INAVSION OF A MAJOR U.S. SEAPORT: COMMUNITY EFFECTS OF AND NOVEL

CONTROL METHODS FOR THE TAWNY CRAZY , FULVA

by

BENJAMIN MICHAEL GOCHNOUR

(Under the Direction of Daniel R. Suiter)

ABSTRACT

Invasive are a growing ecological and economic threat. With greater international trade comes the increased probability that species will be transported along with the goods via ships and seaports. Investigation of their impacts on native fauna is critical to understanding how best to prevent damage to existing habitats and communities by these invaders. Along with this understanding, new methods for control focused on impacting the invading species while leaving native species unharmed will bring about the most effective measures for addressing the issues associated with biological invasions.

INDEX WORDS: ; Invasive ants; Tramp ants; Biological invasion;

Nylanderia fulva; Tawny crazy ant; Semiochemical; Novel control

method; Linepithema humile;

INAVSION OF A MAJOR U.S. SEAPORT: COMMUNITY EFFECTS OF AND NOVEL

CONTROL METHODS FOR THE TAWNY CRAZY ANT, NYLANDERIA FULVA

by

BENJAMIN MICHAEL GOCHNOUR

B.S., University of Central Florida, 2012

A Thesis Submitted to the Graduate Faculty of the University of Georgia in Partial Fulfillment of

the Requirements for the Degree

MASTER OF SCIENCE

ATHENS, GEORGIA

2017

© 2017

Benjamin Michael Gochnour

All Rights Reserved

INAVSION OF A MAJOR U.S. SEAPORT: COMMUNITY EFFECTS OF AND NOVEL

CONTROL METHODS FOR THE TAWNY CRAZY ANT, NYLANDERIA FULVA

by

BENJAMIN MICHAEL GOCHNOUR

Major Professor: Daniel R. Suiter Committee: Brian T. Forschler Joseph V. McHugh Qingguo Huang

Electronic Version Approved:

Suzanne Barbour Dean of the Graduate School The University of Georgia December 2017

iv

DEDICATION

To my mother, whose gentle guidance has led me to find myself.

v

ACKNOWLEDGEMENTS

First and foremost I thank Dr. Dan Suiter for his unwavering dedication and vigilant counsel during this experience. I consider him to be not only an excellent mentor, but a friend. I thank my graduate committee for their support and willingness to accompany me on this journey.

Dr. Brian Forschler: your insight regarding the aspects of academia beyond empirical research has been a grounding force that has only magnified my passion for what I do. Dr. Joe McHugh: you have been a beacon of excellence, the likes of which I can only hope to someday emulate. Dr.

Jack Huang: your enthusiasm for my research and amiable approach to collaboration has provided respite during an incredibly stressful time.

I thank Jerry Davis, whose statistical wizardry was invaluable to the validity of this project. To Christina Crespo: your support and commitment to improving the quality of my writing has been a boon to the arguments made in this thesis. My good friends Brent Phelan and

Mike Arvin: you have provided me with relief from the onslaught of graduate school responsibility. The memories of our many late night gatherings will last a lifetime, as will our friendship.

Finally, I thank my mother, Barbara Moore. Your patience, strength and unyielding love have allowed me to conduct myself with confidence and guile in my quest for knowledge. I owe all that I achieve to you and will be forever grateful for having you in my life.

“The only person you’re truly competing against is yourself.”

-Jean-Luc Picard

vi

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS ...... v

LIST OF TABLES ...... viii

LIST OF FIGURES ...... ix

REVIEW OF THE SCIENTIFIC LITERATURE ASSOCIATED WITH INVASIVE ANT

SPECIES AND THEIR MANAGEMENT ...... 1

Biology of Invasive Ants ...... 1

Argentine Ants ...... 7

Novel Control Methods for Ants ...... 13

Tawny Crazy Ant ...... 16

Ant Fauna of Georgia and the Southeast ...... 20

References Cited ...... 22

A SURVEY OF ANT SPECIES FROM THE PORT OF SAVANNAH, GARDEN CITY,

GEORGIA (CHATHAM COUNTY) ...... 38

Introduction ...... 38

Materials and Methods ...... 39

Results and Discussion ...... 42

References Cited ...... 45

ECOLOGICAL IMPACTS OF THE TAWNY CRAZY ANT, NYLANDERIA FULVA

MAYR, ON THE ANT COMMUNITY AT THE PORT OF SAVANNAH, GEORGIA ...... 61

vii

Introduction ...... 61

Materials and Methods ...... 62

Results ...... 66

Discussion ...... 68

References Cited ...... 74

USE OF ROADSIDE HABITATS AS DISPERSAL CORRIDORS BY THE INVASIVE

TAWNY CRAZY ANT, NYLANDERIA FULVA, AT THE PORT OF SAVANNAH,

GEORGIA ...... 97

Introduction ...... 97

Materials and Methods ...... 99

Results ...... 101

Discussion ...... 104

References Cited ...... 109

POTENTIAL USE OF ARGENTINE ANT, LINEPITHEMA HUMILE MAYR,

SEMIOCHEMICALS FOR THE DELIVERY OF AN ACUTE TOXICANT ...... 127

Introduction ...... 127

Materials and Methods ...... 131

Results ...... 134

Discussion ...... 136

References Cited ...... 139

CONCLUSIONS ...... 150

viii

LIST OF TABLES

Page

Table 2.1. Ant species collected on and around the port of Savannah, Georgia with collection method and habitat type noted...... 47

Table 2.2. Trapping statistics for the 2015 - 2016 survey at the Port of Savannah, Georgia...... 51

Table 2.3. Number of ant species and exotic ant species in Georgia* and adjacent states*...... 52

Table 3.1. Species richness by year plus number of species collected for both years and net change in number of species from 2015-2016 by transect...... 77

Table 3.2. Uniques and Duplicates collected and richness estimator values for each transect by year……………………………………………………………………………………………….78

Table 3.3. Shared and unshared species averages for each transect by year…………………….79

Table 4.1. Ant species collected from roadside habitats (all four sites combined) on the Port of

Savannah in Garden City, Georgia during 2015 and 2016...... 112

Table 4.2. Site one species incidence...... 113

Table 4.3. Site two species incidence...... 114

Table 4.4. Site three species incidence...... 115

Table 4.5. Site four species incidence...... 116

Table 5.1. Mean number of ants inside the border of a 30-mm circle around paper wicks treated with methanol and n-hexane extracts containing cuticular compounds from live pupae...... 144

Table 5.2. Average number of insecticide treated and untreated ant cadavers and paper wicks containing a methylene chloride extract of Argentine ant corpses removed at one, two, and 24 hours...... 145

ix

Table 5.3. Mortality of Argentine ants three days following exposure to insecticide-treated and untreated ant cadavers and paper wicks containing a methylene chloride extract of ant cadavers.

...... 146

x

LIST OF FIGURES

Page

Figure 2.1. Port of Savannah sampling locations...... 53

Figure 2.2. Pitfall sampling...... 54

Figure 2.3. Leaf litter collection...... 55

Figure 2.4. Dissection of course woody debris...... 56

Figure 2.5. Tree trunk sampling...... 57

Figure 2.6. Bait trap utilizing a nine dram vial and its cap along with crushed Pecan Sandie cookie placed in a small clearing on the forest floor...... 58

Figure 2.7. Quadrat sampling...... 59

Figure 2.8. Rarefaction curves for all wooded transect samples combined ...... 60

Figure 3.1. Port of Savannah sampling locations...... 80

Figure 3.2. Pitfall sampling...... 81

Figure 3.3. Leaf litter collection...... 82

Figure 3.4. Dissection of course woody debris...... 83

Figure 3.5. Tree trunk sampling...... 84

Figure 3.6. Bait trap utilizing a nine dram vial and its cap along with crushed Pecan Sandie cookie placed in a small clearing on the forest floor...... 85

Figure 3.7. Number of species collected by transect...... 86

Figure 3.8. Abundance of N. fulva workers in pitfall traps…………………………..………….87

Figure 3.9. Rarefaction curves for transect one by year...... 88

Figure 3.10. Rarefaction curves for transect two by year...... 89

Figure 3.11. Rarefaction curves for transect three by year...... 90

Figure 3.12. Rarefaction curves for transect four by year...... 91

xi

Figure 3.13. Rarefaction curves for transect five by year...... 92

Figure 3.14. Rarefaction curves for transect six by year...... 93

Figure 3.15. Rarefaction curves for transect seven by year...... 94

Figure 3.16. Rarefaction curves for transect eight by year...... 95

Figure 3.17. Rarefaction curves for transect nine by year...... 96

Figure 4.1. Port of Savannah property...... 117

Figure 4.2. Location of bait and quadrat sampling sites...... 118

Figure 4.3. Bait and Quadrat sampling...... 119

Figure 4.4. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits (circles) and in quadrats (squares) at site one during (A) July 2015 and (B) July 2016 at the Port of Savannah,

Garden City, Georgia...... 120

Figure 4.5. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits (circles) and in quadrats (squares) at site two during (A) July 2015, (B) October 2015, (C) March 2016, and (D)

July 2016 at the Port of Savannah, Garden City, Georgia...... 121

Figure 4.6. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits (circles) and in quadrats (squares) at site three during (A) July 2015, (B) October 2015, (C) March 2016, and (D)

July 2016 at the Port of Savannah, Garden City, Georgia...... 122

Figure 4.7. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits (circles) and in quadrats (squares) at site four during (A) July 2015 and (B) July 2016 at the Port of Savannah,

Garden City, Georgia...... 123

Figure 4.8. Incidence of S. invicta and N. fulva by year and sampling method...... 124

Figure 4.9. Subsequent occurrence of Nylanderia fulva in four sample types (pitfall, leaf litter, bait, and tree trunk sampling) along a wooded transect following invasion of the adjacent site 4 roadside habitat in July 2016...... 125

xii

Figure 4.10. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits and in quadrats during (A) July 2015 and (B) July 2016 sampling events at the Port of Savannah, Garden

City, Georgia...... 126

Figure 5.1. Test arena for pupa retrieval and activity measure experiments...... 147

Figure 5.2. Test arena for item removal and mortality experiments...... 148

Figure 5.3. (A) Residual plot for the activity measure experiment...... 149

1

CHAPTER 1

REVIEW OF THE SCIENTIFIC LITERATURE ASSOCIATED WITH INVASIVE ANT

SPECIES AND THEIR MANAGEMENT

Biology of Invasive Ants

“One third of the entire biomass of the Amazonia terra firme rain forest is composed of ants and termites, with each hectare of soil containing in excess of 8 million ants and 1 million termites.” (Hölldobler and Wilson 1990). In one study, 100% of dead on an eastern forest floor were taken by ants (Fellers and Fellers 1982). Ants are undoubtedly a dominant force on the planet. They aerate the ground (Sanford et al. 2009), disperse seeds (Lengyel et al. 2009,

Stuble et al. 2014), and are the most abundant species in almost any natural area (Wilson 1987,

Folgarait 1998). The biological characteristics of ants allow them to be successful competitors for resources in many environments (Suarez et al. 2008, Linksvayer and Janssen 2009, Grangier and

Lester 2014, Yip 2014). Colonies can patrol and defend a territory (Adams 2003), divide labor and reproduction among specialized castes (Jongepier and Foitzik 2016, Villalta et al. 2016), and store energy in workers for overwintering (Yang 2006). These characteristics lead to certain species being particularly adept at surviving and reproducing in novel habitats (King and

Tschinkel 2006).

