This item is only available as the following downloads:
1 ASPECTS OF THE ABUNDANCE, DENSITY, AND MOVEMENT OF APPLE SNAILS RELATIVE TO INVASIVE SUBMERGED AQUATIC PLANTS IN A CENTRAL FLORIDA LAKE By LARA KRISTEN DRIZD A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTI AL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011
2 2011 Lara Kristen Drizd
3 T o the non charismatic mesofauna
4 ACKNOWLEDGEMENTS First and foremost, I would like to thank my advisor, Dr. Wiley Kitchens. His guidance and support were absolutely essential to the development and successful completion of my thesis project. I would also like to thank the other members of my committee, Dr. Julien Martin and Dr. H. Franklin Percival, for offe ring their time and expertise, especially after being asked to participate so far along in the process. I would like to express particular gratitude to Brian Reichert for his help with the analyses. Although I have been lectured to by several professors r enowned for their statistical abilities while at the University of Florida, Brian was the one person who took the time to make sure I fully understood the concepts discussed herein. He was extremely patient with my numerous questions and is ultimately resp onsible for teaching me how to operate Program MARK and offering advice throughout the rest of the analysis process. I greatly appreciate all of the help I received in completing the fieldwork portion of my research, especially from Kristen Aaltonen, Nath an Marcy, Kimberly Schmidt, Thea Hotaling, Melissa DeSa, Jean Olbert, Kyle Pias, and everyone on the snail kite crew It has been my pleasure working with each and every person associated with the Florida Cooperative Fish and Wildlife Research Unit in Gain esville, Florida, and an experience I will never forget. Finally, I thank all of my friends and family m embers for all of their support and encouragement For my e thesis), I am enormously grateful.
5 TABLE OF CONTENTS p age ACKNOWLEDGEMENTS ................................ ................................ ................................ ............ 4 TABLE OF CONTENTS ................................ ................................ ................................ ................ 5 LIST OF TABLES ................................ ................................ ................................ .......................... 7 LIST OF FIGURES ................................ ................................ ................................ ........................ 8 ABSTRACT ................................ ................................ ................................ ................................ .. 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ............... 12 Apple Snails of Florida ................................ ................................ ................................ .. 14 The Hydrilla Invasion ................................ ................................ ................................ .... 18 Study Objectives ................................ ................................ ................................ ............ 20 2 METHODS ................................ ................................ ................................ .......................... 22 Trap Description ................................ ................................ ................................ ............ 22 Snail Handling Procedures ................................ ................................ ............................. 23 Grid Arrays and Study Sites ................................ ................................ .......................... 24 Data Analyses ................................ ................................ ................................ ................ 27 Estimating Apparent Survival ................................ ................................ .................. 27 Covariates and Model Selection ................................ ................................ .............. 28 Effective Trapping Area ................................ ................................ .......................... 29 Superpopulation Estimator and Density ................................ ................................ .. 30 Delineating Size Classes ................................ ................................ .......................... 32 3 RESULTS ................................ ................................ ................................ ............................ 39 Snail Movement and Effective Trapping Area ................................ .............................. 39 Analysis of Respons e to Hydrilla Treatment ................................ ................................ 40 Apparent Survival and Detection Probabilities ................................ ........................ 40 Snail Abundance ................................ ................................ ................................ ...... 41 Snail Density ................................ ................................ ................................ ............ 41 Size Classes ................................ ................................ ................................ .............. 42 Non Target Species ................................ ................................ ................................ ........ 43 4 DISCUSSION ................................ ................................ ................................ ...................... 59
6 Sampling Grid Efficacy ................................ ................................ ................................ 59 Extreme Movements Observed ................................ ................................ ...................... 61 The Effects of Herbicide Treatment ................................ ................................ .............. 62 Apparent Survival and Abundance ................................ ................................ .......... 62 Small Si ze Classes ................................ ................................ ................................ ... 64 Proportion of Native Snails to Exotics ................................ ................................ ........... 65 Non Target Fauna and Unintentional Baiting ................................ ................................ 65 Conclusion ................................ ................................ ................................ ..................... 66 LIST OF REFERENCES ................................ ................................ ................................ .............. 69 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ........ 80
7 LIST OF TABLES Table page 3 1 Estimates of A ( SE) and W ( SE) based on different methods of calculating snail movement. ................................ ................................ ................................ ......................... 44 3 2 Sn ail density ( D ) (snails/m 2 ) calculated using several different methods for estimating snail movement rates and effective sampling area. ................................ .......................... 45 3 3 Models representing hypotheses of effects of hydri lla treatment on apple snail distribution in Lake Jackson. Models also test for potential effects of individual and state covariates. ................................ ................................ ................................ ................. 45 3 4 List of all non target species trapped in funnel t raps between April and June, 2010 in Lake Jackson, Florida. ................................ ................................ ................................ ...... 47
8 LIST OF FIGURES Figure page 2 1 Wire funnel traps modified with chimneys and widened entry holes. .............................. 33 2 2 Diagram of shell measurements. ................................ ................................ ....................... 34 2 3 Tag placement on an apple snail shell. ................................ ................................ ............. 34 2 4 Trapping grid array, 7 by 7 traps encompassing 441 m 2 ................................ ................. 35 2 5 Stage one grid locations in Lake Tohope kaliga. ................................ ............................... 36 2 6 Stage one grid locations in East Lake Tohopekaliga. ................................ ....................... 37 2 7 Stage two grid locations in Lake Jackson, including areas that were treated with endothall. ................................ ................................ ................................ ........................... 38 3 1 Diagram of snail movements. Each color represents the path of an individual snail through time. Looped arrows mean that the individual was caught in the same trap on two consecutive occasions. Snails that were captured on more than three trapping occasions were relatively infrequent. The snails in this depiction were not necessarily fo und in the same grid or even in the same season. ................................ .......................... 44 3 2 Probability for detection ( p ) within control plots in Lake Jackson grids. These values were estimated by averaging most parsimonious CJS models. ................................ ........ 48 3 3 Probability for detection ( p ) within treatment plots in Lake Jackson grids. These values were estimated by averaging most parsimonious CJS models. ................................ ........ 48 3 4 Snail apparent survival ( ) of control plots in Lake Jackson grids. These values were estimated by averaging most parsimonious CJS models. ................................ ................. 49 3 5 Snail apparent survival ( ) of treated plots in Lake Jackson grids. These values were estimated by averaging most parsimonious CJS models. ................................ ................. 49 3 6 Snail apparent survival ( ) response to hydrilla treatment in Lake Jackson. Values based on results from model Phi (before, (immediate effect, delayed effect)* treatment) p ((before, after treatment )*grid). ................................ ................................ ..... 50 3 7 Apparent survival ( ) as a function of shell size. ................................ ............................. 51 3 8 Snail abundance ( N ) in response to hydrilla treatment in Lake Jackson grids. These values were estimated by averaging most parsimonious POPAN models. Each estimate applies to three grids, areas totalling app roximately 9000 m 2 ........................... 52
9 3 9 Snail density ( D ) in response to hydrilla treatment in Lake Jackson grids. These values were calculated using the effective sampling areas based on MMDM (see Table 3 1) and abundance estimates from model averaging top POPAN models. ........... 53 3 10 Proportion of snails smaller than 60 mm captured in treated versus control plots. .......... 54 3 11 The mean snail weight increased over time but was similar in treatment and control plots ................................ ................................ ................................ ................................ 55 3 12 The mean snail height also increased over time b ut was similar in treatment and control plots ................................ ................................ ................................ ...................... 55 3 13 Percentage of snails captured within each size class on Lake Jackson in the 2010 trapping season. Since 98.5% of native apple snails are no greater than 60 mm in height, snails measuring less than 60 mm should be easily consumed by all kites, but those snails larger than 60 mm are expected to be more difficult for juveniles to handle ................................ ................................ ................................ ............................... 56 3 14 Number of snails captured within each size class on Lake Jackson in the 2010 trapping season. Herbicide treatments took place between surveys 7 and 8 and the time period separating those two trapping occasions was roughly three weeks as opposed to three to four days as between each of the other surveys ............................... 57 3 15 Summary of snail sizes captured in control plots throughout the sampling seasons ....... 58 3 16 Summary of snail sizes captured in treated plots throughout the sampling seasons ....... 58
10 Abstract of Thesis Presented to the Graduate School Of the University of Florida i n Partial Fulfillment of the Requirements for the Degree of Master of Science ASPECTS OF THE ABUNDANCE, DENSITY, AND MOVEMENT OF APPLE SNAILS RELATIVE TO INVASIVE SUBMERGED AQUATIC PLANTS IN A CENTRAL FLORIDA LAKE By Lara Kristen Drizd August 2011 Chair: Wiley M. Kitchens Major: Wildlife Ecology and Conservation The apple snails of Florida both native ( Pomacea paludosa ) and exotic ( Pomacea insularum ), have proven to be important prey items for the endangered Everglade snail kite ( Rostrhamus sociabilis pl umbeus ) that relies on them nearly exclusively for food. Although native the availability of the exotics has offered a temporary source of sustenance for these otherwise extrem e dietary sp ecialists The majority of remaining kites are thought to inhabit one area, over 20 lakes that comprise the Kissim mee Chain of Lakes i n central Florida. Researchers have suggest ed that the adult kites may be attracted to this area because of t he abundance of large exotics which are known to utilize hydrilla ( Hydrilla verticillata ) an invasive submerged aquatic plant that dominates throughout many of the lakes. In this study, several areas of Lake Jackson were randomly selected and both species of snail were trapped as part of a mark recapture study to determine the effects of herbicide treatments on snail movement, abundance, and density. Average snail movement rates were used to determine the effective trapping area, and the superpopulation ap proach was used to estimate snail abundance within each area.
11 Over 8,000 apple snails were captured in Lake Jackson between April and June of 2010 Model ing r esults suggest that these snails did not colonize or emigrate from an area in response to the hydr illa treatments. Overall, snail abundance sharply increased over the course of the sampling period. Density increased to a greater degree in the control plots, from 2.04 to 5.86 snails/m 2 Initial estimates were slightly lower in treated plots but increase d from 1.84 to 4.97 snails/m 2 We concluded that attention to timing of sampling events used for estimating lake wide or even site abundance could be very critical to the overall reliability of the estimates. The superpopulation approach proved to be very useful and appropriate for this kind of sampling in which timing was an important element and is recommended for future studies. Because smaller snails were thought to be protected from predation by the dense patches of hydrilla, we expected to find them t o be smaller on average in the plots that were not treated However, we found no support for this hypothesis. When comparing the difference in the number of non target individuals trapped among treated and non treated plots, there was no indication that ve getative density prevented predatory fish from entering and thick mats of hydrilla may actually provide support structure for other avian predators
12 CHAPTER 1 INTRODUCTION Before human settlement and manipulation, Florida was a vast series of interconne cted wetlands, receiving and storing rainfall by flooding the extensive system of marshlands and shallow lakes. Endemic flora and fauna are adapted to the seasonal inundations and fluctuating habitats and have grown to depend on the large quantities of wat er during a portion of the year Unfortunately for many species, a nthropogenic alterations to wetlands in peninsular Florida have caused the degradation of habitat quality (Acosta and Perry 2001, Lorenz 2000) The establishment of water control structures caused innumerable unintended consequences including increased eutrophy and lake succession, diminished water quality, decreased plant diversity, dense vegetation growth and loss of habitat for wildlife Government initiated projects such as the Kissimmee River Restoration Project and the Comprehensive Everglades Restoration Plan have been in progress for almost 20 years and researchers have been committed to monitoring changes in wildlife populations in response to restoration efforts (NAS 2003). The Flor ida apple snail ( Pomacea paludosa ) has been identified as one indicator of potential restoration success in the Everglades (Darby et al. 1999 Kitchens et al. 2002 ). It is an important prey item for numerous wetland species including alligators, redear su nfish, and limpkins (Turner and Mikkelsen 2004). However, interest in the species is largely due to the the endangered snail kite of Florida ( Rostrhamus sociabilis plumbeus ) that relies on them nearly exclusively for food ( Howell 1932, Cottam and Knappen 1939, Stieglitz and Thompson 1967 ). The snail kite is a wetland dependent raptor that inhabits a fragmented series of freshwater wetlands and shallow lakes scattered throughout central and southern Florida (Bennetts and Ki tchens 1997). The historical range of the kite once included over 4000 km 2 that included a portion of the panhandle (Davis and Ogden 1994, Sykes et al.
