EFFECTS OF DIET AND DENSITY ON Lissachatina fulica (STYLOMMATOPHORA: ACHATINIDAE)

By

KATRINA L. DICKENS

A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2016

© 2016 Katrina L. Dickens

To Theodora Dickens

ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. John Capinera, for his mentoring, feedback and advice. I also thank Dr. Trevor Smith for acting as my committee member and for giving me the opportunity to do this research. Also I thank all of the workers at the

Department of Agriculture and Consumer Services, Division of Plant Industry in

Gainesville, Florida that helped in this research and colony maintenance, specifically

Cory Penca, Amy Howe, Jessica McGuire, Shannen Leahy, Shweta Sharma, Addison

Mertz, Cason Bartz, and Steven Rowley.

4

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 7

LIST OF FIGURES ...... 8

ABSTRACT ...... 9

CHAPTER

1 LITERATURE REVIEW ...... 11

Agricultural Pest ...... 11 Nuisance ...... 11 Disease Vector ...... 12 Description ...... 12 Habitat ...... 13 Distribution ...... 14 Control ...... 16 Research Needed ...... 19

2 HOST PLANT SUITABILITY ...... 23

Introduction ...... 23 Methods ...... 24 Results ...... 26 Discussion ...... 27

3 GROWTH AND REPRODUCTION ...... 34

Introduction ...... 34 Methods ...... 37 Rearing Methods ...... 37 Neonate Growth ...... 38 Maturity and Reproduction ...... 38 Pattern of growth ...... 40 First oviposition ...... 40 Sexual reproduction ...... 41 Self-fertilization ...... 42 Results ...... 43 Growth ...... 43 First Oviposition ...... 44 Sexual Reproduction ...... 45 Self-Fertilization ...... 45

5

Discussion ...... 46

4 EFFECTS OF DENSITY AND FOOD DEPRIVATION ON GROWTH AND SURVIVAL ...... 57

Introduction ...... 57 Methods ...... 59 Density Effects on Growth and Reproduction ...... 59 Food Deprivation Effects on Cannibalism...... 61 Results ...... 62 Density Effects on Growth and Reproduction ...... 62 Food Deprivation Effects on Cannibalism...... 63 Discussion ...... 63

LIST OF REFERENCES ...... 69

BIOGRAPHICAL SKETCH ...... 75

6

LIST OF TABLES

Table page

2-1 List of known preferred or susceptible host plants of Lissachatina fulica that occur in the snail-infested area of Florida ...... 30

2-2 Twenty-four diet treatments fed to juvenile Lissachatina fulica to assess suitability ...... 30

3-1 Means (± SD), medians, ranges, and sample sizes (number of snails, n) for height and mass of Lissachatina fulica ...... 51

3-2 Means (± SD), medians, ranges, sample sizes (number of cages) for the number of days from hatching for Lissachatina fulica to lay eggs among diet and density treatments. Means (± SD), medians, ranges, sample sizes (number of snails) for the height of Lissachatina fulica at the first egg laying event ...... 51

3-3 Means (± SD), medians, ranges, and sample sizes (n) for the total number of eggs laid per snail, percent egg viability, and clutch sizes among diet treatments for eggs laid before 241 d for paired Lissachatina fulica ...... 52

3-4 Total eggs produced per snail, mean (± SD) egg viability, and mean (± SD) clutch size for egg produced 0-240 d and 0-540 d for solitary and paired Lissachatina fulica regardless of diet ...... 52

3-5 Mean (± SD) eggs laid per snail, egg viability, and clutch size for eggs produced 0-540 d for solitary and paired Lissachatina fulica for each diet ...... 53

4-1 Mean (± SD) height, mass, percent mortality of Lissachatina fulica 240 d after hatch reared at three different density treatments ...... 66

4-2 Lissachatina fulica reproduction when reared at three different density treatments: mean (± SD) number of days from hatch to initiation of egg production, estimated number of eggs produced per snail (over a 60-day period), eggs per clutch, and percent egg viability ...... 66

7

LIST OF FIGURES

Figure page

1-1 Lissachatina fulica shell ...... 21

1-2 Shells of five snail species native to Florida and one non-native that may be confused with Lissachatina fulica ...... 22

2-1 Mean height and mass of newly hatched Lissachatina fulica after 70 d of feeding on a single diet treatment ...... 32

2-2 Mean percent survival of newly hatched Lissachatina fulica after 70 d of feeding on a single diet treatment ...... 33

3-1 The effects of rearing Lissachatina fulica at 6 constant temperatures ...... 54

3-2 Growth pattern of Lissachatina fulica reared on a diet of lettuce or synthetic diet. A polynomial regression was fitted on individual measurements for height ...... 55

3-3 Individual measurements of Lissachatina fulica peristome thickness taken before (-) and after (+) first oviposition ...... 56

3-4 The mean total Lissachatina fulica eggs laid per snail over time, and the mean percent of eggs that were viable ...... 56

4-1 Lissachatina fulica growth when reared at low, medium, and high densities ...... 67

4-2 Mean percent mortality of Lissachatina fulica for four diet treatments using adults, juveniles, and a mix of the two sizes ...... 68

8

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science

EFFECTS OF DIET AND DENSITY ON Lissachatina fulica (STYLOMMATOPHORA: ACHATINIDAE)

By

Katrina Dickens

May 2016

Chair: John L. Capinera Major: Entomology and Nematology

Effects of diet and density on Lissachatina fulica were investigated to aid in the eradication efforts of the L. fulica infestation in Miami-Dade and Broward Counties. Host plant suitability of 21 plants commonly grown in Miami was tested using snail growth and survival. Tagetes patula produced the largest snails (24 mm in height). Some plants, such as Portulaca oleracea, Helianthus annuus, and Tradescantia spathacea, are not ideal for growth (< 10 mm) but may be able to sustain a snail until it can find better quality food. The results of this study provide insights on the food preferences of this polyphagous snail, which is otherwise understudied.

Several critical aspects of L. fulica growth and reproduction were identified. The optimal temperature for L. fulica development was 30 °C. Snails averaged 195 d old and

97 mm in height when egg laying was initiated. Peristome thickness was inconsistent in predicting full reproductive maturity. Egg production averaged 4,698 eggs/snail laid in the first 540 d after hatch. Oviposition without mating was shown to be possible, but occurred rarely (3% snails) and with lower viability rates and smaller clutch sizes.

The effects among three rearing densities: low (0.29 snails/L), medium (0.88), or high (2.06) on L. fulica biology were determined. High rearing density reduced growth,

9

affected the initiation of egg laying, and reduced the number of eggs laid per snail and per clutch. Mortality did not increase with high rearing density nor was any cannibalism observed, even when snails were deprived of food.

10

CHAPTER 1 LITERATURE REVIEW

Agricultural Pest

The giant African land snail, Lissachatina fulica Bowdich (Stylommatophora:

Achatinidae), has the potential to be a major plant pest because it is a polyphagous feeder, eating hundreds of plant species. Venette & Larson (2004) list several plant species that L. fulica eats that are of agricultural significance. Even in cases where agricultural damage is relatively minor, an agricultural industry may be negatively affected by export restrictions. Countries will not risk accidental transport of the snail hidden in agricultural goods. For this reason "the USA will embargo imports from the

French Antilles" (Civeyrel & Simberloff 1996). This species becomes especially damaging to plants when populations increase rapidly, as is common when first establishing in a new area (Thiengo et al. 2007).

Nuisance

Public panic may be the driving force for agencies to take action rather than the threat of agricultural damage (Civeyrel & Simberloff 1996). This species becomes a general nuisance in urban areas by eating stucco on buildings, clogging air conditioning units, breaking lawn mowers, and damaging ornamental plants (Sturgeon 1971). Large adult snails with very thick and hard shells can be a road hazard (Civeyrel & Simberloff

1996). It is also aesthetically displeasing to see snail feces on the walls of houses, around yards, and inside storage sheds (Sturgeon 1971). Dead snails produce a strong sour and far-ranging smell, in addition to the milder odor associated with the feces. One of the most significant factors related to this pest is its ability to spread parasites and disease.

11

Disease Vector

Lissachatina fulica can act as an intermediate host for Angiostrongylus cantonensis Chen (Raut & Barker 2002; Venette & Larson 2004). For example, a portion of snails were found to be infested by Angiostrongylus cantonensis when samples of L. fulica were tested from a population in South Florida (Smith et al. 2015).

This nematode can cause angiostrongyliasis, which can lead to meningitis in humans if undercooked snail meat, or water or vegetables contaminated with snail feces, are ingested (Alicata 1991; Wang et al. 2008).

Description

The size of L. fulica can vary greatly within a population, and color and patterns are variable as well but to a lesser extent. Some L. fulica can be long and slender and others short and stout, but all have a conical shell (Bequaert 1950). The height of the shell opening is never larger than the height of the spire (Fig. 1-1). Fully mature L. fulica have 7-9 whorls, the last enlarged to fit the majority of the snail's body and internal structures. Whorls are convex and join at moderately deep sutures. The shell color has a yellow to off-white background with vertical (parallel to the columella), dark brown stripes that sometimes merge together. In mature snails, the stripes on the body whorl are less contrasted and thinner, making the body whorl darker overall.

There are at least five snails native to Florida and one non-native that may be confused with L. fulica (Fig. 1-2). The Stock Island tree snail, Orthalicus reses (Say)

(Gastropoda: Orthalicidae), looks very similar to L. fulica, but is smaller and lacks the white columella that curves inward. The columella of O. reses is also continuous with the aperture, while the columella of L. fulica cuts off abruptly at the end. And, unlike L.

12

fulica, this snail also has faint horizontal stripes (Deisler 1983). The Florida tree snail,

Liguus fasciatus (Muller) (Gastropoda: Orthalicidae), has a white columella similar to L. fulica. Liguus fasciatus can be a variety of colors, including yellow and dark brown that form vertical stripes (Deisler 1983). Unlike L. fulica, this snail also has horizontal white, orange, or brown stripes. Horizontal stripes separate the banded tree snail, Orthalicus floridensis Pilsbury (Gastropoda: Orthalicidae), from L. fulica (Deisler 1983). Adult shells of the many-lined tree snail, Drymaeus multilineatus (Say) (Gastropoda: Bulimulidae), are less than 40 mm in height, which is smaller than L. fulica. Also, D. multilineatus shells have a dark apex and a dark stripe at each suture, while L. fulica have a white apex. The rosy wolf snail, Euglandina rosea (Férussac) (Gastropoda: Spiraxidae), is another native snail smaller than L. fulica. The shell of E. rosea can reach up to 76 mm in height, but is narrower than L. fulica and more “fusiform” than conical (Auffenberg &

Stange 2014). The shell is also brownish-pink and has distinct, vertical growth lines but lacks the vertical, dark brown stripes of L. fulica (Auffenberg & Stange 2014). Non- native apple snails, Pomacea maculata (Perry) (Gastropoda: Ampullariidae), are aquatic snails common in Florida, but not a major pest. They can grow to be quite large, but their shells are more spherical than conical in shape than L. fulica (Fasulo 2008). It is important for people to be able to identify the species in order to reduce accidental transportation of L. fulica to new areas. It is also important to know what habitats are the most suitable and therefore susceptible to the establishment of a population.

Habitat

Lissachatina fulica is found widely in East Africa, and is abundant along forest edges (Raut & Barker 2002). They have a wide range, in part, because they tolerate a broad range of temperatures. Optimal temperatures for growth and reproduction are 22 13

- 28 °C (Raut & Ghose 1984). Temperatures above 28 °C are said to initiate aestivation, but the snail can survive temperatures near 45 °C (Raut & Ghose 1984). This snail is tolerant of cold weather, as it is known to be active below 10 °C and to hibernate at 2 °C

(Raut & Barker 2002; Mead 1979; Raut & Ghose 1984). They have been reported to survive temperatures as low as 0 °C (Raut & Ghose 1984). Altitude does not greatly restrict their distribution, as populations have established in Peninsular Malaysia at

1,500 m. As with most snails, L. fulica are generally inactive during the day and active at night. However, temperature and light affect L. fulica activity less than humidity (Raut &

Barker 2002; Takeda & Ozaki 1986). These snails are active when humidity is between

50-80% (Takeda & Ozaki 1986; Raut & Ghose 1984).