Biological invasions can be described as “a species acquiring a competitive advantage following the disappearance of natural obstacles to its proliferation, which allows it to spread rapidly and to conquer novel areas within recipient ecosystems in which it becomes a dominant population” (Valery et al. 2008). In ants, the concept of biological invasion commonly manifests itself when a species is able to reach a new geographic region, often aided by unwitting human intervention (Ward et al. 2005, Ward et al. 2006, Wetterer et al. 2009, Foucaud et al. 2010), survive, and reproduce there. Species that complete this journey often proliferate in new

2 ecological habitats as well as in heavily disturbed or human altered environments (Gotzek et al.

2015). Invasive ant species can be found in agricultural fields, fragmented urban microhabitats and in buildings (Bolger et al. 2000, Roura-Pascual et al. 2011). They tend to be generalists and can disrupt the native communities in ecosystems where they are introduced (Sanders and Suarez

2011).

Ant invasions often occur in two distinct phases. The first phase is arrival and spread, where the species is introduced to a new habitat and over the course of several generations may expand to increase its potential geographic range (Suarez et al. 2001). The second phase involves an increase in abundance on a local scale where the invader increases its colony density and individual abundance in a given area (Suarez et al. 2010). Differences in how these species are able to successfully work through these two phases will determine the impact they have at different geographic scales.

The question of what makes a good invader is complex and has been approached from several angles. Characteristics such as body size (King and Porter 2007, Wills et al. 2014), behavior (Human and Gordon 1999, Suarez et al. 2002), (Holway et al. 2002), reproductive strategy (Aron 2001), and seasonality (Krushelnycky et al. 2004, Heller and Gordon

2006) can all be considered factors in making a species a good invader or not (Tsutsui and Suarez

2003, Phillips and Suarez 2012). Some of the most successful invasive ant species have had workers that are moderately-sized (about 4mm) and often have hundreds of thousands, or even millions of workers in a single colony (Tschinkel 2006). The smaller size may allow the ants to exploit resources or occupy microhabitats that larger species cannot (McGlynn 1999b). Large colony sizes are beneficial for monitoring the surrounding environment for resources and competition as well as for protecting the colony (Tschinkel 2006). With large numbers of workers, invasive ant species can often win a war of attrition with native species although individual fighting prowess is also a significant factor (Adams and Mesterton-Gibbons 2003).

3

Colonies of invading ant species tend to have unusual traits including increased queen density and larger queen body size (Abril et al. 2013), polygyny and queen execution (Inoue et al.

2015), and variable social structure (Ingram 2002) which allow colonies to dominate new environments. Small incipient populations have reduced genetic diversity compared to populations in their home range, allowing them to reduce the cost of territoriality between conspecific colonies (Tsutsui et al. 2000, Cheng et al. 2016). Relative inbreeding in incipient populations leads to colonies that can no longer differentiate themselves from one another (Torres et al. 2007). The queens are so genetically similar that their cuticular hydrocarbon profiles, which are recognition signals, do not differ enough for the workers to discern which are from their colony (Vasquez et al. 2008, Vasquez and Silverman 2008). This kind of non-aggressive behavior between conspecific colonies leads to multi-queen supercolonies that can reproduce and spread at an alarming rate (Abbott 2005). The lack of competition from their own species gives the invaders an edge over native species that often have discrete, intraspecifically competitive colonies (Suarez and Suhr 2012).

The reproductive strategies of invasive ant species can also differ from noninvasive ants to further perpetuate their spread and domination of their new environment. Normally, ants have a seasonal cycle where reproductive castes emerge during a specific time of year (Hart and

Tschinkel 2012). During a mating flight, unmated queens and males fly from the colony and mate in air. The queens then find a suitable place to start their new colony (Tschinkel 2006). This can lead to great distances among individual colonies and can enable the species to disperse long distances in a short time. Fire ants are known to mate 1Km above the ground and be carried by wind currents to the place where they will found their new colony (Fritz et al. 2011). An alternative mating strategy, known as budding, removes the need for a mating flight. Budding leads to colonies that spread from a central location (Buczkowski and Bennett 2009). The new budded colonies are contiguous with the original and the workers and queens may be shared between them. In some cases, like Argentine ants, Linepithema humile (Mayr), the repetition of

4 budding can produce genetically related colonies that stretch over landscapes (Tsutsui and Suarez

2003).

The effect that invasive ant species can have on the native ant community can be significant (Wilder et al. 2013). Invasive ant species can reduce native ant species richness and homogenize the ant community (Holway and Suarez 2006). These actions often extirpate other ant species that overlap with their niche leaving behind specific predictable groups of species that can either tolerate the invader or move to uninfested areas. Lu et al. (2012) found that S. invicta reduced ant species richness by up to 46% in lawn and pasture sites when compared to neighboring uninvaded sites. Sarty et al. (2007) showed that the , gracilipes (Smith), competitively excluded native ant species on Tokelau, a territory of New

Zealand. Anoplolepis gracilipes was shown to reduce the number and abundance of native ant species where it was present. Using several types of baits, it was determined that the native ants’ ability to utilize different food resources facilitated their coexistence with A. gracilipes. Sanders et al. (2003) used seven years of data from Jasper Ridge Biological Preserve in northern

California to demonstrate how an invasion by L. humile contributed to the disassembly of the ant community there. In addition to a reduction in species richness at invaded sites, the community’s compositions were disturbed. While uninvaded sites showed native species segregation governed by competition, invaded sites tended toward a more random community assembly. Lessard et al.

(2009) conducted a meta-analysis of community level ant invasion studies and found that invasions often not only reduce species richness but also disrupt the phylogenetic community structure. Uninvaded communities were generally evenly dispersed phylogenetically which suggests that they are structured by competition. Invaded communities were clustered, with a few closely related taxa remaining after invasion. This clustering suggests that the invaders have a nonrandom effect on the members of the community that may persist alongside the invader.

Beyond the ant community, invasive ant species can have a great impact on the survival of ground nesting avian, reptilian and mammalian species (Suarez and Case 2002, Suarez et al.

5

2005a, Diffie et al. 2010). Eggs, new born or injured individuals are more susceptible to predation by aggressive species of ants. Suarez et al. (2005a) showed experimentally that the Argentine ant preyed on more artificial quail nests with actual eggs, than did native species. This predation can exacerbate problems with conservation efforts for species that may be threatened or endangered.

Horned lizards underwent a diet shift after invasion by Argentine ants. The lizards went from eating large native ant species to other groups when the native ants were replaced with

Argentine ants. The horned lizards would not eat the invasive ants, disrupting the food web in this ecosystem (Suarez et al. 2000). Allen et al. (2001) found increasing numbers of S. invicta in loggerhead, green and freshwater turtle nests in Florida. Solenopsis invicta was shown to prey on up to 70% of the freshwater turtle hatchlings either during pipping or shortly after hatching. Long et al. (2015) used enclosures to measure the interaction of S. invicta and the hispid cotton rat,

Sigmodon hispidus (Say). Rat mortality increased by two to four times in the presence of S. invicta when compared to enclosures with no S. invicta. Holtcamp et al. (1997) showed that the deer mouse, Peromyscus maniculatus (Wagner), altered its foraging behavior in the presence of S. invicta. Mice foraging in plots with S. invicta tended to prefer more productive patches to maximize their foraging efforts. The mice would also abandon a patch sooner in favor of more productive patches in the presence of S. invicta.

Invasive ant species are also known to form mutualistic symbiotic relationships with producing plants and producing hemipteran (Lach et al. 2010, Wilder et al. 2011). These relationships can lead to the proliferation of specific exotic plants in their new range or the demise of native plants due to increased numbers of sap feeding insects in their environment.

Invasive ant species can have great economic impacts in their introduced ranges

(Tschinkel 2006). Just like in natural communities, invasive ants can have a negative impact on agriculture by tending sap feeding hemipterans that feed on important crop plants. The ants can also become pests in factories and buildings causing loss of production efficiency or

6 contamination of products (Oi et al. 1994). In urbanized areas these species can be household pests, costing millions of dollars in treatment of residential homes and businesses.

Recent studies of Tawny Crazy ant, Nylanderia fulva (Mayr), populations on the island of St. Croix have shown that at about 6 years the populations decline (Wetterer et al. 2014). Their numbers are greatly reduced and previously excluded species may begin recolonizing previous habitats. This “boom and bust” cycle has also been shown in southern fire ants in the United

States (Tschinkel 2006).

Efforts to eradicate invasive ant species have largely been unsuccessful (Sanders and

Suarez 2011). Considering the difficulty of killing even one mature colony of modern invaders, it is easy to see how unlikely eradication would be on a large scale. In addition, there is always the possibility of reintroduction. Control or suppression of these species is likely the best and most reasonable option combined with preventative measures to reduce the likelihood of future introductions.

Recently, biocontrol methods have been utilized to aid in the fight against the red imported , Solenopsis invicta (Buren). Flies in the family Phoridae, which are known to parasitize S. invicta workers, use the workers as oviposition sites (Porter et al. 1995). In this manner a single female fly may kill many adult worker ants in its lifetime. The efficacy of this parasitoid has been the subject of further studies that have shown that many other factors must be considered outside of host-parasitoid interactions when determining an effective biological control agent (Morrison and Porter 2005). Morrison and Porter (2005) showed no selective pressure from phorid flies on fire ants on a population level. They concluded that one species may not have enough of an impact on the ants or that the phorid fly populations may need more time to develop.

In addition to being the target of biocontrol measures the fire ant has also been considered as a biocontrol agent (Sterling 1978). According to Sterling (1978) fire ants were allowed to proliferate in cotton fields that were affected by the cotton boll weevil. In some plots

7 the fire ants were able to consume up to 85% of the weevils. The ants showed no interest in the cotton plants and work as a deterrent to herbivores by patrolling the plants near their colony. This negative relationship is non-specific and was observed in 16 of 16 herbivore taxa in cotton fields in (Eubanks 2001). This relationship was also observed between fire ants and 22 of 24 herbivore predators. The fire ants preferably foraged on plants with higher densities of cotton but were found to reduce lady beetle and green lacewing larvae by over 90% and 80% respectively, reducing their utility as a biocontrol agent (Kaplan and Eubanks 2002). Fire ants are something to be considered when found in agricultural fields as they can have both a cost and benefit under different conditions.

Overall, the diversity of ants has allowed them to thrive in almost all terrestrial habitats on the planet. Some of these species have acquired the traits that allow them to take advantage of resources in novel ecosystems and out-compete competitor species. Their ability to survive and reproduce while under the pressure of extermination by humans and native species is a tribute to their evolutionary success. With increasing disruption of natural communities by anthropogenic forces, we will continue to witness the spread and proliferation of invasive exotic ant species with dire consequences for the remaining biodiversity.