13 1995), but they are now restricted to small areas within the Everglades, Lake Okeechobee, Loxahatchee Sl ough, the Kissimmee River, and the Upper St. Johns River (Reichert et al. 2011). Their population has been reduced from an estimated 3400 individuals in 1999 to roughly 826 in 2010 (Reichert et al. 2011). Many potential causes have been implicated in this decline, but each is related back to one of three important habitat features: distribution of emergen t vegetation, nesting substrate, or apple snail presence. Researchers believe that snail density estimates are an essential component of understanding kite demography and habitat use. They calculate that a snail density of 0.14 individuals per m 2 is the minimum needed to support foraging snail kites (Darby et al. 2006). Observations suggest that snail density is correlated with foraging success rates, but a lack of kite presence does not necessarily imply a lack of snails (Darby et al. 2006). Kites are nomadic creatures that move frequently in search of a more abundant and available food supply (Bennetts and Kitchens 2000). Collins and Rawlings have noticed an apparent decline or even disappearance of native apple snails in some Florida wetlands (Rawlings et al. 2007). DeSa (2008) however, noted an increase in snail occupancy after restoration efforts on Lake Tohopekaliga (from here on referred to as Lake T oho ). Ap ple snails are very susceptible to changes in habitat such as extended dry downs (Kushlan 1975, Turner and McCaffree 1994) exposure to agricultural chemicals (Hoang and Rand 2009, Rogevich et al. 2009) and the removal of invasive plants (C orrao e t al. 2006). Because the snail kite population is tightly linked to apple snail abundance and availability to kites, factors that negatively affect the snail population could increase the risk of extinction of the snail kite
14 Apple Snails of Florida The Fl orida apple snail is the largest freshwater snail native to North America (Pennak 1989) and can be found in lakes, rivers, and swamps in Florida, Georgia (Thompson 1984), and Alabama (Hubricht 1962). It s distribution throughout this area makes it the north ernmost species in the Pomacea genus while the other members are found naturally distributed in South and Central America and parts of the Caribbean (Haldeman 1845). Unnaturally, howeve r, many of these other species have escaped and become established in parts of Southeast Asia, Sri Lanka, Guam, Papua New Guinea, the Dominican Republic, Hawaii, and the mainland U.S. (Cowie 2002, Lai et al. 2004, Joshi and Sebastian 2006). The Florida apple snail is the only member of the apple snail family (Ampullariidae) that is native to th is continent (Burch 1982), but at least three other species of Pomacea have become established in Florida: P. diffusa (Thompson 1984, Rawlings et al. 2007), P. haustrum (Winner 1991), and P. insularum (Rawlings et al. 2007). For the pur poses of this study, I will be focusing on P. paludosa and P. insularum as both are important food sources for the Florida population of snail kites. The Ampullariidae family refers to those freshwater snails that are capable of aestivation. They are dist inguished from other families by the presence of both a ctenidium (gill like structure) and a lung for respiration (Little 1983). Both species, P. paludosa and P. insularum are found in a wide range of wetland habitats including graminoid marshes, sloughs and lake littoral zones, and the dual respiratory system is a necessary adaptation for surviving seasonal fluctuations in water depth as well as dissolved oxygen content (McClary 1964). The lung is also used when the snails leave the water to deposit e gg masses on emergent vegetation (Winner 1989) and other non living strata (Wallace et al. 1956). Similarities end there, however, as the exotic P. insularum is significantly larger. P. paludosa adults reach approximately 40 70 mm in
15 length (Thompson 1984) and the exotics grow about three times longer and four to five times the mass (Darby et al. 2007). Additionally, the exotic lays clutches consisting of approximately 1,600 to 2000 eggs (Boland et al. 2007, Barnes et al. 2008) while the native typically la ys an average of 30 eggs per mass (Hanning 1979). Since P. insularum are able to lay eggs weekly throughout the breeding season, Barnes et al. (2007) estimates that a single female, even with a conservative 1% hatchling success rate, could produce 252 via ble hatchlings each reproductive season. The lifespan of P. insularum is unknown but it is thought to be similar to that of P. canaliculata which live 3 to 6 years (Estebenet and Martin 2002). P. paludosa live a year to a year and a half (Hanning 1979, Fer rer et al. 1990) and generally die post reproduction in May and June (Darby et al. 2003). Native to South America, the exotic P. insularum snails (known as island apple snails) were introduced to Florida possibly as early as 1991 (Winner 1991) or earlier ( Thompson 1997) through the aquarium and ornamental trade ( Britton 1991, Ho wells 200 2 ), industr ies which are most destructive aquatic invasive species (Padilla and Williams 2004). In the mid 1900s, Florida began development of its fish farm culture, producing tropical aquarium fishes, mollusks, and plants easily in the warm subtropical climate ( Howells et al. 200 6 ). During this same time period, aquarium trade experimentation with the various as vivaparids which are a related family of operculate freshwater snails that bear live young inste ad several of which have also been found released in wetlands within the state (USGS 2009). Gastropods represent 4% of the exotic aquatic species that have been introduced to Florida making it the fourth largest group after fish, plants, and turtles (USGS 2011).
16 Until genetic analyses were conducted, P. insularum snails were incorrectly referred to as channeled apple snails, P. canaliculata (Rawlings et al. 2007). We now know that although the two snail s have similar shell morphologies, they should be con sidered separate species and they are lumped into a group of monophyletic sisters with channeled sutures called the Pomacea canaliculata complex (Howells et al. 2006). In other countries where members of this group have been introduced especially within southern and eastern Asia, ma jor agricultural losses have resulted (Anderson 1993, Naylor 1996, Cowie 2002) but researchers have found it much more difficult to estimate their ecological impacts (Hayes et al. 2008). In wetlands in Thailand, Carlsson et al (2004) found that the presence of more than two P. canaliculata per m 2 resulted in a decrease in macrophytes and vegetative diversity and an increase in nutrients and algal blooms Wood et al. (2005) discovered that these snails graze selective ly on freshwater bryozoans in Southeast Asia and they have caused the bryozoans to become so rare that they are now only able to exist o n float ing structures that do no t come in contact with the shore or bottom sediment. Halwart (1994) blamed them for the decline of native ampullariids in the Philippines It has also been suggested that they be introduced as a biocontrol agent against invasive planorbid snails due to their effectiveness at consuming eggs and young (Cazzaniga 1990) However, i t has b een well established that their voracious appetites are troublesome at best. They were listed among most invasive species in 2000 (Lowe et al. 2000). Within the United States, P. insularum has also been found established in Texa s and Georgia and th e number of invaded states may increase as more people are reportedly seeing snails with channeled sutures each year particularly in Louisiana, Mississippi, Alabama, Georgia, and South Carolina (Rawlings et al. 2007 USGS 2009 ). The potential impacts of t hese snails are numerous. It has been suggested that further spread of P. insularum may facilitate the
17 spread of vertebrate parasites like Parastongylus spp. which can cause eosinophilic meningoencephalitis and abdominal angiostrongyliasis, both treatable diseases that infect humans (Hollingsworth and Cowie 2006) and are known to be present in Florida already (Duffy et al. 2004, Miller et al. 2006). Burlakova et al. (2009) suspects that P. insularum can impact wetland restoration efforts because of specific feeding preferences for plants such as Hymenocallis and Sagitaria spp Others wonder if the decline in abundance of P. paludosa individuals following P. insularum introduction is a direct result of egg consumption by the exotics (Rawlings et al. 2007) as occurred when P. bridgesi was introduced to a congeneric species ( Aditya and Raut 2001, 2002). Rawlings et al. (2007) even hypothesize extinction of P. paludosa by hybridization since the two snails are so closely related phylogenetically. Perhaps the most disturbing effect is the potential for this exotic to cause the extinction of the already endangered snail kite. Despite the substantial difference in size, adult snail kites successfully forage for and consume both species of snail (Darby et al. 2007, Ca ttau et al. 2009). This has been fortunate for (Darby et al. 2005 Rawlings et al. 2007 ) and the availability of the exotics has offered a temporary source o f sustenance for these otherwise extreme dietary specialists (Bennetts et al. 2000). Unfortunately, juvenile kites struggle to handle the larger snails and waste a high amount of energy on snails they are never able to consume (Cattau et al. 2009). Despit e this fact, the majority of remaining kites are thought to inhabit one area, over 20 lakes that comprise the Kissimmee Chain of Lakes area (KCOL) in central Florida (Cattau et al. 2009) where the exotic snails proliferate (DeSa 2008). In 2010, 51% of suc cessful nesting attempts range wide occurred in the KCOL, two thirds of those on Lake Toho alone (Reichert et al. 2011). This dependence on
18 Lake Toho is disconcerting. Juvenile kite s that fledge in the KCOL will be forced to feed on the exotic snails and t heir inefficiency in doing so may negatively impact their survival (Cattau et al. 2010). Researchers suggest that the adult kites may be attracted to this area because of the abundance of large exotics which are known to utilize hydrilla ( Hydrilla verticil lata ) (Reichert et al. 2011), an aquatic invasive plant that dominates throughout many of the lakes (Howells 2002, Gettys et al. 2008, Baker et al. 2010) The Hydrilla Invasion Native to the warmer regions of Asia (Cook and Lnd 1982), hydrilla has been r ecognized as a major submersed aquatic weed in Florida since 1967 (Mackenzie and Hall 1967). It was planted by aquatic plant dealers in the southeast in 1959 (Blackburn et al. 1969) and is now found in at least seventy percent of Florida's freshwater drain age basins, including the KCOL (FLDEP 1988 94). Hydrilla is an extremely competitive colonizer of aquatic habitats and is known to displace important native aquatic plants such as eelgrass ( Vallisneria americana ) once established (Langeland 1996). It can grow to such high densities that it reduces flow in drainage canals and around water control structures (Blackburn et al. 1969, Langeland 1996). Additionally, fishermen and other recreational boaters find that hydrilla can greatly limit the areas where the y are able to navigate (Langeland 1996). It is for these reasons that 174 million dollars were spent to control the noxious weed in Florida from 1980 to 2005 and it continues to be a pressing issue (Schmitz 2007). Corrao et al. (2006) discovered that her bicide applications to water lettuce in Wacissa Springs resulted in drowned snail eggs, but hydrilla treatments in the KCOL are considered to be fairly harmless when conducted properly. Two different herbicides are currently used to control the hydrilla in Osceola County: endothall and imazamox (Interagency Aquatic Plant Management 201 1 ).
19 Endothall is the preferred herbicide and is sprayed as a liquid or distributed in pellet form to selectively target and eliminate hydrilla without damaging other aquatic p lant species Imazamox is sometimes added to the entire body of water after the endothall treatment to limit excessive regrowth and produce longer lasting results Within the KCOL, approximately 5700 acres were planned for treatment by one or both herbicid es in 2010 (Interagency Aquatic Plant Management 201 1 and Zach Welch, per s. comm ). This is a decrease from over 8000 acres that were treated in 2009 (Interagen cy Aquatic Plant Management 2011) but Lake Toho was responsible for the bulk of that decrease because the hydrilla there did not expand to the same extent as in previous years (Ed Harris, pers comm ). Little is known about the nutritional requirements and feeding behavior of apple snails, but both species are known to consume a variety of macrophytes ( Howells et al. 2006, Baker et al. 2010) and macrophyte assemblages (Sharfstein and Steinman 2001). The exotic snails have been found to flourish within submerged hydrilla, feeding on the leaves and stems (Howells et al. 2006; Bak er et al. 2010), whereas the natives appear to simply graze the epiphytic complexes attached to the surface of the plant eating very lit t le of the actual hydrilla ( pers obs ). Aquarium observations suggest that the two species are equally attra cted to hydrilla especially when other food sources are limited suggesting that it is likely that they would be attracted to specific areas where hydrilla is present in natural systems as well Cattau et al. (Appendix D and E, 2009) observed kites success fully foraging in areas of dense hydrilla and suggested that they may be able to capture smaller size classes in those areas. Because of its ability to serve as a refuge from aquatic predators, thick mats of hydrilla might be able to concentrate higher den sity populations of younger exotic snails that would be more suitable for juvenile kites to handle. Therefore, by altering an important food source and climbing structure required for breathing, herbicide
20 treatments could profoundly affect snail distribut ions, potentially decreasing their availability to the kites. This study address es these issues and it is hoped that an improved management strategy for invasive plants in the KCOL and the apple snail will be produced as a result. Study Objectives This stu dy was intended to answer the following research questions and hypotheses. 1.) What is the efficacy of deploying grid arrays of snail traps to determine the snail presence, occupancy, and densities in deeper water habitats typically or routinely dominated by h ydrilla in a reas of historical kite nesting? Hypothesis 1: Snail densities will be lower in the deeper lacustrine habitats than in shallo w wetlands sampled by Valentine Darby et al. ( 200 8 ) ) Telemetry research suggests that snails tend to move toward shall ower areas when they reach depths greater than 50 cm (Darby et al. 2002). Biologists who have caught apple snails in the past have limited their sampling to water that is no deeper than 60 cm. 2.) What is the effect of herbicide treatment on apparent snail d ensities in the vicinity of historically high use nesting sites? Hypothesis 2 : Apparent s nail survival (the probability that the snail is alive and remains on the study area, referred to as phi or ) will decrease in treatment plots after herbicide treatments have been completed. Since we believe that the snails are attracted to the hydrilla as a food source, removing it should cause them to disperse to the surrounding area. Hypothesis 3 : Grids with hydrilla will yield smaller than average size classes of exotic snails. Because smaller snails should be provided refugia by dense hydrilla, they are
21 expected to be smaller on aver age due to decreased predation. In areas without hydrilla, only those large r snails that are no longer vulnerable to so many aquatic predators will be present in high abundance. Once the treatment grids are treated, we expect to the average snail size to increase. 3.) Can the superpopulation approach to estimating abundance be used to det ermine density for species such as aquatic snails? This method has been discussed as an option in other trapping grid studies (Conroy et al. 2008), but, to my knowledge, has not been attempted. 4.) What are the differential distributions of native and exotic apple snails in these locations? Hypothesis 4 : The number of exotics will greatly e xceed that of the native snails. Although both species are likely to be found among the hydrilla r esearchers have been observing the expansion of the exotic snails in the area for at least ten years (Rawlings et al 2007, DeSa 2008, Cattau et al. 2010). 5.) What additional faunal species are attracted to the funnel traps and are these species associated with snail captures? Will they be affected by the elimination of the hydril la? Hypothesis 5 : Snails will not be attracted to the traps due the presence of non target species and detection rates will not be influenced as a result. In a pilot study conducted by Darby et al. (2001), similar funnel traps were baited with and without dead fish and snail capture rates were unaffected.