Distribution

Lissachatina fulica is native to the coastal areas of East Africa (Bequaert 1950), but has since spread throughout Africa to Morocco, the Ivory Coast, and Ghana (Raut &

Barker 2002). Outside of Africa, it spread to the Indian Ocean Islands (1800s), then to

India (1847), Sri Lanka (1900), Malaysia (1911), and to most of East Asia by 1940 (Raut

& Barker 2002). From East Asia, it spread to the Pacific, reaching Papua New Guinea

(1946) and then French Polynesia (1978). It spread to Hawaii in 1936 via Japan and

Taiwan, and from Hawaii to Florida in 1966 (Raut & Barker 2002). After being eradicated from Florida in 1975 it was reintroduced in 2011. Thus, it is highly invasive and adaptable.

Like many pests, these snails can be moved unintentionally while hidden in plants, produce or other cargo. What is fairly unique about snails is that they are sold in many parts of the world as escargot. Lissachatina fulica was distributed to many parts of

Asia due to the snail meat trade (Raut & Barker 2002). It was imported to Brazil in 1988 14

with the same purpose; however, the industry never flourished, but the snail remains

(Thiengo et al. 2007).

Another major reason for the importation of this snail is the pet trade, because people are fascinated with its large size. Lissachatina fulica has been deliberately smuggled into the United States and sold in pet stores in Florida (U.S. Congress Office of Technology Assessment 1993). Many others have brought snails home as pets, not knowing the risks or restrictions. For example, a boy brought two snails from Hawaii to the U.S. in 1958. Luckily, the snails remained captive and were identified and culled when the boy gave them to a wild animal farm in Arizona (Mead 1959). In June 1966, a boy and his mother vacationed in Hawaii where he took interest in the giant snails he saw there. He stowed three in his pockets and released them outside his North Miami home. Heavy summer rains and a mild winter created the perfect environment for the snails to lay eggs and rapidly increase in number. However, the snails went unreported for several reasons. First, the snail population was only dense near the boy’s home and were otherwise evenly spread out. Also, plant damage isn't uncommon in south Florida and residents didn't take notice or realize that it was caused by this new invader. The snails were not identified until September 1969 and finally eradicated in 1975 (Poucher

1975; Sturgeon 1971).

Another L. fulica infestation is believed to have originated from snails smuggled into Miami-Dade County for use in religious rituals. This population was detected in

September of 2011 and has yet to be eradicated, as it is much larger than the L. fulica population discovered decades earlier. It is important to identify the places that L. fulica

15

has been introduced as quickly as possible because once they have established they can be very hard to eradicate.

Control

The three major ways to control this pest involve chemical, physical, and biological methods. Starting in the 1930s, gastropods were successfully controlled using baits containing metaldehyde, which along with calcium arsenate, were the first chemicals used to control L. fulica (Civeyrel & Simberloff 1996; Raut & Barker 2002).

There have been a few minor improvements in chemical control technology since the introduction of these chemicals. Baits using iron phosphate, sodium ferric EDTA, or boric acid are considered environmentally safe and effective against L. fulica, but none are as effective as metaldehyde (Smith et al. 2013). Biologically based molluscicides

(usually botanical in origin) have mostly been designed for aquatic gastropods, but

Panigrahi and Raut (1994) have found that an extract from the yellow oleander fruit,

Thevetia peruviana (Pers.) Schumann (Gentianales: Apocynaceae), is an effective control agent for L. fulica. However, this plant is very toxic and can cause mortality in vertebrates (Langford & Boor 1996). It can be challenging to find a safe and effective chemical because these snails produce so much mucus as a defense mechanism, and can slough off many chemicals when they come in contact with a toxicant (Panigrahi &

Raut 1994). Metaldehyde continues to be the preferred choice, but it is dangerous and toxic to non-target organisms, and is not recommended for use around pets or children

(Flint & Wilen 2009), Panigrahi & Raut 1994). For this reason, bittering agents have been added to most commercially available metaldehyde products to prevent ingestion by mammals and other non-target organisms.

16

Field testing is also important, because many chemicals break down in the rain

(Raut & Barker 2002). In a place that receives high precipitation like Miami, these molluscicides would need to be reapplied often. Researchers continue to try to formulate more durable baits that do not degrade in the rain, and to find more effective environmentally friendly alternatives to synthetic molluscicides.

Physically collecting and destroying snails and eggs has been a successful means of eradication for early stages of infestations in Japan and Australia (Raut &

Barker 2002). Physical barriers around plants can prevent damage (Raut & Barker

2002), but this would not help eradicate a snail population from an area. These methods are most effective when used alongside chemical controls.

Physically collecting snails and using chemical controls can have a large monetary cost (Civeyrel & Simberloff 1996). The giant African land snail was introduced and established a population in Miami, FL in 1966. This introduction led to an eradication program conducted by the Florida Department of Agriculture, Division of

Plant Industry. During the first four years of this program, almost 128 tons of arsenate and metaldehyde were applied and 17,000 snails were hand collected from 133 properties spread over 40 square blocks (Sturgeon 1971; Poucher 1975). Although it was declared successfully eradicated on April 13, 1975, about $700,000 (not adjusted for inflation) was spent to eliminate this pest (Poucher 1975). Civeyrel and Simberloff

(1996) give other examples of countries and their eradication costs.

Because hand removal and chemical control are costly and often do not reduce

L. fulica populations adequately (Civeyrel & Simberloff 1996), biological control agents against L. fulica have been used in many places, including parts of India, Asia, and the

17

Pacific Islands. Raut and Barker (2002) and Civeyrel and Simberloff (1996) list invertebrate predators found in and outside of Africa that have been, and could be, used in biological control programs. The main agents that have been used are two African predatory snails, Gonaxis kibweziensis E.A. Smith and Gonaxis quadrilateralis Preston

(Gastropoda, Streptaxidae), one predatory snail found in Florida, E. rosea, and a

Hawaiian flatworm, Platydemus manokwari. Prior to their use in Hawaii in the 1960s, the two Gonaxis species were tested in the lab and field and were not found to be very effective against L. fulica (Civeyrel & Simberloff 1996). They mostly feed on L. fulica eggs and small snails (< 35mm) (Civeyrel & Simberloff 1996). Euglandina rosa was chosen for testing because it had not evolved with the L. fulica and, in theory, L. fulica does not have an adaption to be able to avoid and escape the novel predator species.

Euglandina rosea was not tested, however, against native Hawaiian species. Tests of potential biological control agents to be released in Hawaii and elsewhere focused on the effectiveness of the agents against L. fulica, and did not adequately consider the effects of these biological control agents on native fauna. Thus, when introduced, these predatory snails caused many native land snail extinctions, particularly in the Pacific

Islands (Hopper & Smith 1992; Cowie 1992; Hadfield et al. 1993).

The intentional introduction of L. fulica predators in the Pacific Islands put additional pressure on native species, whose populations were already in decline

(Hopper & Smith 1992; Cowie 1992; Hadfield et al. 1993). Islands are particularly vulnerable to invasive species because native species have not evolved to adapt to new competition or predation, and confined populations are small and often localized; thus, they cannot replenish quickly (Hadfield et al. 1993). Partulidae, a large snail family

18

endemic to the Pacific Islands, and Achatinellinae, a tree snail genus endemic to

Hawaii, are in decline due to many causes: habitat destruction, shell collectors, and accidental non-native predators, such as rats and ants (Hopper & Smith 1992; Cowie

1992; Hadfield et al. 1993). Partulidae naturally have no known predators (Cowie 1992), but now is considered extinct in the Mariana Islands, largely due to P. manokwari

(Hopper & Smith 1992), and on Moorea and Tahiti, due to E. rosea (Clarke et al. 1984;

Murray et al. 1988). Both E. rosea and rats cause native snail populations to decline, but unlike E. rosea, rats don't eat snails to extinction. Rats often move to another area before decimating a snail population. Also, rats eat only large native snails, whereas E. rosea eats all native snail size ranges (Hadfield et al. 1993). However, because native snails don’t get as large as L. fulica, E. rosea eats mostly 15-30 mm L. fulica, and therefore is not very effective against L. fulica as a biological control agent (Civeyrel &

Simberloff 1996). In fact, L. fulica remains established in areas where E. rosea was introduced (Clarke et al. 1984; Civeyrel & Simberloff 1996). Predators are usually generalists and, due to their effect on non-target populations, are not recommended for use in future biological control programs (Clarke et al. 1984; Civeyrel & Simberloff

1996).

Research Needed

In order to effectively manage present and future infestations of L. fulica in the

U.S., we need to gain a better basic understanding of the biology of this invasive species. Lissachatina fulica is said to reach sexual maturity in 5-15 mo (Bequaert 1950;

Pawson & Chase 1984; Tomiyama 1993). It would be helpful to narrow this age range, if possible, and to know the size at which snails are able to produce eggs. Also, it would be helpful to be able to estimate the date of introduction when an infestation is found, 19

which might be possible if snail ages could be determined based on shell size. This might give government officials a general idea of how long snails have inhabited a specific area. Lissachatina fulica is estimated to live for 3-5 yr in the field (Tomiyama

1993) and as long as 9 yr in captivity (Bequaert 1950). This potentially means that a snail can be sexually reproductive for as long as 8.5 yr, though some sources say that

L. fulica becomes less fecund with age (Pawson & Chase 1984; Raut & Barker 2002).

Even so, the number of offspring is quite remarkable, because they can produce as many as 442 eggs in a single clutch (Kekauoha 1966). However, because fecundity reports of this species vary depending on the region (van der Meer Mohr 1949; Ghose

1959; Kekauoha 1966a; Upatham et al. 1988), it would be helpful to know the reproduction rates exhibited by the population in Miami, Florida. In addition, because L. fulica infestations have declined in the absence of a biological control agent (Clarke et al. 1984; Civeyrel & Simberloff 1996), it would be interesting to understand if this is related to over-crowded conditions. Finally, although Venette & Larson (2004) list a large number of plant species that L. fulica eats, it would be useful to know what L. fulica prefers to eat. Knowing more about L. fulica will create a good basis for combating the invasive population in South Florida.

20

Figure 1-1. Lissachatina fulica shell. Photo courtesy of Katrina Dickens.

21

A B

C D

E F

Figure 1-2. Shells of five snail species native to Florida and one non-native that may be confused with Lissachatina fulic. A) Orthalicus reses. B) Liguus fasciatus. C) Orthalicus floridensis. D) Drymaeus multilineatus. E) Euglandina rosea. F) Pomacea maculata. Photos courtesy of Katrina Dickens.

22

CHAPTER 2 HOST PLANT SUITABILITY

Introduction

Pests, whether direct crop pests or plant disease vectors, are a major concern in agriculture. For example, it is estimated that $179 million in Florida and $120 billion nation-wide are lost annually to invasive pest species (Pimentel et al. 2005).

Lissachatina fulica is a generalist, with documented consumption of over 158 plant species and 152 total plant genera, including both agricultural and ornamental plants

(Lange 1950; Rao & Singh 2002; Raut & Barker 2002; Sturgeon 1971). The ability of L. fulica to accept so many types of food sources has made it adaptable to environments all over the world, where it has established and become a pest (Raut & Barker 2002).

To combat a new infestation in South Florida, government officials are applying molluscicide treatments. However, the most effective molluscicides are harmful to pets and some homeowners in the area are resistant to their use. As an alternative, or in addition to using chemical treatments, efforts are underway to design traps to catch L. fulica. For these trap designs, it will be useful to know what L. fulica prefers to eat so those plants can be used as baits or bait components. Knowing their preferred food plants would also help to know where to search for L. fulica, making it easier to find and remove them from infested areas. This information could also be used to avoid creating preferred habitats for the establishment of these snails. Homeowners could avoid preferred plants when selecting ornamental plants for planting in residential settings.

While we know L. fulica is capable of having a diverse diet, there is very little known about the food preferences and suitability in this species. Rao and Singh (2002) found 4 out of 20 plants that they tested to be susceptible and preferred. However, the

23

focus of this study is to aid in the management of the L. fulica population in South

Florida. Of the plants mentioned in Lange (1950), Sturgeon (1971), Rao and Singh

(2002), and Raut & Barker (2002), only ten (Table 2-1) reportedly grow well in Miami

(Haynes et al. Undated). Plants likely to be found in Miami landscaping should be tested for host suitability, as this is most applicable to the South Florida population of L. fulica.