Argentine Ants

Argentine Ant Biology

The Argentine ant, Linepithema humile (Mayr), is a successful invasive ant species, having spread to six continents with the aid of human-mediated jump dispersal (Suarez et al.

2001). The ants are adept at invasion due to evolved traits developed in their home range.

Argentine ants nest ephemerally, have a high growth rate, can reproduce year round, and have a temporally and spatially fluid colony structure (Suarez et al. 2008). Their year round polygynous nature increases the likelihood of dispersal when a portion of the colony is transported by humans

(Aron 2001). Upon introduction, Argentine ants often undergo a genetic bottleneck, thereby reducing heterozygosity and allelic diversity in the founding population. This often leads to an

8 absence of competition among conspecific colonies, which produces super colonies that can extend for miles (Suarez et al. 2008). Behaviorally, the Argentine ant outcompetes native ant species by dominating resources as well as using chemical warfare. In a study by Holway (1999),

Argentine ants located and recruited to baits as quickly, or more quickly, than native ants. They also displaced native ant species from baits using defensive chemical compounds and physical aggression. Argentine ants were found to initiate aggressive encounters more often than native species, while native species were more likely to retreat from the encounter, leading to resource domination by Argentine ants (Human and Gordon 1999).

The ability of Argentine ants to dominate food resources can be attributed, in part, to their large colony size. The phenomenon known as the supercolony can be attributed to a lack of genetic diversity which causes an inability of ants to differentiate between workers and queens from nearby colonies. Akino et al. (2004) showed that cuticular hydrocarbons are the source of workers ability to differentiate between nestmates and workers from unrelated colonies. Isolates from the cuticle of Formica japonica (Motschoulsky) applied to a glass dummy evoked either an aggressive or non-aggressive response based on whether the isolate was collected from a foreign worker or nestmate, respectively. These hydrocarbon signal molecules are often the same between colonies but are produced in different proportions allowing for colony specific blends

(Sturgis and Gordon 2012). The cuticular hydrocarbons that comprise the chemical profile are often redundant, having a synonymous signal, where several similar hydrocarbons will communicate the same type of signal. This allows the profiles to be plastic and leads to context dependent interpretation of the profile (van Wilgenburg et al. 2010). The level at which the ants are able to detect these signals can be as low as the equivalent of 10-4 worker (Ichinose and

Lenoir 2010). Differences in the diversity of these signal molecules can be correlated to genetic and evolutionary distance, with more closely related species having similar hydrocarbon profiles and more distantly related species having greater differences between their profiles (van

9

Wilgenburg et al. 2011). The combination of life history traits, mentioned earlier, and the inability of colonies to differentiate among themselves makes controlling Argentine ants difficult.

Argentine Ant Conventional Control

Current control methods utilize toxic baits and sprays to control the Argentine ants’ presence in and around structures. Laboratory studies by Hooper-Bui and Rust (2000) showed that would be an effective control measure when delivered in a 25% sucrose solution. In this study, fipronil concentrations of 1 x 10-4% provided complete mortality of workers and queens that were baited for 24 hours.

Klotz et al. (2002) used 0.0001% fipronil in a 25% sucrose solution at bait stations around homes and were able to achieve >70% reduction in Argentine ants after five weeks of exposure. The fipronil baits had comparable control to the spray treatments in the same experiment, but required weekly maintenance whereas the spray treatments required only one application. In another study by Klotz et al. (2007), a spot treatment using a 0.06% fipronil solution applied to active ant trails with a backpack sprayer was effective at reducing Argentine ant numbers by 90% at eight weeks. Klotz et al. (2008) achieved a 93% reduction in Argentine ants at eight weeks using a combination of a 0.06% fipronil perimeter spray and a 0.20% granular formulation. Since fipronil sprays are effective at reducing Argentine ant numbers in urban settings, Klotz et al. (2010) conducted a study to determine whether volume or application method had any effect on the efficacy of a fipronil treatment. In this experiment either one gallon or 0.5 gallon volume of 0.06% fipronil solution was applied by either a pin-stream applicator in a 5 cm thick band at the base of the foundation or a fan spray nozzle applicator covering 30 cm up and away from the foundation around the entire perimeter of the house. These variables were chosen to minimize the amount of applied insecticide and to reduce runoff of residual insecticide. The pin-stream application of one gallon of fipronil solution was the most effective, resulting in an 85% reduction of Argentine ants around the structure and a 63% reduction in Argentine ants in the yard at six weeks after the single treatment.

10

Wiltz et al. (2009) conducted a series of tests to evaluate the effectiveness of bifenthrin, chlorfenapyr, thiamethoxam, and fipronil at killing Argentine ants in a laboratory colony.

Workers were topically treated with the chemicals and mobility impairment as well as median lethal time was recorded. Time to mobility impairment was slowest in fipronil treated colonies although mortality was also highest in fipronil treated colonies. Tests using fipronil killed ant corpses as donors were conducted at 10, 20, and 30°C, where fipronil, again showed the highest mortality due to contact transfer of the pesticide, and a positive correlation with temperature was observed.

Costa and Rust (1999) tested the effects of fipronil and diazinon soil treatments on

Argentine ant foraging rates and mortality in potted plants. Twenty four laboratory colonies of

Argentine ants were each given one potted oleander plant in a 15 cm diameter pot. Colonies were given one of three rates, 5, 10, and 20 ppm, of a fipronil soil mixture consisting of granular fipronil mixed with moist soil or a broadcast treatment consisting of dry soil treated with 14 and

28 g of fipronil broadcast and added to the potting soil. After one week, foraging rates in all treatments but the control had dropped to zero. The soil mix treatment of fipronil had killed >90% of workers after one week, and had killed all queens after four weeks. Fipronil broadcast treatments killed <50% of workers at one week, and took eight weeks to kill all queens.

Choe and Rust (2008) tested eight insecticides for their ability to be spread by contact, and mortality in Argentine ants. After exposure to insecticide treated sand for one minute, ants were placed into a colony of untreated ants. Only fipronil, at rates of 0.002% and 0.004%, resulted in adequate contact transfer and subsequent high mortality rates within four days.

Mortality rates for the remaining insecticides were not significantly different from the control. A second experiment using fipronil treated ant corpses showed that necrophoresis played an important role in the contact transfer of the insecticide. Fipronil treated corpses placed closer to the nest resulted in higher mortality compared to corpses placed 30 cm away from the nest.

11

Vega and Rust (2003) marked foraging Argentine ants with a fluorescent brightener

(FB28) to determine the origin of resurgence after treatment with fipronil baits and sprays. Baits containing 0.0001% fipronil and sprays formulated at 0.06% fipronil achieved significant reduction in Argentine ant workers by four weeks. The percentage of marked ants in these treatments decreased throughout the experiment due to immigration from the surrounding areas, suggesting that to effectively control Argentine ants, a much larger treatment area may be necessary.

Ecological Considerations of Argentine Ant Control

In a study by Hayasaka et al. (2015), the susceptibility to fipronil baits of four Argentine ant supercolonies was tested along with four native ant species, a native isopod, and two species of cockroach. Species sensitivity distributions were used to compare susceptibility to fipronil among the tested species. Only one of the supercolonies of Argentine ants showed significantly higher sensitivity to fipronil, suggesting that efforts to control the other supercolonies using fipronil would result in a negative impact on non-target arthropod species. These results indicate that fipronil has no increased effect on Argentine ants when compared to other arthropod species and that broad application of fipronil will likely have unintended effects on native species in the treatment area.

Jiang et al. (2014) evaluated fipronil toxicity to Argentine ants and run-off potential on concrete surfaces after exposure to summer weather conditions and simulated precipitation.

Fipronil treated surfaces killed >50% of Argentine ants, exposed for one minute, within 16 hours of exposure. Toxicity to Argentine ants was lost on these surfaces after 20 days while run-off water from simulated precipitations contained detectable amounts of insecticide 89 days after treatment.

Gan et al. (2012) collected run-off water from large, residential communities, each comprised of hundreds of single family homes in Orange County and Sacramento County,

California. The water samples were tested for fipronil and its biologically active derivatives over

12 a 26 month period. The amount of fipronil in the samples from Orange County was greater than ten times the amount of that in the samples from Sacramento County. In Orange County, fipronil load in water was positively correlated with higher insecticide use. A temporal pattern was also observed, with higher levels of fipronil in the samples from April to October, while levels decreased from October through March. This pattern reflects the higher activity of Argentine ants in the warmer months resulting in greater use of insecticides and coincides with more frequent rain events leading to more insecticide being washed away in the run-off water. Fipronil and its oxidative derivative, fipronil sulfone, comprised 70% of the total concentration of insecticides in the samples, while fipronil’s photolytic derivative, desulfinyl fipronil, accounted for another 25% of the concentration. The levels of fipronil and fipronil derivatives found in the water samples often exceeded the LC50 values for local, aquatic arthropods.

In another study evaluating fipronil and two of its derivatives, fipronil sulfide and desulfinyl fipronil, Mahler et al. (2009) collected indoor and outdoor dust samples from 24 residences in Austin, Texas. Every sample collected in the study contained at least one of the three compounds. All indoor samples had a higher concentration of fipronil and its derivatives when compared to its corresponding outdoor sample. The increased concentrations of fipronil and its derivatives indoors may be due to a lower rate of degradation compared to outdoors, or an indoor source of fipronil. Three homes containing a pet using a flea control product containing fipronil had elevated levels of fipronil and its derivatives compared to homes without a pet.

Greenberg et al. (2010) evaluated fipronil and bifenthrin levels in run-off and irrigation water directly following treatment at residential homes in 2007 and 2008. The 2007 treatment consisted of a spray application of the two insecticides using standard practices. The 2008 treatment applied product as a pin stream, utilized spray-free zones, and restricted the application of insecticides to the foundation of the homes. After one week the level of fipronil in the run-off water from the 2007 treatment was enough to cause acute aquatic toxicity to sensitive organisms.

At eight weeks post-treatment, the same was true for levels of bifenthrin in the run-off water. In

13 contrast, the water samples from the 2008 treatment contained no detectable levels of fipronil at one week post-treatment and contained greatly reduced levels of bifenthrin when compared to the

2007 treatment.

Novel Control Methods for Ants

Given the dangers of continuously depositing insecticides into the ground water via run- off water from residential use, several alternatives have been explored in an attempt to reduce or eliminate the need for such insecticides. Guerra et al. (2011) conducted a laboratory study to determine the effectiveness of nine essential oils at killing the black , Camponotus pennsylvanicus (De Geer). Extracts of basil, lemon, citronella, clove, eucalyptus, peppermint, rosemary, tea tree, and thyme were used in this study at a 10% concentration in acetone. A 0.03% bifenthrin solution, an acetone only, and an untreated control were also assayed. The treatments were applied topically to the ants and mortality was recorded daily for seven days. At seven days post-treatment, none of the essential oil extracts were able to achieve mortality comparable to the bifenthrin positive control. Citronella and tea tree extracts had mortality that was significantly higher than the acetone only, and untreated controls, but did not exceed 33% mortality, compared to the 96% ant mortality in the bifenthrin treatment.

