Apalachee Bay- St. Marks, 36)Tampa Bay, 37)Manatee, 38)Charlotte Harbor, 39)Peace, 40)Cape Canaveral, 41)Blackwater, 42)Little Manatee, 43)Nassau, 44)Perdido Bay, 45)Choctawhatchee Bay, 46)Chipola, 47)Econfina- Steinhatchee, 48)Upper St. Johns, 49)Vero Beach, 50)Lake Okeechobee, 51)Waccasassa, 52)Santa Fe, 53)Lower Flint, 54)Florida Southeast Coast

3.0 Results 3.1 Base SDMs The performance of all models was better than random (AUC>0.5) at predicting habitat suitability for training and test locations with the given variables for all species (Table 5 and 6). The base SDMs forecasted 9%, 12 %, and 3% of the study area as highly suitable habitat for P. canaliculata (n=21), P. diffusa (n=35), and P. maculata (n=157) respectively (Table 7, Figure 3-5). The marginal response curves, a measure of the singular effect of one variable with all others set to the sample average values, suggests P. canaliculata’s presence is positively correlated to maximum temperature of the warmest month, and negatively correlated with annual precipitation (Table 8 and 9, Appendix 3-5). Response curves suggest a positive correlation between the minimum temperature of the coldest month and annual precipitation for the presence of P. diffusa (Table 8 and 10, Appendix 6-8). For P. maculata, maximum temperature of the warmest month and minimum temperature of the coldest month was positively correlated to species presence, precipitation of the driest month suggest a parabolic correlation to species presence, and precipitation of the warmest quarter suggests a constant relation to species presence from 1.5-5.5 mm and a negative correlation from 5.5-7.5 (Table 8 and 11, Appendix 9-13).

Table 5: AUC values of base SDM's Training Training standard Test Test standard Species AUC Deviation AUC Deviation P. canaliculata 0.827 0.023 0.821 0.120 P. diffusa 0.858 0.023 0.856 0.303 P. maculata 0.868 0.003 0.783 0.225

Table 6: AUC values of extended SDM's Training Training Standard Test Test Standard Species AUC Deviation AUC Deviation P. canaliculata 0.966 0.013 0.896 0.109 P. diffusa 0.911 0.019 0.858 0.284 P. maculata 0.868 0.002 0.837 0.192

Table 7: Percent of forecasted habitat suitability in Florida for selected invasive Pomacean snails High Moderate Low Species SDM Suitability Suitability Suitability P. canaliculata Base 9 38 53 Expanded 15 67 18 P. diffusa Base 12 53 34 Expanded 8 69 23 P. maculata Base 3 37 60 Expanded 9 50 41

Table 8: Strongest indicators of invasion success based on jackknife and marginal response curves BIO BIO BIO BIO BIO GEO Species SDM NLCD* pH 5 6 12 17 18 SOIL** P. Base 33- - 0------canaliculata 34 130 Expanded 32- - - - - 21, 23, 7 5.7- 34 95 7.2 P. diffusa Base - 13- 140------20 180 Expanded - 12- 140- - - 22, 23, 2, 12 - 20 170 24, 82 P. maculata Base 32- 7- - 140- 530- - - - 35 13 180 580 Expanded 32.7- - - 140- - 11, 21, 1, 2 5.7- 33 280 23,24, 7.6 95 Values listed represent the predicted probability (>0.5) for the indicated variable based on the marginal response curves. See Appendix 1 for a description of WorldClim variables and codes. * See Appendix 2 for land cover classification (Homer et al. 2015) ** See Appendix 14 for soil type classification (USGS2 2017)

Figure 3: Base SDM of P. canaliculata

Figure 4: Base SDM of P. diffusa

Figure 5: Base SDM of P. maculata

Table 9: Base SDM variable importance for P. canaliculata Before Parsimony Parsimony Variable Percent Permutation Percent Permutation Contribution Importance Contribution Importance BIO5 78.3 70.5 78.3 80.1 BIO6 0.0 0.0 - - BIO12 21.7 29.2 21.7 19.9 BIO17 0.0 0.4 - - BIO18 0.0 0.0 - -

Table 10: Base SDM variable importance for P. diffusa Before Parsimony Parsimony Variable Percent Permutation Percent Permutation Contribution Importance Contribution Importance BIO5 2.0 0.0 - - BIO6 79.9 99.4 83.4 99.9 BIO12 14.3 0.0 16.6 0.1 BIO17 3.2 0.2 - - BIO18 0.6 0.5 - -

Table 11: Base SDM variable importance for P. maculata Before Parsimony Parsimony Variable Percent Permutation Percent Permutation Contribution Importance Contribution Importance BIO5 47.6 47.0 47.4 44.0 BIO6 19.8 35.0 22.7 32.1 BIO12 4.6 0.4 - - BIO17 20.8 11.2 22.9 22.6 BIO18 7.2 6.5 7.0 1.3

3.2 Extended SDMs The extended SDMs forecasted 15%, 8%, and 9% of the studied area as highly suitable habitat for P. canaliculata, P. diffusa, and P. maculata respectively (Table 7, Figure 6-8). The marginal response curves indicate P. canaliculata’s presence is positively correlated to maximum temperature of the warmest month and negatively correlated with pH. A ‘beach sand’ ground type and land cover types ‘developed (medium intensity)’, ‘emergent herbaceous wetlands’, and ‘developed (open space)’ had the greatest probability of occurrence (Table 8 and 12, Appendix 15-19). For P. diffusa, response curves suggest a positive correlation with minimum temperature of the coldest

25 month and annual precipitation. Ground types ‘calcarenite’ and ‘limestone’, and land cover types ‘developed (high intensity)’, ‘developed (low intensity)’, ‘developed (medium intensity)’, and ‘cultivated crops’ had the greatest probability of occurrence (Table 8 and 13, Appendix 20-24). The response curves for P. maculata suggest a parabolic correlation with maximum temperature of the warmest month and precipitation of the driest month, and a constant correlation with a 5.7-7.6 pH but negative from 7.6- 8.2 pH. Ground types ‘sand’ and ‘limestone’, and land cover types ‘developed (high intensity)’, ‘open water’, ‘developed (medium intensity)’, ‘emergent herbaceous wetlands’, and ‘developed (open space)’ had the greatest probability of occurrence (Table 8 and 14, Appendix 25-30).

Figure 6: Extended SDM of P. canaliculata

Figure 7: Extended SDM of P. diffusa

Figure 8: Extended SDM of P. maculata

Table 12: Extended SDM variable importance for P. canaliculata Before Parsimony Parsimony Variable Percent Permutation Percent Permutation Contribution Importance Contribution Importance BIO5 33.0 22.5 34.5 14.3 BIO6 0.0 0.0 - - BIO12 3.2 0.3 - - BIO17 0.0 0.0 - - BIO18 0.0 0.0 - - NLCD 39.9 61.7 41.1 85.7 PH 5.8 6.2 6.0 0.0 GEO_SOIL 18.0 9.3 18.4 0.0

Table 13: Extended SDM variable importance for P. diffusa Before Parsimony Parsimony Variable Percent Permutation Percent Permutation Contribution Importance Contribution Importance BIO5 0.1 0.0 - - BIO6 54.9 79.3 59.8 78.9 BIO12 9.2 3.0 8.4 0.0 BIO17 0.6 1.6 - - BIO18 0.1 0.1 - - NLCD 21.0 14.7 20.0 20.3 PH 2.4 1.0 - - GEO_SOIL 11.7 0.4 11.7 0.8

Table 14: Extended SDM variable importance for P. maculata Before Parsimony Parsimony Variable Percent Permutation Percent Permutation Contribution Importance Contribution Importance BIO5 23.5 37.4 27.3 39.2 BIO6 3.1 5.8 - - BIO12 0.7 1.0 - - BIO17 15.1 11.5 22.1 13.6 BIO18 4.5 2.8 - - NLCD 42.4 31.4 41.6 35.6 PH 5.6 6.0 5.7 7.6 GEO_SOIL 5.2 4.0 3.2 4.0

3.3 Risk Analysis Maps The risk analysis maps describe the potential invasion risk for each Pomacea species modeled based on the extended SDMs (i.e, calculated based on the conditions described in Table 4). Approximately 1%, 1%, and 29% of the study area was forecasted high risk of invasion under one-hundred year flood conditions for P. canaliculata, P. diffusa, and P. maculata respectively (Table 15, Figure 9-11).