Identification of ‘preferred’ host plants can be challenging, and often varies with plant species, age, and region (Raut & Barker 2002). Furthermore, if one plant is preferred but unavailable another can become relatively susceptible, even shrubs and grasses (Raut & Barker 2002). Although plant preference is affected by food availability, there is a good correlation between host plant preference and suitability (Mody et al.

2015; Mulkern 1967). Suitability, the quality of the plant to allow the herbivore to grow, is not affected by food availability and was used in this study as an index of preference. To discover what plants will have the most impact on the Miami L. fulica infestation, identification of plants suitable to sustain growth of young L. fulica was investigated.

Methods

Plant host suitability was tested by comparing the survival and growth rates of juvenile L. fulica while rearing them on one of 21 plants (Table 2-2). Plants selected included some from the 10 plants already identified as preferred host plants growing in this area (Table 2-1). Other plants were selected, mostly from a list of 350 plants described as low-maintenance plants for Miami landscaping (Haynes et al. Undated). In addition, romaine lettuce (Lactuca sativa)), wheat germ-based synthetic insect diet

(gypsy moth diet, BioServ, Frenchtown NJ), and soil only treatments were included as controls. Juvenile snails were chosen for testing their development on different host

24

plants because if plants are not able to promote growth and survival of juveniles, the snail population would not persist.

Snails used in this experiment were F1 generation snails derived from the wild population in Miami-Dade County, Florida. Snails (1-2 d after hatch) were reared ten to a 4 L cage for 70 d. There were five cages of each dietary treatment (21 natural diets and 3 control diets) (Table 2-2). Each of the five cages contained snails originating from one of 5 different clutches. Cages were maintained in The Florida Biological Control

Laboratory (FBCL) quarantine in Gainesville (29°338’06’’N, 82°22’15’’W), Florida, with a photoperiod of 16:8 h L:D. Temperature and humidity ranged from 21-25 °C and 28-

66%, respectively. The calcium needed for shell growth was provided as lawn lime suspended in agar (75 mL 54% pelletized lime, 10 g agar, and 1 L boiling water). Cages were cleaned as needed and 500 mL of potting soil (Metro Mix 930 Sun Gro

Horticulture, Agawam, MA) was provided in each cage. Snails were fed ad libitum on each diet treatment, except for the control that consisted of soil only. All snails were measured for shell height (length) and snail mass every 10 days using a Fisher

Scientific 15-077-958 caliper and a Mettler Toledo, ML1502e balance, respectively.

Growth and survival across treatments were analyzed using R statistical computing software (R Core Team 2014) and the lawstat package (version 2.4.1.tar.gz) and dunn.test package. Because heteroscedasticity was detected with the Levene’s test, the Kruskal-Wallis and Dunn tests were used to test snail height, mass, and survival 70 d after hatch among 11 annual host plants and 2 controls, and among 9 perennial host plant treatments and 2 controls. Soil and Aloe ciliaris were removed from the height and mass statistical analyses, because less than three snails survived to 70 d

25

post treatment. The relationship between height versus mass was tested using all heights and masses, then using heights and masses from annuals and controls and then perennials and controls. These tests and the relationship between mean height versus percent survival for each cage were analyzed with Pearson’s correlation coefficient.

Results

Rearing snails under different dietary treatments, including annual host plant and control treatments, affected snail height (H = 347.96; df = 12; P < 0.01) and mass (H =

346.22; df = 12; P < 0.01) (Fig. 2-1-A). Snails fed Tagetes patula, L. sativa, or synthetic diet had mean heights > 20 mm and masses > 2 g, and were the largest snails among the 22 diet treatments tested that allowed survival (Fig. 2-1). Cosmos bipinnatus, Salvia splendens, Zinna elegans, Leucophyllum frutescens, and Helianthus debilis were somewhat less suitable annuals for snail growth, resulting in snails with mean heights of

10-20 mm and masses of 0.3 - 2 g. Mean heights < 10 mm and masses < 0.3 g were found in snails fed Helianthus annuus, Portulaca oleracea, Centaurea cineraria,

Solenostemon scutellarioides, and Antirrhinum majus.

Perennial host plant treatments significantly affected snail height (H = 245.42; df

= 10; P < 0.01) and mass (H = 247.61; df = 10; P < 0.01) (Fig. 2-1-B). Codiaeum variegatum Linnaeus resulted in the largest snail height (15 mm) of the perennial host plants, but not larger than L. sativa and the synthetic diet. The perennial treatments where snail height ranged 10-15 mm were Lantana camara, Kalanchoe blossfeldiana, and Asclepias tuberosa. Diets of Tradescantia spathacea, Philodendron bipinnatifidum,

Persea americana, Callicarpa americana, and Bougainvillea glabra resulted in snail heights < 10 mm. 26

Overall, there was a very strong positive linear correlation between snail height and mass (r = 0.87). This is also true when comparing height and mass among annuals and controls (r = 0.90) and among perennials and controls (r = 0.89).

Data analysis suggested that a strong positive linear correlation exists between survival and host plant suitability, as judged by snail height (r = 0.89). As with height, survival varied among the different annual plant and control treatments (H = 39.61; df =

13; P < 0.01) and the perennial plant and control treatments (H = 45.47; df = 12; P <

0.01) (Fig. 2-2). The 10 most suitable annual plant and control diet treatments for snail survival were L. sativa, C. cineraria, T. patula, synthetic diet, Z. elegans, P. oleracea, H. annuus, S. splendens, C. bipinnatus, and H. debilis. The second best survival rates occurred in snails fed A. majus, S. scutellarioides, and L. frutescens, which were statistically similar to survival rates from Z. elegans, P. oleracea, H. annuus, S. splendens, and C. bipinnatus, but also statistically similar to the survival rate of zero that was the result from the soil treatment. Of the control and perennial plant treatments, L. sativa, K. blossfeldiana, synthetic diet, L. camara, A. tuberosa, C. variegatum, and T. spathacea were the most suitable for snail survival. The second best survival rates occurred in snails fed C. americana, B. glabra, P. americana, and P. bipinnatifidum, which were statistically similar to survival rates from A. tuberosa, C. variegatum, and T. spathacea, but also statistically similar to the lowest survival rates from A. ciliaris and soil.

Discussion

Host plant preference and suitability are often found to be positively correlated.

Host preference involves herbivore selection and is thought to be principally affected by region, age, availability, and host chemistry (Raut & Barker 2002; An et al. 2007). Host 27

suitability reflects on the herbivore responses after selecting a host plant, and is thought to be most influenced by the quality of plant (nutrition and chemical defenses) which can vary by plant and leaf age (An et al. 2007). With multiple factors influencing host plant preference and suitability, some studies have found the 2 to be independent or negatively correlated (An et al. 2007). Even so, many studies have found that the plants most preferred by grasshoppers were also the most suitable for survival, growth, and reproduction (Mulkern 1967). Similarly, apple cultivars that were less preferred by

Anthonomus pomorum Linnaeus (Coleoptera: Curculionidae) were associated with A. pomorum that had lower mass and later emergence times than A. pomorum found on more preferred cultivars (Mody et al. 2015).

The growth and survival of L. fulica found in this study is generally consistent with other studies. For example, Cosmos bipinnatus, Z. elegans, and K. blossfeldiana were plants previously cited as damaged by L. fulica (Raut & Barker 2002) and they also supported high survival rates and relatively high growth rates in the present study. In general, annual plants were more suitable for growth and survival than were perennials.

It is likely that a perennial would be less suitable because perennials are generally vulnerable to herbivores for longer periods of time, and therefore have evolved stronger defense mechanisms whereas annuals generally allocate more energy towards fast growth than defense against herbivores (Brinker & Frank 1998).

Tagetes patula has been documented in other studies by listing to be a preferred host plant by gastropods. For example, the brown slug, Mariella dussumieri Gray, 1855

(Gastropoda: Ariophantidae), has been reported to eat most parts of this plant (Onkara

Naik et al. 2014). Zachrysia provisoria Pfeiffer (Gastropoda: Pleurodontidae),

28

Bradybaena similaris Férussac (Gastropoda: Bradybaenidae), Deroceras laeve Müller

(Gastropoda: Agriolimacide) and Deroceras reticulatum Müller (Gastropoda:

Agriolimacide) are other molluscs reported to readily consume T. patula (White-Mclean

2012). Lissachatina fulica has previously been observed to cause significant damage to

T. patula, especially to young plants (Sridhar et al. 2012). Tagetes patula reportedly was eaten by L. fulica at the same rate as lettuce (Raut & Ghose 1983). In the same study, the snails ate more marigold when tested against gourd, cabbage, castor, papaya, tomato, lady's finger, and cotton (Raut & Ghose 1983). Tagetes patula is so attractive to

L. fulica that it has been recommended as a trap crop (Raut & Ghose 1983).

Growth and survival were correlated in this study, but not entirely predictable. As reported by Capinera & Rodrigues (2015), a diet of either P. oleracea or S. scutellarioides each resulted in marginal growth (mass) for Leidyula floridana (Leidy)

(Gastropoda: Veronicellidae) and in this study, resulted in marginal growth (height and mass) in L. fulica. Helianthus annuus, a plant reported as damaged by L. fulica (Raut &

Barker 2002), was found in this study to be unsuitable for growth, but suitable for survival. Plants such as P. oleracea, H. annuus, and T. spathacea may not be ideal for growth, but capable of sustaining a snail until it can find better quality food (Capinera &

Rodrigues 2015; Yadav & Singh 2003).

This information can be used to identify what plants are most suitable to sustain a population of L. fulica in South Florida and where populations are most likely to be found. This will also supply future studies with a list of suitable host plants that can be tested against each other in preference choice tests.

29

Table 2-1. List of known preferred or susceptible host plants of Lissachatina fulica that occur in the snail-infested area of Florida. Common Scientific Name Order: Family Reference Name Aloe indica Asparagales: Aloe Raut & Barker (2002) Xanthorrhoeaceae Annona squamosa Magnoliales: Annonaceae Sugar apple Rao & Singh (2002) Cosmos spp. Asterales: Asteraceae Cosmos Raut & Barker (2002) Helianthus annuus Asterales: Asteraceae Sunflower Raut & Barker (2002) Ipomoea pes-caprae Solanales: Railroad vine Raut & Barker (2002) & Convolvulaceae Lange (1950) Kalanchoe Saxifragales: Kalanchoe Raut & Barker (2002) blossfeldiana Crassulaceae Mangifera indica Sapindales: Mango Rao & Singh (2002) Anacardiaceae Portulaca oleracea Caryophyllales: Purslane Lange (1950) Portulacaceae Swietenia mahagoni Sapindales: Meliaceae Mahogany Raut & Barker (2002) Zinnia elegans Asterales: Asteraceae Zinnia Raut & Barker (2002)

Table 2-2. Twenty-four diet treatments fed to juvenile Lissachatina fulica to assess suitability. Common Perennial/ Scientific Name Order: Family Name Annual Antirrhinum majus Lamiales: Plantaginaceae Snapdragon Annual Centaurea cineraria Asterales: Asteraceae Dusty Miller Annual Cosmos bipinnatus Asterales: Asteraceae Cosmos Annual Helianthus annuus Asterales: Asteraceae Sunflower Annual Helianthus debilis Asterales: Asteraceae Beach Annual sunflower Leucophyllum frutescens Lamiales: Texas sage Annual Scrophulariaceae Portulaca oleracea Caryophyllales: Purslane Annual Portulacaceae Salvia splendens Lamiales: Lamiaceae Salvia Annual Solenostemon scutellarioides Lamiales: Lamiaceae Coleus Annual Tagetes patula Asterales: Asteraceae French Annual Marigold Zinnia elegans Asterales: Asteraceae Zinnia Annual Aloe ciliaris Asparagales: Aloe Perennial Xanthorrhoeaceae Asclepias tuberosa Gentianales: Butterfly weed Perennial Apocynaceae Bougainvillea glabra Caryophyllales: Bougainvillea Perennial Nyctaginaceae

30

Table 2-2. Continued Common Perennial/ Scientific Name Order: Family Name Annual Callicarpa americana Lamiales: Lamiaceae Beautyberry Perennial Codiaeum variegatum Malpighiales: Petra croton Perennial Euphorbiaceae Kalanchoe blossfeldiana Saxifragales: Kalanchoe Perennial Crassulaceae Lantana camara Lamiales: Verbenaceae Lantana Perennial Persea americana Laurales: Lauraceae Avocado Perennial cv Choquette Philodendron bipinnatifidum Alismatales: Araceae Philodendron Perennial Tradescantia spathacea Commelinales: Oyster Plant Perennial Commelinaceae Soil Soil Control Synthetic diet Synthetic diet Control Lactuca sativa Asterales: Asteraceae Romaine Control lettuce

31

A

B

Figure 2-1. Mean height and mass of newly hatched Lissachatina fulica after 70 d of feeding on a single diet treatment. A) Annual host plants. B) Perennial host plants. Means topped by the same lowercase letters are not significantly different (P > 0.05; Kruskal-Wallis rank sum test and Dunn’s test). Error bars indicate standard error.