22 CHAPTER 2 METHODS Trap Description Wire funnel traps like those used in Darby et al (2001) and Cattau et al. (2010) were modified and used to capture the apple snails Such traps are sold commercially f or catching crayfish They have also been used in numerous field studies for capturing other aquatic fauna such as sirens, amphiumas, and small turtles (Johnson and Barichivich 2004) The funnel traps are made from green vinyl coated hardware cloth with " by " square mesh and have three funnels at the base where individuals can enter but are unable to exit (Fig. 2 1) In order to catch the much larger invasive snails in a lacustrine habitat, two modifications were made to each trap: (1) The tapered ends of th e funnels were cut and widened from 5 cm to approximately 8 cm in diameter. P. insularum can sometimes exceed 9 cm in height (Darby et al. 2007), but in an effort to decrease the likelihood for smaller snails to escape we chose to exclude the very largest snails. It is not known what portion of the population exceed 9 cm, but judging from empty shells that can be found along the shoreline of lakes in the KCOL, the number of these snails is thought be negligible Due to the design of the trap, we do not exp ect to see higher escape rates than previously experienced with smaller holes. (2) coated wire extensions, were attached to existing traps in order to allow the snails and other non target species access to the water surface f or aerial respiration. Modified tr aps stand approximately 1.5 m, capable of sampling more than two times the depth of other sampling methods. Traps will not be baited, but trapped females may attract additional males (Darby 1998). Snails may inadvertently be attracted to bycatch as well as some species (Florida and Longnose gar in
23 particular) occasionally become injured and die while trapped. However, Darby et al. (2001) found that dead fish did not help to attract snails. Snail Handling Procedures Each app le snail obtained from a trap was identified as native or exotic and then measured and tagged. Four measurements were collected from each individual every time it was captured. Wet weight was measured using one of a series of spring scales (all Pescola br and ) Because of the drastic difference in snail sizes and our desire to achieve the most accurate reading, maximum capacities ranged from 10 g for the very smallest snails up to 2 00 g for the largest. Shells were also measured for total height, aperture h eight, and width using vernier calipers Total height was measured linearly along the spiral axis from the apex to the outer edge of the aperture lip while aperture height was measured from the top of the aperture to the same outer edge of the aperture (Fi g 2 2) Width was measured perpendicular to the axis from the edge of the aperture to the opposite side of the shell body. Upon initial capture, we individually mark ed each snail with a laminated 3 mm by 5 mm plastic tag (Floy Tag & Manufacturing, Inc., Se attle, WA) applied with marine epoxy glue (Evercoat) that is formulated to resist damage from exposure to water, sun, chemicals, and fuels and sets in less than 10 minutes. Tags were attached to the shell approximately 15 mm up from the aperture on the ri ght side of the body whorl adjacent to the suture (Fig. 2 3 ) Prior to attaching each tag, snail shells were dried and then lightly sanded to ensure a tight bond. After allowing epoxy adequate drying time, snails were measured and released within a meter o f the trap where discovered Recaptures were noted, measured, and rereleased as well.
24 Grid Arrays and Study Sites Valentine Darby et al. (2008) deployed and tested trap grid arrays to estimate snail densities in the Everglades. For this study, we used a similar capture technique. Reproductive behaviors and egg production peak between April and May in South Florida wetlands (Darby et al. 1999). Cooler temperatures in central Florida cause this pe ak to extend into June (Stevens et al. 2002) For that reas on, a series of similar grid arrays w ere used to determine snail presence, abundance and densities between April and June of 2009 and 2010 The objectives were addressed in two stages. Stage 1: Grids for the first stage include d 64 funnel traps spaced 3. 5 m apart (8 by 8 square) covering an area of 600 m 2 Traps were held in place with PVC poles which were pounded into the substrate. Vege tation was pushed aside to allow traps to come in contact with the lake bottom but none was removed. Traps were checked and emptied every three to four days, the period used by Valentine Darby et al. (2008), to ensure adequate capture rates for statistical analysis. As Otis et al. (1978) recommend s that sampling take place no fewer than five separate occasions, grids were checked a minimum of six times. Trapping took place in March through June 2009. The first stage was a preliminary test that allowed us to determine the effectiveness of using funnel traps to sample for apple snails in deeper water (greater than 60 cm) and took place in Lake Toho (Fig. 2 5 ) and East Lake Toho (Fig. 2 6 ). Lake Toho is frequently foraged by snail kites and known to support a high number of invasive apple snails (Cattau et al. 2009) while containing dense mats of hydrilla ( pers. obs .). East La ke Toho is one of the few lakes in the KCOL to maintain a healthy population of native apple snails despite invasion by the larger exotic snails and allowed us to determine whether the modified traps we re capable of catching smaller snails. This stage al so test ed the efficiency of using the grid array formation as opposed
25 to other arrangements such as transects (DeSa 2008, Muench 2004) or randomly scattering traps throughout the study area (Cattau 2008). Stage 2: The second stage took place at six sites r andomly selected from the littoral zone of Lake Jackson (Fig. 2 7) in May and June 2010 Lake Jackson is a much smaller lake (413 ha) just east of Lake Kissimmee that is frequented by a relatively high number of snail kites and has also been invaded by hyd rilla. Just prior to sampling in Lake Jackson, kite surveys revealed that the number of kites on the lake had increased from zero in March to 14 individuals in April 20 10 1 From 1992 to 1997, Lake Jackson underwent a large scale habitat enhancement project that involved three extreme drawdowns and several phases of muck removal and mechanical extraction of vegetation mostly cattails (Hulon et al. 1999). Lake Toho went through a similar process in 2004 and it resulted in the exotic snails spreading from on e small cove (Cattau 2010) to 83% of the entire lake (DeSa 2008) Fish, birds, mammals, reptiles, and amphibians were sampled and surveyed for before and after the Lake Jackson Habitat Enhancement Project (Hulon et al. 1999). 94 species were discovered in all but mollusks were entirely forgotten so we have no record of changes to the apple snail population s or species present However, no snail kites were found on the lake during that time period s uggesting that the snails may have only been present at ver y low densities. Sampling in Lake Jackson required slight changes to the sampling protocol established the year before. 7 by 7 square grids were used rather than 8 by 8 in order to allow us to set up a greater number of sampling grids with the limited num ber of traps (300), encompassing areas 441 m 2 instead (Fig. 2 4 ). We continued to check and empty the traps twice a week for two 1 Th ese numbe rs are based purely on field observations and have not been corrected for detection probability.
26 months. After four weeks of trapping, three of the six sites were chosen at random to be treated with endothall, applied as aqu atic herbicide Aquathol. Two weeks after the application t rapping resumed for an additional four weeks in all six of the grid arrays after the hydrilla had dropped All non target species were identified and counted during snail sampling as well. A water depth was collected from each trap point. Water temperatures were taken twice a week throughout the sampling period. Three temperatures were recorded from 0.5 m below the surface of the water in random locations throughout each of the grids using a waterp roof digital thermometer with a plumb bob. The temperatures were then averaged. Vegetation sampling was completed before and after the herbicide treatment to determine the extent of changes in the density of hydrilla present in all of the grids Twelve 0.2 5 m 2 samples were taken from each of the grids. We chose to limit the sampling size to decrease the potential effects of our plant sampling on apple snail occupancy. Sample locations were stratified to three samples per quarter of each grid and randomly se lected. Vegetation was cut at the ground level for those species that were rooted mostly belowground such as E g yptian panicgrass ( Paspalidium geminatum ). These species were included based on their root locations as opposed to where their stems protruded f rom the water surface. Floating and mat forming species like waterlettuce ( Pistia stratiotes ) and alligatorweed ( Alternanthera philoxeroides ) were included if they were found at the surface within the quadrat but their roots were removed prior to being w eighed. Submerged species like hydrilla and Utricularia spp. were cut and collected from the entire water column within the square. All s pecies were identified in the field and weighed wet. For species like Eqyptian panicgrass or alligatorweed with adequa te structure ( unlike hydrilla), stems were also counted.
27 Data Analyse s Estimating Apparent Survival The statistical modeling program Program MARK ( Cooch and White 2010 ), was used to analyze the mark recapture dat a for snail detection rates and survival Each time the grids were checked and emptied even though it took two full days to process all six of the grids is referred to as a single survey. The before and after treatment periods were considered two separate seasons. To determine survival Program MARK uses a list of encounter histories for each individual where each number in the series of values corresponds with a survey. I f a particular snail w as discovered during a survey, then a 1 was recorded. If not, then 0. Had we been unable to complete a survey or the data was otherwise lost, a dash would have been recorded. For example, an encounter history for a snail on Lake Jackson on which we completed 14 surveys (7 before treatment and 7 after) would look something like this: 0000100 0011000. T his would indicate that the snail was first captured in the first season during the fifth survey and then recaptured on the third and fourth surveys of the second season. Encounter histories for all individuals that were given a unique tag identification n umber were entered into Microsoft Excel and then transported into Program MARK as a n .inp file. Extensions of the Cormack Jolly Seber (CJS) model (Cormack 1964, Jolly 1965, Seber 1965) that are time specific and control for treatment were used to estimate apparent survival ( ) and detection rates ( p ). Estimating detection the probability a snail was present but not detected is an extremely important step in the process. By ignoring detection probabilities, we would be unable to produce accurate estimates for any of the ot her variables. Apparent survival from the population. Apparent survival is an important factor because it allows us to test for
28 demographic closure within the grids. A closed population is one that is free of unknown changes in abundance, particularly births, deaths, immigration, and emigration. If a population can be proven to be closed, then it is possible to account for known changes and estimate occupan cy. In planning this study, a variety of options were discussed but we decided to not attempt to achieve within grid population closure physically because it was deemed co ntrary to our goal of observing immigration or emigration in response to herbicide tr eatments. Since the results of the CJS model showed that the apparent survival wa s not equal to 1 (meaning that the assumption of closure had been violate d) an occupancy model could not be completed. Covariates and Model Selection Variables in sampling or in the environment that might potentially influence detection probabilities colonization, and extinction rates are called covariates. The CJS model discussed above was used to test for potential effects of covariates. Site covariates that were considered are grid number (location on the lake), trap location within the grid, and water depth at each trap. Sample covariates included water temperature and dead fish presence. Continuous covariates that can vary widely like water depth and temperature were stan dardized to perform more efficiently in Program MA RK The categorical covariates were coded with ones and zeros so that each had a unique sequence. Covariates were included in the models to determine if apparent survival and detection were significantly af fected by these other variables. Models were chosen based on the Akaike Information Criterion (AIC) method. This method determines the likelihood of each model and applies penalties for including too many parameters and covariates (MacKenzie et al. 2006). Simple yet parsimonious a re the ultimate goals for model selection. T he difference s between resulting AIC values ( ) were examined. Generally, those s that fall within 2 of the top model (that which has the lowest AIC ) are
29 considered and the one tha t makes the most biological sense is chos en (Burnham and Anderson 2002). However, it is not always necessary to make this choice. Model averaging takes the top competing models and weights them according to their AIC values. This method was used to calcula te estimates of abundance described in greater detail below. Effective Trapping Area The nave density estimator, where density ( D ) is a function of abundance ( N ) divided by the area ( A ) of the trapping grid, is recognized as a biased method of calculatin g density because Edge effect can be caused by two potential sources: (1) animals outside of the grid being attracted to the grid and (2) by animals near the edge whose h ome ranges include a portion of the boundary of the grid thereby increasing the trapping area (Dice 1938, Tanaka 1972). Darby (1998) found that the idea of a well defined home range does not apply to apple snails but since we are unsure of what does motiv ate the snails to climb inside the traps, the first source could be an issue. Animal density is now estimated usi ng the following equation (1) : where W represents the width of an additional boundary strip (Wilson and Anderson 1985). T o estimate W Valentine Darby et al. (2008) attempt to us e the mean distance moved (MDM) by snails between trap occasions. This was contrary to the method developed by Wilson and Anderson (1985) in which half of the maximum mean distance moved (MMDM) wa s used instead. Valentine Darby et al. (2008) were concerned that the MMDM estimate would be inflated because they had noticed a small portion of snails traveling extraordinary distances in a separate study However, they compared the two methods and reported that t he adjustment resulted in only small differences in density estimations. Since ther e was no way on knowing
30 true snail density within their grids, they recommend reporting both types of adjusted density estimates. In this study, t he MMDM ( SE ) was calculat ed based on only those snails that were captured on t wo or more occasions but only the distance between the site of the initial capture and the trap located the greatest distance away were used to calculate the mean. Each measurement was divided by the to tal number of days from first to last encounter. The mean net distance moved (MNDM) was also included for comparison. This refers to the distance between the trap where the individual was first encountered and the trap where it was last encountered. Supe r p opulation Estimator and Density The superpopulation approach (Crosbie and Manly 1985, Schwartz and Arnason 1996) was used to estimate snail abundance within each of the grids, both before and after herbicide treatments. This is a robust design method which allows for movement between surveys and has been applied to snail kites in previous studies ( see Dreitz et al. 2002 Martin et al. 2007). A et al. 2002) A b undance is the total number of snails available for capture at any time during the study with B i the number of new ly recruited individuals in the population at period i +1 that were not present at time i as in the following equation (2) : individuals captured, i.e. B 0 = N 1 A multinomial distribution is used to model the random variables, B i in which individuals from the superpopulation are assumed to enter the sampled population at different times according to entry parameters i In other words, recruitment ( B 0 B K 1 ) over the course of the season is distributed as a multinomial with parameters ( N 0 K 1 ).