Table 15: Forecasted percent cover of Florida at risk of Pomacea spp. invasion Species High Risk Moderate Risk Low Risk P. canaliculata 1 65 34 P. diffusa 1 56 43 P. maculata 29 57 14

Figure 9: P. canaliculata risk analysis map using FEMA one-hundred year flood data

Figure 10: P. diffusa risk analysis map using FEMA one-hundred year flood data

Figure 11: P. maculata risk analysis map using FEMA one-hundred year flood data

4.0 Discussion and Conclusion According to the models developed herein, locations where P. canaliculata have already been observed in the Lower St. Johns watershed were forecasted as the most suitable for this species within the state of Florida. Secondarily, the majority of south and central Florida were forecasted as moderately suitable, and northwest Florida was found to be the least suitable (Figure 3, 6, and 9). Locations within the Southeast Coast- Biscayne Bay watershed were forecasted as the most suitable for P. diffusa, the majority of south Florida was forecasted as moderately suitable, and central and northern Florida were mostly forecasted as the least suitable. The portion of central Florida that appeared as moderately suitable for P. diffusa, however, is likely a result of generalizing all missing Florida county flood risk data from the FEMA one-hundred year flood dataset as moderately suitable (Figure 4, 7, and 10). Locations within Everglades West Coast, Southeast Coast- Biscayne Bay, Upper East Coast, Upper St. Johns, Charlotte Harbor, Withlacoochee, Tampa Bay, Ocklawaha, Ochlocknee- St. Marks, and Choctawhatchee- St. Andrews watersheds (see Appendix 31 for watershed delineations) were found to be most suitable for P. maculata. Other watersheds in south and central Florida were generally forecasted as moderately suitable for P. maculata, and northern and northwestern Florida were forecasted as least suitable (Figure 5, 8, and 11). On the balance of probabilities, each respective apple snail species in this study, particularly P. maculata, may have already been introduced to suitable areas as forecasted by the SDMs, but records of observation at these locations were absent in the data used for this study. Therefore, areas of moderate suitability should be assumed to be at a high risk of invasion. This observation highlights the importance of implementing early detection, prevention, and eradication plans within areas of high suitability. Temperature is one of the most important biological factors limiting the spread of the invasive apple snails according to this study. P. canaliculata reportedly tolerates temperatures of 13-37 °C (Byers et al. 2013, Hayes et al. 2015), with temperatures above 30 °C or inhibiting the viability and production of eggs respectively (Seuffert and Martín 2017), and they aestivate at 10-17.5 °C (Estebenet and Martin 2002), whereas P diffusa tolerates 18-44.9 °C (Howell et al. 2006, Mu et al. 2015), and P. maculata tolerates 0- 36.6 °C, but the latter’s embryonic development has an lower thermal limit of 15°C

(Martin et al. 2001, Ramakrishnan 2007, Byers et al. 2013, Yoshida 2014). Low temperatures can decrease activity and many snails will withdraw into their shells. Snails aestivate when temperatures remain low, and can remain this way up to a few months. However, many snails will die if lower temperatures persist (Ito 2002, Steven et al. 2002, Byers et al. 2013). The lower thermal tolerances of the three species studied here are therefore likely to limit the distribution of Pomacea species in the continental US to states with warmer temperatures. However, a population of P. canaliculata does persist in temperate regions of Japan. This population shows seasonal fluctuations of cold hardiness, but the physiological mechanism of how this is developed is still relatively unknown. In general, cold-acclimation, decreased water body content- which could result in increased osmolarity, production of low molecular weight compounds (i.e. glycerol) (Matsukura and Wada 2007, Wada and Matsukura 2007, Wada and Matsukura et al. 2008, Matsukura et al. 2009). Furthermore, some studies have shown a linkage between the mechanisms for cold hardiness and desiccation tolerance, with cold tolerant snails more tolerant to desiccation than cold-intolerant snails (Matsukura 2011). Still, consistent high temperatures such as found in Florida can increase feeding rates, increased activity (reproduction and movement), reduce life span, increase rate to maturation, and allow for continuous reproduction (Estebenet and Cazzanga 1992, Ramakrishnan 2007, McAskill 2015). This can be problematic for the sympatric, native P. paludosa as the presence of non-native apple snails has been shown to decrease the growth rate, decreased survivability, and delayed sexual maturity- which affects the size and age of first copulation and recruitment (Posch et al. 2013). Increased populations of non-native apples snails can also alter macrophyte community structure (Conner et al. 2008, Horgan et al. 2014, Monette 2014). Apple snails are sensitive to changes in hydrology (age and size dependent) and Florida is similar to the snails’ native South American habitat which experiences hydrologic cycles (Glass and Darby 2009). In their native range, many species aestivate during the dry season and breed in the rainy season once ideal temperatures are reached (Scott 1957, Andrews 1964, Coelho et al. 2012). Under suitable conditions copulation and spawning can occur year-round (Estebenet and Martin 2002). While Florida does not experience a true dry season, it does undergo dry down events. During dry downs, apple

36 snails aestivate up to 25 months. If dry conditions persist, mortality rates likely increase which would affect recruitment in the following years (Darby et al. 2002, Darby et al. 2008, Glasheen et al. 2017). On the other hand, prolonged inundation can degrade habitat structures such as emergent vegetation, which are important structures for snails life history particularly reproduction and respiration. Newly hatched snails (<5 mm shell) can only tolerate dry down conditions for a few days. In contrast adult snails (>25 mm shell) can tolerate dry conditions and show high survival rates (up to 75%) after 3 months (USGS1 2017). Furthermore, a number of Florida’s water bodies have a suitable pH for apple snail survival. Prolonged exposure to acidic waters below or at the respective lower range of tolerance may inhibit apple snail spread (Table 1, Ramakrishnan 2007, Byers et al. 2013). The pH tolerance of these species is important for predicting invasion risk as apple snails pose a threat to both natural and agricultural wetlands (Howells and Smith 2002). Regulation of water levels may help manage invasive Pomacea populations as submersion has been shown to decrease hatching rate and increase duration of development as a result of physiological stresses (Pizani et al. 2005). Timing of such natural control efforts would be crucial as the age at which submersion occurs affects hatching rate and duration of development. Furthermore, submersion has also been reported to have similar, if not more detrimental effects on embryonic growth and viability of native P. paludosa embryos (Turner 1998). These differences can be explained by their respective habitats and spawning seasons. Florida’s tropical climate undergoes periodic wet and dry seasons and the spawning of P. paludosa overlaps with periods of decreased water levels, primarily in April and May. In contrast, the Southern Pampas is semiarid and water levels are more unpredictable, which may explain the higher tolerance to submersion in P. canaliculata (Pizani et al. 2005). Aerial exposure is less likely to affect Pomacea spp. as most have truly cleidoic (i.e., gas exchange with the environment) calcareous eggs. This has been reported to provide an effective barrier against water loss. Though some studies show that most hatchlings and many immature snails can be managed by repeated drainage of paddy fields in which exposing the soil for short periods of time (less than four days) (Litsinger and Estano 1993, Wada et al. 1999). Dry-downs would also open up new predation (e.g., fire ants) opportunities that may aid in natural biological control of the snails (Yusa 2001, Stevens et al. 1999).