32

A

B

Figure 2-2. Mean percent survival of newly hatched Lissachatina fulica after 70 d of feeding on a single diet treatment. A) Annual host plants. B) Perennial host plants. Means topped by the same lowercase letters are not significantly different (P > 0.05; Kruskal-Wallis rank sum test and Dunn’s test). Error bars indicate standard error.

33

CHAPTER 3 GROWTH AND REPRODUCTION

Introduction

Lissachatina fulica is a major pest in regions where it is established, probably because it has a wide host range and is incredibly prolific. The strong reproductive potential of this generalist feeder has helped this species to establish populations throughout the world (Raut & Barker 2002). However, despite being a nearly worldwide problem for more than 100 years, information on its biology is limited. Growth rates, age of sexual maturity, and fecundity of this species need to be better understood in order to effectively manage and prevent L. fulica infestations.

Snail growth and developmental rates are temperature dependent. Depending on the location, L. fulica populations have been known to survive over a wide range of temperatures, from 0-45 °C (Raut & Barker 2002; Mead 1979; Raut & Ghose 1984).

Optimal temperatures for this species, when the snail is active and growing, have been reported to be 22-28 °C (Raut & Barker 2002; Mead 1979; Raut & Ghose 1984).

However, the maximum development rate within that temperature range is unknown.

Growth studies on L. fulica show shell heights ranging from 35-63 mm after 60 days of age and ranging from 83-104 mm after 150 days of age (Bequaert 1950; Kondo

1964; Pawson & Chase 1984). Because new population centers continue to be discovered in South Florida, regulatory personnel are interested in distinguishing new infestations from older infestations as they are found. An estimate of a snail’s age based on shell size could possibly be used to give an estimate of how long snails have inhabited a specific site, because snails become larger over time. Predictability likely will be reduced in later months of life due to the known tendency of snail growth rate to

34

taper off as snails age. Nevertheless, establishing the age of younger infestations has value in assessing dispersiveness.

The validity of existing literature on L. fulica is uncertain. For example, sample sizes were never mentioned by Bequaert (1950) or Pawson & Chase (1984), and Kondo

(1964) used a sample size of only five snails. The diversity of results reported for growth rates in L. fulica could be due to inadequate sample size, dissimilar rearing and experimental designs, or genetic differences found in snails in different geographic regions.

An accurate method for visually identifying reproductive maturity in this species is not known. However, the thickness of the peristome has been suggested by Tomiyama

(1993) to be a useful tool to assess sexual maturity. Increases in weight and shell length are said to slow as the snail becomes mature, while the snail allocates more growth to the thickness of the shell; thus, the peristome grows thicker (Pawson & Chase 1984;

Tomiyama 1993). Tomiyama’s compared peristome thickness and maturity based on the assessment of 52 field-collected snails with peristomes less than 0.5 mm thick and an additional 52 larger snails that had peristomes greater than 0.8 mm thick. Snails with thick peristomes had oocytes in ovotestes and large protein glands, which are required for egg laying, whereas snails with thin peristomes had no oocytes and smaller protein glands (Tomiyama 1993). It would be very useful to know if the peristome thickens around the time of the first egg-laying event, as this would help in determining snail and snail infestation age.

Reproductive success can be measured by determining the age at which egg laying begins, the number of viable eggs laid per clutch, and the number of viable eggs

35

laid per year. Lissachatina fulica snails are reported to be hermaphroditic, but the male reproductive system develops first and the male reproductive system is capable of copulation before the female reproductive system develops completely (Mead 1949;

Tomiyama 1993). Reports in the literature suggest that oviposition commences at 5-15 months of age (Bequaert 1950; Ghose 1959; Civeyrel & Simberloff 1996; Raut & Barker

2002; Venette & Larson 2004). Egg production also varies, with reports of 160-1,817 eggs produced annually, and with 10-500 eggs per clutch (Kekauoha 1966; Raut &

Barker 2002; Venette & Larson 2004; CLEAPSS 2006). A maximum of 442 eggs laid per clutch was found in a study that monitored egg laying in snails isolated for seven months after copulation, and found fecundity to decrease over time, possibly due to the depletion of the sperm supply (Kekauoha 1966). Because snails are not often isolated in the field, sperm depletion is unlikely, so higher rates of reproduction might be expected.

Thus, more study is needed to narrow the estimate of when the minimum and average time to maturity occurs, and to estimate egg production rates.

Because L. fulica is hermaphroditic, reproduction without mating with another snail would greatly increase the ability of the species to invade and spread to new areas. However, studies report both the ability and inability to reproduce without mating with another snail (van der Meer Mohr 1949; Ghose 1959; Kondo 1964; CLEAPSS

2006). Sources that have observed reproduction without mating have assumed the process to be self-fertilization (van der Meer Mohr 1949; Ghose 1959).

The purpose of this study was to determine critical aspects of L. fulica biology that affect regulatory activities. Specific objectives were to determine:

1. temperature-dependent development.

36

2. growth rates at various densities and diet regimes.

3. the age and size at first oviposition.

4. if peristome thickness is a reliable indicator of sexual maturity.

5. short-term and long-term reproductive rates.

6. whether this species will reproduce without mating.

Methods

Rearing Methods

The experimental animals were F1 generation snails derived from the wild population in Miami-Dade County, Florida and maintained in FDACS Florida Biological

Control Laboratory (FBCL) quarantine in Gainesville, Florida. They were fed a synthetic gypsy moth diet or a diet of organic romaine lettuce. Lettuce was used as a representative of an optimal “natural” diet (Capinera 2013; Estebenet & Cazzaniga

1992). The synthetic diet was a wheat germ-based diet used for gypsy moth rearing

(Bio-Serv Flemington, NJ) and was selected as a suitable but sub-optimal diet, as it will allow for the growth and reproduction of the species, but has been previously observed

(unpublished) to produce smaller snails with higher rates of mortality. Calcium needed for shell growth was provided in each experiment as lawn lime suspended in agar (75 mL 54% pelletized lime, 10 g agar, and 1 L boiling water). All cages contained potting soil (Metro Mix 930, Sun Gro Horticulture, Agawam, MA), and the amount of soil provided in a cage increased as the snail grew; this ensured proper habitat for egg laying but still allowed for easy access to the snail when it was small.

37

Neonate Growth

Temperature-dependent growth was studied using young snails. We divided

1,200 neonates, 9-10 d in age, into different temperature regimes, and reared them until they attained 15 mm in height. Two hundred snails were divided into twenty 525 mL clear plastic cups with unvented lids. They were provided with lettuce and calcium ad libitum, and checked for moisture, snail mortality, and growth every 2-3 d. This was repeated for 6 different constant temperatures (10, 15, 20, 25, 30, and 32.5 °C).

Constant temperatures were maintained using a Percival Scientific 1-30BLL incubator.

The number of days required for each snail to reach 15 mm was recorded and used to calculate the development rate (days-1) at each temperature. The height of each snail was measured using a Fisher Scientific 15-077-958 caliper. Percent survival was calculated at the end of 70 d. If snails had not attained 15 mm in height by 70 d after hatching, they were culled and removed from development analysis, but used in the survival analysis.

Data were analyzed using R statistical computing software (R Core Team 2014), the lawstat package (version 2.4.1.tar.gz), and the dunn.test package. Growth rate and percent survival at each temperature treatment were compared using the Kruskal–

Wallis and Dunn’s tests because heteroscedasticity was detected with the Levene’s test.

Maturity and Reproduction

In order to evaluate the age of sexual maturity, reproductive rates, and to determine if mating with another snail is required for egg production, 4 treatments were compared: both solitary and paired snails when fed synthetic diet (a sub-optimal diet) or a diet of organic romaine lettuce (a more suitable diet). Snails were monitored for egg 38

production for up to 930 d. Previous experience had demonstrated that these snails grow well on romaine lettuce, and that the synthetic diet would provide slower rate of growth than lettuce. The sub-optimal synthetic diet would show how diet affected the onset of reproduction.

This experiment was conducted using 4-6 week-old neonates with heights of 15-

20 mm, as these were old enough to avoid damaging their soft shells during handling; this reduced early mortality. Also, these snails were young enough to assure that snails were unmated. Cages consisted of 4 L plastic containers and were initially filled with

300 mL soil, then cleaned and soil replaced each month thereafter using 800 mL of soil.

The cages were maintained in a photoperiod of 16:8 h L:D; temperature and humidity ranged from 21-25 °C and 28-66%, respectively (Raut & Ghose 1984; Raut & Barker

2002). Snails were provided with a constant surplus of food and calcium. This experiment was conducted with 3 replicates, with replicates based on different starting dates. Each replicate consisted of 20 cages containing snails that hatched on the same date (± 24 h). A total of 360 snails were used in these studies.

Height and mass were measured every 30 d for 240 d. Shell heights were measured using a Fisher Scientific 15-077-958 caliper, and the mass of each snail was determined using a Mettler Toledo ML1502e balance. In some cases, the spire of the snail shell was damaged, making the height of the shell artificially smaller; these were removed from analyses.

A two-way ANOVA was used to determine the effects of diet and density on height and mass 240 d after hatch. Height and mass were analyzed within each diet and density treatment using the two-tailed Student’s t-test. When the unequal variance

39

assumption was violated according to the Levene’s test, the nonparametric Mann-

Whitney rank sum test was performed. The Mann-Whitney rank sum test was used to compare mass of snails fed synthetic diet between paired and solitary treatments. To further test the quality between diets, a two-way ANOVA was used to test if diet and density affected mean percent survival.

Pattern of growth

The pattern of growth (height) for snails from the aforementioned reproduction studies was analyzed with polynomial regressions.

After the growth studies were terminated at 240 days, additional measurements were continued on a subset of the snails. Snails from the first replication were monitored for long-term growth for 600 d (n = 60 snails; 10 cages each from the paired and solitary treatments, and the romaine and synthetic diets).

First oviposition

When snails in the aforementioned growth studies first produced eggs, they were measured for height. After 240 d, many snails in replicate 3 had not deposited eggs, which was in strong contrast to the previous 2 replicates. For this replicate, monitoring was extended until 360 days after hatch. However, only data from the first 300 days were used in the reproductive rate analysis of these snails.

A two-way ANOVA was used to determine the effects of diet and density on age and height of the adults at the first egg laying event. Student’s t-test was used to compare mean age and height of the adults at the first egg laying event within each diet treatment and within each density treatment.

Peristome thickness was measured using a Marathon CO 030150 caliper every

30 d to determine whether the peristome thickened around the time eggs were laid, as

40

this potentially could be used as an indicator of maturity. Peristome thickness was plotted against the time that the measurements were taken relative to the first oviposition. Negative values were used to indicate the time period before first oviposition, and positive values indicate the time period after first oviposition. A repeated measures ANOVA was used to analyze peristome thicknesses found in snails before and after first oviposition.

Sexual reproduction

As egg laying commenced, the numbers of eggs in each clutch were counted.

Clutches laid between hatch and 241 d by paired snails were randomly chosen to test egg viability, including 86 clutches from snails fed synthetic diet and 91 clutches from snails fed lettuce. For determination of egg viability, the eggs were buried under a thin layer of moist soil in a six-inch diameter petri dish. The number of live hatchlings that emerged was counted after 3 weeks, and clutch viability was calculated. Cages were checked for eggs every 3 to 5 days. The total number of eggs found in each cage (n =

59 cages for lettuce and 58 for synthetic diet paired treatments), between hatch and 241 d, was divided by the number of snails in each cage to determine the mean number of eggs produced per snail. The analysis does not include eggs produced after 300 d by late-developing snails in replicate 3. In order to address the unequal variance assumption violation detected by the Levene’s test, the Mann-Whitney rank sum test was used to test the significance of differences found between diet treatments in the number of eggs laid per snail, the percent viability of eggs laid, and clutch size.