31 The superpopulation model makes t wo assumptions (Williams et al. 2002) First, there must be h omogeneity among probabilities of entry. In other words, all members of the superpopulation that have been unavailable for capture during one sampling occasion a t time i are assumed to have the same probability of capture as other snails at time i + 1 It is also assumed that survival as well as entry of additional individuals into the sampling area population. These assumptions can be tested using a goodness of fit test for homogeneity of survival, detection, and lack of independence between survival and captur e events (Williams et al. 2002) Snail encounter histories were entered i nto Program MARK us ing the POPAN model which emulates the Jolly Seber model to estimate apparent survival ( ) and probability of capture ( p ), but then provides additional estimates for the probability of new individuals entering the population for each occasion ( pent ) and superpopulation size ( N ) ( Schwartz and Arnason 1996). While it is capable of calculating e stimates that the CJS model cannot, the POPAN model is incapable of utilizing individual covariates. Because it includes individuals that were never captured, it does not have covariate measurements for each of the animals and therefore cannot model popula tion size as a function of those covariates. Several different link functions were necessary to estimate the parameters using the POPAN model. Link functions are used to take the beta values provided by the model and transform them into real parameter esti mates. Logit was used to estimate and p which means that the parameter probabilities were constrained to fall between 0 and 1. M logit links were used for pent so that the sum of the pent s from each group (before and after hydrilla treatment within contro l and treated plots) would total 1. To constrain estimates of N to be positive, the l og link function was used. The most parsim onious POPAN models were averaged to determine values of
32 N for treated and control plots which were then used to estimate snail d ensities with equation (1) discussed earlier. The effect size of the abundance estimates was also determined to indicate the strength of the relationship between the variables tested. Delineating Size Classes Since juvenile kites struggle to handle the lar ger exotic snails, size classes were considered to be a very important factor of this study. Snails were grouped into five different size which measured less than 30 mm and would most likely be ignored by foraging snail kites (Sykes 1987) but could potentially be even more susceptible to aquatic predators. Snails that fell size class in which both species of snail, native and exotic, over lap. Sixty mm was used as a cutoff for juvenile kite accessibility. Since 98.5% of native apple snails selected by kites are no greater than 60 mm in height (Sykes 1987) snails measuring less than 60 mm should be easily consumed by all kites, but those s nails larger than 60 mm are expected to be more difficult for juveniles to handle. Snails measuring between 61 and 75 mm The largest snail ever documented to be extracted by an adult snail kite was 75 mm (Kyle Pias, pers c omm ). For this reason, those snails that measured inaccessible to all kites.
33 Figure 2 1 Wire funnel traps modified with chimneys and widened entry holes.
34 Figure 2 2 Diagram of shell measurements. Figure 2 3 Tag placement on an apple snail shell.
35 Figure 2 4 Trapping grid array, 7 by 7 traps encompassing 441 m 2
36 Figure 2 5. Stage one grid locations in Lake Tohopekaliga. Figure 2 5 Stage one grid locations in Lake Tohopekaliga.
37 Figure 2 6 Stage one grid locations in East Lake Tohopekaliga.
38 Figure 2 7 Stage two grid locations in Lake Jackson, including areas that were treated with endothall.
39 CHAPTER 3 RESULTS Snail Movement and Effective Trapping Area Apple snails were observed moving between 1. 45 and 1.63 meters per day. Movement s appeared random and varied widely (Fig. 3 1). Two individuals actually switched grids over the course of the sampling period. Both were recaptured about a month after th eir initial capture. One was found in Grid #6 after being caught in Grid #1 originally a distance of 0.84 k m The other was found in Grid #5 after being caught in Grid #4 a distance of 0.61 km Excluding th ose two outliers, t he maximum speed observed for an ind ividual was 7.47 m/ day. This was consistent with telemetry data published by Darby et al. (2002) in which they observed that several snails travelled approximately 50 m in a week. The values for the width of the boundary strip ( W ) and the total area sampled ( A ) which resulted from the various methods of determining snail movement rates did not differ significantly (Table 3 1). This was expected based on results found in Valentine Darby et al. (2008) Because the majority of marked snail s that were rec aptured were caught only one additional time, many of the movement rates that went into calculating the means were the same for each method. By excluding those snails which were only recaptured once, the MMDM would have d ecrease d to 1.07 m/day instead. T he nave trapping area which represents only the space covered by the grid array is 441 m 2 Because the calculations for MDM, MMDM, and MNDM indicated that snails could potentially travel 3 3 to 3 7 m during the sampling period, the effective trapping area w as estimated to be 3091 to 3286 m 2 (Table 3 1).
40 Analysis of Response to Hydrilla Treatment Apparent Survival and Detection Probabilities Models that ignored treatment effects but constrained detection probabilities by season (before or after hydrilla t reatment) and grid number (location on the lake) outperformed those which held detection constant or were constrained only by season and treatment ( Table 3 3 ). Model selection based on AIC value comparisons suggested that grid specific characteristics (such as vegetation density or wate r depth) may explain more variation in detection probabilities than hydrilla presence or absence. Detection probabilities were relatively low in general, ranging from 3 to 14 % (Figs. 3 2 and 3 3) Detection in two of the three grids in which the hydrilla w as not treated appears to have decreased over the course of the sampling period (Fig. 3 2) whereas detection in all of the treated grids showed a slight increase over time (Fig. 3 3). When modeling to best estimate apparent survival, model selection sugge sted similar preferences as for detection probabilities. Analyzing each grid separately always resulted in more parsimonious models than grouping according to treatment (see models #8 and #12) Apparent survival rates were estimated to range between 83 and 98 % (Figs. 3 4 and 3 5). Since these estimates fell short of 100 % indicating that the assumption of population closure had been violated snail occupancy could not be modeled. The results of these analyses indicated that apparent snail survival decrease d to a greater extent in treatment plots after herbicide treatments had been completed (Fig. 3 6). However, the confidence intervals of these estimates overlapped to some degree and the results cannot be considered signif ic ant. The most parsimonious model t hat had the lowest AIC value included snail shell height as an additive individual covariate affecting apparent survival rates (Table 3 3 ) This model received 41 % of the AICc weight versus 16 % model weight which was given to the interactive
41 snail size mod el (model #3) Models which incorporated shell size represented 93% of the total model weight. The top model was 8.50 AIC points lower than the model which ignored grid and time factors (model #8) A difference of 4 points or more is considered significan t (Burnham and Anderson 2002) Analyzing the beta parameter s revealed a significantly negative additive relationship between apparent survival and snail size ( SEASONxGRID+SIZE = 0.00 70, 95% CI = 0.0127 to 0.0012). This indicates that apparent sur vival decreased as size increased (Fig. 3 7) contrary to what might be expected The smallest snails (14 mm) had a 5.79% greater chance of Snail location within the grid (models # 2, 5, 10 ) and average grid water depth (models #4, 7 ) had no significant effect on detection probability or survival. Models which tested for the influence of temperature and the presence of dead fish consistently resulted in much higher AIC values and wer e therefore ex cluded from the table of resulting models Snail Abundance after hydrilla treatment) of sampling was roughly three times what it was Model averaging of the most parsimonious POPAN models revealed that a bundance was higher in the grids which were not treated both before and after treatment ( N CONTROLxBEFORE = 6 359; N CONTROLxAFTER = 18 291) However, these results are again insignificant due to overlapping c onfidence intervals (Fig. 3 8 ) and an insignificant effect size (0.00006 0.02782) Snail abundance estimates for treated plots were only slightly lower ( N TREATEDxBEFORE = 5,752; N TREATEDxAFTER = 15,514). Snail Density Similar to the changes noticed in abundance, snail density sharply increased between April and June sampling (Fig. 3 9 ). Using the MMDM method for calculating snail movement,
42 density appears to have increased to a greater degree in the control plots, from 2.04 to 5.86 snails/m 2 (see Table 3 2 for a complete list). Initial estimates were slightly lower in treated plots but increased from 1.84 to 4.97 snails/m 2 Size Classes Researchers have suggested that the presence of hydrilla may protect the smaller snails from predation but the resul ts gathered from trapping in Lake Jackson did not support that hypothesis. The proportion of snails that were smaller than 60 mm in height the cutoff between did not differ signifi cantly between the treated and control plots. If anything, grids in which the hydrilla was removed actually supported a greater percentage of smaller size classes in the later sampling occasions (Fig. 3 10 ). Mean snail shell height s and wet weight s increas ed over time but did not differ significantly between treated and control plots either (Figs. 3 1 1 and 3 1 2 ). Excluding small enough for juvenile kites to handle decreased from 42.4 to 13 .6% between the end of April and the end of June (Fig. 3 1 3 ). Although the proportion of small snails dropped, in general the number of snails captured greatly increased over the course of the sampling period. The number of snails that were captured that w ould be suitable for juvenile kites increased from 50 individuals in the initial trapping occasion to 124 individuals in the final occasion (Fig. 3 1 4 that would be unsuitable for juvenile k ites grew by nearly 1000% while the category of snails that were presumably extract increased by even more. The quantity of snails captured was higher in the treated plots
43 (Fig. 3 1 6 ) than in the control plots (Fig. 3 1 5 ) but relative proportions among size classes did not differ significantly. 8,261 unique individuals were captured in all. Exotic snails represent 99.75% of that number as only 22 native snails were found. Exactly half of all native snails were captured in Gri d #5. The rest were caught randomly in each of the five other grids. Non Target Species A number of fish species were regularly found in the funnel traps on Lake Jackson, especially blue spotted sunfish ( Enneacanthus gloriosus ) warmouths ( Lepomis gulosus ) and bluegills ( Lepomis macrochirus ) (Table 3 4). Fish represented 96.1% of the non target species captured. The list of o ther notable non fish species trapped includes sirens, frogs, snakes, and crayfish. The number of non target species captured greatly decreased over the course of the sampling period by 88.6% Differences between the numbers of individuals caught in the treated and control plots were insignificant. Of the 75 individual gar that were captured 84% suffered mortalities, by far the highest mortality rate seen among any of the non target species. Snails were frequently found in traps alongside the decomposing fish, but results from the CJS model showed that the presence of dead fish did not correlate wi th detection and unintentional baiting was not a significant issue.