Using the ‘climate matching’ method via MaxEnt is one preliminary step in evaluating the risk of invasion of apple snails in Florida and is well-suited to prioritizing regions for future surveys by estimating potential distributions. Given successful model validation, the model can be used to fill in the gaps between known snail occurrences and their forecasted distributions in new regions in the future (Pearson 2010). However, it should be noted that this method still has several limitations that must be acknowledged. For example, the variables used for modelling in this study only represent a subset of the possible environmental variables that may influence apple snail distribution (Araújo and Luoto 2007, Pearson 2010). MaxEnt modeling is not able to consider other biotic and abiotic interactions (i.g. competition, predator-prey interactions) that are not included within the environmental variables used to predict the species distribution (Peterson and Vieglais 2001). Therefore, additional environmental variables should be added to increase the predictive power of the models and the areas identified herein should be re- evaluated with the availability of new data for future study of apple snail spread. Additionally, comprehensive surveys of apple snail distribution in Florida would also increase model predictability as sampling extent and intensity are typically biased to areas easily accessible (Anderson 2003). Other errors may arise due to a lack of sufficient geographic detail, species misidentification, low number of occurrence localities, or the set of environmental variables that were used may not be sufficient to describe all parameters of a species fundamental niche. SDMs can also be problematic for species with few occurrence records, as may be the case for P. canaliculata and P. diffusa. Ironically, sparsely-documented species are frequently those which need predictive models the most (Shcheglovitova and Anderson 2013). Still, SDM’s are a useful tool that can identify and assess species specific ranges, screen for likely pests, and of interest to this study, invasive potential before an invasion takes place (Peterson 2003, Pearson et al. 2006, Herborg et al. 2007). Such information could be used to focus management efforts for prevention, control, and in some cases eradication. The SDM’s produced in this study show that much of Florida is at risk of invasion by at least one species of apple snail. Municipalities located in areas with the greatest risk should inform and educate the public of the potential impacts of this species to prevent new introductions. In some cases, eradication plans should be considered. For moderate

38 and low risk areas, additional surveys should be conducted to detect if additional populations of invasive apple snails exist. Education and a strict monitoring plan should be created to prevent newly introduced populations from establishing, followed by an eradication plan when necessary.

Acknowledgements I thank my advisor Dr. Matthew Johnston for his guidance and encouragement throughout this study. His expertise in invasive species and computer modelling were an invaluable resource throughout the course of my work. I also thank my committee members Dr. Kenneth A. Hayes whose expertise on apple snails and continued guidance was instrumental in the success of this project, and Dr. Bernhard Riegl for his continued time and patience throughout this process. Special thanks to Dr. Jennifer Bernatis and John Slapcinsky who evaluated all recorded observation data. I also express my gratitude to my family for their continued love, support, and encouragement through this long and sometimes frustrating journey. Without you I would not be where I am today. Additional thanks to Nolan Padilla and Marissa Skinner, whose love and encouragement have kept me sane as I embark down this career path. Lastly, thank you to my professors and fellow students at the University of Tampa and Nova Southeastern University’s Halmos College of Natural Sciences and Oceanography for all their imparted knowledge, support, advice, and skill in pursuit of my Master’s Degree and future.

References

Akinyemiju O.A. (1987). Invasion of Nigerian Waters by Water Hyacinth. Journal of Aquatic Plant Management, 25, 24-26.

Anderson R. P. (2003). Real vs. artefactual absences in species distributions: tests for Oryzomys albigularis (Rodentia: Muridae) in Venezuela. Journal of Biogeography, 30(4), 591-605.

Andrews E.B. (1964). The functional anatomy and histology of the reproductive system of some pilid gastropod molluscs. Proceedings of Malacological Society of London, 36, 121-138.

Allan J.D. and Johnson L.B. (1997). Catchment-scale analysis of aquatic ecosystems. Freshwater Biology, 37, 107-111.

Araújo M.B. and Luoto M. (2007). The importance of biotic interactions for modeling species distributions under climate change. Global Ecology and Biogeography, 16, 743–753.

Bargeron C.T. and Moorhead, D.J. (2007). EDDMapS- Early Detection and Distribution Mapping System for the Southeast Exotic Pest Plant Council. Wildland Weeds, 4- 7.

Bernatis J.L. (2008). Shell morphology and physiological tolerances of the exotic apple snails Pomacea insularum and Pomacea canaliculata, in Florida.

Bernatis J.L. (2014). Morphology, ecophysiology, and impacts of nonindigenous Pomacea in Florida. University of Florida.

Bernatis J.L. and Warren G.L. (2014). Effectiveness of a Hand Removal Program for Management of Nonindigenous Apple Snails in an Urban Pond. Southeastern Naturalist, 13(3), 607-318.

Bradley B.A. and Marvin D.C. (2011). Using expert knowledge to satisfy data needs: mapping invasive plant distributions in the western United States. Western North American Naturalist, 71(3), 302-315.

Burlakova L.E., Karatayev A.Y., Padilla D.K., Cartwright L.D., and Hollas D.N. (2009). Wetland Restoration and Invasive Species: Apple snail (Pomacea insularum) Feeding on Native and Invasive Aquatic Plants. Restoration Ecology, 17(3), 433-440.

Byers J.E., McDowel W.G., Dodd S.R., Haynie R.S., Pintor L.M., and Wilde S.B. (2013). Climate and pH Predict the Potential Range of the Invasive Apple Snail

(Pomacea insularum) in the Southeastern United States. PLOS One. 8(2): e56812. doi:10.1371/journal.pone.0056812.

Capiner J.L. and White J. 2011. Terrestrial snails affecting plants in Florida. Department of Entomology, University of Florida. Publication number EENY-497.

Carlsson N.O.L., Bronmark C., and Hansson L. (2004). Invading Herbivory: The Golden Apple Snail Alters Ecosystem Functioning in Asian Wetlands. Ecology, 85(6), 1575-1580.

Carlsson N.O.L. and Lacoursiére J.O. (2005). Herbivory on aquatic vascular plants by the introduced golden apple snail (Pomacea canaliculata) in Lao PDR. Biological Invasions, 7, 233-241.

Carlton J.T. and Geller J.B. (1993). Ecological Roulette: The Global Transport of Nonindigenous Marine Organisms. Science, 261, 78-82.

Coelho A.R., Calado G.J.P., and Dinis M.T. (2012). Freshwater snail Pomacea bridgesii (Gastropoda: Ampullariidae), life history traits and aquaculture potential. AACL Bioflux, 5(3), 168-181.

Conner S.L., Pomory C.M., and Darby P.C. (2008). Density Effects of Native and Exotic Snails on Growth in Juvenile Apple Snails Pomacea paludosa (Gastropoda: Ampullariidae): A Laboratory Experiment. Journal of Molluscan Studies, 74(4), 355-362.

Cowie R.H. (2002). Apple snails (Ampullariidae) as agricultural pests: their biology, impacts and management. Molluscs as Crop Pests, 145-170

Cowie R.H. and Hayes, K.A. (2012). Apple snails. A Handbook of Global Freshwater Invasive Species. Earthscan, Oxon, 207-221.

Darby P.C., Bennetts R.E., and Percival H.F. (2008). Dry Down Impacts on Apple Snail (Pomacea paludosa) Demography: Implications for Wetland Water Management. Wetlands, 28(1), 204-214.

Darby P.C., Bennetts R.E., Miller S.J., and Percival H.F. (2002). Movements of Florida Apple Snails in Relation to Water Levels and Drying Events. Wetlands, 22(3), 489-498.