As with growth, egg production monitoring continued on a subset of snails from the first replication. However, egg production was monitored longer than growth, for a total of 930 d after hatch. The data reported only includes surviving snails from the 41

lettuce treatment (n = 16 snails) because only one snail survived to 930 d in the synthetic diet treatment. All eggs that were laid were counted and egg viability was estimated as stated above (n = 128 clutches). The number of eggs found in each of the

8 paired-snail cages were totaled every 30 d after hatching and then divided by the number of snails in each cage to find the number of eggs laid per month per snail.

Snails that did not produce eggs that month were included in the analysis.

Self-fertilization

In addition to the eggs produced by snails in the aforementioned reproduction experiment, we were able to access the egg production data from additional solitary and paired snails cultured. Added to the above experiment were egg production data from

134 solitary snails (total n = 246 including the previous replications), of which 65 were fed lettuce and 69 fed synthetic diet, and data from 132 snails raised in pairs (total n =

366 including the previous replications). These snails were reared similarly, but 116 were kept in a greenhouse with ambient lighting and humidity, with temperature ranging from 23-26 °C. After some died or were culled at 240 d, 143 solitary and 92 paired snails were still being monitored through 540 d after hatch. Dull, soft, non-viable eggs were frequently found in solitary snail cages. Eggs clutches with only these dull eggs were not considered in fecundity data. Otherwise, clutches (including 8 clutches from snails fed synthetic diet and 7 clutches from snails fed lettuce) were tested for viability following the same methods as used for paired treatments.

A 2-way ANOVA was used to test if diet and density affected the total number of eggs laid per snail, the percent viability of eggs laid, and clutch size from cages monitored for 540 d. The Student’s t-test was used to test the significance of differences

42

found between diet and density treatments. The Mann-Whitney rank sum test was used when heteroscedasticity was detected with the Levene’s test.

Results

Growth

The development rate of the young snails varied across the temperatures tested, with no growth at 10 and 15 °C, growth at 20, 25, 30, and 32.5 °C, and a maximum growth rate of 0.037 day-1 at 30 °C (Fig. 3-1A). At 10 °C, all snails died within a week.

Many snails survived to 70 d at 15 °C, but snail growth was not detected. Among the temperatures where snails grew, development rate differed by temperature (H = 460.15; df = 3; P < 0.01). For the 32.5 °C temperature treatment, 97 snails were culled after 60 d, before they attained a height of 15 mm, but probably eventually would have attained

15 mm at a later date, so the true average development rate is somewhat lower than what is reported in Figure 3-1A.

Percent survival also was affected by temperature (H = 78.69; df = 5; P < 0.01)

(Fig. 3-1B). Snails in the 15 °C and 32.5 °C temperature treatments experienced some mortality, whereas over 90% of snails survived when reared at 20, 25, and 30 °C.

Growth was affected by diet and density after 240 d, while survival was not. From the average of solitary and paired snails, the lettuce diet produced snails with greater height (F = 200.06; df = 1, 341; P < 0.01) and mass (F = 231.39; df = 1, 346; P < 0.01) as compared to the synthetic diet treatment (Table 3-1). In comparing the solitary snail treatment with the paired snail treatment, regardless of diet treatment, the solitary snail treatment produced snails with greater mean height (F = 181.81; df = 1, 341; P < 0.01) and mass (F = 127.25; df = 1, 346; P < 0.01) (Table 3-1). No interaction effect was found on height (F = 2.57; df = 1, 341; P = 0.11), but an interaction between diet and 43

density was noted with respect to mass (F = 5.13; df = 1, 346; P = 0.02). Diet (F = 2.52; df = 1, 8; P = 0.15) and density (F = 0.33; df = 1, 8; P = 0.58) did not significantly affect survival. Nor was an interaction effect found (F = 0.91; df = 1, 8; P = 0.37).

The growth plots of paired and solitary snails produced negative exponential curves (Fig. 3-2). According to the height growth regression for paired snails, if a snail is

20 mm in height, it is likely to be 41 d old. In the same way, 34 and 85 mm snails would be aged 60 and 150 d, respectively. The trend that height was greater among solitary snails than among paired snails further supports a density effect on growth. Data from long-term growth studies show that snails, on average, did not continue to grow after

240 d (Fig. 3-2).

First Oviposition

The minimum time from hatch to egg laying was 124 d (Table 3-2). The number of days from hatch until the first egg laying event did not differ between diet treatments

(F = 0.01; df = 1, 117; P = 0.93), nor between paired and single treatments (F = 0.26; df

= 1, 117; P = 0.61), nor was there any interaction effect (F = 0.06; df = 1, 117; P = 0.81)

(Table 3-2). Diet affected the height of snails when oviposition commenced (F = 42.53; df = 1, 228; P < 0.01); however, density did not (F = 1.28; df = 1, 228; P = 0.26) and no interaction was found between diet and density (F = 0.235; df = 1, 228; P = 0.63) (Table

3-2). Eight solitary virgin snails laid large clutches, 5 of which proved to be viable, attesting that oviposition without mating with another snail is possible. Snails fed lettuce were larger at the first oviposition than were snails fed synthetic diet, but only among paired snails (t = 6.18; df = 222; P < 0.01), not among solitary snails (t = 2.27; df = 6; P

= 0.06).

44

Peristome thickness increased as snails oviposited for the first time (F = 1021; df

= 1, 932; P < 0.01) (Fig. 3-3). Despite the significant relationship found between peristome thickness and the onset of oviposition, outliers were present as several peristome thicknesses were measured < 0.1 mm after eggs were first laid.

Sexual Reproduction

The average total number of eggs laid per snail over the course of the experiment from paired snails was higher from the lettuce treatment than the synthetic diet treatment (U = 2,168; P = 0.01) (Table 3-3). Also, paired snails fed lettuce laid eggs with higher viability (U = 5,059; P < 0.01) and laid larger clutches (U = 35611.5; P <

0.01) than paired snails fed synthetic diet (Table 3-3).

Egg laying continued while monitoring long-term paired snails and the mean number of eggs laid per month per snail and percent egg viability are shown in Figure 3-

4.

Self-Fertilization

In the original experiment with 120 solitary snails, 116 snails survived and were monitored from hatch to between 240-300 d. Of these, 8 solitary virgin snails (7%) laid 1 clutch each. Of the total 246 solitary snails monitored for 240 d, 7 snails (3%) laid 1 clutch each (Table 3-4). When eggs from one clutch laid by a virgin snail were reared to adulthood, they produced viable offspring.

Six of the 143 solitary snails (4%) monitored for 540 d laid eggs that appeared viable. One snail laid 5 egg clutches, while the remainder laid just 1 each. The actual viabilities of the eggs laid ranged from 0 to 98% per clutch (Table 3-4). The number eggs laid per snail (F = 8.37; df = 1, 185; P < 0.01) (F = 280.16; df = 1, 185; P < 0.01), egg viability (F = 13.74; df = 1, 468; P < 0.01) (F = 12.81; df = 1, 468; P < 0.01), and

45

clutch size (F = 40.12; df = 1, 860; P < 0.01) (F = 47.43; df = 1, 860; P < 0.01) were affected by diet and density, respectively (Table 3-5). The interaction between diet and density was significant for eggs per snail (F = 19.48; df = 1, 185; P < 0.01), but not egg viability (F = 0.91; df = 1, 468; P = 0.34) or clutch size (F = 3.13; df = 1, 860; P = 0.08).

There were no differences between diets in the number of eggs laid per snail among paired (U = 335; P = 0.11) or solitary snails (U = 2745; P = 0.28). Paired snails laid more eggs per snail than solitary snails fed lettuce (U = 1779.5; P < 0.01) and fed synthetic diet (U = 1467; P < 0.01). Egg production without mating also resulted in eggs with lower percent viability (U = 2036; P = 0.01) (U = 1934.5; P = 0.02) and smaller clutch size (U = 5018.5; P < 0.01) (U = 4507.5; P < 0.01) in lettuce and synthetic diet treatments, respectively.

Discussion

The temperature-dependent development rates obtained in these studies provide estimates of appropriate rearing temperatures for L. fulica and the physiological time needed for a snail to develop to 15 mm in height. The optimal temperature for growth in this study was 30 °C, which is higher than the optimal temperature range of 22-28 °C previously reported by Raut & Ghose (1984). Mead (1979) suggested that 26 °C was the optimal temperature for feeding and, therefore, growth. Also, temperatures over 28

°C have been reported to initiate aestivation (Raut & Ghose 1984). These temperatures likely are specific for initiation of aestivation by older snails. According to our observations, young snails less than 25 mm do not aestivate.

Other studies previously demonstrated that diet can significantly affect snail height (Upatham et al. 1988) and mass (Egonmwan 2007; Ejidike 2007). Upatham et al.

46

(1988) found that a combination of a synthetic diet and lettuce promoted increases in both height and mass as compared to lettuce alone. Height was less affected. This may indicate that lettuce is lacking in a beneficial nutrient or that snails benefit from dietary changes. The synthetic diet used in their study may have provided a missing nutrient, whereas our wheat germ-based synthetic diet did not. Upatham et al. (1988) never tested their synthetic diet on its own to see if it alone could support growth. They also did not specify the type of lettuce used for rearing. Iceberg lettuce, for example would likely be less nutritious than romaine.

The growth and reproduction results of this study support the hypothesis that the synthetic diet was a suboptimal diet. By including snails fed a suboptimal diet in this study, we obtained a better representation of the variation in growth and reproduction found in wild snail populations. The growth equation developed in this study does well at estimating juvenile age based on size. However, as a snail becomes larger it is more difficult to estimate its age due to a slower growth rate. Thus, size is likely a reliable indicator of age only in the range of heights less than 95 mm.

The patterns of snail growth observed herein are similar to other reports

(Bequaert 1950; Kondo 1964; Pawson & Chase 1984), though Pawson & Chase (1984) reported higher juvenile height and Kondo (1964) reports higher adult height. The observed reduction in growth rates after 150 d is consistent with other species and with the theory that growth slows around the time maturity is reached (Bequaert 1950;

Kondo 1964; Pawson & Chase 1984; Tomiyama 1992, 1993). However, as also reported by Tomiyama (1996), growth did not stop once snails became sexually mature.

47

It is not certain why solitary snails grew significantly larger than the paired snails, though density generally affects animal growth in this manner (Dupont-Nivet et al. 2000;

Lasaridou-Dimitriadou et al. 1998; Mangal et al. 2010). Possibly, there was a cage effect and the paired snails were crowded. Alternatively, there may have been competition for food, though food availability was always adequate. Chemical exudates are commonly postulated to inhibit growth (Mangal et al. 2010; Garr et al. 2011).

The earliest initiation of egg laying found in the present study is about a month earlier than some other studies (Bequaert 1950; Civeyrel & Simberloff 1996; Raut &

Barker 2002). It is unknown why snails in replicate 3 took longer to lay eggs than the other replications. All replications were kept in the same windowless room, under the same environmental conditions, but during different times of the year. It is possible the snails perceived the change in seasons, which affected their development. Replication 3 ran May 4th through December 30th. The other two replicates were conducted January

23rd through September 20th and August 17th through April 14th. Alternatively, the differences may simply indicate a natural variation in development often found in this species (Ghose 1959; Raut & Barker 2002; Venette & Larson 2004).

Peristome thickness may suggest full reproductive maturity, as originally reported by Tomiyama (1993). However, peristomes were much thinner in the present study than those observed by Tomiyama (1993). Perhaps this reflects a difference between snails reared in a laboratory and field collected snails. The presence of outliers in this study demonstrates that the peristome does not consistently thicken as the snail develops into a mature state. Thus, peristome thickness is not a definitive measure of maturity.

48

On average, egg production rates were much higher in these studies than many others (Kekauoha 1966; Civeyrel & Simberloff 1996; Raut & Barker 2002; Venette &

Larson 2004). It is possible that their reported egg production only included viable eggs.

Viability was lower on average than some studies (van der Meer Mohr 1949; Kekauoha

1966). In this case, with an average of 55% viability rate, snails in the long-term study laid a mean of 2,358 viable eggs during their first year of egg production. This value is still much higher than previously reported. The higher egg production seen in our study is a result of more eggs being laid per clutch than in other studies (van der Meer Mohr

1949; Ghose 1959; Kekauoha 1966; Raut & Barker 2002; Venette & Larson 2004).