Wiltz et al. (2007) evaluated the deterrent and toxic properties of six essential oils including tea tree, peppermint, lemon, citronella, and basil oil in a choice test. Argentine ants crossed treated barriers less often than corresponding control barriers, with the exception of the eucalyptus extract which had no deterrent effect. When continuously exposed to the extracts, citronella was the only treatment to kill 100% of argentine ants after 24 hours, while peppermint oil killed 89.8% of exposed ants and teatree oil killed 85.7%.

In a study to determine the repellency of essential oils to Argentine ants, Scocco et al.

(2012) tested five extracts along with two controls at three different concentrations. Clove, cinnamon, wintergreen, spearmint, and peppermint oils were applied at 10%, 1%, and 0.1% concentrations along with a hexane negative control and a Cinnamite positive control. The

14 extracts were used in a choice test involving two harborages, one treated and one untreated. The harborages containing extract were aged for either two hours or seven days to test the residual effects of the extracts. All extracts were repellent when testing two hour old harborages, with significantly less ants entering the harborages containing an extract. At seven days, the 10% and

1% extracts were still repellent, while only spearmint remained repellent at the 0.1% concentration.

Buczkowski (2016) developed a novel delivery method for insecticides involving species specific food resources to control the Asian needle ant, Brachyponera chinensis (Emery). Using a single eastern subterranean termite, Reticulitermes flavipes (Kollar), topically treated with fipronil and then placed in small laboratory colonies consisting of 100 worker ants, all Asian needle ant workers were killed within nine hours of exposure. A field study involving the distribution of fipronil treated termites on the forest floor in an area infested with Asian needle ants was able to achieve a 98% reduction in the ants within 28 days. The specificity of the bait allows for extremely effective control of the target species while reducing the amount of insecticide deposited into the environment.

In a series of studies by Suckling et al. (Suckling et al. 2008, Suckling et al. 2010,

Suckling et al. 2011), the usefulness of Argentine ant trail , (Z)-9-hexadedcenal, as a deterrent was explored. Initially Suckling et al. (2008) found that a point source of the trail pheromone reduced the ability of the ants to trail meaningfully and find resources in the surrounding area. It was also observed that the number of workers in the pheromone treated areas increased as recruitment exceeded departures from the area. A subsequent field experiment,

Suckling et al. (2010) treated 400 m2 plots with a sprayable, micro-encapsulated particle of the pheromone and observed a reduction in Argentine ant trailing and recruitment to baits in the plots. Suckling et al. (2011) also found that the concentration of existing trails in an area can have an effect on efficacy of the trail pheromone. Higher concentrations of foraging trails reduced the disruptive ability of the trailing pheromone treatments.

15

Westermann et al. (2014) used trail pheromone to manipulate competitive interactions between Argentine ants and several other ant species in New Zealand. Baits were treated with the trail pheromone and were placed into a 10 x 10 m area. With increasing concentrations of the trail pheromone, the three competing species were able to dominate an increasingly larger proportion of 50 baits in the 10 x 10 m area.

Greenberg and Klotz (2000) Showed that addition of the Argentine ant trail pheromone

Z9-16:Ald to sucrose solutions enhanced feeding by Argentine ants in both laboratory and field tests. In the lab experiment the pheromone was combined with the sucrose solution and resulted in a >150% increase in ants feeding on the solution compared to sucrose alone. In the field experiments the pheromone was applied to a plastic membrane covering the end of a vial containing sucrose solution. Small holes in the membrane allowed ants to feed through the plastic membrane. Argentine ant feeding was enhanced at vials with the pheromone by 29% at four hours and by 33% at 24 hours when compared to control vials without pheromone.

Choe et al. (2014) combined the use of Argentine ant pheromone (Z)-9-hexadecenal with a common fipronil spray application and found that mortality and control of Argentine ants was increased with the addition of the pheromone to the pesticide spray. Workers were diverted from existing trails and nest entrances and were more likely to be exposed to the pesticide.

In a study by Choe et al. (2009), the chemistry behind necrophoresis, the removal of dead bodies of nestmates from the colony, was explored. Previous work implicated decomposition products as the signal for necrophoresis (Wilson et al. 1958). In Choe’s study, it appears that the molecules that elicit the necrophoretic response, a combination of cuticular hydrocarbons, are always present whether the ants are alive or dead. The signal is instead masked by a combination of other highly volatile compounds, dolichodial and iridomyrmecin, preventing necrophoretic behavior toward live workers by their conspecifics. The masking molecules are considerably more volatile and so, when the ant dies and ceases production of the masking molecules, the signal molecules for necrophoresis are revealed, leading to faster response times than would be

16 possible using decomposition products. Since these necrophoretic signal molecules are always present on live ants, it is possible to extract the compounds and elicit a similar necrophoretic response when they are placed on an inert object other than a nest mate’s corpse. The extract of one hour old worker corpses containing hydrocarbons led to the pupae being taken to the refuse pile, while those treated with extracts from freshly killed workers were left alone where they were placed. Pupae from the control group, solvent treated pupae, were promptly returned to the nest.

The remaining chemicals from the extract of one hour old worker corpses elicited an aggressive response and concluded with the pupae being carried to the refuse pile. This fraction mainly consisted of triglycerides, the main internal lipid component in argentine ants.

Choe et al. (2012) also found that dolichodial and iridomyrmecin were major components of the Argentine ants trail composition, while (Z)-9-hexedecenal was a minor but additive component when mixed with the former two compounds.

Tawny Crazy Ant

Taxonomy

The earliest records of the Tawny Crazy ant, Nylanderia fulva (Mayr), in North America are museum specimens from Texas in 1938 (Trager 1984). It was subsequently rediscovered in

2002 near Houston, Texas (Meyers 2008). Over the next several years N. fulva became well established and spread to surrounding counties. A large population was also found in Florida, but it is unknown how long it has been there due to confusion with a similar species, Nylanderia pubens (Forel) (Wetterer and Keularts 2008, Wetterer et al. 2014). Between 2009 and 2014, N. fulva was reported from Alabama, Georgia, Louisiana and Mississippi although it is unclear where they dispersed from, be it Texas, Florida or South America (Klotz et al. 1995, Lach and

Hooper-Bui 2010, MacGown and Layton 2010). Multiple studies using molecular and morphological characters have now concluded that the populations from both Texas and Florida are the same species, N. fulva, and that it is likely to have spread to all of the gulf coast states

(Gotzek et al. 2012, Zhao et al. 2012).

17

Ecology

The Tawny Crazy ant is an ecological menace in its introduced range. It reduces species richness and homogenizes the co-occurring ant community where it is present (LeBrun et al.

2013). LeBrun et al. (2013) found that following invasion by N. fulva, the species richness of ants and non-ant arthropods was reduced. The effect on the ant community in their study was non- random and favored the exclusion of larger-bodied and regionally distributed species, thereby favoring other exotic ant species.

The tawny crazy ant can outcompete its biggest rival the , Solenopsis invicta, where they both occur (Horn 2010). In this study, individual as well as colony level mortality was evaluated after aggressive encounters between tawny crazy ants and fire ants. In individual encounters, tawny crazy ants had higher mortality, while in colony level encounters, fire ants had higher mortality. These results suggest that the tawny crazy ant’s dominance over fire ants is not due to greater individual competitive ability but more likely due to their greater abundance in the environment where they are both present.

The abundance of the tawny crazy ant in its invasive range can be partially attributed to the increased reproductive capacity of a polygyne colony (Arcila et al. 2002, McDonald 2012).

Arcila et al. (2002) found that with increasing numbers of queens, brood production increased while brood mortality decreased. McDonald (2012) observed an increase in brood production per queen correlated with an increase in the number of queens in a colony. These attributes have a synergy that not only allows the tawny crazy ant to increase in abundance at a high rate but also make it much more formidable opponent in competitive interactions with other ant species.

Carbohydrate levels can change the aggressive behavior in tawny crazy ants, lowering their mortality during competitive interactions when compared to fire ants when carbohydrate levels are low (Horn et al. 2013). The tawny crazy ant’s dominance in its introduced range has also been hypothesized to be attributed to its ability to take advantage of and dominate disturbed habitats (Calcaterra et al. 2016). Calcaterra et al. (2016) examined several globally invasive ant

18 species in their native range, including N. fulva. It was found that N. fulva was not a competitively dominant species in its native range, and that its success in its introduced range may be due to its ability to exploit disturbed and anthropic habitats that native species do not.

The potential distribution of the tawny crazy ant in North America has been modeled, and is comparable to that of S. invicta (Kumar et al. 2015). Sharma et al. (2013) found a positive correlation between the number of N. fulva and 22 honeydew producing hemipteran species on 15 host plant species in Florida, adding to the potential negative impact this species has on its new habitat. On the US Virgin Islands, several populations of N. fulva were monitored and it was found that there was a boom and bust cycle for the populations (Wetterer and Keularts 2008). The initial population expansion was rapid but over the course of a decade, two of the three populations had completely crashed while the third remained robust. This boom and bust cycle may be a characteristic of this species in its introduced range as these cycles have been reported from Texas as well (Wetterer et al. 2014).

Physiology

Tawny crazy ants have a suite of physiological traits that make them excellent invaders of novel environments. These ants have been observed in their invasive range displacing the red imported fire ant, Solenopsis invicta, where they co-occur (LeBrun et al. 2013). This may be in part due to the ability of tawny crazy ants to detoxify the fire ants with a combination of secretions from the abdominal exocrine glands (LeBrun et al. 2014). The formic acid from the venom gland, combined with compounds from the Dufours gland, not only have a fumigant toxicity to S. invicta, but also have a synergistic detoxifying effect on the fire ant’s venom (Zhang et al. 2015, Wang and Henderson 2016). Tawny crazy ants also produce these chemicals in quantities several orders of magnitude higher than in comparable formicine ant species (Chen et al. 2013). At these rates the chemicals have a significant lethal effect on fire ants and may contribute to the tawny crazy ant’s initial dominance over S. invicta. Tawny Crazy ants have been observed spreading the chemicals over their cuticle as a defensive barrier to the opponent’s

19 venom (LeBrun et al. 2014). This adaptation may be in part due to the shared evolutionary history between the two species in their native range where there is high species richness and competition

(Kronauer 2014). LeBrun et al. (2015) has also shown that these secretions and associated behaviors are widespread in the subfamily , providing increased survivorship for these ants after conflicts with S. invicta.

Tawny crazy ants have a preference for a diet consisting of a 2:1 ratio of carbohydrates to protein when presented with a choice of several ratios (Cook et al. 2012). Colony mortality was lowest when given food at this ratio.

Thermal tolerance of the tawny crazy ant has been shown to be narrower than that of S. invicta, possibly limiting the potential range of the tawny crazy ant to one smaller than that of S. invicta in North America (Bentley et al. 2016). While thermal tolerance affects tunneling performance in both species, S. invicta increases tunneling performance at lower temperatures while N. fulva decreases tunneling performance at lower temperatures (Bentley et al. 2015).