44 Table 3 1 Estimates of A ( SE) and W ( SE) based on different methods of calculating snail movement. Mean Movement (m/day) Boundary Width ( W ) Total Area Sampled (m 2 ) ( A ) MDM 1.58 0.0 49 36.33 1.13 3286.95 133.58 MMDM 1.52 0.05 1 34.86 1.17 3120.79 134.93 M NDM 1.50 0.05 1 34.60 1.17 3091.41 134.51 Figure 3 1 Diagram of snail movements. Each color represents the path of an individual snail through time. Looped arrows mean that the individual was caught in the same trap on two consecutive occasions. Snails that were captured on more than three trapping occasions were relatively infrequent. The snails in this depiction were not necessarily fou nd in the same grid or even in the same season.
45 Table 3 2 Snail density ( D ) (snails/m 2 ) calculated using several different methods for estimating snail movement rates and effective sampling area. Treatment Sampling Period M DM MMDM MNDM Control Grids Before Treatment 1.9347 2.0377 2.0571 After Treatment 5.5648 5.8610 5.9167 Treated Grids Before Treatment 1.7501 1.8433 1.8608 After Treatment 4.7200 4.9713 5.0185 Table 3 3 Models represen ting hypotheses of effects of hydrilla treatment on apple snail distribution in Lake Jackson. Models also test for potential effects of individual and state covariates. # Model # Parm Akaike Weight Biological Hypothesis (1) Phi(t(b, d, a)*g + Size) p(t(b, a)*g) 31 0 0.41 Apparent survival varies by sampling period (before, immediately following, and after hydrilla treatment) and location on the lake (grid #) with an additive ef fect for snail shell size; detection probabilities vary by sampling period (before and after hydrilla treatment) and location on the lake. (2) Phi(t(b, d, a)*g + Size) p(t(b, a)*g + Loc) 32 1.6 6 0.1 8 Apparent survival varies by sampling period (before, i mmediately following, and after hydrilla treatment) and location on the lake with an additive effect for snail shell size; detection probabilities vary by sampling period (before and after hydrilla treatment) and location on the lake with an additive effec t for trap location within each grid. (3) Phi(t(b, d, a)*g*Size) p(t(b, a)*g) 45 1.8 5 0.16 Apparent survival varies by sampling period (before, immediately following, and after hydrilla treatment) and location on the lake with an interactive effect for s nail shell size; detection probabilities vary by sampling period (before and after hydrilla treatment) and location on the lake. (4) Phi(t(b, d, a)*g + Size) p(t(b, a)*g + Depth) 32 2.0 2 0.15 Apparent survival varies by sampling period (before, immediate ly following, and after hydrilla treatment) and location on the lake with an additive effect for snail shell size; detection probabilities vary by sampling period (before and after hydrilla treatment) and location on the lake with an additive effect for th e mean water depth of each grid. (5) Phi(t(b, d, a)*g) p(t(b, a)*g + Loc) 31 5.23 0.03 Apparent survival varies by sampling period (before, immediately following, and after hydrilla treatment) and location on the lake; detection probabilities vary by samp ling period (before and after hydrilla treatment) and location on the lake with an additive effect for trap location within each grid.
46 Table 3 3 Continued # Model # Parm Akaike Weight Biological Hypothesis (6) Phi(t(b, d, a)*g) p(t(b, a)*g + S ize) 31 5.4 8 0.0 3 Apparent survival varies by sampling period (before, immediately following, and after hydrilla treatment) and location on the lake; detection probabilities vary by sampling period (before and after hydrilla treatment) and location on the lake with an additive effect for snail shell size. (7) Phi(t(b, d, a)*g) p(t(b, a)*g + Depth) 31 5.5 8 0.0 3 Apparent survival varies by sampling period (before, immediately following, and after hydrilla treatment) and location on the lake; detection proba bilities vary by sampling period (before and after hydrilla treatment) and location on the lake with an additive effect for the mean water depth of each grid. (8) Phi(t(b, d, a)*g) p(t(b, a)*g) 29 8.50 0.0 1 Apparent survival varies by sampling period (be fore, immediately following, and after hydrilla treatment) and location on the lake; detection probabilities vary by sampling period (before and after hydrilla treatment) and location on the lake. (This is the top model excluding covariates.) (9) Phi(t(b, d, a)*g) p(t(b, d, a)*g) 36 8.85 0.0 1 Apparent survival varies by sampling period (before, immediately following, and after hydrilla treatment) and location on the lake; detection probabilities vary by sampling period (before, immediatedly following, and after hydrilla treatment) and location on the lake. (10) Phi(t(b, d, a)*g) p(t(b, a)*g*Loc) 42 14.69 0.00 Apparent survival varies by sampling period (before, immediately following, and after hydrilla treatment) and location on the lake; detection proba bilities vary by sampling period (before and after hydrilla treatment) and location on the lake with an interactive effect for trap location within each grid. (11) Phi(g) p(t(b, a)*g) 18 17.78 0.00 Apparent survival varies by location on the lake (grid); detection probabilities vary by sampling period (before and after hydrilla treatment) and location on the lake (12) Phi((t(b); t(d, a)*Trt) + Size) p(t(b, a)*g) 18 32.3 5 0 Apparent survival varies by sampling period (immediately following hydrilla treat ment or surveys after that, before treatment surveys are held constant among all sites) and treatment (either treated or control plots) with an additive effect for snail shell size; detection probabilities vary by sampling period (before and after hydrilla treatment) and location on the lake.
47 Table 3 3. Continued # Model # Parm Akaike Weight Biological Hypothesis (1 3 ) Phi(t(b, d, a)*Trt) p(t(b, a)*g) 18 41.92 0 Apparent survival varies by sampling period (before, immediately following, and after hydrilla treatment) and treatment (either treated or control plots); detecti on probabilities vary by sampling period (before and after hydrilla treatment) and location on the lake. (1 4 ) Phi(Size) p(t(b, a)*g) 14 43.50 0 Apparent survival varies by snail shell size; detection probabilities vary by sampling period (before and afte r hydrilla treatment) and location on the lake. Table 3 4 List of all non target species trapped in funnel traps between April and June, 2010 in Lake Jackson, Florida. Scientific Name Common Name Total Number Trapped Ennea canthus gloriosus Blue Spotted Sunfish 2475 Lepomis gulosus Warmouth 786 Lepomis macrochirus Bluegill 251 Astacoidea spp. Crayfish 91 Lepisosteus spp. Gar (Florida & Alligator) 75 Lepomis microlophus Redear Sunfish 73 Lepomis punctatus Spotted Sunfi sh 70 Ameiurus nebulosus Brown Bullhead Catfish 58 Lepomis marginatus Dollar Sunfish 41 Erimyzon sucetta Chubsucker 37 Vivaparous georgianus Banded Mysterysnail 25 Dytiscidae spp. Predaceous Diving Beetle 23 Notemigonus crysoleucas Golden Shiner 22 Esox niger Chain Pickerel 20 Hoplosternum littorale Armored Catfish 12 Siren lacertina Greater Siren 10 Micropterus salmoides Largemouth Bass 9 Nerodia floridana Florida Green Watersnake 9 Amia calva Bowfin 5 Unknown Tadpole 4 Rana grylio Pig Fro g 3 Poecilia latipinna Sailfin Molly 3 Fundulus chrysotus Golden Topminnow 2 Nerodia fasciata Banded Water Snake 2 Hyla cineria Green Tree Frog 1
48 Figure 3 2 Probability for detection ( p ) ( SE ) within control plots in Lake Jackson grids. These values were estimated by averaging mo st parsimonious CJS models. Figure 3 3 Probability for detection ( p ) ( SE ) within treatment plots in Lake Jackson grids. These values were estimated by avera ging most parsimonious CJS models. 0.00 0.05 0.10 0.15 0.20 0.25 Before Treatment After Treatment Detection Probability Sampling Period Grid 1 Grid 2 Grid 4 0.00 0.05 0.10 0.15 0.20 0.25 Before Treatment After Treatment Detection Probability Sampling Period Grid 3 Grid 5 Grid 6
49 Figure 3 4 Snail apparent survival ( ) ( SE ) of control plots in Lake Jackson grids. These values were estimated by averagin g most parsimonious CJS models. Figure 3 5 Snail apparent survival ( ) ( SE ) of treated plots in Lake Jackson grids. These values were estimated by averaging most parsimonious CJS models. 0.6 0.7 0.8 0.9 1.0 Before Treatment Immediately Following After Treatment Apparent Survival Sampling Period Grid 1 Grid 2 Grid 4 0.6 0.7 0.8 0.9 1.0 Before Treatment Immediately Following After Treatment Apparent Survival Sampling Period Grid 3 Grid 5 Grid 6
50 Figure 3 6 Snail apparent survival ( ) ( SE ) response to hydrilla treatment in Lake Jackson. Values based on results from model Phi (before, (immediate effect, delayed e ffect)*treatment) p ((before, after treatment )*grid). 0.80 0.82 0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 Before Immediately Following After Sampling Period Relative to Hydrilla Treatment Control Treated
51 Figure 3 7 Apparent survival ( ) as a function of shell size. 0.80 0.82 0.84 0.86 0.88 0.90 0.92 0.94 0.96 14 22 29 37 44 52 59 67 74 82 89 Apparent Survival Probability Shell Size (mm) Phi LCI UCI
52 Figure 3 8 Snail abundance ( N ) ( SE ) in response to hydrilla treatment in Lake Jacks on grids. These values were estimated by averaging most parsimonious POPAN models. Each estimate applies to three grids, areas totalling approximately 9000 m 2 0 2 4 6 8 10 12 14 16 18 20 22 Control Grids Treated Grids Control Grids Treated Grids Before Treatment After Treatment Abundance Estimate (Thousands)
53 Figure 3 9 Snail density ( D ) ( SE ) in response to hydrilla t reatment in Lake Jackson grids. These values were calculated using the effective sampling areas based on MMDM (see Table 3 1) and abundance estimates from model averaging top POPAN models. 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 Before After Snails/m 2 Sampling Period Relative to Hydrilla Treatment Control Grids (MMDM) Treated Grids (MMDM)
54 Figure 3 10 Proportion of snail s smaller than 60 mm captured in treated versus control plots. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Percentage of Total Captured within Each Area Survey Number Control Treated
55 Figure 3 11 The mean snail weight increased over time but was similar in treatment and control plots. Figure 3 12 The mean sna il height also increased over time but was similar in treatment and control plots. 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Mean Snail Weight Survey Number Control Treated 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Mean Snail Height Survey Number Control Treated
56 Figure 3 13 Percentage of snails captured within each size class on Lake Jackson in the 2010 trapping season. Since 98.5% of native apple sn ails are no greater than 60 mm in height, snails measuring less than 60 mm should be easily consumed by all kites, but those snails larger than 60 mm are expected to be more difficult for juveniles to handle. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Percentage of Snails Captured Survey Number >75 mm 61-75 46-60 30-45 <30 mm
57 Figure 3 14 Number of snails captured within each size class on Lake Jackson in the 2010 trapping season. Herbicide treatments took place between surveys 7 and 8 and the time period separating those two trapping occasions was roughly three weeks as opposed to three to four days as between each of the other surveys. 0 100 200 300 400 500 600 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Number of Snails Captured Survey Number <30 mm 30-45 46-60 61-75 >75 mm
58 Figure 3 15 Summary of snail sizes captured in control plots throughout the sampling seasons. Figure 3 16 Summary of snail sizes captured in treate d plots throughout the sampling seasons. 0 50 100 150 200 250 300 350 400 450 1 3 5 7 Treatment Period 8 10 12 14 Number of Snails Captured Survey Number <60 mm 60-75 >75 mm 0 50 100 150 200 250 300 350 400 450 1 3 5 7 Treatment Period 8 10 12 14 Number of Snails Captured Survey Number <60 mm 60-75 >75 mm
59 CHAPTER 4 D ISCUSSION Sampling Grid Efficacy Logistically, the use of grid arrays to conduct a mark recapture study of apple snails in central Florida lakes in f ested with hydrilla was very effective. The modification s made to each trap to sample the larger exotic species did appear to successfully trap smaller snails. The chimney extensions proved to be extremely necessary. The only fatalities seen among amphibians and reptiles occurred when water levels on Lake Toho rose very suddenly covering several traps located in a n especially deep area of the lake and one stinkpot turtle ( Sternotherus odoratus ) and a greater siren ( Siren lacertina ) drowned as a result. Snails captured during the same time period were found plast ered to the mesh lids, attempting to escape through the exit in the top, but were not harmed. Physically, the amount of time required to collect, measure, and mark trapped snails greatly exceeded what was expected. Sampling similar grid arrays in shallow w etland habitats, Valentine Darby et al. (2008) estimated that a team of two was able to set up and process a 100 trap grid array in approximately 15 to 40 hours over the course of 21 to 28 days. Preliminary sampling in Lake Toho and East Lake Toho verified this estimate. During the second half of the sampling season in Lake Jackson when catching upwards of 300 snails each day, a single grid consisting of 49 traps took between 4 0 and 6 0 hours over a 23 day period. The drastic difference in snail abundance be tween the various sites sampled was the primary factor causing this disparity. The only problem with the number of hours required is that trapping is difficult to stop once it has begun. Problems such as i nclement weather or mechanical failures can result in loss of data and greater mortality among non target species (see Desa 2008) Care should be taken to
60 avoid deploying so many traps that they become unmanageable when snail abundance increases as the season progresses. Despite potential issues the trap ping grid array should be considered the best viable option for determining snail abundance and density in lacustrine habitats Deploying a full grid array to determine only snail presence would be unnecessary. Trapping transects, like those used in Muench (2004) and DeSa (2008) would provide equally valuable data but would be much less time consuming. Unless a physical barrier is constructed to maintain t he assumption of geographic closure, the grid array design would not be recommended for estimating occ upancy either. The results of this study were unable to support Hypothesis 1 which predicted that snail densities would be lower in the deeper lacustrine habitats than in shallow wetlands sampled by Valentine Darby et al. (2008). Using the MMDM method fo r determining the effective sampling area, they calculated densities of native snails in southern Florida wetlands to be no greater than 0.32 snails/m 2 The lowest density of exotics observed in this study was 1.84 snails/m 2 Considering that the differenc e in water depth which suggests that a density that takes the cubic nature of the environment into consideration may be more appropriate, these estimates may not actually be that different. The depth of their traps was unknown but they state that the funne l traps may be used in water that is at least 15 cm. Assuming that their traps were located in 16 cm of w ater and adjusting the vertical depth to equal the mean of Lake Jackson grids would increase the density of snails per m 3 to approximately 1.8. Of cou rse, this adjustment ignores obvious biological differences between the two species. As mentioned previously, t elemetry research suggests that native snails tend to move toward shallower areas when they reach depths greater than 50 cm (Darby et al. 2002). Trapping