Reductions in production of viable eggs a year after egg laying began are consistent with the literature (Kekauoha 1966; Pawson & Chase 1984; Raut & Barker

2002). However, it is important to note that egg production continued throughout the course of this experiment. The reproductive potential of this species is an important aspect when considering eradication. Snails that each produce more than 100 eggs per month even after two years, as in this study, present considerable obstacles to population elimination.

Oviposition without mating with another snail was shown to occur, but only rarely and with lower viability rates and smaller clutch sizes. In some cases, egg production seemed to be delayed in virgin snails, occurring after 240 days. Thus, apparent self- fertilization may occur more often than is perceived based on short-term studies, and perhaps as many as 3% of solitary snails can produce viable eggs.

The results of this study have shed light on some aspects of L. fulica basic biology that have considerable implications for eradication efforts. Firstly, the optimal

49

conditions for growth have been identified, which can give us an indication of when to expect sharp increases in snail activity in the field, and it provides researchers with the information they need to guide future studies in the laboratory. Secondly, predicting infestation age may be difficult based strictly on juvenile size and age, and these estimates may be altered by ecological factors, such as density. Also, sexual maturity cannot be precisely determined without dissection, because peristome thickness was slightly variable. Additionally, high potential reproductive rates accentuate the speed at which control efforts must address new infestations. Lastly, the ability for the rare snail to oviposit without mating means that every single snail must be removed, because even one can start another population.

50

Table 3-1. Means (± SD), medians, ranges, and sample sizes (number of snails, n) for height and mass of Lissachatina fulica when fed lettuce or synthetic diet and reared alone or in pairs for 240 d after hatch. Density Diet Height (mm) Mass (g) Mean Range n Mean Range n Paired Lettuce 107 ± 8aB 75-121 117 139 ± 32aB 53-264 118 Synthetic diet 96 ± 9bB 73-133 113 92 ± 28bB 43-170 116

Solitary Lettuce 121 ± 8aA 103-134 60 190 ± 41aA 115-278 60 Synthetic diet 107 ± 9bA 90-121 55 125 ± 35bA 76-192 56 Means in a column, within each density, followed by the same lowercase letters are not significantly different (P > 0.05; two-tailed Student’s t-test test). Means in a column, within each diet, followed by the same uppercase letters are not significantly different (P > 0.05; two-tailed Student’s t-test test).

Table 3-2. Means (± SD), medians, ranges, sample sizes (number of cages) for the number of days from hatching for Lissachatina fulica to lay eggs among diet and density treatments. Means (± SD), medians, ranges, sample sizes (number of snails) for the height of Lissachatina fulica at the first egg laying event. Only 8 solitary virgin snails laid large, viable looking clutches. Density Diet Pre-reproductive period (d) Height (mm) Mean Range n (cages) Mean Range n (snails) Paired Lettuce 194 ± 57aA 127-339 56 102 ± 11aA 78-124 112 Synthetic diet 195 ± 56aA 124-339 57 92 ± 12bA 73-122 112

Solitary Lettuce 201 ± 59aA 164-314 6 107 ± 8aA 100-121 6 Synthetic diet 214 ± 49aA 179-249 2 93 ± 5aA 90-97 2 Means in a column, within each density, followed by the same lowercase letters are not significantly different (P > 0.05; two-tailed Student's t-test test). Means in a column, within each diet, followed by the same uppercase letters are not significantly different (P > 0.05; two-tailed Student's t-test test).

51

Table 3-3. Means (± SD), medians, ranges, and sample sizes (n) for the total number of eggs laid per snail, percent egg viability, and clutch sizes among diet treatments for eggs laid before 241 d for paired Lissachatina fulica. Diet Eggs/snail Viability (%) Eggs/clutch Meana Range n Meana Range n Meana Range n (cages) (clutches) (clutches) Lettuce 738 ± 526a 0-1863 59 68 ± 28a 0-100 91 322 ± 151a 13-1033 264 Synthetic 486 ± 294b 0-1034 58 52 ± 33b 0-99 86 261 ± 91b 28-584 209 diet aMeans in a column followed by the same lowercase letters are not significantly different (P > 0.05; Mann-Whitney rank sum test).

Table 3-4. Total eggs produced per snail, mean (± SD) egg viability, and mean (± SD) clutch size for egg produced 0-240 d and 0-540 d for solitary and paired Lissachatina fulica regardless of diet. Duration Treatment Snails Cages Cages with Eggs/snail Viability (%) Eggs/clutch (days) monitored Monitored eggs (%) 240 Paired 366 183 67 843 52 ± 35 330 ± 160 Solitary 246 246 3 1.24 25 ± 27 44 ± 31

540 Paired 92 46 98 4698 49 ± 36 332 ± 162 Solitary 143 143 4 10.18 27 ± 31 132 ± 136

52

Table 3-5. Mean (± SD) eggs laid per snail, egg viability, and clutch size for eggs produced 0-540 d for solitary and paired Lissachatina fulica for each diet. Density Diet Eggs/snaila Viability (%)b Eggs/clutchb Paired Lettuce 5816 ± 4141aA 53 ± 36aA 354 ± 164aA Synthetic diet 3367 ± 2090aA 42 ± 35bA 289 ± 151bA

Solitary Lettuce 2 ± 8aB 19 ± 32aB 75 ± 122aB Synthetic diet 21 ± 144aB 22 ± 23aB 121 ± 112aB aMeans in a column, within each density, followed by the same lowercase letters are not significantly different (P > 0.05; Mann-Whitney rank sum test). Means in a column, within each diet, followed by the same uppercase letters are not significantly different (P > 0.05; Mann-Whitney rank sum test). bMeans in a column, within each density, followed by the same lowercase letters are not significantly different (P > 0.05; two-tailed Student's t-test test). Means in a column, within each diet, followed by the same uppercase letters are not significantly different (P > 0.05; Mann-Whitney rank sum test).

53

A

B

Figure 3-1. The effects of rearing Lissachatina fulica at 6 constant temperatures A) Mean (± SD) rate to develop to 15 mm in height. B) Mean (±SD) percent survival. Means topped by the same letter are not significantly different (P > 0.05; Kruskal-Wallis rank sum test and Dunn’s test). Error bars indicate standard deviation.

54

A

B

Figure 3-2. Growth pattern of Lissachatina fulica reared on a diet of lettuce or synthetic diet. A polynomial regression was fitted on individual measurements for height. A) Paired snails. B) Solitary snails.

55

Figure 3-3. Individual measurements of Lissachatina fulica peristome thickness taken before (-) and after (+) first oviposition. First oviposition occurred on day 0.

Figure 3-4. The mean total Lissachatina fulica eggs laid per snail over time, and the mean percent of eggs that were viable. Error bars indicate standard error.

56

CHAPTER 4 EFFECTS OF DENSITY AND FOOD DEPRIVATION ON GROWTH AND SURVIVAL

Introduction

Although importation of L. fulica is outlawed in the United States and many other countries where it can easily become a major agricultural and urban pest, L. fulica has been reared successfully as a human food source in Africa and Europe (Raut & Barker

2002). It also is sold by several companies in the United Kingdom as a pet or for educational purposes (CLEAPSS 2006). The cold climate of northern Europe reduces the risk of accidental establishment of a snail population, but it is a hazard in warm weather climates such as Florida. Population density commonly affects snail biology and population dynamics (Conner et al. 2008; Mangal et al. 2010) and is an important ecological factor to consider in any pest control program. It is from the L. fulica food industry that information is available on rearing captive populations of these snails. For example, some snails are reared in arenas at 975-1610 g/m2 (Thompson & Cheney

1996) or 0.26-0.33 snails/L (Upatham et al. 1988). Under such conditions, about 30 percent of the eggs laid will mature into adults within five and a half months after hatching (Upatham et al. 1988).

High-density rearing in the laboratory has been shown to affect growth and reproduction in many snail species. Survival rates, growth rates (shell diameter and weight), egg viability, and eggs produced per clutch have all been shown to decrease in

Helix aspersa Müller, (Stylommatophora: Helicidae) reared under high-density conditions compared to a low-density environment (Lasaridou-Dimitriadou et al. 1998).

Helix aspersa reared in low density treatments have also resulted in the snails consuming more food, growing larger, and reaching maturity in higher numbers than

57

snails in high density treatments (Dupont-Nivet et al. 2000). Density also negatively affected growth, survival, and the number of eggs laid per snail in Biomphalaria alexandrina (Ehrenberg), (Hygrophila: Planorbidae) (Mangal et al. 2010). But in other cases, a more complicated relationship exists between density and growth, with some levels of density actually increasing growth rates. Archachatina marginata (Swainson),

(Stylommatophora: Achatinidae) reared at the highest of 4 treatment densities experienced the lowest weight gain and shell growth, but snails in the third highest density gained the most weight and had the greatest shell growth of all density treatments (Ademolu et al. 2006). There are also species that are especially sensitive to density. Pomacea paludosa (Say), (Gastropoda: Ampullariidae) juveniles experienced decreased shell growth beginning at the second lowest density treatment used by

Conner et al. (2008) and thus were determined to need large habitat areas for the species to thrive. Clearly, the effects of density can vary from species to species.

High population density may also affect more than snail growth and reproduction.

Cannibalism may be another outcome of high density populations, especially when resources such as space and food are limited (Capinera 2008; Fox 1975; Polis 1981).

Instances of cannibalism due to overcrowding are known in many species in the animal kingdom and can play an important role in population regulation (Capinera 2008; Fox

1975; Polis 1981). Lissachatina fulica has been documented preying on veronicellid slugs in Hawaii (Meyer et al. 2009), and observations of cannibalism among L. fulica have been made in the Florida Biological Control Laboratory quarantine in Gainesville,

Florida, in 2012, when L. fulica were observed consuming the flesh of dead snails, but it is not known if the dead snails died due to predation or other causes, such as disease.

58

The purpose of this study was to determine the effects of population density on L. fulica biology. Specific objectives include:

1. To determine the effects of density on growth, mortality, and egg production.

2. To determine whether L. fulica is capable of cannibalism.

Methods

Density Effects on Growth and Reproduction

The experimental animals were F1 generation snails derived from the wild population in Miami-Dade County, Florida, and maintained in The Florida Department of

Agriculture and Consumer Services’ Florida Biological Control Laboratory containment facility in Gainesville (29°338’06’’N, 82°22’15’’W), Florida. Culture containers were 16 L in volume, measuring 25 x 38 x 17 cm (L x W x H), initially filled with 800 mL of soil

(Metro Mix 930 Sun Gro Horticulture, Agawam, MA), and were cleaned and soil replaced monthly. At the second cleaning, the amount of fresh soil was increased to

2,000 mL. The amount of soil was increased as the snail grew to ensure a proper habitat for egg laying, but still allowed for easy access to the snail when it was small.

Holes were drilled in the bottom of the cages to allow for the drainage of liquid waste, and solid waste was removed every 3 days. Snails were maintained in a greenhouse with ambient lighting and humidity, and with temperature ranging from 23-26 °C.

Adequate supplies of organic romaine lettuce, a wheat germ-based synthetic insect diet, and lawn lime suspended in agar (as a calcium source) were provided every 3 to 5 days. The synthetic diet was a gypsy moth diet (BioServ, Flemington, NJ) and was added to ensure an adequate and diverse food supply (Upatham et al. 1988). The lime- agar suspension was made by adding 75 mL of 54% pelletized lime to 10 g agar mixed in 1 L boiling water.

59

The effects of density on mortality, growth, and egg production were evaluated.

There were 3 treatment densities: 5, 15, and 35 snails per cage or 0.29, 0.88, or 2.06 snails/L. Snails 4-7 weeks old, with heights between 15 and 25 mm, were randomly assigned to one of the 3 treatments. This experiment was repeated in three trials with three containers of each density class for a total sample size of 9 cages. Within each trial, snails with similar hatch dates (±1 day) were used.

Height and mass of each snail were measured every 30 d, using a Fisher

Scientific 15-077-958 caliper and a Mettler Toledo, ML1502e balance, respectively.