Darby P.C., Valentine-Darby P.L., Percival H.F., and Kitchens W.M. (2004). Florida apple snail (Pomacea paludosa Say) responses to lake habitat restoration activity. Archi Fϋr Hydrobiologie, 161(4), 561-575.

Dong S., Zheng G., Yu X., and Fu C. (2012). Biological control of golden apple snail, Pomacea canaliculata by Chinese soft-shelled turtle, Pelodiscus sinensis in the wild rice, Zizania latifolia field. Scientia Agricola, 69(2), 142-146.

Dubey S., Pike D.A., and Shine R. (2013). Predicting the impacts of climate change on genetic diversity in an endangered lizard species. Climatic Change, 117, 319- 327.

Elith J. (2013). Predicting distributions of invasive species. arXiv preprint arXiv:1312.0851.

Elith J., Graham, C.H., Anderson, R.P., Dudık, M., and Ferrier, S. (2006). Novel methods improve prediction of species’ distribution models. Ecography, 32, 66-77.

Elith J., Kearney, M., and Phillips, S. (2010). The art of modelling range-shifting species. Methodsin Ecology and Evolution, 1, 330–342.

Elith J. and Leathwick J. (2009). The contribution of species distribution modelling to conservation prioritization. In: Moilanen A, Wilson KA, Possingham HP, Editors. Spatial conservation prioritization: Quantitative methods & computational tools. pp. 70–93. Oxford University Press.

Elith J., Phillips S.J., Hastie T., Dudík M., Chee Y.E., and Yates C.J. (2011). A statistical explanation of MaxEnt for ecologists. Diversity and Distributions, 17, 43-57.

Emerton L. and Howard G. (2008). A Toolkit for the Economic Analysis of Invasive Species. Global Invasive Species Programme, Nairobi.

Endries M. (2011). Aquatic Species Mapping in North Carolina Using Maxent. US Fish and Wildlife Service, Ecological Services Field Office, Ashville, NC.

ESRI 2011. ArcGIS Desktop: Release 10. Redlands, CA: Environmental Systems Research Institute.

Estebenet A.L., and Cazzaniga, N.J. (1992). Growth and demography of Pomacea canaliculata (Gastropoda: Ampullariidae) under laboratory conditions. Malacological Review, 25(1-2), 1-12.

Estebenet A.L., and Martin P.R. (2002). Pomacea canaliculata (Gastropoda: Ampullariidae): life history traits and their plasticity. Biocell, 26, 83-89.

Falk W. and Mellert K.H. (2011). Species distribution models as a tool for forest management planning under climate change: risk evaluation of Abies alba in Bavaria. Journal of Vegetation Science, 22, 621-634.

Ficetola G.F., Thuiller W. and Miaud C. (2007). Prediction and Validation of the Potential Global Distribution of a Problematic Invasive Species: The American Bullfrog. Diversity and Distributions, 13(4), 476-485.

Florida Fish and Wildlife Conservation Commission. (2017). Florida’s Exotic Fish and Wildlife. http://myfwc.com

FEMA. (2017). FEMA GeoPlatform. https://fema.maps.arcgis.com/home/index.html

Ferreira A.C. and Capítulo A.R. (2017). Growth and survival of juvenile Pomacea canaliculata (Gastropoda: Ampullariidae) in plain streams associated to different land uses. Studies on Neotropical Fauna and Environment, 52(2), 95-102.

Fick S.E. and Hijmans R.J. (2017). Worldclim 2: New 1-km spatial resolution climate surfaces for global land areas. International Journal of Climatology. Florida Department of Environmental Protection. (2016). Watershed Management. www.dep.state.fl.us

Frakes R.A., Bargar T.A., and Bauer E.A. (2008). Sediment copper bioavailability to freshwater snails in south Florida: risk implications for the Everglade snail kite (Rostrhamus sociabilis plumbeus). Ecotoxicology, 17, 598-604.

Glasheen P.M., Calvo C., Meerhoff M., Hayes K.A. and Burks R.L. (2017). Survival, recovery, and reproduction of apple snails (Pomacea spp.) following exposure to drought conditions. Freshwater Science, 36(2), 000-000.

Glass N.H. and Darby P.C. (2009). The effect of calcium and pH on Florida apple snail Pomacea paludosa (Gastropoda: Ampullariidae), shell growth and crush weight. Aquatic Ecology, 43, 1085-1093.

Guo J., Zhang C., Xiang Y., Zhang J.E. and Liang K. (2017). Biological control of the exotic invasive snail Pomacea canaliculata with the indigenous medicinal leech Whitmania pigra. Biocontrol Science and Technology, 27(9), 1071-1081.

Guo J., Zhang J.E., Zhao B., Luo M. and Zhang C. (2016). The role of spotted green pufferfish Tetraodon nigroviridis in controlling golden apple snail Pomacea canaliculata: an effective biological control approach involving a new agent. Biocontrol Science and Technology, 26(8), 1100-1112.

Hanelt B., Grother L.E., Janovy J. and Jr. (2001). Physid Snails as Sentinels of Freshwater Nematomorphs. The Journal of Parasitology, 87(5), 1049-1053.

Harman W.N., (1972). Benthic Substrates: Their Effect on Fresh-Water Mollusca. Ecology, 53(2), 271-277.

Hayes K.A., Joshi R.C., Thiengo S.C., and Cowie R.H. (2008). Out of South America: multiple origins of non‐native apple snails in Asia. Diversity and distributions, 14(4), 701-712.

Hayes K.A., Cowie R.H., Jorgensen A., Schultheiβ R., Albrecht C., and Thiengo S.C. (2009). Molluscan Models in Evolutionary Biology: Apple Snails (Gastropoda: Ampullariidae) as a System for Addressing Fundamental Questions. American Malacological Bulletin, 27(1/2), 47-58.

Hayes K.A., Cowie, R.H., Thiengo, S.C., and Strong, E.E. (2012). Comparing apples with apples: clarifying the identities of two highly invasive Neotropical Ampullariidae (Caenogastropoda). Zoological Journal of the Linnean Society, 166(4), 723-753.

Hayes K.A., Burks R.L., Castro-Vazquez A., Darby P.C., Heras H., Martin P.R., Qiu J., Thiengo S.C., Vega I.A., Wada T., Yusa Y., Burela S., Cadierno M.P., Cueto J.A., Dellagnola F.A., Dreon M.S., Frassa M.V., Giraud-Billoud M., Godoy M.S., Ituarte S., Koch E., Matsukura K., Pasquevich M.Y., Rodriguez C., Saveanu L., Seuffert M.E., Strong E.E., Sun J., Tamburi N.E., Tiecher M.J., Turner R.L., Valentine-Darby P.L., and Cowie R.H. (2015). Insights from an Integrated View of the Biology of Apple Snails (Caenogastropoda: Ampullariidae). Malacologia, 58(1-2), 245-302.

Herbold B. and Moyle P.B. (1986). Introduced Species and Vacant Niches. The American Naturalist, 128(5), 751-760.

Herborg L.M., Jerde C.L., Lodge D.M., Ruiz G.M., and MacIsaac H.J. (2007). Predicting invasion risk using measures of introduction effort and environmental niche models. Ecological applications, 17(3), 663-674.

Hijmans R.J. (2012). Cross-validation of species distribution models: removing spatial sorting bias and calibration with a null model. Ecology, 93(3), 679-688.

Hijmans R.J., Cameron S. & Parra J. (2005) WorldClim, version 1.3. University of California, Berkeley http://biogeo.berkeley.edu/worldclim/worldclim.htm.