Control

Current control methods for the tawny crazy ant involve preventative action, reducing the suitable habitats and dispersal opportunities by keeping property properly maintained and being wary of opportunistic human aided dispersal (Meyers 2008). Studies of the chemical compounds secreted by N. fulva have led to the discovery of a synergy between the ants’ venom and aggregation signals that may lead to a viable means of attracting a significant number of the ants to a bait station. Other studies have focused on microbial and viral pathogens that may be used to reduce or eliminate populations in the invasive range. Plowes et al. (2015) have described a new genus and species of microsporidian, Myrmecomorba nylanderiae, from a population found in

North American populations of N. fulva. A virus, Nylanderia fulva virus 1 (NfV-1), was discovered and phylogenetically placed in an unclassified clade of viruses (Valles et al. 2016).

This virus was not found in any other closely related species of ant. Both of these pathogens were found in all life stages of the ants and may help elucidate new approaches to controlling this

20 invasive species. Additionally, the nature of the tawny crazy ants’ invasion may provide its own control measure in the form of a boom and bust cycle which has been observed in the US Virgin

Islands, where populations initial explosion is followed by a crash in which the ant all but disappears (Wetterer et al. 2014).

Ant Fauna of Georgia and the Southeast

The Georgia ant fauna is a large one, containing 185 species in 52 genera (Ipser et al.

2004). Of those, 23 species across 16 genera are considered exotic, and originate from Asia,

Australia, Africa, Europe, Central America, and South America (Deyrup et al. 2000). The

Southeastern fauna, including Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi,

North Carolina, South Carolina, and Tennessee, contains 332 species in 54 genera, with 68 of those species being exotic, across 27 genera (MacGown 2014 web citation).

Ipser et al. (2004) conducted a survey of the Major Land Resource Areas in Georgia which consists of Atlantic Coast Flatwoods, Southern Coastal Plains, Carolina and Georgia Sand

Hills, Black Lands, Southern Piedmont, Southern Appalachian Ridges and Valleys, Sand

Mountains, and Blue Ridge. There were 29 collection sites utilizing four collection techniques including pitfall trapping, leaf litter extraction using a Berlese funnel, active search, and baiting.

Ninety six ant species were found representing 33 genera.

MacGown and Forster (2005) compiled a list of species from Alabama using recent collections, material from Auburn University Entomological Museum and Mississippi

Entomological Museum, and from literature records. The combined records bring the number of species in Alabama to 154 in 37 genera.

Deyrup (2003) updated his original list of Florida ants, bringing the number of ants in the state to 218 representing 49 genera. Of those species, 65 are considered exotic. Florida has the highest proportion and number of exotic ant species of any eastern state due to its unique position and climate, acting as a corridor of passage for species from Central and South America.

21

General and Thompson (2009) surveyed seven natural areas in five counties in Arkansas.

Collection techniques included pitfall trapping, leaf litter extraction, dissection of coarse woody debris, vegetation beating, and active search. Their collection captured 60 species in 27 genera, including twelve new records for the state and 4 species putatively new to science.

Dash and Hooper-Bùi (2008) conducted an extensive search for the ant taxa of Louisiana, utilizing literature review and museum records to supplement their field collection. Thirty seven separate sampling areas were surveyed over the course of three years. Survey techniques included a five hour period at each site consisting of leaf litter extraction using a Berlese funnel, net sweeping, arboreal and ground bait transect samples, and opportunistic collecting. One hundred and thirty two species from 40 genera were found.

MacGown and Brown (2006) surveyed the Tombigbee National Forest in Mississippi.

Samples were taken at 21 sites, over a five year period, in the forest using multiple trapping methods including pitfall traps, hand collection of ants from leaf litter samples, leaf litter extraction using a Berlese funnel, dissection of coarse woody debris, vegetation beating, arboreal and ground baiting, malaise traps, black light traps and active search. The collection yielded 70 species in 27 genera.

Using only pitfall trapping and records from several other surveys in South Carolina,

Davis (2009) was able to list 121 species of ants collected in the state. The species list was notably missing subterranean and arboreal species that will likely be found once a more diverse set of trapping methods is applied to these habitats.

22

References Cited

Abbott, K. L. 2005. Supercolonies of the invasive Yellow Crazy ant, Anoplolepis gracilipes, on

an oceanic island: forager activity patterns, density and biomass. Insectes Sociaux 52:

266-273.

Abril, S., M. Diaz, M. L. Enriquez, and C. Gomez. 2013. More and bigger queens: a clue to the

invasive success of the Argentine ant (: Formicidae) in natural habitats.

Myrmecological News 18: 19-24.

Adams, E. S. 2003. Experimental analysis of territory size in a population of the fire ant

Solenopsis invicta. Behavioral Ecology 14: 48-53.

Adams, E. S., and M. Mesterton-Gibbons. 2003. Lanchester's attrition models and fights

among social . Behavioral Ecology 14: 719-723.

Akino, T., K. Yamamura, S. Wakamura, and R. Yamaoka. 2004. Direct behavioral evidence

for hydrocarbons as nestmate recognition cues in Formica japonica (Hymenoptera:

Formicidae). Applied Entomology and Zoology 39: 381-387.

Allen, C. R., E. A. Forys, K. G. Rice, and D. P. Wojcik. 2001. Effects of fire ants

(Hymenoptera: Formicidae) on hatching turtles and prevalence of fire ants on sea turtle

nesting beaches in Florida. Florida Entomologist 84: 250-253.

Arcila, A. M., L. A. Gomez, and P. Ulloa-Chacon. 2002. Immature development and colony

growth of crazy ant fulva under laboratory conditions (Hymenoptera:

Formicidae). Sociobiology 39: 307-321.

Aron, S. 2001. Reproductive strategy: an essential component in the success of incipient colonies

of the invasive Argentine ant. Insectes Sociaux 48: 25-27.

Bentley, M., F. Oi, S. Gezan, and D. Hahn. 2015. Tunneling performance increases at lower

temperatures for Solenopsis invicta (Buren) but not for Nylanderia fulva (Mayr). Insects

6: 686.

23

Bentley, M. T., D. A. Hahn, and F. M. Oi. 2016. The thermal breadth of Nylanderia fulva

(Hymenoptera: Formicidae) is narrower than that of Solenopsis invicta at three thermal

ramping rates: 1.0, 0.12, and 0.06 degrees C min(-1). Environmental Entomology 45:

1058-1062.

Bolger, D. T., A. V. Suarez, K. R. Crooks, S. A. Morrison, and T. J. Case. 2000. Arthropods

in urban habitat fragments in southern California: area, age, and edge effects. Ecological

Applications 10: 1230-1248.

Buczkowski, G. 2016. The Trojan horse approach for managing invasive ants: a study with Asian

Needle ants, Pachycondyla chinensis. Biological Invasions 18: 507-515.

Buczkowski, G., and G. Bennett. 2009. Colony budding and its effects on food allocation in the

highly polygynous ant, Monomorium pharaonis. Ethology 115: 1091-1099.

Calcaterra, L., S. Cabrera, and J. Briano. 2016. Local co-occurrence of several highly invasive

ants in their native range: are they all ecologically dominant species? Insectes Sociaux

63: 407-419.

Chen, J., T. Rashid, G. L. Feng, L. M. Zhao, D. Oi, and B. M. Drees. 2013. Defensive

chemicals of Tawny Crazy ants, Nylanderia fulva (Hymenoptera: Formicidae) and their

toxicity to Red Imported Fire ants, Solenopsis invicta (Hymenoptera: Formicidae).

Toxicon 76: 160-166.

Cheng, D. F., G. W. Liang, and Y. J. Xu. 2016. Meta-analysis reveals asymmetric reduction in

the genetic diversity of introduced populations of exotic insects. Biological Invasions 18:

1163-1175.

Choe, D.-H., and M. K. Rust. 2008. Horizontal transfer of insecticides in laboratory colonies of

the Argentine ant (Hymenoptera: Formicidae). Journal of Economic Entomology 101:

1397-1405.

24

Choe, D.-H., J. G. Millar, and M. K. Rust. 2009. Chemical signals associated with life inhibit

necrophoresis in Argentine ants. Proceedings of the National Academy of Sciences of the

United States of America 106: 8251-8255.

Choe, D. H., D. B. Villafuerte, and N. D. Tsutsui. 2012. Trail pheromone of the Argentine ant,

Linepithema humile (Mayr) (Hymenoptera: Formicidae). Plos One 7: 7.

Choe, D. H., K. Tsai, C. M. Lopez, and K. Campbell. 2014. Pheromone-assisted techniques to

improve the efficacy of insecticide sprays against Linepithema humile (Hymenoptera:

Formicidae). Journal of Economic Entomology 107: 319-325.

Cook, S. C., R. A. Wynalda, R. E. Gold, and S. T. Behmer. 2012. Macronutrient regulation in

the (Nylanderia sp nr. pubens). Insectes Sociaux 59: 93-100.

Costa, H. S., and M. K. Rust. 1999. Mortality and foraging rates of Argentine ant

(Hymenoptera: Formicidae) colonies exposed to potted plants treated with fipronil.

Journal of Agricultural and Urban Entomology 16: 37-48.

Dash, S. T., and L. M. Hooper-Bùi. 2008. Species diversity of ants (Hymenoptera: Formicidae)

in Louisiana. Annals of the Entomological Society of America 101: 1056-1066.

Davis, T. S. 2009. The ants of South Carolina, vol. ProQuest.

Deyrup, M. 2003. An updated list of Florida ants (Hymenoptera: Formicidae). Florida

Entomologist 86: 43-48.

Deyrup, M., L. Davis, and S. Cover. 2000. Exotic ants in Florida. Transactions of the American

Entomological Society: 293-326.

Diffie, S., J. Miller, and K. Murray. 2010. Laboratory observations of Red Imported Fire ant

(Hymenoptera: Formicidae) predation on reptilian and avian eggs. Journal of

Herpetology 44: 294-296.

Eubanks, M. D. 2001. Estimates of the direct and indirect effects of Red Imported Fire ants on

biological control in field crops. Biological Control 21: 35-43.

25

Fellers, G. M., and J. H. Fellers. 1982. Scavenging rates of invertebrates in an eastern

deciduous forest. American Midland Naturalist 107: 389-392.

Folgarait, P. J. 1998. Ant biodiversity and its relationship to ecosystem functioning: a review.

Biodiversity and Conservation 7: 1221-1244.

Foucaud, J., J. Orivel, A. Loiseau, J. H. C. Delabie, H. Jourdan, D. Konghouleux, M.

Vonshak, M. Tindo, J. L. Mercier, D. Fresneau, J. B. Mikissa, T. McGlynn, A. S.

Mikheyev, J. Oettler, and A. Estoup. 2010. Worldwide invasion by the Little Fire ant:

routes of introduction and eco-evolutionary pathways. Evolutionary Applications 3: 363-

374.