Growth data for this experiment was monitored until 240 d after hatch. However, additional data were collected from 5 cages of each treatment to monitor snail growth for 300 d. Egg production was monitored by counting the number of clutches laid per cage for 60 d after egg laying began for each cage. The number of eggs produced was estimated, based on actual counts from randomly selected clutches throughout the experiment, resulting in n = 89, 110, and 120 clutches with actual egg number determinations for the low, medium and high density treatments, respectively. The numbers of eggs per clutch were averaged (mean) for each cage and then multiplied by the total clutches laid for each cage to get the estimated egg production per cage. Half of the eggs from counted clutches were allowed to hatch to determine viability rates.

Snail height and mass, mortality, the total number of clutches, and estimated number of eggs produced were analyzed across density treatments using R statistical computing software (R Core Team 2014), the lawstat package (version 2.4.1.tar.gz), and the dunn.test package. When unequal variances were identified using the Levene’s test, the nonparametric Kruskal-Wallis rank sum test was performed. Snail height,

60

mass, and mortality at 240 days after hatch were tested across treatments using the

Kruskal-Wallis rank sum test and Dunn’s test. Density effects were also analyzed using an analysis of variance (ANOVA) and Tukey's HSD to compare the number of days from hatching to the first eggs laid, the total estimated eggs laid, and percent egg viability, for each treatment. The data for the number of eggs laid per clutch was analyzed across treatments using the Kruskal-Wallis rank sum test and Dunn’s test.

Food Deprivation Effects on Cannibalism

Lissachatina fulica were also tested for cannibalistic activity under conditions of different dietary availability. The 4 dietary regimes were:

 No Food - receiving no food or calcium,

 Calcium – receiving calcium supplement (lime suspended in agar), but no other

food,

 Food – receiving food (lettuce plus synthetic diet), but no calcium supplements,

and

 Food and calcium – receiving a complete diet

For each dietary regime, there were three age classes (based on snail height):

 Juveniles (20-30 mm),

 Adults (50-75 mm), and

 A mixture of the two snail sizes (5 juveniles and 5 adults).

There were 10 snails in each of the 12 diet and age treatments, and each was repeated three times (n = 360 snails). As in the density study, the snails were maintained in 16 L containers. A petri dish was filled with wet cotton in each cage, and cages were sprayed with water in all treatments daily to avoid aestivation. Cages were housed in ambient

61

light and humidity, with temperature ranging from 23-26 °C. Food and calcium were provided, and all cages examined at 48 h intervals. The experiment ran for 50 d. A 2- way ANOVA was used to test if food deprivation and snail size affected mean percent mortality (n = 3).

Results

Density Effects on Growth and Reproduction

Mean height at 240 d after hatch differed significantly across treatments (H =

309.70; df = 2; P < 0.01) (Table 4-1). Mean mass at 240 d after hatch also differed significantly across treatments (H = 263.70; df = 2; P < 0.01) (Table 4-1). No significant differences in mortality were found among the different density treatments (H = 5.18; df

= 2; P = 0.08) (Table 4-1).

Rearing density affected L. fulica reproduction by reducing the number of eggs laid per snail, egg viability and the number of eggs laid per clutch. The mean number of days from hatch until the first egg laying event were different among treatments (F =

4.23; df = 2, 22; P = 0.03). Snails in the high density treatment took significantly longer to begin egg laying than did the medium density treatment, though there was no difference between low and high densities nor between low and medium densities

(Table 4-2). There was a decrease in egg production as density increased. Eggs per snail (F = 86.04; df = 2, 22; P < 0.01) and eggs per clutch (H = 77.43; df = 2; P < 0.01) decreased (Table 4-2). The variances for the number of eggs laid per clutch are large, with clutch size ranging from 21 to 940 (Table 4-2). Additionally, mean egg viability differed across treatments (F = 3.63; df = 2, 308; P = 0.03). Significantly fewer eggs hatched in the high density treatment than in the low density treatment (Table 4-2).

62

Food Deprivation Effects on Cannibalism

Snails died (Fig. 4-2), but no deaths could be attributed to cannibalism. Food deprivation (F = 83.36; df = 3, 24; P < 0.01), snail size (F = 62.71; df = 2, 24; P < 0.01), and the interaction between the two factors (F = 32.56; df = 6, 24; P < 0.01) affected mortality. For juvenile snails, the no food and calcium only treatments experienced higher mortality than the food only and food and calcium diets. The mixed size treatment was only higher than other size treatments among snails fed food only.

Discussion

The decrease in shell height and snail mass as rearing density increased is consistent with other research reporting that high density levels can decrease snail growth rates (Conner et al. 2008; Ademolu et al. 2009; Mangal et al. 2010; Lasaridou-

Dimitriadou et al. 1998). Because density affects size, size cannot be used to predict the age of a snail.

The density treatment levels used in the present study did not affect mortality, which is similar to density studies of other snail species (Ademolu et al. 2009; Conner et al. 2008). Mangal et al. (2010), however, found mortality rates of B. alexandrina to range from 20 to 55%, with the lowest density treatment experiencing the lowest mortality rate. However, the mortality reported in the Mangal et al. (2010) study, could have been a result of natural juvenile mortality because such young snails (4-7 mm in height) were used in that study.

Surprisingly, the food deprivation experiment was consistent with the density experiment in that cannibalism was not observed. Also, after snails died due to starvation, often their flesh liquefied within a day. This rapid decomposition likely deterred any scavenging. As the experiment progressed, snails in both age groups were

63

inactive and sometimes aestivated, although the cages were sprayed with water daily.

This suggests that snails utilize dormancy to conserve energy rather than resort to cannibalism. The lack of observations of cannibalism in our colony also indicates that, although this species may scavenge, it is not inclined to cannibalism.

The reduced reproduction rates of L. fulica found in the high-density rearing experiment have been reported in other snail species (Conner et al. 2008; Ademolu et al. 2009; Mangal et al. 2010; Lasaridou-Dimitriadou et al. 1998). Like growth, the medium density treatment produced an intermediate effect on fecundity for L. fulica.

This differs from Ademolu et al. (2006), where a spike in A. marginata growth and reproduction was recorded at medium density, indicating that a certain level of density has a positive effect on that species. It also seems to differ from Conner et al. (2008), where a steep drop in P. paludosa growth and reproduction was observed at the medium density level, suggesting that this species is highly sensitive to density.

Lissachatina fulica does not seem to be as sensitive, although this could simply be a difference in what was selected as an intermediate density level.

Clutch size can also be affected by density. Clutch sizes of Pomacea canaliculata (Lamarck) (Gastropoda: Ampullariidae) (Mozzer et al. 2015) were not as variable (mean ± SD: 160 ± 12.14) as clutches from the present study. However, high variability in clutch size (15-210 eggs) has been found in Physella acuta (Draparnaud)

(Gastropoda: Physidae) (De Castro-Català et al. 2013). Clutch size for L. fulica has also been shown to vary from 10-500 eggs per clutch in other studies (Kekauoha 1966; Raut

& Barker 2002; CLEAPSS 2006).

64

High-density rearing effects are most likely due to the release of chemicals that inhibit the growth of other snails (Mangal et al. 2010; Garr et al. 2011). Even with precautions of proper drainage and waste removal in this experiment, toxic waste products accumulated quickly, especially in high density treatments. Inhibited growth and fecundity is not likely caused by limited food, calcium, or oxygen because these were adequately provided. Other density effects include biting, stress from immobility, and limited egg laying sites.

When selecting a rearing density, one’s goals must be taken into consideration.

For example, in our L. fulica rearing program, our goals are to keep egg production high and to raise the most adults as quickly as possible to use in research. However,

Dupont-Nivet et al. (2000) noted in discussing the results of their density experiment that although snails in low densities grow larger, the total biomass produced is the highest in high densities. Thus, if rearing snails for snail meat is the priority, high density may be the most economical choice.

65

Table 4-1. Mean (± SD) height, mass, percent mortality of Lissachatina fulica 240 d after hatch reared at three different density treatments. Density Height (mm)a Mass (g) a Mortality (%)a Low 105 ± 5a 136.5 ± 26.7a 0 ± 0a Medium 90 ± 4b 97.4 ± 20.2b 0.8 ± 2a High 76 ± 7c 60.5 ± 16.2c 1.6 ± 2a aMeans in a column followed by the same lowercase letters are not significantly different (P > 0.05; Kruskal-Wallis rank sum test and Dunn’s test).

Table 4-2. Lissachatina fulica reproduction when reared at three different density treatments: mean (± SD) number of days from hatch to initiation of egg production, estimated number of eggs produced per snail (over a 60-day period), eggs per clutch, and percent egg viability. Pre-reproductive Egg viability (% Density Eggs/snaila Eggs/clutchb period (d)a hatching)a Low 157 ± 14ab 777 ± 143a 396 ± 173a 50 ± 35a Medium 151 ± 13a 271 ± 110b 310 ± 147b 41 ± 37ab High 170 ± 12b 125 ± 37c 211 ± 90c 33 ± 33b aMeans in a column followed by the same lowercase letters are not significantly different (P > 0.05; ANOVA and Tukey's HSD test). bMeans in a column followed by the same lowercase letters are not significantly different (P > 0.05; Kruskal-Wallis rank sum test and Dunn’s test).

66

A

B

Figure 4-1. Lissachatina fulica growth when reared at low, medium, and high densities. A) Mean (± SD) height B) Mean (± SD) mass. Error bars indicate standard deviation.

67

Figure 4-2. Mean percent mortality of Lissachatina fulica for four diet treatments using adults, juveniles, and a mix of the two sizes. Means topped by the same lowercase letters are not significantly different within each size treatment, and the same uppercase letters are not significantly different within each diet treatment (P > 0.05; ANOVA and Tukey HSD test). Error bars indicate standard error.

68

LIST OF REFERENCES

Ademolu K, Idowu A, Agbelusi O. 2006. Effect of stocking density on the growth and haemolymph biochemical value of Archachatina marginata. African Journal of Biotechnology 8: 2908–2910.

Alicata JE. 1991. The discovery of Angiostrongylus cantonensis as a cause of human eosinophilic meningitis. Parasitology Today 7: 151–3.

An XC, Ren SX, Hu QB. 2007. Assessment of herbivore performance on host plants. Environmental Entomology 36: 694–699.

Auffenberg K, Stange LA. 2014. Snail-eating snails of Florida. Featured Creatures EENY251: 1–5. University of Florida, http://entomology.ifas.ufl.edu/creatures/misc/gastro/snail_eating_snails.htm (last accessed 8 Feb 2016)

Bequaert JC. 1950. Studies on the Achatinidae, a group of African land snails. Bulletin of the Museum of Comparative Zoology 105: 1–216.

Brinker T, Frank T. 1998. The palatability of 78 wildflower strip plants to the slug Arion lusitanicus. Annals of Applied Biology 133: 123–133.

Capinera J. 2008. Cannibalism, pp. 710-714 In Capinera J [ed.], Encyclopedia of Entomology. Springer Science & Business Media, Heidelberg, Germany.

Capinera J. 2013. Cuban brown snail, Zachrysia provisoria (Gastropoda): damage potential and control. Crop Protection 52: 57-63.

Capinera JL, Rodrigues CG. 2015. Biology and control of the leatherleaf slug Leidyula floridana (Mollusca: Gastropoda: Veronicellidae). Florida Entomologist 98: 243– 253.

De Castro-Català N, López-Doval J, Gorga M, Petrovic M, Muñoz I. 2013. Is reproduction of the snail Physella acuta affected by endocrine disrupting compounds? An in situ bioassay in three Iberian basins. Journal of Hazardous Materials 263P: 248–255.

Civeyrel L, Simberloff D. 1996. A tale of two snails: is the cure worse than the disease? Biodiversity and Conservation 5: 1231–1252.

Clarke B, Murray J, Johnson M. 1984. The extinction of endemic species by a program of biological control. Pacific Science 38: 97–104.

CLEAPSS. 2006. Giant African land snails. Brunel University, http://www.cleapss.org.uk/attachments/article/0/L197.pdf?Primary/Resources/Gui des/ (last accessed 8 Feb 2016)

69

Conner SL, Pomory CM, Darby PC. 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.

Cowie RH. 1992. Evolution and extinction of Partulidae, endemic pacific island land snails. Philosophical Transactions of the Royal Society B: Biological Sciences 335: 167–191.