61 in Lake Jackson has suggested that this is not true for exotic snails. Densities of natives in shallower littoral regions of Lake Jackson may be similar to those in shallow wetlands, but further trapping would be required to determine that. Extrem e Movements Observed The two snails that were observed in different grids are thought to be an omalies. For a snail to simply use pedal wave locomotion, the stand method of movement, to travel from one grid to another is considered impossible. Judging by th e large size of the sn be conceivable for an avian predator to carry them from one grid to another and then discard them after failing to remove the operculum Adult snail kites frequently perched on poles in the grids, favoring the l arger diameter poles that were often used for the corners. It would not be unreasonable for a kite to catch a snail in or near one grid and then fly to another to extract and consume it. Human error is another option for the snail that moved from Grid #4 t o #5. It is possible, although unlikely, that a switch was made during sampling. The snails were held in water in lidded plastic containers while waiting to be processed, and containers from each grid were kept in different areas on the boat to avoid confu sion. When outdoor temperatures got very warm, we kept the containers open slightly to prevent the snails from overheating in the sun. Snails often hitting the bottom of the boat worked like an alarm to summon immediate attention, but a mix up would not have been impossible. However, this explanation is not an option for the snail that moved from Grid #1 to #6 because those grids were always sampled on different days. Snails have also been observed in aquariums travelling by suctioning to the surface tension of the water and floating upside down from one side of the tank to the other, allowing
62 their body to be pushed by the current and sometimes undulating to prop el their bodies further This method has been observed in other aquatic snails (see Lee et al. 2008) but has never before been reported in reference to apple snails This could have been another way for a snail to tra vel from Grid #1 to #6, but it would b e extremely unlikely for a snail to use this method to cross the center of a lake. Maintaining suction through even slight wave action would be nearly impossible over such a great distance. It would be even less likely for a snail to travel using this method between G rid s #4 and #5 because there was fairly dense vegetation between the two grids and a spoil island. The Effects of Herbicide Treatment Apparent Survival and Abundance Apparent survival did seem to decrease to a greater extent in treated plots than control plo ts, but according to model variances the differences were insignificant. Hypothesis 2 which stated that apparent snail survival would decrease in treatment plots after herbicide treatments had been completed could be neither confirmed nor denied Grids # 1 and #2, both control plots on the eastern side of the lake, showed no obvious differences throughout the season. Temperature readings, water depths, and vegetation samples all appeared to be similar, yet increased in Grid #1 and decreased in Grid #2 wh ile the opposite was observed for p in both grids (Figs. 3 2 and 3 4) All of the treated grids showed a slight decrease in but the confidence intervals indicate that there is little reason to believe that those values are exact. There may or may not be an explanation for the changes in apparent survival but they cannot be attributed to treatment effects in this study The rationale for believing that snails would leave the treated grids post treatment was based on the assumption that hydrilla is a major factor attracting snails to a particular area and
63 removing it should cause snail dispersion. However, the exotic snails are generalists known for their voracious appetites. The limited number of vegetation samples indicated an increase in biomass ( both hydrill a and non hydrilla species) over time. Removing the hydrilla would not necessarily ha ve caused the snails to starve or seek nourishment elsewhere. However, without the structural lofting quality provided by the hydrilla which would allow the snails to spen d more time near the water surface it may be true that they were less a vailable to snail kites. Bennetts et al. (2005) observed kites catching snails more frequently in areas of South Florida with more abundan t submersed macrophytes and sparse emergent ve getation. Future studies may want to include an aspect of kite foraging observations within ea ch grid to confirm the lofting hypothesis. The relationship between apparent survival and snail shell size was unexpected but not unreasonable. Apparent survival includes all changes in the observed population, incorporating births, deaths, immigrants and emigrants. Although true survival was probably very low for small snails, apparent survival was estimated to be lowest among the largest snails. This observation was most likely a response to the ability of the large snails to move at a faster pace. Simply put, they more frequently emigrated from the sampling grids. Further ana lyses to verify and support this hypothesis are being conducted. Speculations about snail abundance or density throughout the entire lake should be avoided. Snail distribution is known to be patchy in general (Darby et al. 2001). Very early preliminary sampling in Lake Jackson yielded zero captures in other areas. Kite behavior also supports this concept. If they locate an extremely good foraging patch, snail kites will focus their efforts on that one area continuously and may even exhibit territorial behavior to defend it (Snyder and Snyder 1970, Bennetts and Kitchens 2000). Apple snails are not known to be
64 extremely mobile but they have shown a propensity to move in search of more favorable habitat (Darby et al. 2002). A study which spans several years and includes larger treatment areas may produce more definitive results as the long term pr esence or absence of hydrilla could have a greater effect on snail abundance. Small Size Classes There was no evidence to support Hypothesis 3 which predicted that the proportion of smaller snails would be lower in the treated grids post treatment. The sma llest snails became almost non existent in all of the grids in the last sampling occasions. Adult snails continued to lay egg clusters throughout the entire season and although most of the observed difference was due to the fact that the snails were contin uing to grow, individuals in the smaller size classes were not being replaced. Hydrilla may not provide protection from predators as expected. When comparing the difference in the number of non target individuals trapped among treated and non treated plots there was no indication that vegetative density prevent ed predatory fish from entering Thick mats of hydrilla may actually provide support structure upon which other avian predators (e.g. boat tailed grackles ( Quiscalus major ) and limpkins ( Aramus guara una )) might perch and utilize to forage Since snail activity and rates of surface inspiration increase as temperatures increase (McClary 1964), foraging would not be difficult for these species even with limited water depth capabilities. Additionally, it is sus pected that the design of this study may have created a bias against the smaller size classes that cannot be ignored As mentioned earlier, apple snails deposit their egg masses on emergent vegetation (Winner 1989) and other non living strata (Wallace et a l. 1956). Throughout the entire season, the snails displayed an obvious preference for the PVC poles which were used to hold the traps in place. Farmers in the Philippines place stakes near
65 drainage canals which invasive P. canaliculata snails use for egg deposition and the farmers can then collect large batches of egg clusters with little physical exertion (Halwart 1994). It does not eliminate the snails, but it is one of many control methods employed. While we were not intentionally trying to influence sn ail demographics, scraping fresh egg clusters from the sides of the poles as we pulled each trap out of the water was unavoidable. However, clusters were also attached to nearby emergent vegetation and care was taken to avoid disturbing them when maneuveri ng the airboat so a proportion of young snails still should have still been entering the population. It is recommended that future studies take this speculation into consideration, especially if sampling is planned to continue for more than a few weeks. P roportion of Native Snails to Exotics As expected, the P. insularum individuals greatly outnumbered P. paludosa The actual difference was 375 to 1. This support ed Hypothesis 4 which stated that the number of the exotics would exceed that of the native sna ils. Researchers have been observing the expansion of the exotic snails in the KCOL for at least ten years (Rawlings et al 2007, DeSa 2008 Cattau et al. 2010). It has been suggested that kites are attracted to Lake Toho specifically because of the presenc e of exotics (Cattau 2008 ). Considering the increase of kites nesting on Lake Jackson in recent years, the same is likely to be true there. Non Target Fauna and Unintentional Baiting Model results showed that there was no strong correlation between non ta rget species captures and apple snails. This supported Hypothesis 5 which stated that snails would not be attracted to the traps due to the presence of non target species and detection rates would not be influenced as a result. Of the three most common by catch species, only bluegills ( Lepomis macrochirus ) have been known to consume juvenile snails (less than 10 mm) (Darby et al.
66 2009). It is assumed that adult b luegills were predating small snails within the grids, but most of the bluegills captured were s mall juveniles themselves and would not have capable of consuming such large prey within the traps, thereby a ffecting detection. Trapped apple snails, the obvious target species, greatly outnumbered the non target species. Darby et al. (2001) suggested tha t funnel traps may be effective in capturing other species commonly found in Florida wetlands, but such a projection would not apply to the lacustrin e habitats sampled in Lake Jackson P reliminary sampling on Lake Toho and East Lake Toho did yield a more d iverse list of by catch that included striped mud turtles ( Kinosternon bauri ) stinkpots ( Sternotherus odoratus ), and much more numerous greater sirens ( Siren lacertina ) and banded mysterysnails ( Vivaparous georgianus ) suggesting that those species were ju st not abundant on Lake Jackson or perhaps they prefer dif ferent habitat qualities such as shallower areas or different vegetative structure However, even in the preliminary plots, apple snails were still the most frequent species encountered. Using modified funnel traps could be a useful t ool for determining the presence of other species, but unless the species is known to be have a high abundance or limited home range, the funnel traps may not be recommended for a mark recapture study as small sample size could prove to be a serious issue. Conclusion Given then obvious trends in abundance estimated in this study, attention to timing of sampling events used for estimating lake wide or even site abundance could be very critical to the overall reliability of the estimates. The accuracy of abun dance estimates depend s entirely on trapping success which is a result of snail movements. Snail activity has been shown to be random ( also see Darby 1998) and c hanges in habitat such as the vegetation structure water depth, and food availability are only part of the equation. Stevens et al. (2003) argue that
67 attempts to compare relative snail abundance will be confounded if temperatures vary between sampling sites. Temperature is known to play a major role in snail activity, but it can differ according to habitat type. Shallow wetlands in which thick vegetation restricts water current will generally be warmer than the littoral zone of an open lake. Temporal scale should be greatly considered when conducting range wide surveys for apple snails in central an d southern Florida. Any comparisons of snai l abundance or density are meaning less if care is not taken to adjust sampling in accordance with these other factors. We have developed a process for determining more accurate estimates of snail abundance. Previo us grid based studies have estimated N by obtaining the mean number of captured individuals per trapping occasion and dividing by the detection probability (Valentine Darby 2008). This method is oversimplified and ignores the recruitment variable B i and in stead conditions on just the number of unmarked individuals trapped at each sampling occasion. The superpopulation approach provides an explicit link between the observable data and unknown parameters thought to be important. It was designed to be a gener alization and reparameterization of the Jolly Seber model that takes the complex structure of open populations into account, incorporating births, deaths, emigration, and immigration (Crosbie and Manly 1985). Using Program MARK, the POPAN model quickly tra nsformed this issue of complexity into simple and straightforward results The superpopulation approach proved to be very useful and appropriate for this kind of sampling in which timing was an important element. It is highly recommended for future grid ba sed studies. Unfortunately, t he effects of hydrilla treatment on apple snail abundance yielded inconclusive results. As discussed earlier, more significant evidence may be produced from a study which spans several years and includes larger treatment areas. It is still unclear whether
68 apple snails show a preference for areas dominated by the invasive submerged aquatic plant species. We know only that the snail kites have been observed foraging successfully in hydrilla dominated habitats and that this behavio r could have negative consequences for the future health of the snail kite population in Florida.