Fritz, G. N., A. H. Fritz, and R. K. V. Meer. 2011. Sampling high-altitude and stratified mating

flights of Red Imported Fire ant. Journal of Medical Entomology 48: 508-512.

Gan, J., S. Bondarenko, L. Oki, D. Haver, and J. X. Li. 2012. Occurrence of fipronil and its

biologically active derivatives in urban residential runoff. Environmental Science &

Technology 46: 1489-1495.

General, D. M., and L. C. Thompson. 2009. New distributional records of ants in Arkansas for

2008. Journal of the Arkansas Academy of Science 63: 182-184.

Gotzek, D., S. G. Brady, R. J. Kallal, and J. S. LaPolla. 2012. The importance of using

multiple approaches for identifying emerging invasive species: the case of the Rasberry

Crazy ant in the United States. Plos One 7: 10.

Gotzek, D., H. J. Axen, A. V. Suarez, S. H. Cahan, and D. Shoemaker. 2015. Global invasion

history of the Tropical Fire ant: a stowaway on the first global trade routes. Molecular

Ecology 24: 374-388.

Grangier, J., and P. J. Lester. 2014. Carbohydrate scarcity increases foraging activities and

aggressiveness in the ant Prolasius advenus (Hymenoptera: Formicidae). Ecological

Entomology 39: 684-692.

26

Greenberg, L., and J. H. Klotz. 2000. Argentine ant (Hymenoptera: Formicidae) trail

pheromone enhances consumption of liquid sucrose solution. Journal of Economic

Entomology 93: 119-122.

Greenberg, L., M. K. Rust, J. H. Klotz, D. Haver, J. N. Kabashima, S. Bondarenko, and J.

Gan. 2010. Impact of ant control technologies on insecticide runoff and efficacy. Pest

Management Science 66: 980-987.

Guerra, M. D., D. R. Suiter, and C. M. Scocco. 2011. Topical toxicity of nine essential oils to

Camponotus pennsylvanicus (Hymenoptera: Formicidae). Sociobiology 58: 419-426.

Hart, L. M., and W. R. Tschinkel. 2012. A seasonal natural history of the ant, Odontomachus

brunneus. Insectes Sociaux 59: 45-54.

Hayasaka, D., N. Kuwayama, A. Takeo, T. Ishida, H. Mano, M. N. Inoue, T. Nagai, F.

Sanchez-Bayo, K. Goka, and T. Sawahata. 2015. Different acute toxicity of fipronil

baits on invasive Linepithema humile supercolonies and some non-target ground

arthropods. Ecotoxicology 24: 1221-1228.

Heller, N. E., and D. M. Gordon. 2006. Seasonal spatial dynamics and causes of nest movement

in colonies of the invasive Argentine ant (Linepithema humile). Ecological Entomology

31: 499-510.

Hölldobler, B., and E. O. Wilson. 1990. The ants, vol. Harvard University Press.

Holtcamp, W. N., W. E. Grant, and S. B. Vinson. 1997. Patch use under predation hazard:

effect of the Red Imported Fire ant on deer mouse foraging behavior. Ecology 78: 308-

317.

Holway, D. A. 1999. Competitive mechanisms underlying the displacement of native ants by the

invasive Argentine ant. Ecology 80: 238-251.

Holway, D. A., and A. V. Suarez. 2006. Homogenization of ant communities in mediterranean

California: The effects of urbanization and invasion. Biological Conservation 127: 319-

326.

27

Holway, D. A., L. Lach, A. V. Suarez, N. D. Tsutsui, and T. J. Case. 2002. The causes and

consequences of ant invasions. Annual Review of Ecology and Systematics 33: 181-233.

Hooper-Bui, L. M., and M. K. Rust. 2000. Oral toxicity of abamectin, boric acid, fipronil, and

hydramethylnon to laboratory colonies of Argentine ants (Hymenoptera: Formicidae).

Journal of Economic Entomology 93: 858-864.

Horn, K. 2010. Examining competitive interaction between Rasberry Crazy ants (Paratrechina

sp. nr. pubens) and Red Imported Fire ants (Solenopsis invicta) using laboratory and field

studies. Rice University.

Horn, K. C., M. D. Eubanks, and E. Siemann. 2013. The effect of diet and opponent size on

aggressive interactions involving Caribbean Crazy ants (Nylanderia fulva). Plos One 8: 7.

Human, K. G., and D. M. Gordon. 1999. Behavioral interactions of the invasive Argentine ant

with native ant species. Insectes Sociaux 46: 159-163.

Ichinose, K., and A. Lenoir. 2010. Hydrocarbons detection levels in ants. Insectes Sociaux 57:

453-455.

Ingram, K. K. 2002. Flexibility in nest density and social structure in invasive populations of the

Argentine ant, Linepithema humile. Oecologia 133: 492-500.

Inoue, M. N., F. Ito, and K. Goka. 2015. Queen execution increases relatedness among workers

of the invasive Argentine ant, Linepithema humile. Ecology and Evolution 5: 4098-4107.

Ipser, R. M., M. A. Brinkman, W. A. Gardner, and H. B. Peeler. 2004. A survey of ground-

dwelling ants (Hymenoptera: Formicidae) in Georgia. Florida Entomologist 87: 253-260.

Jiang, W. Y., A. Soeprono, M. K. Rust, and J. Gan. 2014. Ant control efficacy of pyrethroids

and fipronil on outdoor concrete surfaces. Pest Management Science 70: 271-277.

Jongepier, E., and S. Foitzik. 2016. Fitness costs of worker specialization for ant societies.

Proceedings of the Royal Society B-Biological Sciences 283: 7.

28

Kaplan, I., and M. D. Eubanks. 2002. Disruption of cotton (Homoptera: Aphididae) -

natural enemy dynamics by Red Imported Fire ants (Hymenoptera: Formicidae).

Environmental Entomology 31: 1175-1183.

King, J. R., and W. R. Tschinkel. 2006. Experimental evidence that the introduced fire ant,

Solenopsis invicta, does not competitively suppress co-occurring ants in a disturbed

habitat. Journal of Ecology 75: 1370-1378.

King, J. R., and S. D. Porter. 2007. Body size, colony size, abundance, and ecological impact of

exotic ants in Florida's upland ecosystems. Evolutionary Ecology Research 9: 757-774.

Klotz, J. H., M. K. Rust, L. Greenberg, and M. A. Robertson. 2010. Developing low risk

management strategies for Argentine ants (Hymenoptera: Formicidae). Sociobiology 55:

779-785.

Klotz, J. H., J. R. Mangold, K. M. Vail, L. R. Davis, and R. S. Patterson. 1995. A survey of

the urban pest ants (Hymenoptera: Formicidae) of peninsular Florida. The Florida

Entomologist 78: 109-118.

Klotz, J. H., M. K. Rust, H. S. Costa, D. A. Reierson, and K. Kido. 2002. Strategies for

controlling Argentine ants (Hymenoptera: Formicidae) with sprays and baits. Journal of

Agricultural and Urban Entomology 19: 85-94.

Klotz, J. H., M. K. Rust, L. Greenberg, H. C. Field, and K. Kupfer. 2007. An evaluation of

several urban pest management strategies to control Argentine ants (Hymenoptera:

Formicidae). Sociobiology 50: 391-398.

Klotz, J. H., M. K. Rust, H. C. Field, L. Greenberg, and K. Kupfer. 2008. Controlling

Argentine ants in residential settings (Hymenoptera: Formicidae). Sociobiology 51: 579-

588.

Kronauer, D. J. C. 2014. Invasive species: old foes meet again. Current Biology 24: R372-R374.

29

Krushelnycky, P. D., L. L. Loope, and S. M. Joe. 2004. Limiting spread of a unicolonial

invasive and characterization of seasonal patterns of range expansion. Biological

Invasions 6: 47-57.

Kumar, S., E. G. LeBrun, T. J. Stohlgren, J. A. Stabach, D. L. McDonald, D. H. Oi, and J.

S. LaPolla. 2015. Evidence of niche shift and global invasion potential of the Tawny

Crazy ant, Nylanderia fulva. Ecology and Evolution 5: 4628-4641.

Lach, L., and L. M. Hooper-Bui. 2010. Consequences of ant invasions. Ant ecology. Oxford

University Press, Oxford: 261-286.

Lach, L., C. V. Tillberg, and A. V. Suarez. 2010. Contrasting effects of an invasive ant on a

native and an invasive plant. Biological Invasions 12: 3123-3133.

LeBrun, E. G., J. Abbott, and L. E. Gilbert. 2013. Imported crazy ant displaces imported fire

ant, reduces and homogenizes grassland ant and arthropod assemblages. Biological

Invasions 15: 2429-2442.

LeBrun, E. G., N. T. Jones, and L. E. Gilbert. 2014. Chemical warfare among invaders: a

detoxification interaction facilitates an ant invasion. Science 343: 1014-1017.

LeBrun, E. G., P. J. Diebold, M. R. Orr, and L. E. Gilbert. 2015. Widespread chemical

detoxification of venom by formicine ants. Journal of Chemical Ecology 41:

884-895.

Lengyel, S., A. D. Gove, A. M. Latimer, J. D. Majer, and R. R. Dunn. 2009. Ants sow the

seeds of global diversification in flowering plants. Plos One 4: 6.

Lessard, J. P., J. A. Fordyce, N. J. Gotelli, and N. J. Sanders. 2009. Invasive ants alter the

phylogenetic structure of ant communities. Ecology 90: 2664-2669.

Linksvayer, T. A., and M. A. Janssen. 2009. Traits underlying the capacity of ant colonies to

adapt to disturbance and stress regimes. Systems Research and Behavioral Science 26:

315-329.

30

Long, A. K., L. M. Conner, L. L. Smith, and R. A. McCleery. 2015. Effects of an invasive ant

and native predators on cotton rat recruitment and survival. Journal of Mammalogy 96:

1135-1141.

Lu, Y. Y., B. Q. Wu, Y. J. Xu, and L. Zeng. 2012. Effects of Red Imported Fire ants

(Solenopsis invicta) on the species structure of ant communities in South China.

Sociobiology 59: 275-285.

MacGown, J., and B. Layton. 2010. The invasive Rasberry crazy ant, Nylanderia sp. near

pubens (Hymenoptera: Formicidae), reported from Mississippi. Midsouth Entomol 3: 44-

47.

MacGown, J. A., and J. A. Forster. 2005. A preliminary list of the ants (Hymenoptera:

Formicidae) of Alabama, USA. Entomol. News 116: 61-74.

Macgown, J. A., and R. L. Brown. 2006. Survey of ants (Hymenoptera: Formicidae) of the

Tombigbee National Forest in Mississippi. Journal of the Kansas Entomological Society

79: 325-340.

Mahler, B. J., P. C. Van Metre, J. T. Wilson, M. Musgrove, S. D. Zaugg, and M. R.

Burkhardt. 2009. Fipronil and its degradates in indoor and outdoor dust. Environmental

Science & Technology 43: 5665-5670.