Deisler JE. 1983. A key to the tree snails of Florida (Gastropoda: Bulimulidae). Florida Department of Agriculture and Consumer Services, Division of Plant Industry, Entomology Circular No. 246, http://syndication.freshfromflorida.com/content/download/23928/486395/ent246.p df (last accessed 8 Feb 2016)

Dupont-Nivet M, Coste V, Coinon P, Bonnet JC, Blanc JM. 2000. Rearing density effect on the production performance of the edible snail Helix aspersa Muller in indoor rearing. Annales de Zootechnie 49: 447-456.

Egonmwan RI. 2007. Food utilisation in a laboratory colony of the giant African land snail, Archachatina marginata (Swainson) (Pulmonata: Achatinidae). Journal of Zoology 31: 265–270.

Ejidike BN. 2007. Influence of articial diet on captive rearing of African giant land snail Archachatina marginata Pulmonata: Stylommatophora. Journal of Animal and Veterinary Advances 6: 1028–1030.

Estebenet AL, Cazzaniga NJ. 1992. Growth and demography of Pomacea canaliculata (Gastropoda: Ampullariidae) under laboratory conditions. Malacology Review 25: 1-12. Fasulo TR. 2008. Applesnails of Florida Pomacea spp. (Gastropoda: Ampullariidae). Featured Creatures ENY323: 1–5. University of Florida, http://entnemdept.ufl.edu/creatures/misc/gastro/apple_snails.htm (last accessed 8 Feb 2016)

Flint ML, Wilen CA. 2009. Snails and slugs: integrated pest management for home gardeners and landscape professionals. Pest Notes: 1–4. University of California, http://www.ipm.ucdavis.edu/PMG/PESTNOTES/pn7427.html (last accessed 8 Feb 2016)

Fox LR. 1975. Cannibalism in natural populations. Annual Review of Ecology and Systematics 6: 87–106.

Fretter V, Peake J [eds.]. 1979. Pulmonates Vol. 2B. Economic malacology with particular reference to Achatina fulic. Academic Press, London, United Kingdom.

70

Garr AL, Lopez H, Pierce R, Davis M. 2011. The effect of stocking density and diet on the growth and survival of cultured Florida apple snails, Pomacea paludosa. Aquaculture 311: 139–145.

Ghose KC. 1959. Observations on the mating and oviposition of two land Plumonates, Achatina fulica Bowdich and Macrochlamys indica Godwin-Austen. Bombay Natural History Society 56: 183–187.

Hadfield MG, Miller SE, Carwile AH. 1993. The decimation of endemic Hawaiian tree snails by alien predators. American Zoology 33: 610–622.

Hopper DR, Smith BD. 1992. Status of tree snails (Gastropoda: Partulidae) on Guam with a resurvey of sites studied by H. E. Crampton in 1920. Pacific Science 46: 77–85.

Haynes, J., J. McLaughlin, L. Vasquez, and A. Hunsberger. Undated. Low-Maintenance Landscape Plants for South Florida. University of Florida IFAS Extension, ENH854, http://miami-dade.ifas.ufl.edu/old/programs/fyn/publications/PDF/low- maintenance-plants.PDF (last accessed 8 Feb 2016).

Kekauoha W. 1966. Life history and population studies of Achatina fulica. The Nautilus 80: 3–10, 39–46.

Kondo Y. 1964. Growth rates in Achatina fulica Bowdich. The Nautilus 78: 6–15.

Lange WH. 1950. Life history and feeding habits of the giant African snail on Saipan. Pacific Science 4: 323–335.

Langford SD, Boor PJ. 1996. Oleander toxicity: an examination of human and animal toxic exposures. Toxicology 109: 1–13. Lazaridou-Dimitriadou M, Alpoyanni E, Baka M, Brouziotis T, Kifonidis T, Mihaloudi E, Sioula D, Vellis G. 1998. Growth, mortality and fecundity in successive generations of Helix aspersa Muller cultured indoors and crowding effects on fast-, medium- and slow-growing snails of the same clutch. Journal of Molluscan Studies 64: 67-74.

Mangal TD, Paterson S, Fenton A. 2010. Effects of snail density on growth, reproduction and survival of Biomphalaria alexandrina exposed to Schistosoma mansoni. Journal of Parasitology Research 2010: 1–6.

Mead AR. 1949. The giant snails. Atlantic Monthly 184: 38–42. Mead AR. 1959. The appearance of the giant African snail in Arizona. Proceedings of the Hawaiian Entomological Society 17: 85–86.

71

Mead AR. 1979. Chapter 3: Evaluation of dispersal, pp. 18-25 In Fretter V, Peake J. [eds.], Pulmonates, Vol. 2B. Economic malacology with particular reference to Achatina fulica. Academic Press, London, United Kingdom.

Meyer WM, Hayes KA, Meyer AL. 2009. Giant African snail, Achatina fulica,as a snail predator. American Malacological Bulletin 24: 117-119.

Mody K, Collatz J, Dorn S. 2015. Plant genotype and the preference and performance of herbivores: cultivar affects apple resistance to the florivorous weevil Anthonomus pomorum. Agricultural and Forest Entomology 17: 337–346.

Mozzer LR, Coaglio AL, Dracz RM, Ribeiro VM., Lima WS. 2015. The development of Angiostrongylus vasorum (Baillet, 1866) in the freshwater snail Pomacea canaliculata (Lamarck, 1822). Journal of Helminthology 89: 755–759.

Mulkern GB. 1967. Food selection by grasshoppers. Annual Review of Entomology 12: 59–78.

Murray J, Murray E, Johnson MS, Clarke B. 1988. The extinction of Partula on Moorea [French Polynesia]. Pacific Science 42: 150–153.

Onkara Naik S, Jayashankar M, Sridhar V, Chakravarthy AK. 2014. Record of the brown slug, Mariella dussumieri Gray, 1855 (Gastropoda: Ariophantidae) in marigold (Tagetes sp.). Journal of Plant Protection and Environment 11: 125– 126.

Panigrahi A, Raut S. 1994. Thevetia peruviana (family: Apocynaceae) in the control of slug and snails pests. Memórias do Instituto Oswaldo Cruz 89: 247–250.

Pawson PA, Chase R. 1984. The life cycle and reproductive activity of Achatina fulica (Bowdich) in laboratory culture. Journal of Molluscan Studies 50: 85-91. Pimentel D, Zuniga R, Morrison D. 2005. Update on the environmental and economic costs associated with alien-invasive species in the United States. Ecological Economics 52: 273–288.

Polis GA. 1981. The evolution and dynamics of intraspecific predation. Annual Review of Ecology and Systematics 12: 225–251.

Poucher C. 1975. Eradication of the giant African snail in Florida. Florida State Horticultural Society 88: 523–524.

R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org/ (last accessed 30 Jan 2016).

72

Rao IG, Singh DK. 2002. Toxic effect of single and binary treatments of synthetic and plant-derived molluscicides against Achatina fulica. Journal of Applied Toxicology 22: 211–5.

Raut SK, Barker GM. 2002. Achatina fulica Bowdich and other Achatinidae as pests in tropical agriculture, pp. 55–114 In Barker GM [ed.], Molluscs as Crop Pests. CAB International, Wallingford, United Kingdom.

Raut SK, Ghose KC. 1984. Pestiferous Land Snails of India. Zoological Survey of India No. 11, Bani Press, Calcutta, India. 151 pp.

Smith TR, Howe AC, Dickens K, Stanley JD, Brito JA, Inserra RN. 2015. First comprehensive molecular and morphological identification of the rat lungworm, Angiostrongylus cantonensis Chen, 1935 (Nematoda: Strongylida: Metastrongylida) in association with the giant African snail, Lissachatina fulica (Bowdich, 1822). Nematropica 45: 20–33. Smith TR, White-Mclean J, Dickens K, Howe AC, Fox A. 2013. Efficacy of four molluscicides against the giant African snail, Lissachatina fulica (Gastropoda: Pulmonata: Achitinidae). Florida Entomologist 96: 396-402.

Sridhar V, Jayashankar M, Vinesh LS, Verghese A. 2012. Severe occurrence of the giant African snail, Achatina fulica (Bowdich) (Stylommatophora: Achatinidae) in Kolar District, Karnataka. Pest Management in Horticultural Ecosystems 18: 228– 230.

Sturgeon RK. 1971. Achatina fulica infestation in North Miami, Florida. The Biologist 53: 93–103.

Takeda N, Ozaki T. 1986. Induction of locomotor behavior in the giant African snail, Achatina fulica. Comparative Biochemistry and Physiology 83A: 77–82.

Thiengo SC, Faraco FA, Salgado NC, Cowie RH, Fernandez MA. 2007. Rapid spread of an invasive snail in South America: the giant African snail, Achatina fulica, in Brasil. Biological Invasions 9: 693–702.

Thompson R, Cheney S. 1996. Raising snails. Special reference briefs series no. SRB 96-05. U.S. Department of Agriculture. Agricultural Research Service National Agricultural Library, Beltsville, Maryland, USA, http://pubs.nal.usda.gov/sites/pubs.nal.usda.gov/files/srb96-05_0.html (last accessed 30 Jan 2016).

Tomiyama K. 1992. Homing Behaviour of the Giant African Snail, Achatina fulica (Ferussac) (Gastropoda; Pulmonata). Journal of Ethology 10: 139–147.

Tomiyama K. 1993. Growth and maturation pattern in the African giant snail, Achatina fulica (Ferussac) (Stylommatophora: Achatinidae). Venus 52: 87–100.

73

Tomiyama K. 1996. Mate-choice criteria in a protandrous simultaneously hermaphroditic land snail Achatina fulica (Ferussac) (Stylommatophora: Achatinidae). Journal of Molluscan Studies 62: 101–111.

U.S. Congress Office of Technology Assessment. 1993. Harmful non-indigenous species in the United States. OTA-F-565. Government Printing Office, Washington, DC, USA.

Upatham ES, Kruatrachue M, Baidikul V. 1988. Cultivation of the giant African snail, Achatina fulica. Journal of the Science Society of Thailand 14: 25–40. van der Meer Mohr JC. 1949. On the reproductive capacity of the African or giant snail, Achatina fulica (Fer). Treubia 20: 1–10.

Venette RC, Larson M. 2004. Mini risk assessment giant African snail, Achatina fulica Bowdich [Gastropoda: Achatinidae]. University of Minnesota, http://wwx.inhs.illinois.edu/files/4713/4013/9195/afulicapra.pdf (last accessed 8 Feb 2016)

Wang QP, Lai DH, Zhu XQ, Chen XG, Lun ZR. 2008. Human angiostrongyliasis. The Lancet Infectious Diseases 8: 621–630.

White-Mclean J. 2012. Identification of agriculturally important molluscs to the U.S. and observations on select Florida species. University of Florida.

Yadav RP, Singh A. 2003. Effect of sub-lethal concentrations of Codiaeum variegatum latex on freshwater target snail Lymnaea acuminata and non-target fish Channa punctatus. Nigerian Journal of Natural Products and Medicine 7: 20-24.

74

BIOGRAPHICAL SKETCH

Katrina Dickens has had a lifelong passion for ecology and conservation and over the years she has developed a particularly strong interest in the area of invasive species control. Katrina was born and raised in Tucson, Arizona. She earned her bachelor’s degree in Environmental Science at Northern Arizona University. While attending NAU, she worked for the Pinyon Ecology Research Group studying the growth and reproduction rates of pinyon trees under the pressures of insect herbivory, drought and poor soils. She continued to pursue her love of nature and conservation in the Summer of 2008 with the California Department of Fish and Wildlife. In this position, she worked in the mountains of the Sierra Nevada removing non-native fish from the lakes and streams that are habitat to the endangered mountain yellow legged frogs and the Lahontan cutthroat trout. She derived great satisfaction from seeing how her direct actions were restoring the quality of the habitat and she was inspired to further pursue a career in invasive species control.

Katrina moved to Gainesville, Florida in 2009 and in February of 2010 she began working for the Division of Plant Industry of the Florida Department of Agriculture and

Consumer Services. She started in the sterile insect technique program for the control of the invasive cactus moth. This job introduced her to the pest control issues that abound in Florida and tied in neatly with her interest in invasive species control. Her hard work was recognized and she was promoted and moved into the Giant African

Land Snail Emergency Program. Soon after she developed an idea for a master’s thesis involving the snail and began taking classes at the University of Florida where she enrolled as a student in the Department of Entomology and Nematology.

75