Hoang T.C., Rogevich E.C., Rand G.M., Gardinali P.R., Rakes R.A., and Bargar T.A. (2008). Copper desorption in flooded agricultural soils and toxicity to the Florida apple snail (Pomacea paludosa): Implication in Everglades restoration. Environmental Pollution, 154, 338-347.

Hoddle M.S. (2004). Restoring balance: using exotic species to control invasive exotic species. Conservation Biology, 18(1), 38-49.

Hollingsworth R.G., Cowie R.H., Joshi,R.C., & Sebastian L.S. (2006). Apple snails as disease vectors. Global advances in ecology and management of golden apple snails, 121-132.

Homer C.G., Dewitz J.A., Yang L., Jin S., Danielson P., Xian G., Coulston J., Herold N.D., Wickham J.D., and Megown K. (2015). Completion of the 2011 National Land Cover Database for the conterminous United States-Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, 81(5), 345-354.

Horgan F.G., Stuart A.M., and Kudavidange E.P. (2014). Impact of invasive apple snails on the functioning and services of natural and managed wetlands. Acta Oecologica, 54, 90-100.

Howells R.G., Burlakova L.E., Karatayev A.Y., Marfurt R.K., and Burks R.L. (2006). Native and introduced Ampullaridae in North America: History, status, and ecology. Global Advances in the Ecology and Management of Golden Apple Snails, 73-112.

Howells R.G., and Smith J.W. (2002). Status of the applesnail Pomacea canaliculata in the United States. In Proceedings of the Special Working Group on the Golden Apple Snail (Pomacea spp.). The Seventh International Congress on Medical and Applied Malacology (7th ICMAM). Los Banos, Laguna, Philippines, SEAMEO Regional Center for Graduate Study and Research in Agriculture (SEARCA) (Vol. 8696).

Hutchinson G.E. (1957). Concluding remarks. Cold Spring Harbor Symposia on Quantitative Biology, 22, 415-427.

Huxel G.R. (1999). Rapid displacement of native species by invasive species: effects of hybridization. Biological Conservation, 89, 143-152.

Ito K. (2002). Environmental factors influencing overwintering success of the golden apple snail, Pomacea canaliculata (Gastropoda: Ampullariidae), in the northernmost population of Japan. Application of Entomological Zoology, 37(4), 655-661.

Johnston M.W. and Purkis S.J. (2011). Spatial analysis of the invasion of lionfish in the western Atlantic and Caribbean. Marine Pollution Bulletin, 62, 1218-1226

Jordan P.J. and Deaton L.E. (1999). Osmotic regulation and salinity tolerance in the freshwater snail Pomacea bridgesi and the freshwater clam Lampsilis teres. Comparative Biochemistry and Physiology Part A, 122, 199-205.

Kappes H. and Haase P. (2012). Slow, but steady: dispersal of freshwater molluscs. Aquatic Science, 74, 1-14.

Karraker N.E. and Dudgeon D. (2014). Invasive apple snails (Pomacea canaliculata) are predators of amphibians in South China. Biological invasions, 16(9), 1785-1789.

Karunaratne L.B., Darby P.C., and Bennetts R.E. (2006). The Effects of Wetland Habitat Structure on Florida Apple Snail Density. Wetlands, 26(4), 1143-1150.

Kim J.R., Hayes K.A., Yeung N.W., and Cowie R.H. (2014). Diverse Gastropod Hosts of Angiostrongylus cantonensis, the Rate Lungworm, Globally and with a Focus on the Hawaiian Islands. PLos ONE, 9(5), e94969. doi:10.1371/journal.pone.0094969.

Klug P.E., Reed R.N., Mazzotti F.J., McEachern M.A., Vinci J.J., Craven K.K., and Adams A.A.Y. (2015). The influence of disturbed habitat on the spatial ecology of Argentine black and white tegu (Tupinambis merianae), a recent invader in the Everglades ecosystem (Florida, USA). Biological Invasions, 17(6), 1785- 1797.

Kozak K.H., Graham C.H., and Wiens J.J. (2008). Integrating GIS-based environmental data into evolutionary biology. Trends in Ecology & Evolution, 23(3), 141-148.

Kwong K, Chan R.K.Y. and Qiu J. (2009). The Potential of the Invasive Snail Pomacea canaliculata as a Predator of Various Life-Stages of Freshwater Snails. Malacologia. 51(2), 343-356.

Litsinger J.A., and Estano D.B. (1993). Management of the golden apple snail Pomacea canaliculata (Lamarck) in rice. Crop Protection, 12(5), 363-370

Lockwood J.L., Hoopes M.F., and Marchetti M.P. (2013). Invasion ecology. Chichester, West Sussex. Wiley-Blackwell

Lowe S., Browne M., Boudjelas S., and De Poorter, M. (2000) 100 of the world's worst invasive alien species: a selection from the global invasive species database. The Invasive Species Specialists Group of the Species Survival Commission of the World Conservation Union. First published in Aliens 12, December 2000. Reprinted November 2004. Hollands Printing, Auckland, New Zealand

Lv S., Ahang Y., Liu H., Hu L., Liu Q., Wei F., Guo Y., Steinmann P., Hu W., Zhou X., and Utzinger J. (2013). Phylogenetic evidence for multiple and secondary introductions of invasive snails: Pomacea species in the People’s Republic of China. Diversity and Distributions, 19, 147-156.

MacDougall A.S. and Turkington R. (2005). Are Invasive Species the Drivers or Passenger of Change in Degraded Ecosystems? Ecological Society of America, 86(1), 42-55.

Mantyka-Pringle C.S., Martin T.G., Moffatt D.B. Linke S. and Rhodes J.R. (2014). Understanding and predicting the combined effects of climate change and land- use change on freshwater macroinvertebrates and fish. Journal of Applied Ecology, 51, 572-582.

Martin C.W., Bayha K.M., and Valentine J.F. (2012). Establishment of the Invasive Island Apple Snail Pomacea insularum (Gastropoda: Ampullariidae) and Eradication Efforts in Mobile, Alabama, USA. Gulf of Marine Science, 30, 30-38.

Martin C.W. and Valentine J.F. (2014). Tolerance of embryos and hatchlings of the invasive apple snail Pomacea maculata to estuarine conditions. Aquatic Ecology, 48, 321-326.

Martín P.R., Estebenet A.L., and Cazzaniga N.J. (2001). Factors affecting the distribution of Pomacea canaliculata (Gastropoda: Ampullariidae) along its southernmost natural limit. Malacologia, 43, 13-23.

Matsukura K., Tsumuki H., Izumi Y. and Wada T. (2008). Changes in chemical components in the freshwater apple snail, Pomacea canaliculata (Gastropoda: Ampullariidae), in relation to the development of its cold hardiness. Cryobiology, 56(2), 131-137.

Matsukura K., Tsumuki H., Izumi Y. and Wada T. (2009). Temperature and water availability affect decrease of cold hardiness in the apple snail, Pomacea canaliculata. Malacologia, 51(2), 263-269.

Matsukura K. and Wada T., 2007. Environmental factors affecting the increase of cold hardiness in the apple snail Pomacea canaliculata (Gastropoda: Ampullariidae). Applied entomology and zoology, 42(4), 533-539.

McAskill S.C. (2015). Interactive effects of environmental stressors and the invasive apple snail, Pomacea maculata, on tapegrass, Vallisneria Americana. (Master’s Thesis). Florida Gulf Coast University.

Monette D.J. (2014). Resource Use, Competition, Grazing Behavior, and. Waterbirds, 20(2), 324- 329.

Monette D., Ewe S. and Markwith S.H. (2016). Effects of the Consumption Behavior of Adult Pomacea maculata and Pomacea paludosa on Vallisneria americana. Southeastern Naturalist, 15(4), 689-696.