McDonald, D. L. 2012. Investigation of an invasive ant species: Nylanderia fulva colony

extraction, management, diet preference, fecundity, and mechanical vector potential.

Texas A&M University.

McGlynn, T. P. 1999. Non-native ants are smaller than related native ants. American Naturalist

154: 690-699.

Meyers, J. M. 2008. Identification, distribution and control of an invasive. Texas A&M

University.

31

Morrison, L. W., and S. D. Porter. 2005. Testing for population-level impacts of introduced

Pseudacteon tricuspis flies, phorid parasitoids of Solenopsis invicta fire ants. Biological

Control 33: 9-19.

Oi, D. H., K. M. Vail, D. F. Williams, and D. N. Bieman. 1994. Indoor and outdoor foraging

locations of pharaoh ants (Hymenoptera: Formicidae) and control strategies using bait

stations. Florida Entomologist 77: 85-91.

Phillips, B., and A. V. Suarez. 2012. The role of behavioural variation in the invasion of new

areas, vol. Oxford Univ Press, New York.

Plowes, R. M., J. J. Becnel, E. G. LeBrun, D. H. Oi, S. M. Valles, N. T. Jones, and L. E.

Gilbert. 2015. Myrmecomorba nylanderiae gen. et sp nov., a microsporidian parasite of

the Tawny Crazy ant Nylanderia fulva. Journal of Invertebrate Pathology 129: 45-56.

Porter, S. D., R. K. Vandermeer, M. A. Pesquero, S. Campiolo, and H. G. Fowler. 1995.

Solenopsis (Hymenoptera: Formicidae) fire ant reactions to attacks of pseudacteon flies

(Diptera: Phoridae) in Southeastern Brazil. Annals of the Entomological Society of

America 88: 570-575.

Roura-Pascual, N., C. Hui, T. Ikeda, G. Leday, D. M. Richardson, S. Carpintero, X.

Espadaler, C. Gomez, B. Guenard, S. Hartley, P. Krushelnycky, P. J. Lester, M. A.

McGeoch, S. B. Menke, J. S. Pedersen, J. P. W. Pitt, J. Reyes, N. J. Sanders, A. V.

Suarez, Y. Touyama, D. Ward, P. S. Ward, and S. P. Worner. 2011. Relative roles of

climatic suitability and anthropogenic influence in determining the pattern of spread in a

global invader. Proceedings of the National Academy of Sciences of the United States of

America 108: 220-225.

Sanders, N. J., and A. V. Suarez. 2011. Elton's insights into the ecology of ant invasions:

lessons learned and lessons still to be learned, vol. Wiley-Blackwell, Malden.

32

Sanders, N. J., N. J. Gotelli, N. E. Heller, and D. M. Gordon. 2003. Community disassembly

by an invasive species. Proceedings of the National Academy of Sciences of the United

States of America 100: 2474-2477.

Sanford, M. P., P. N. Manley, and D. D. Murphy. 2009. Effects of urban development on ant

communities: implications for ecosystem services and management. Conservation

Biology 23: 131-141.

Sarty, M., K. L. Abbott, and P. J. Lester. 2007. Community level impacts of an ant invader and

food mediated coexistence. Insectes Sociaux 54: 166-173.

Scocco, C. M., D. R. Suiter, and W. A. Gardner. 2012. Repellency of five essential oils to

Linepithema humile (Hymenoptera: Formicidae). Journal of Entomological Science 47:

150-159.

Sharma, S., D. H. Oi, and E. A. Buss. 2013. Honeydew-producing hemipterans in Florida

associated with Nylanderia fulva (Hymenoptera: Formicidae), an invasive crazy ant.

Florida Entomologist 96: 538-547.

Sterling, W. L. 1978. Fortuitous biological suppression of the boll weevil by the Red Imported

Fire ant. Environmental entomology 7: 564-568.

Stuble, K. L., C. M. Patterson, M. A. Rodriguez-Cabal, R. R. Ribbons, R. R. Dunn, and N.

J. Sanders. 2014. Ant-mediated seed dispersal in a warmed world. Peerj 2: 15.

Sturgis, S. J., and D. M. Gordon. 2012. Nestmate recognition in ants (Hymenoptera:

Formicidae): a review. Myrmecological News 16: 101-110.

Suarez, A. V., and T. J. Case. 2002. Bottom-up effects on persistence of a specialist predator:

ant invasions and horned lizards. Ecological Applications 12: 291-298.

Suarez, A. V., and E. L. Suhr. 2012. Ecological and evolutionary perspectives on

"supercolonies": a commentary on Moffett. Behavioral Ecology 23: 937-938.

33

Suarez, A. V., J. Q. Richmond, and T. J. Case. 2000. Prey selection in horned lizards following

the invasion of Argentine ants in southern California. Ecological Applications 10: 711-

725.

Suarez, A. V., D. A. Holway, and T. J. Case. 2001. Patterns of spread in biological invasions

dominated by long-distance jump dispersal: insights from Argentine ants. Proceedings of

the National Academy of Sciences of the United States of America 98: 1095-1100.

Suarez, A. V., P. Yeh, and T. J. Case. 2005. Impacts of Argentine ants on avian nesting success.

Insectes Sociaux 52: 378-382.

Suarez, A. V., D. A. Holway, and N. D. Tsutsui. 2008. Genetics and behavior of a colonizing

species: the invasive Argentine ant. American Naturalist 172: S72-S84.

Suarez, A. V., T. P. McGlynn, and N. D. Tsutsui. 2010. Biogeographic and taxonomic patterns

of introduced ants, pp. 233-244. In L. Lach, C.L. Parr, and K.L. Abbott (eds.), Ant

Ecology. Oxford University Press, Oxford, NY.

Suarez, A. V., D. A. Holway, D. S. Liang, N. D. Tsutsui, and T. J. Case. 2002. Spatiotemporal

patterns of intraspecific aggression in the invasive Argentine ant. Animal Behaviour 64:

697-708.

Suckling, D. M., L. D. Stringer, and J. E. Corn. 2011. Argentine ant trail pheromone disruption

is mediated by trail concentration. Journal of Chemical Ecology 37: 1143-1149.

Suckling, D. M., R. W. Peck, L. D. Stringer, K. Snook, and P. C. Banko. 2010. Trail

pheromone disruption of Argentine ant trail formation and foraging. Journal of Chemical

Ecology 36: 122-128.

Suckling, D. M., R. W. Peck, L. M. Manning, L. D. Stringer, J. Cappadonna, and A. M. El-

Sayed. 2008. Pheromone disruption of Argentine ant trail integrity. Journal of Chemical

Ecology 34: 1602-1609.

34

Torres, C. W., M. Brandt, and N. D. Tsutsui. 2007. The role of cuticular hydrocarbons as

chemical cues for nestmate recognition in the invasive Argentine ant (Linepithema

humile). Insectes Sociaux 54: 363-373.

Trager, J. C. 1984. A revision of the genus Paratrechina (Hymenoptera: Formicidae) of the

continental United States. Sociobiology 9: 51-162.

Tschinkel, W. R. 2006. The fire ants, vol. Harvard University Press.

Tsutsui, N. D., and A. V. Suarez. 2003. The colony structure and population biology of invasive

ants. Conservation Biology 17: 48-58.

Tsutsui, N. D., A. V. Suarez, D. A. Holway, and T. J. Case. 2000. Reduced genetic variation

and the success of an invasive species. Proceedings of the National Academy of Sciences

of the United States of America 97: 5948-5953.

Valery, L., H. Fritz, J. C. Lefeuvre, and D. Simberloff. 2008. In search of a real definition of

the biological invasion phenomenon itself. Biological Invasions 10: 1345-1351.

Valles, S. M., D. H. Oi, J. J. Becnel, J. K. Wetterer, J. S. LaPolla, and A. E. Firth. 2016.

Isolation and characterization of Nylanderia fulva virus 1, a positive-sense, single-

stranded RNA virus infecting the Tawny Crazy ant, Nylanderia fulva. Virology 496: 244-

254. van Wilgenburg, E., M. R. E. Symonds, and M. A. Elgar. 2011. Evolution of cuticular

hydrocarbon diversity in ants. Journal of Evolutionary Biology 24: 1188-1198. van Wilgenburg, E., R. Sulc, K. J. Shea, and N. D. Tsutsui. 2010. Deciphering the chemical

basis of nestmate recognition. Journal of Chemical Ecology 36: 751-758.

Vasquez, G. M., and J. Silverman. 2008. Queen acceptance and the complexity of nestmate

discrimination in the Argentine ant. Behavioral Ecology and Sociobiology 62: 537-548.

Vasquez, G. M., C. Schal, and J. Silverman. 2008. Cuticular hydrocarbons as queen adoption

cues in the invasive Argentine ant. Journal of Experimental Biology 211: 1249-1256.

35

Vega, S. Y., and M. K. Rust. 2003. Determining the foraging range and origin of resurgence

after treatment of Argentine ant (Hymenoptera: Formicidae) in urban areas. Journal of

Economic Entomology 96: 844-849.

Villalta, I., F. Amor, X. Cerda, and R. Boulay. 2016. Social coercion of larval development in

an ant species. Science of Nature 103: 8.

Wang, C., and G. Henderson. 2016. Repellent effect of formic acid against the Red Imported

Fire ant (Hymenoptera: Formicidae): a field study. Journal of Economic Entomology

109: 779-784.

Ward, D. F., R. J. Harris, and M. C. Stanley. 2005. Human-mediated range expansion of

Argentine ants Linepithema humile (Hymenoptera: Formicidae) in New Zealand.

Sociobiology 45: 401-407.

Ward, D. F., J. R. Beggs, M. N. Clout, R. J. Harris, and S. O'Connor. 2006. The diversity and

origin of exotic ants arriving in New Zealand via human-mediated dispersal. Diversity

and Distributions 12: 601-609.

Westermann, F. L., D. M. Suckling, and P. J. Lester. 2014. Disruption of foraging by a

dominant invasive species to decrease its competitive ability. Plos One 9: 8.

Wetterer, J. K., and J. L. Keularts. 2008. Population explosion of the Hairy Crazy ant,

Paratrechina pubens (Hymenoptera: Formicidae), on St. Croix, US Virgin Islands.

Florida Entomologist 91: 423-427.

Wetterer, J. K., O. Davis, and J. R. Williamson. 2014. Boom and bust of the Tawny Crazy ant,

Nylanderia fulva (Hymenoptera: Formicidae), on St. Croix, US Virgin Islands. Florida

Entomologist 97: 1099-1103.

Wetterer, J. K., A. L. Wild, A. V. Suarez, N. Roura-Pascual, and X. Espadaler. 2009.

Worldwide spread of the Argentine ant, Linepithema humile (Hymenoptera: Formicidae).

Myrmecological News 12: 187-194.

36

Wilder, S. M., D. A. Holway, A. V. Suarez, E. G. LeBrun, and M. D. Eubanks. 2011.

Intercontinental differences in resource use reveal the importance of mutualisms in fire

ant invasions. Proceedings of the National Academy of Sciences of the United States of

America 108: 20639-20644.