69 LIST OF REFERENCES Acosta, C.A., and S.S. Perry. 2001. Impact of hydropattern disturbance on crayfish population dynamics in the seasonal wetlands of Everg lades National Park, USA. Aquatic Conservation: Marine and Freshwater Ecosystems 11:45 57. Aditya G. and S K Raut. 2001. Food of the snail Pomacea bridgesi introduced i n India. Current Science 80:919 921. Aditya G. and S K Raut 2002. Destruction of I noplanorbis exustus (Planorbidae) eggs by Pomacea bridgesi (Ampullariidae). Molluscan Research 22:87 90. Aizaki, K., and Y. Yusa. 2008. Field observations of the alarm response to crushed conspecifics in the freshwater snail Pomacea canaliculata : effects of habitat, vegetation, and body size. Journal of Ethology 27(1): 175 180. Anderson, B. 1993. The Philippine Snail Disaster. The Ecologist 23(2): 70 72. Angehr, G. R. 1999. Rapid long distance colonization of Lake Gatun, Panama by sn ail kites. Wilson Bulle tin 111: 265 268. Baker, P., F. Zimmanck, S.M. Baker. 2010. Feeding rates of an introduced freshwater gastropod Pomacea insularum on native and nonindigenous aquatic plants in Florida. J ournal of Molluscan Studies 00: 1 6. Barnes, M.A., R.K. Forham, R.L. B urks, and J.J. Hand. 2008. Fecundity of the exotic applesnail, Pomacea insularum Journal of the North American Benthological Society 27(3):738 745. Beissinger, S.R. 1983. Hunting behavior, prey selection, and energetic of Snail kites in Guyana: consumer choice by a specialist. Auk 100:84 92. Bennetts, R.E. and W.M. Kitchens. 1997. The demography and movements of snail kites in Florida. U.S.G.S Biological Resource Division, Florida Cooperative Fish and Wildlife Research Unit. Technical Report No. 56. 169 pp. Bennetts, R.E. and W.M. Kitchens. 2000. Factors influencing movement probabilities of a nomadic food specialist: proximate foraging benefits or ultimate gains from exploration? Oikos 91(3):459 467. Bennetts, R.E., P.C. Darby, L.B. Karunaratne. 2006 Foraging patch selection by snail kites in response to vegetation structure and prey abundance and availability. Waterbirds 29:88 94.
70 Blackburn, R.D., L.W. Weldon, R.R. Yeo, and T. M. Taylor. 1969. Identification and distribution of certain similar app earing submersed aquatic weeds in Florida. Hyacinth Control Journal 8:17 23. Boland, B.B., M. Meerhof f, C. Fosalba, N. Mazzeo, M.A. B arnes, and R.L. Burks. 2007. Juvenile snails, adult appetites: Contrasting resource consumption between two species of app lesnails ( Pomacea ). Journal of Mulluscan Studies 74:47 54. Britton, J.C. 1991. Pathways and consequences of the introduction of non indigenous freshwater, terrestrial, and estuarine mollusks in the United States. Unpublished report to the Office of Techno logy Assesment, United States Congress. Contract H3 5750.0. Burela, S., and P. B. Martn. 2007. Nuptial feeding in the freshwater snail Pomacea canaliculata (Gastropoda: Ampullariidae ). Malacologia 49: 465 470. Burlakova, L.E., A.Y Karatayev, D.K. Padilla L.D. Cartwright and D.N Hollas. 2009. Wetland restoration and invasive species: applesnail ( Pomacea insularum ) feeding on native and invasive aquatic plants. Restoration Ecology 17(3):433 440. Burnham, K.P., and D.R. Anderson. 2002. Model selection and mu ltimodel inference: a practical information theoretic approach Springer Science + Business Media, New York. Carlsson, N.O.L., C. Bronmark, and L.A. Hansson. 2004. Invading herbivory: the golden apple snail alters ecosystem functioning in Asian wetlands Ecology 85:1575 1580. Carlsson, N.O.L. and J.O. Lacoursiere. 2005. Herbivory on aquatic vascular plants by the introduced golden apple snail ( Pomacea canaliculata ) in Lao PDR. Biological Invasions 7:233 241. Cattau, C.E. 2009. Effects of the invasive exotic apple snail ( Pomacea insularum ) on the snail kite ( Rostrhamus sociabilis plumbeus ) in Florida, USA [thesis]. [Gainesville (FL)]: University of Florida. Cattau, C.E., J. Martin and W.M. Kitchens. 2009. Effects of an exotic prey species on a native specialist: example of the snail kite. Biological Conservation 143:513 520. Cazzaniga, N.J. 1990. Predation of Pomacea canaliculata on adult Biomphalaria peregrina Annals of Tropical Medicine and Parasitology 84:25 27, 29. Conner, S.L, C. M. Pomory and P.C. Darby. 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 74:355 362.
71 Cooch, E. and G. White. 20 10 Program MARK: A Gentle Introduction, 9th ed. [Online.] Available at www.phidot.org/software/mark/docs/book/ Cormack, R.M. 1964. Estimates of survival from the sighting of marked animals. Biometrika 51:429 438. Cottam, C., and P. Knappen. 1939. Food of some uncommon Nort h American birds. Auk 56:138 169. Cowie, R.H. 2002. Apple snails (Ampullariidae) as agricultural pests: their biology, impacts and management. Molluscs as crop pests (ed. by G.M. Barker), pp. 145 192. CAB International, Wallingford, UK. Corrao, N.M., P.C Darby, and C.M. Pomory. 2006. Nitrate impacts on the Florida apple snail. ( Pomacea paludosa Say). Hydrobiologia 568:135 143. Crosbie, S.F., and B.F.J. Manly. 1985. Parsimonious modeling of capture mark recapture studies. Biometrics 41:429 438. Darby, P .C., P. L. Valentine Darby, and R.E. Bennetts. 1995. Spatial relationships of foraging and roost sites used by snail kites at Lake Kissimmee and Water Conservation Area 3A, Florid a. Florida Field Naturalist 24: 1 9. Darby, P.C. 1998. Florida apple snail ( Po macea paludosa Say) lif e history in the context of a hydrologically fluctuating environment [dissertation]. [Gainesville (FL)]: University of Florida. Darby, P.C., P.L. Valentine Darby, R.E. Bennetts, J.D. Croop, H.F. Percival, and W.M. Kitchens. 1998. Ec ological studies of apple snails ( Pomacea paludosa Say). Prepared under joint contract with South Florida Water Management District and St. Johns River Water Management District (SJRWMD), submitted December 1997. SJRWMD Special Pub. SJ98 SP6, Palatka, FL. Darby, P.C. 1999. Radio transmitter retrieval in wet lands using a magnetic probe. Journal of Field Ornithology. 70: 587 590. Darby, P.C., R. E. Bennetts, J. D. Croop, P. Valentine Darby and W.M. Kitchens. 1999. A comparison of sampling techniques for quant ifying abundance of the Florid a apple snail ( Pomacea paludosa Say). Journal of Molluscan Studies 65: 195 208. Darby, P.C., P.L. Valentine Darby, H.F. Percival, and W. M. Kitchens. 2001. Collecting apple snails ( Pomacea paludosa Say) from wetland habitats u s ing funnel traps. Wetlands 21: 308 311.
72 Darby, P.C., R.E. Bennetts, S.J. Miller, and H. F. Percival. 2002. Movements of Florida apple snails in relation to water levels and drying events. Wetlands 22:489 98. Darby, P.C., P.L. Valentine Darby, and H. F. Perc ival. 2003. Dry season survival in a Florida apple snail ( Pomacea paludosa Say) population. Malacologia 45:179 84. Darby, P .C., P.L. Valentine Darby, H. F. Percival and W.M. Kitchens. 2004. Florida apple snail responses to lake habitat restoration activi ty. Archiv Fr Hydrobiologie. 161:561 575 Darby P.C., L.B. Karunaratne and R.E Bennetts. 2005. The influence of hydrology and associated habitat structure on spatial and temporal patterns of apple snail abundance and recruitment. Final report submitted to USGS, Gainesville, FL. Darby, P.C., R.E Bennetts and L.B. Karunaratne. 2006. Apple snail densities in habitats used by foraging snail kites. Florida Field Naturalist 34:37 47. Da rby, P.C., D.J. Mellow, and M. L. Watford. 2007. Food handling difficul ties for snail kites capt uring non native apple snails. Florida Field Naturalist 35:79 85 Darby, P. C. 2008. Baseline monitoring of apple snails as an indicator of restoration success. Submitted to the St. Johns River Water Management District, Palataka, FL. January 28, 2008. 15 pp. Darby, P.C., R.E. Bennetts, and H.F. Percival. 2008. Dry down impacts on apple snail demography: implications f or wetlands water management. Wetlands 28:204 214. Dice, L.R. 1938. Some census methods for mammals. Journal of Wildlife Management 2:119 130. Dreitz, V.J., J.D. Nichols, J.E. Hines, R.E. Bennetts, W.M. Kitchens, and D.L. DeAngelis. 2002. The use of resighting data to estimate the rate of population growth of the Snail Kite of Florida. Journal of Applied Statistic s 29:609 623. Duffy M S C.L. Miller, J.M. Kinsella, and A. de Lahunta. 2004. Parastrongylus cantonensis in a nonhuman primate, Florida. Emerg Infect Dis 2004, 10:2207 2210. Estebenet, A.L. and P.R. Martin. 2002. Pomacea canaliculata (Gastropoda: Ampu llariidae): Life history traits an d their plasticity. Biocell 26: 83 89. Ferrer, J.R., G. Perera, and M. Yong. 1990. Life tables of Pomacea paludosa (Say) in natural conditions. Florida Scientist, Supplement 53. 15 pp. Gettys, L.A., W.T. Haller, C.R. Mudg e, and T.J. Koschnick. 2008. Effects of temperature and feeding preference on submerged plants by the island apple snail, Pomacea insularum 254.
73 Glass, N.H. and P. C. Darby. 2008. The effect of calcium and pH on Florida apple snail, Pomacea paludosa (Gastropoda: Ampullariidae), shell growth and crush weight. Aquatic Ecology DOI: 10.1007/s10452 008 9226 3. Haldeman, S.S. 1845. A monograph of the freshwater univalve mollusca of the United States, including no tices of species in other parts of North America. No. 8: Turbidae. E.G. Dorsey, Philadelphia, PA. 11 pp. Haller, W.T., and D. L. Sutton. 1975. Community structure and competition between Hydrilla and Vallisneria Hyacinth Control Journal 13 :48 50. Halwart M. 1994. The golden apple snail Pomacea canaliculata in Asian rice farming systems: present impact and future threat. International Journal of Pest Management, 40:199 206. Hanning, G.W. 1979. Aspects of reproduction in Pomacea paludosa (Mesogastropoda: Pilidae) [thesis]. [Tallahassee (FL)]: Florida State University. Hayes, K.A., R.C. Joshi, S.C. Thiengo, and R.H. Cowie. 2008. Out of South America: multiple origins of non native apple snails of Asia. Diversity and Distributions 14:701 712. Hayes, K.A., R.H. Cowie, A. Jrgensen, R. Schulthei, C. Albrecht, S.C. Thiengo. 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. Hoang, T.C., and G.M. Rand. 2009. Exposure routes of copper: Short term effects on survival, weight, and uptake in Florida apple snails ( Pomacea paludosa ). Chemosphere 76(3): 407 414. Hollingsworth R G ., and R H Cowie. 2006. Apple snails as disease vec tors. In Ecology and management of golden apple snails: a global scenario. Edited by Joshi RC. Muoz, Nueva Ecija: Philippine Rice Research Institute, p. 121 132. Howell, A.H. 1932. Florida Bird Life. Tallahassee, Florida, USA: Florida Department of Game and Fresh Water Fish. Howells, R.G. 2002. Comparative feeding of two species of applesnails ( Pomacea ). Ellipsaria 4(1):14 16. Howells, R.G., L.E. Bulakova, A.Y. Karatayev, R.M. Marfurt and R.L. Burks. 2006. Native and introduced Ampullariidae in North Am erica: history, status, and ecology. In: Global advancement in ecology and management of golden apple snails (R. C. Joshi & L. S. Sebastian eds), pp. 73 112. Philippine Rice Research Institute, Science City of Muoz, Nueva Ecija.