Moreno R., Zamora R., Molina J.R. Vasquez A. Herrera A.M. (2011). Predictive modelling of microhabitats for endemic birds in Southern Chilean temperate forest using Maximum entropy (Maxent). Ecological Informatics, 6(6) 364-370.

Millennium Ecosystem Assessment. (2005). Ecosystems and human well- being. Washington, DC.

Morrison W.E. and Hay M.E. (2011). Feeding and growth of native, invasive, and non- invasive alien apple snails (Ampullariidae) in the United States: Invasives eat more and grow more. Biological Invasions, 13, 945-955.

Mu H., Sun J., Fang L., Luan T., Williams G.A., Cheung S.G., Wong C.K.C., and Qiu J. (2015). Genetic Basis of Differential Heat Resistance between Two Species of Congeneric Freshwater Snails: Insights from Quantitative Proteomics and Base Substitution Rate Analysis. Journal of Proteome Research, 14(10), 4296-4308.

Naylor R. (1996). Invasions in Agriculture: Assessing the Cost of the Golden Apple Snail in Asia. Ambio. 25(7), 443-448.

Pan D., Zhang J., Long J., Zhang J.E., Zhao B. and Luo M., 2014. Impacts of golden apple snail (Pomacea canaliculata) on water quality and microbes. Zhongguo Shengtai Nongye Xuebao/Chinese Journal of Eco-Agriculture, 22(1), 58-62.

Pearce J., and Ferrier S. (2000). Evaluating the predictive performance of habitat models developed using logistic regression. Ecological modelling, 133(3), 225-245.

Pearson R.G. (2007). Species’ distribution modeling for conservation educators and practitioners. Synthesis. American Museum of Natural History, 54-89.

Pearson R.G., Thuiller W., Araújo M.B., Martinez‐Meyer E., Brotons L., McClean C., Miles L., Segurado P., Dawson T.P., and Lees D.C. (2006). Model‐based uncertainty in species range prediction. Journal of Biogeography, 33(10), 1704- 1711.

Pease K.M., Freedman A.H., Pollinger, J.P. McCormack J.E., Buermann W., Rodzen J., Banks J., Meredith E., Bleich V.C., Schaefer R.J., Jones K, and Wayne R.K. (2009). Landscapes genetics of California mule deer (Odocoileus hemionus): the role of ecological and historical factors in generating differentiation. Molecular Ecology, 18, 1848-1862.

Peh K.S.H. (2010). Invasive species in Southeast Asia: the knowledge so far. Biodiversity and Conservation, 19, 1083-1099.

Peterson A.T. (2003). Predicting the Geography of Species’ Invasions via Ecological Niche Modeling. The Quarterly Review of Biology, 78(4), 419-433.

Peterson A.T. and Nakazawa Y. (2008). Environmental data sets matter in ecological niche modelling: an example with Solenopsis invicta and Solenopsis richteri. Global ecology and Biogeography, 17(1), 135-144.

Peterson A.T. and Vieglais D.A. (2001). Predicting species invasions using Ecological niche modeling: new approaches from bioinformatics attack a pressing problem. Bioscience, 51, 5.

Phillips S.J., Anderson R.P., and Schapire R.E. (2006). Maximum entropy modeling of species geographic distributions. Ecological modelling, 190(3), 231-259.

Phillips S.J. and Dudík M. (2008). Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography, 31(2), 161-175.

Phillips S.J. and Elith J. (2010). POC plots: calibrating species distribution models with presence‐only data. Ecology, 91(8), 2476-2484.

Pias K.E., Welch Z.C., and Kitchens W.M. (2012). An Artificial Perch to Help Snail Kites Handle an Exotic Apple Snail. The Waterbird Society, 35(2), 347-351.

Pimentel D., Zuniga R., and Morrison D. (2005). Update on the environmental and economic costs associated with alien-invasive species in the United States. Ecological Economics, 52(3), 273-288.

Pimm S.L. (1989). Theories of predicting success and impact of introduced species. Biological Invasions: a Global Perspective, 351-367.

Pizani N., Estebenet A.L., and Martín P.R. (2005) Effects of submersion and aerial exposure on clutches and hatchlings of Pomacea canaliculata (Gastropoda: Ampullariidae). American Malacological Bulletin, 20, 55–63.

Plant Health Australia. (2009). Golden Apple Snail. Prepared by the Ricegrowers’ Association of Australia Inc.

Posch H., Garr A.L., and Reynolds E. (2013). The Presence of an Exotic Snail, Pomacea maculata, Inhibits Growth of Juvenile Florida Apple Snails, Pomacea paludosa. Journal of Molluscan Studies, 79(4), 383-385.

Prabhakaran G., Bhore S.J. and Ravichandran M. (2017). Development and Evaluation of Poly Herbal Molluscicidal Extracts for Control of Apple Snail (Pomacea maculata). Agriculture, 7(3), 22.

Raes N. and Steege H. (2007). A null-model for significance testing of presence-only species distribution models. Ecography, 30, 727-736.

Ramakrishnan V. (2007). Salinity, pH, Temperature, Desiccation, and Hypoxia Tolerance in the Invasive Freshwater Apple Snail Pomacea insularum (Doctoral Dissertation). University of Texas at Arlington.

Rawlings T.A., Hayes K.A., Cowie R.H., and Collins T.M. (2007). The identity, distribution, and impacts of non-native apple snails in the continental United States. BMC Evolutionary Biology, 7: 97.

Rawlins K.A., Griffin J.E., Moorhead D.J., Bargeron C.T., and Evans C.W. (2011). EDDMapS: Invasive Plant Mapping Handbook. The University of Georgia. Center for Invasive Species and Ecosystem Health, Tifton GA. BW-2011-02, 3- 3.

Ray G.L. (2005). “Invasive estuarine and marine animals of the Gulf of Mexico,” ANSRP Technical Notes Collection (ERDC/TN ANSRP-05-4), U.S. Army Engineer Research and Development Center, Vicksburg, MS. http://el.erdc.usace.arm.mil/ansrp/

Robertson S.M. (2012). Potential Threats of the Exotic Apple Snail Pomacea insularum to Aquatic Ecosystems in Georgia and Florida (Graduate Thesis). University of Georgia.

Saatchi S., Buermann W., Steege H.T., Mori S., and Smith T.B. (2008). Modeling distribution of Amazonian tree species and diversity using remote sensing measurements. Remote Sensing of Environment, 11(2), 200-2017.

Sandlund O.T., Schei P.J., and Viken A. (2001). Invasive species and biodiversity management (Vol. 24). Springer Science & Business Media.

Sarell M., Haney A., and Tolkamp C. (2003). Sensitive Ecosystems Inventory: Central Okanagan, 2000-2001. Volume 3: Wildlife Habitat Mapping. Victoria: BC Ministry of Environment.

Savidge J.A. (1987). Extinction of an Island Forest Avifauna by an Introduced Snake. Ecology, 68(3), 660-668.

Savrick C.E., Burks R.L., and Hayes K.A. In Review. Moving from Florida: Spread of invasive Pomacea maculata Perry, 1810 (Ampullariidae) in the southern United States. Biological Invasions

Sawasdee. B., and Kӧhler H.R. (2009). Embryo toxicity of pesticides and heavy metal to the Ramshorn snail, Marisa cornuarietis (Prosobranchia). Chemosphere, 75, 1539-1547.

Scott M.I.H. (1957). Estudio morfologica y taxonomica de los Ampullariidos de la Republica Argentina. Revista del Museo de Argentino de Ciencias Naturales, 3, 233-333

Seuffert M.E. and Martín P.R. (2017). Thermal limits for the establishment and growth of populations of the invasive apple snail Pomacea canaliculata. Biological Invasions, 19(4), 1169-1180.