74 Hulon, M., C.K. McDaniel A.S. Furukawa, J. Buntz, C.S. Michael, D.C. Arwood, and A.C. Jacent. 1999. 1994 1999 Completion report for Kissimmee Chain of Lakes studies. Study IV. Lake Jackson and Marion investigation. Florida Game and Freshwater Fish Commission, Tallahassee, FL. J ohnson, S.A., and W.J. Barichivich. 2004. A simple technique for trapping Siren Iacertina Amphiuma means and other aquatic vertebrates. Journal of Freshwater Ecology 19(2):263 269. Jolly, G.M. 1965. Explicit estimates from mark recapture data with both death and immigration stochastic model. Biometrika 52:225 247. Joshi, R.C., and L S Sebastian. 2006. Global advances in the ecology and management of golden apple snails. Munoz Nueva Ecija: Philippine Rice Research Institute. Karatayev, A.Y., L.E. Burla kova, V.A. Karatayev, and D.K. Padilla. 2008. Introduction, distribution, spread, and impacts of exotic freshwater gastropods in Texas. Biome dical and Life Sciences 619(1): 181 194. Karunaratne, L.B., P.C. Darby, and R.E. Bennetts. 2006. The effects of hab itat structure on apple snail density and implications for wetland restoration and management of Florida wetlands. Wetlands 26:1143 1150. Kitchens, W.M., R.E. Bennetts, and D.L. DeAngelis. 2002. Linkages between the Snail Kite population and wetland dynam ics in a highly fragmented South Florida hydroscape. Pages 183 203 in Porter, J.W. and K.G. Porter, editors. Linkages between ecosystems: the south Florida hydroscape. CRC/St. Lucie Press, Delray Beach, Florida, USA. Kushlan, J.A. 1975. Population changes of the apple snail ( Pomacea paludosa ) in the southern Everglades. Nautilus 89(1):21 23. Lai, P.Y., Y F Chang, and R H Cowie. 2005. Proceedings APEC symposium on the management of the golden apple snail. September 2004 Pingtung, Chinese Taipei [Taiwa n]. National Pingtung University of Science and Technology. Langeland, K. A. 1996. Hydrilla verticillata (L.F.) Royle (Hydrocharitaceae), "The Perfect Aquatic Weed". Castanea 61 :293 304. Lee S., J.W.M. Bush, A. E. Hosoi and E. Lauga 2008. Craw ling bene ath the free surface: w ater snail locomotion Physics of Fluids 20:082106. Little, C. 1983. The colonisation of land: origins and adaptations of terrestrial animals. University Press, Cambridge.
75 Lorenz, J.J. 2000. The impact of water management on roseat e spoonbills and their piscine prey in the coastal wetlands of Florida Bay [dissertation]. [Coral Gables (FL)]: University of Miami. Lowe, S., M. Browne S. Boudjelas, and M. de Poorter alien species: a selection f rom 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. Hollands Printing, Auckland, New Zealand. Mackenzie, D. I., J. D. Nichols, J.A. Royle, K.H. Pollock, L.L. Bailey, and J.E. Hines. 2006. Occupancy estimation and modeling: inferring patterns and dynamics of species occurrence. Academic Press, Burlington, MA. McClary, A. 1964. Surface inspiration and ciliary feeding in Pomacea paludosa (Prosobranchia: Mesogastropoda: Ampullariidae). Malacologia 2:87 104. Miller C L J.M. Kinsella M.M. Garner S. Evan, P.A. Gullett and R E Schmidt. 2006. Endemic infections of Parastrongylus ( Angiostrongylus ) costaricenesis in two sp ecies of nonhuman primates, raccoons, and an opossum from Miami, Florida. J ournal of Parasitol ogy 92:406 408. Moxley, D. J. and F. H. Langford. 1982. Beneficial effects of hydrilla on two eutrophic lakes in central Florida. Proceedings of the Annual Confer ence Southeastern Association of Fish and Wildlife Agencies 36:280 86. National Academy of Sciences (NAS). 2003. Science and the greater Everglades ecosystem restoration: an assessment of the critical ecosystem studies initiative panel to review the criti cal ecosystem studies initiative. Washington, DC: National Research Council, NAS. Naylor, R. 1996. Invasions in agriculture: assessing the cost of the go lden apple snail in Asia. Ambio 25:443 448. Otis, D.L., K. P. Burnham, G.C. White, and D.R. Anderson. 1978. Statistical inference from capture data on closed animal populations. Wildlife Monographs 62:1 135. Padilla, D.K., and S.L. Williams. 2004. Beyond ballast water: aquarium and ornamental trades as sources of invasive species in aquatic ecosystems. Fr ontiers in the Ecology and Environment 2:131 138. Perry, M. 1973. Ecological studies of the apple snail at Lake Woodruff National Wildlife Refuge. Florida Scientist 36:22 30.
76 Rawlings, T.A., K.A. Hayes, R.H. Cowie, and T. M. Collins. 2007. The identity, distribution and impacts of non native apple snails in the continental United States. BMC Evolutionary Biology 7:97. Reed, W.L., and F.J. Janzen. 1999. Natural selection by avian predators on size and colour of a freshwater snail ( Pomacea flagellata ). Bio logical Journal of the Linnean Society 67(3):331 342. Reichert, B., C. Cattau, W. Kitchens, R. Fletcher, J. Olbert, K. Pias, C. Zweig and J. Wood. 2011. Snail Kite Demography Annual Report 2010. Submitted to the U.S. Army Corps of Engineers Environmental Branch, Jacksonville, FL. March 2011. 111 pp. Rhymer, J.M ., and D. Simberloff. 1996. Extinction by hybridization and introgression. Annual Review of Ecology and Systematics 27:83 109. Richardson, R.J. 2008. Aquatic plant management and the impact of eme rging herbicide resistance issues. Weed Technology 22(1):8 15. Rogevich, E .C., T.C. Hoang, and G.M. Rand. 2009. Effects of s ublethal c hronic c opper e xposure on the g rowth and r eproductive s uccess of the Florida a pple s nail ( Pomacea paludosa ). Archives of Environmental Cont amination and Toxicology 56(3): 450 458. Scheffer, M., S.H. Hosper, M.L. Meijer, B. Moss, and E. Jeppesen. 1993. Alternative equilibria in shallow lakes. Trends in Ecology and Evolution 8:275 279. Schwartz, C.J., and A.N. Arnason. 1996. A general methodology for the analysis of capture recapture experiments in open populations. Biometrics 52:860 873. Seber, G.A.F. 1965. A note on the multiple recapture census. Biometrika 52:249 259. Seber, G.A.F. 1982. The estimation of animal abundance and related parameters. MacMillan, New York, 654 pp. Sharfstein, B. and A.D. Steinman. 2001. Growth and survival of the Florida apple snail ( Pomacea paludosa ) fed 3 naturally occurring macrophyte assemblages. Journal of the North American Benthological Society 20:84 95. Snyder, N.F. R. and H. A. Snyder. 197 0. Feeding territories in the Everglade Kite. Condor 72:492 493. Snyder, N.F. R. and H. A. Snyder. 1971. Defenses of the Florida apple snail Pomacea paludosa Behaviour 40: 175 215.
77 Stevens, A.J., Z.C Welch, P.C. Darby, and H. F. Percival. 2002. Temperature effects on Florida applesnail activity: implications for snail kite foraging success and distribution. Wildlife Society Bulletin 30:75 81. Stevens, A.S., N.M. Stevens, P.C. Darby, and H. F. Percival 1999. Observations of fire ants attack ing aestivating apple snails. Journal of Molluscan Studies 65: 507 510. Stickel, L.F. 1954. A comparison of certain methods of measuring home range of small mammals. Journal of Mammalogy 35:1 15. Stieglitz, W. O and R.L. Thompson. 1967. Status and life history of the Everglade Kite in the United States. Special Scientific Report, Wildlife 109. Washington, District of Columbia, USA: U. S. Fish and Wildlife Service. Strange, L.A. 1998. The applesnails of Florida ( Gastropoda: Prosobranchia: Pilidae). Florida Department of Agriculture and Consumer Services, Entomology Circular 388. 2 pp. Sykes, P.W. 1987. The feeding habits of the Snail Kite in Florida, USA. Colonial Waterbirds 10:84 92. Sykes, P.W ., J.A. Rogers, a nd R.E. Bennetts. 1995. Snail kite ( Rosthamus sociablilis ). p 1 32 i n A. Poole and F. Gill (eds.) The Birds of North America, No. 171. The Academy of DC, USA. Tanaka, R. 1 972. Investigation into the edge effect by use of capture recapture in a vole population. Researches in Population Ecology 13:127 480. Tanaka, M.O., A.L.T. Souza, and E.S. Mdena. 2006. Habitat structure effect on size selection of snail kites ( Rostrhamus sociabilis ) and limpkins ( Aramus guarauna ) when feeding on apple snails ( Pomacea spp. ). Acta Oecologica 30: 88 96. Thompson, F.G. 1984. An identification manual for the freshwater snails of Florida. University of Florida Press, Gainesville, 94p. Thompson F.G. 1997. Pomecea canaliculata (Lamarck 1822) (Gastropoda, Prosobrancha, Pilidae): a freshwater snail introduced into Florida. Walkerana 10:1 96. Turner, R.L. 1996. Use of stems of emergent plants for oviposition by the Florida applesnail, Pomacea palu dosa, and implications for marsh management. Florida Scientist 59:34 49. Turner, R.L., and S.E. McCaffree. 1994. Desiccation tolerance of adult and hatching Florida apple snails. Florida Scientist 57(suppl. 1):18.
78 USGS. 2009. NAS Nonindigenous Aquatic Species, Species List of Nonindigenous Mollusks. U.S. Department of the Interior, U.S. Geological Survey http://nas.er.usgs.gov/queries/SpeciesList.aspx?Group=Mollusks USGS. 2011. NAS Graphs and Charts: Aquatic Species by State U.S. Department of the In terior, U.S. Geological Survey. http://nas.er.usgs.gov/graphs/State.aspx Valentine Darby, P.L., P.C. Darby, R. E. Bennetts, W.M. Kitchens and H. F. Percival. 2008. The use of mark recapture to estimate Florida apple snail ( Pomacea paludosa Say) density in wetland habitats. Florida Scientist 71:115 127. Van, T. K., G. S. Wheeler and T. D. Center. 1998. Competition between Hydrilla verticillata and Vallisneria americana as influenced by soi l fertility. Aquatic Botany 62: 225 233. Van, T.K., G. S. Wheeler and T. D. Center. 1998. Competitive interactions between Hydrilla ( Hydrilla verticillata ) and Vallisneria ( Vallisneria americana ) as influenced by insect he rbivory. Biological Control 11: 185 192. Van, T.K., and K. K. Steward. 1990. Longevity of monoecious hydr illa propagules. Journal of Aquatic Plant Management 28 :74 76. Wada, T. and K. Matsukura. 2007. Seasonal changes in cold hardiness of the invasive freshwater apple snail, Pomacea canaliculata (Lamarck) (Gastropoda: Ampullariidae). Malacologia 49: 383 392. Wallace, H.E., C.M. Loveless, F.J. Ligas, and J.A. Powell. 1956. Wildlife investigation of the central and southern Florida flood control project. Annual progress report for Investigations Project as required by Federal Aid in Fish and Wildlife Restorati on Acts. Florida Game and Freshwater Fish Commission, Tallahassee, FL. 36 pp. (Unpublished.) Wheeler, G. S. and T. D. Center. 2001. Impact of the biological control agent Hydrellia pakistanae (Diptera: Ephydridae) on the submersed aquatic weed Hydrilla ver ticillata (Hydrochari taceae). Biological Control 21: 168 181. Williams, B.K., J.D. Nichols, and M.J. Conroy. 2002. Analysis and management of animal populations. Academic Press, San Diego, California. Wilson, K.R., and D.R. Anderson. 1985. Evaluation of t wo d ensity e stimators of s mall m ammal p opulation s ize Journal of Mammalogy 66(1):13 21. Winner, B.E. 1989. A comparison of three species of Pomacea and their spawn. American Conchologist 17:15 17. Winner, B.E. 1991. A field guide to mulluscan spawn. II. North Palm Beach.
79 Wood, T., P. Anurakpongsatorn, R. Chaichana, J. Mahujchariyawong and T. Satapanajaru. 2005. Predation on freshwater bryozoans by the apple snail, Pomacea canaliculata Ampulariidae, an invasive species in Southeast Asia: a summary repor t. Denisia 16:283 286.
80 BIOGRAPHICAL SKETCH Lara Kristen Drizd was born in St. Petersburg, Florida in 1982. The younger of two children, she grew up in the Tampa Bay area, graduating from Countryside High Sch ool in 2001. She obtained a Bachelor of Arts in environmental s tudies at New College of Florida in 2005 after Liguus fasciatus ) in the Everglades In 2006, Lara began working in the Everglades as a field technician for Florida Atlantic University, Davie on several projects including the investigations into: the growth and expansion of Lygodium microphyllum into the Everglades, determining the hydrology of tree islands in South Florida, and crayfish population dynamics in respons e to hydrological influences. A year later she moved to Gainesville to work for the Florida Fish and Wildlife Cooperative Research Unit and continued collecting data important to wildlife species endemic to the state by helping to launch a pilot study whi ch established a series of sites to be used to monitor the health of nesting habitat of the Everglade Snail Kite. She was then accepted to the Wildlife Ecology and Conservation P rogram as a graduate assistant at the University of Florida to pursue research into the species of apple snails which inhabit the Kissimmee Chain of Lakes.