Sharfstein B. and Steinman A.D. (2001). Growth and survival of the Florida apple snail (Pomacea paludosa) fed 3 naturally occurring macrophyte assemblages. Journal of the North American Benthological Society, 20(1), 84-95.

Shcheglovitova M. and Anderson R.P. (2013). Estimating optimal complexity for ecological niche models: a jackknife approach for species with small sample sizes. Ecological Modelling, 269, 9-17.

Shukia S. (2004). Watersheds- Functions and Management (IFAS ABE350). Gainesville: University of Florida Institute of Food and Agricultural Sciences. Retrieved December 19, 2017. http://edis.ifas.ufl.edu.

Sin T.S. (2003). Damage potential of the golden apple snail Pomacea canaliculata (Lamarck) in irrigated rice and its control by cultural approaches. International Journal of Pest Management, 49(1), 49-55.

Sin T.S. (2006). Evaluation of different species of fish for biological control of golden apple snail Pomacea canaliculata (Lamarck) in rice. Crop Protection, 25(9), 1004-1012.

Solhjouy-Fard S., Sarafrazi A. Moeini M.M., and Ahadiyat A. (2013). Predicting habitat distribution of five heteropteran pest species in Iran. Journal of Insect Science, 13, 116. Available online: www.insectscience.org/13.116

Stevens A.J., Stevens N.M., Darby P.C., and Percival H.F. (1999). Observations of the fire ants (Solenopsis invicta Buren) attacking apple snails (Pomacea paludosa) exposed during dry down conditions. Journal of Molluscan Studies, 65, 507- 510.

Stevens A.J., Welch Z.C., Darby P.C., and Percival H.F. (2002). Temperature Effects on Florida Applesnail Activity: Implications for Snail Kite Foraging Success and Distribution. Wildlife Society Bulletin (1973-2006), 30(1), 75-81.

Teem J. and Gutierrez J.B. (2008). Human Health Risks Associates with Channeled Apple Snails in the GSARP Region. Florida Department of Agriculture & Florida State University.

Teem J.L., Qvarnstrom Y., Bishop H.S., Silva H.J., Carter J., White-Mclean J., and Smith T. (2013). The Occurrence of the Rat Lungworm, Angiostrongylus cantonensis, in Nonindigenous Snails in the Gulf of Mexico Region of the United States. Hawai’i Journal of Medicine and Public Health, 72(6), 11-14.

Teo S.S. (2001). Evaluation of different duck varieties for the control of the golden apple snail (Pomacea canaliculata) in transplanted and direct seeded rice. Crop Protection, 20(7), 599-604.

Teo S.S. (2004). Biology of the golden apple snail, Pomacea canaliculata (Lamarck, 1822), with emphasis on responses to certain environmental conditions in Sabah, Malaysia. Molluscan Research, 24, 139-148.

Thuiller W., Richardson D.M., Pyšek P., Midgley G.F., Hughes G.O., and Rouget M. (2005). Niche‐ based modelling as a tool for predicting the risk of alien plant invasions at a global scale. Global Change Biology, 11(12), 2234-2250.

Turner R.L. (1996). Use of stems of emergent plants for oviposition by the Florida applesnail, Pomacea paludosa, and implications for marsh management. Florida Scientist, 34-49.

Turner R.L. (1998). Effects of submergence on embryonic survival and developmental rate of the Florida applesnail, Pomacea paludosa: implications for egg predation and marsh management. Florida Scientist, 118-129.

United States Environmental Protection Agency. (2017). Water Quality Data (WQX). https://www.epa.gov/waterdata/water-quality-data-wqx

US Federal Executive Order 13112. (1999). Invasive species. 12 April 1999. http://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1024&context=icwdm other. Viewed 22 September 2017.

USGS1. (2017). Apple Snail. http://www.usgs.gov

USGS2. (2017). Florida geologic map data. https://mrdata.usgs.gov/geology/state/state.php?state=FL

USGS3. (2017). Hydrography. https://nhd.usgs.gov/

Usher M.B., Kornberg H., Horwood J.W., Southwood R., and Moore P.D. (1986). Invasibility and Wildlife Conservation: Invasive Species on Nature Reserves [and Discussion]. Royal Society, 314, 695-710.

Václavík T. and Meentemeyer R.K. (2009) Invasive species distribution modeling (iSDM): Are absence data and dispersal constraints needed to predict actual distributions? Ecological Modeling, 220, 3248-3258.

Wada T., Ichinose K., and Higuchi H. (1999). Effect of drainage on damage to direct- sown rice by the apple snail Pomacea canaliculata (Lamarck) (Gastropoda: Ampullariidae). Applied entomology and zoology, 34(3), 365-370.

Wada T. and Matsukura K. (2007). Seasonal changes in cold hardiness of the invasive freshwater apple snail, Pomacea canaliculata (Lamarck)(Gastropoda: Ampullariidae). Malacologia, 49(2), 383-392.

Wada T. and Matsukura K., 2011. Linkage of cold hardiness with desiccation tolerance in the invasive freshwater apple snail, Pomacea canaliculata (Caenogastropoda: Ampullariidae). Journal of Molluscan Studies, 77(2), 149-153.

Wang A. and Pei Y. (2012). Ecological risk resulting from invasive species: a lesson from riparian wetland rehabilitation. Procedia Environmental Sciences, 13, 1798- 1808.

Wallace R.D., Bargeron C.T, Ziska L. and Dukes J. (2014). Identifying invasive species in real time: early detection and distribution mapping system (EDDMapS) and other mapping tools. Invasive species and global climate change, 4, 219-230.

Ward D.F. (2007). Modelling the potential geographic distributions of invasive ant species in New Zealand. Biological Invasions, 9, 723-735.

Warren D.L., and Seifert S.N. (2011). Ecological niche modeling in Maxent: the importance of model complexity and the performance of model selection criteria. Ecology Applications, 21(2), 335–342.

Yoshida K., Matsukura K., Cazzaniga N., and Wada T. (2014). Tolerance to low temperature and desiccation in two invasive apple snails, Pomacea canaliculata and P. maculata (Caenogastropoda: Ampullariidae), collected in their original distribution area (northern and central Argentina). Journal of Molluscan Studies, 80, 62-66.

Yusa Y. (2001). Predation on eggs of the apple snail Pomacea canaliculata (Gastropoda: Ampullariidae) by the fire ant Solenopsis geminata. Journal of Molluscan Studies, 67(3), 275-279.

Appendices

Appendix 1: WorldClim bioclimatic variables and codes BioClim Codes Bioclimatic Variables BIO1 Annual Mean Temperature BIO2 Mean Diurnal Range (Mean of monthly (max temp- min temp) BIO3 Isothermality (BIO2/BIO7)(*100) BIO4 Temperature Seasonality (standard deviation *100) BIO5 Max Temperature of Warmest Month BIO6 Min Temperature of Coldest Month BIO7 Temperature Annual Range (BIO5-BIO6) BIO8 Mean Temperature of Wettest Quarter BIO9 Mean Temperature of Driest Quarter BIO10 Mean Temperature of Warmest Quarter BIO11 Mean Temperature of Coldest Quarter BIO12 Annual Precipitation BIO13 Precipitation of Wettest Month BIO14 Precipitation of Driest Month BIO15 Precipitation Seasonality (Coefficient of Variation) BIO16 Precipitation of Wettest Quarter BIO17 Precipitation of Driest Quarter BIO18 Precipitation of Warmest Quarter BIO19 Precipitation of Coldest Quarter Fick and Hijmans 2017

Appendix 2: National land cover database 2011 (NLCD_2011) classification system Value Class Classification Description 11 Water Open Water- area of open water, less than 25% cover of vegetation or soil 12 Water Perennial Ice/Snow- area characterized by perennial cover of ice and/or snow, generally greater than 25% of total cover 21 Developed Developed, Open Space- areas with a mixture of some constructed materials, but mostly vegetation in the form of lawn grasses. Impervious surfaces account for less than 20% of total cover. These areas most commonly include large-lot single-family housing units, parks, golf courses, and vegetation planted in developed settings for recreation, erosion control, or aesthetic purposes. 22 Developed Developed, Low Intensity- areas with a mixture of constructed materials and vegetation. Impervious surfaces account for 20% to 49% percent of total cover. These areas most commonly include single-family housing units. 23 Developed Developed, Medium Intensity -areas with a mixture of constructed materials and vegetation. Impervious surfaces

account for 50% to 79% of the total cover. These areas most commonly include single-family housing units. 24 Developed Developed High Intensity-highly developed areas where people reside or work in high numbers. Examples include apartment complexes, row houses and commercial/industrial. Impervious surfaces account for 80% to 100% of the total cover. 31 Barren Barren Land (Rock/Sand/Clay) - areas of bedrock, desert pavement, scarps, talus, slides, volcanic material, glacial debris, sand dunes, strip mines, gravel pits and other accumulations of earthen material. Generally, vegetation accounts for less than 15% of total cover. 41 Forest Deciduous Forest- areas dominated by trees generally greater than 5 meters tall, and greater than 20% of total vegetation cover. More than 75% of the tree species shed foliage simultaneously in response to seasonal change. 42 Forest Evergreen Forest- areas dominated by trees generally greater than 5 meters tall, and greater than 20% of total vegetation cover. More than 75% of the tree species maintain their leaves all year. Canopy is never without green foliage. 43 Forest Mixed Forest- areas dominated by trees generally greater than 5 meters tall, and greater than 20% of total vegetation cover. Neither deciduous nor evergreen species are greater than 75% of total tree cover. 51 Shrubland Dwarf Scrub- Alaska only areas dominated by shrubs less than 20 centimeters tall with shrub canopy typically greater than 20% of total vegetation. This type is often co- associated with grasses, sedges, herbs, and non-vascular vegetation. 52 Shrubland Shrub/Scrub- areas dominated by shrubs; less than 5 meters tall with shrub canopy typically greater than 20% of total vegetation. This class includes true shrubs, young trees in an early successional stage or trees stunted from environmental conditions. 71 Herbaceous Grassland/Herbaceous- areas dominated by gramanoid or herbaceous vegetation, generally greater than 80% of total vegetation. These areas are not subject to intensive management such as tilling, but can be utilized for grazing. 72 Herbaceous Sedge/Herbaceous- Alaska only areas dominated by sedges and forbs, generally greater than 80% of total vegetation. This type can occur with significant other grasses or other grass like plants, and includes sedge tundra, and sedge tussock tundra.

73 Herbaceous Lichens- Alaska only areas dominated by fruticose or foliose lichens generally greater than 80% of total vegetation. 74 Herbaceous Moss- Alaska only areas dominated by mosses, generally greater than 80% of total vegetation. 81 Planted/Cultivated Pasture/Hay-areas of grasses, legumes, or grass-legume mixtures planted for livestock grazing or the production of seed or hay crops, typically on a perennial cycle. Pasture/hay vegetation accounts for greater than 20% of total vegetation. 82 Planted/Cultivated Cultivated Crops -areas used for the production of annual crops, such as corn, soybeans, vegetables, tobacco, and cotton, and also perennial woody crops such as orchards and vineyards. Crop vegetation accounts for greater than 20% of total vegetation. This class also includes all land being actively tilled. 90 Wetlands Woody Wetlands- areas where forest or shrubland vegetation accounts for greater than 20% of vegetative cover and the soil or substrate is periodically saturated with or covered with water. 95 Wetlands Emergent Herbaceous Wetlands- Areas where perennial herbaceous vegetation accounts for greater than 80% of vegetative cover and the soil or substrate is periodically saturated with or covered with water. Homer et al. 2015

Appendix 3: MaxEnt jackknife of variable importance of P. canaliculata base SDM

Appendix 4: MaxEnt marginal response curve of maximum temperature of the warmest month to habitat suitability for P. canaliculata base SDM

Appendix 5: MaxEnt marginal response curve of annual precipitation to habitat suitability for P. canaliculata base SDM

Appendix 6: MaxEnt jackknife of variable importance of P. diffusa base SDM

Appendix 7: MaxEnt marginal response curve of minimum temperature of the coldest month to habitat suitability for P. diffusa base SDM

Appendix 8: MaxEnt marginal response curve of annual precipitation to habitat suitability for P. diffusa base SDM

Appendix 9: MaxEnt jackknife of variable importance of P. maculata base SDM

Appendix 10: MaxEnt marginal response curve of maximum temperature of the warmest month to habitat suitability for P. maculata base SDM

Appendix 11: MaxEnt marginal response curve of minimum temperature of the coldest month to habitat suitability for P. maculata base SDM

Appendix 12: MaxEnt marginal response curve of precipitation of the driest quarter to habitat suitability for P. maculata base SDM

Appendix 13: MaxEnt marginal response curve of precipitation of the warmest quarter to habitat suitability for P. maculata base SDM

Appendix 14: Classification of Florida's ground cover types Value Class 1 Sand 2 Limestone 3 Delta 4 Alluvium 5 Sandstone 6 Dolostone (dolomite) 7 Beach sand 8 Claystone 9 Water 10 Clay or mud 11 Mixed clastic/carbonate 12 Calcarenite 13 Dune sand USGS2 2017

Appendix 15: MaxEnt jackknife of variable importance of P. canaliculata expanded SDM

Appendix 16: MaxEnt marginal response curve of maximum temperature of the warmest month to habitat suitability for P. canaliculata expanded SDM

Appendix 17: MaxEnt marginal response curve of pH to habitat suitability for P. canaliculata expanded SDM

Appendix 18: MaxEnt marginal response curve of land cover types to habitat suitability for P. canaliculata expanded SDM

*see Appendix 2 for clarification of land cover type codes

Appendix 19: MaxEnt marginal response curve of ground type to habitat suitability for P. canaliculata expanded SDM

*see Appendix 14 for clarification of land cover type codes

Appendix 20: MaxEnt jackknife of variable importance of P. diffusa expanded SDM

Appendix 21: MaxEnt marginal response curve of minimum temperature of the coldest month to habitat suitability for P. diffusa expanded SDM

Appendix 22: MaxEnt marginal response curve of annual precipitation to habitat suitability for P. diffusa expanded SDM

Appendix 23: MaxEnt marginal response curve of land cover types to habitat suitability for P. diffusa expanded SDM

*see Appendix 2 for clarification of land cover type codes

Appendix 24: MaxEnt marginal response curve of ground types to habitat suitability for P. diffusa expanded SDM

*see Appendix 14 for clarification of ground type codes

Appendix 25: MaxEnt jackknife of variable importance of P. maculata expanded SDM

Appendix 26: MaxEnt marginal response curve of maximum temperature of the warmest month to habitat suitability for P. maculata expanded SDM

Appendix 27: MaxEnt marginal response curve of precipitation of the driest quarter to habitat suitability for P. maculata expanded SDM

Appendix 28: MaxEnt marginal response curve of land cover types to habitat suitability for P. maculata expanded SDM

*see Appendix 2 for clarification of land cover type codes

Appendix 29: MaxEnt marginal response curve of ground types to habitat suitability for P. maculata expanded SDM

*see Appendix 14 for clarification of land cover type codes

Appendix 30: MaxEnt marginal response curve of pH to habitat suitability for P. maculata expanded SDM

Appendix 31: Watershed delineations for Florida

Shukia 2004