Wilder, S. M., T. R. Barnum, D. A. Holway, A. V. Suarez, and M. D. Eubanks. 2013.

Introduced fire ants can exclude native ants from critical mutualist-provided resources.

Oecologia 172: 197-205.

Wills, B. D., C. S. Moreau, B. D. Wray, B. D. Hoffmann, and A. V. Suarez. 2014. Body size

variation and caste ratios in geographically distinct populations of the invasive big-

headed ant, megacephala (Hymenoptera: Formicidae). Biological Journal of the

Linnean Society 113: 423-438.

Wilson, E. O. 1987. Causes of ecological success - the case of the ants - the 6th Tansley lecture.

Journal of Animal Ecology 56: 1-9.

Wilson, E. O., N. I. Durlach, and L. M. Roth. 1958. Chemical releaser of necrophoric behavior

in ants. Psyche 65: 108-114.

Wiltz, B. A., D. R. Suiter, and W. A. Gardner. 2007. Deterrency and toxicity of essential oils to

Argentine and Red Imported Fire ants (Hymenoptera: Formicidae). Journal of

Entomological Science 42: 239-249.

Wiltz, B. A., D. R. Suiter, and W. A. Gardner. 2009. Activity of bifenthrin, chlorfenapyr,

fipronil, and thiamethoxam against Argentine ants (Hymenoptera: Formicidae). Journal

of Economic Entomology 102: 2279-2288.

Yang, A. S. 2006. Seasonality, division of labor, and dynamics of colony-level nutrient storage in

the ant Pheidole morrisi. Insectes Sociaux 53: 456-462.

Yip, E. C. 2014. Ants versus spiders: interference competition between two social predators.

Insectes Sociaux 61: 403-406.

37

Zhang, Q. H., D. L. McDonald, D. R. Hoover, J. R. Aldrich, and R. G. Schneidmiller. 2015.

North American invasion of the Tawny Crazy ant (Nylanderia fulva) is enabled by

pheromonal synergism from two separate glands. Journal of Chemical Ecology 41: 853-

858.

Zhao, L. M., J. Chen, W. A. Jones, D. H. Oi, and B. M. Drees. 2012. Molecular comparisons

suggest Caribbean Crazy ant from Florida and Rasberry Crazy ant from Texas

(Hymenoptera: Formicidae: Nylanderia) are the same species. Environmental

Entomology 41: 1008-1018.

38

CHAPTER 2

A SURVEY OF ANT SPECIES FROM THE PORT OF SAVANNAH, GARDEN CITY,

GEORGIA (CHATHAM COUNTY)

Introduction

Seaports act as a critical interface between land and sea for the international transportation of goods (Blonigen and Wilson 2006). Moreover, a seaport serves as a key node within international production and distribution networks (Carbone and Martino 2003).

Approximately 80% of the world’s trade is conducted by vessels traveling between seaports across the globe (United Nations Conference on Trade and Development 2015). A correlation between increasing gross domestic product (GDP) and the importation of exotic species means that the inexorable economic growth brought about by increasing international trade will also bring with it an increasing number of exotic species (Hulme 2009). In this way, ports play an important role in their surrounding ecosystem, serving as an interface between ecology and economics and being hot spots for the introduction of exotic species (Stohlgren and Schnase

2006).

As Suarez et al. (2005b) has shown, port cities have played an important role in the introduction of exotic ant species many times over, with species such as the red imported fire ant,

Solenopsis invicta Buren, the Argentine ant, Linepithema humile Mayr, the tawny crazy ant,

Nylanderia fulva Mayr, the big-headed ant, (Fabricius), and its congener,

Pheidole obscurithorax Naves, arriving first at or near port cities along the Gulf Coast of the

United States. The Port of Savannah is the fourth busiest maritime port in North America, and has the greatest potential for future growth among the five busiest ports in North America. Its location, approximately 30 kilometers inland from the Atlantic coast on the Savannah River, would allow invading species to bypass potentially hostile coastal habitats.

39

Monitoring species occurrences is a difficult and ongoing task handled by the cooperative efforts between the Georgia Ports Authority and several government agencies including the U.S

Department of Agriculture, State Departments of Agriculture, and U.S Customs and Border

Protection. While these agencies monitor the cargo arriving at and leaving from the Port of

Savannah, the habitable areas surrounding the port remain uninspected.

This objective of this survey was to identify the ant species residing on and immediately around the Port of Savannah property. The survey focused on two types of habitats, wooded areas and roadsides, which are conducive to habitation by ant species. This survey also aimed to identify exotic ant species with the potential to be invasive, should they develop a stable population in the area.

Materials and Methods

Site Description

The Port of Savannah in Garden City, Georgia (32°08’00.00 N, 81°09’00.00 W), is located approximately 30 kilometers inland from the Atlantic shore, on the Savannah River. The property consists of a 3 km docking area along the edge of the Savannah River. Immediately adjacent to the docking area are large paved areas for container storage. Beyond this are several forested/green areas fragmented by paved and dirt roads. Furthest from the river is the rail yard accompanied by more paved container storage.

Thirteen sampling sites within a survey area ≈1 km2 were chosen based on a preliminary visual assessment of the port property in June 2015 (Figure 2.1). Nine sites located in wooded areas, both on and surrounding the port property, were selected for transect sampling.

Additionally, four sites located along paved and unpaved roadsides were selected to assess the ant community in highly disturbed habitats.

Transect Sampling in Wooded Areas

The forested fragments within the survey area were sampled using one to three 90 m transects, depending on fragment size. Transects were sampled once during June or July in 2015

40 and 2016. Samples were collected using five methods: pitfall trapping, leaf litter extraction, standardized dissection of coarse woody debris, standardized active searching on tree trunks, and baiting. Sampling was conducted over the course of three days to allow for 48 hours of pitfall activity as described by Agosti and Alonso (2000b). Day one consisted of the placement of all pitfall traps along three transects and the sampling of coarse woody debris, tree trunks, and baits on one transect. On the second day, coarse woody debris, tree trunk, and bait samples were taken on the two remaining transects. The third day entailed the collection of all leaf litter samples and removal of pitfall traps that were installed on the first day. This three day protocol was repeated two more times to complete the sampling of all nine transects.

Pitfall traps were placed individually along each transect at ten meter intervals, for a total of ten traps per transect. Pitfall traps were 9 dram plastic vials (Bioquip Products Inc., Rancho

Dominguez, CA) filled to two centimeters from the bottom with propylene glycol (Figure 2.2 A).

Propylene glycol acts as a preservative and does not readily evaporate when left in the field for several days. A small area was cleared of leaf litter and debris to expose the ground underneath.

The traps were installed in the ground with the opened top flush with the soil. Traps were covered by a plastic plate, suspended approximately five centimeters from the ground by three nails, to prevent them from filling with rainwater (Figure 2.2 B). Traps remained in the ground for 48 hours before they were collected.

Four additional sample types, around each pitfall trap, were collected at least five meters from the trap and each other. Leaf litter samples were hand collected by filling a one gallon bag with leaves and the top layer of soil below the leaves from a single location (Figure 2.3). Bags were then placed in a cool environment until they were returned to the laboratory. In the laboratory, ants were extracted from litter by placing the samples in Berlese funnels, which were allowed to run for one week.

Ants were collected from randomly selected coarse woody debris (≈ 5-10-cm ID and no longer than 100-cm in length) dissected in a large white pan and searched for ants by two people

41 for five minutes (Figure 2.4). The ants were collected using aspirators and then transferred to vials of 75% ethanol.

Tree trunks ≈ 5-20-cm ID at eye level were selected and litter, debris, and vegetation removed from the base of the tree (Figure 2.5 A). Each tree were visually searched for ants by two people for 5 minutes. Trees were searched from the ground associated with the root zone to six feet high on the tree’s trunk. Flashlights were used to enhance visualization of ants (Figure 2.5

B). Ants were aspirated from the trunk and then transferred to vials of 75% ethanol.

Baits were made by placing crushed Pecan Sandie cookies (Keebler Company, Battle

Creek, MI) in 9 dram vials, filling each vial to one centimeter from the bottom. A bait vial was upended allowing the cookie to fall onto the vial’s cap and the vial left on its side next to the cap on the ground (Figure 2.6). After one hour, ants on the bait were collected using aspirators and then transferred to vials of 75% ethanol.

Roadside Sampling

Baiting and quadrat sampling were used to collect ants from four roadside habitat sites within the 1 km2 survey area. The four sites selected surrounded the forested fragments that were sampled using transects. Each site was sampled once during June or July of 2015 and again in

2016. Two of the sites were also sampled in October 2015 and March 2016 as part of a concurrent study (Chapter 3). Baiting and quadrat sampling were alternated along the roadside at a minimum of ten meter intervals and maximum of 25 meters. Twenty baiting samples and 20 quadrat samples were collected from each of the four sites at each sampling date. However, the small size of one of the sites only allowed for the collection of ten baiting and ten quadrat samples.

The baiting method described for collecting ants in the forested sites was also used for roadsides. The baits were placed in grassy areas within a meter of the road’s edge. Additionally, quadrat sampling was conducted by placing 0.5 x 0.5 m quadrats half on the paved/dirt road and half on the grassy area flanking the road (Figure 2.7). A flashlight was used to enhance the

42 visualization of ants in and amongst the vegetation inside the quadrat. Vegetation and grass was gently moved and/or removed to highlight ant occurrence. All ants seen in the area of the quadrat during a five minute period were collected using an aspirator and then transferred to vials containing 75% ethanol.

All ants were identified by BMG using Joe MacGown’s “Ants (Formicidae) of the

Southeastern United States” key (MacGown 2017). Voucher specimens reside at the University of Georgia Collection of Arthropods (UGCA) in Athens, Georgia.

Results and Discussion

Sixty-five thousand four hundred and twenty ants were collected and identified over the course of this study. Forty-six ant species representing 20 genera were collected including 14 exotic species from ten genera (Table 2.1). Two species had not been previously reported from

Georgia, while 31 species were new records for Chatham County, Georgia (Wheeler 1913, Ipser et al. 2004). Two exotic species were new to the state of Georgia; Brachymyrmex obscurior Forel, and Cardiocondyla venustula Wheeler.

The number of species identified for the first time in Chatham County may be due to a dearth of previous surveys. The most recent survey to include Chatham County was conducted by

Ipser et al. (2004). In that survey, 27 species from Chatham County were collected, including 15 that were also collected in this survey. However, Ipser et al. (2004) focused on ground dwelling species, whereas this survey employed several sampling methods aimed at collecting ants from other habitats. A list of ants from Georgia was published in 1913 (Wheeler 1913). Beyond these two publications, the ant fauna from Georgia and Chatham County are known only from museum records and collections.

This survey employed 1300 individual samples to achieve adequate detection of species, determined by rarefaction estimators, at the Port of Savannah (Figure 2.8). Each collection method on the wooded transects totaled 180 samples over the two years for a combined 900 samples. Each collection method on the roadsides totaled 200 samples over the two years for a

43 combined 400 samples. In only 30 samples were no ants collected, 21 from wooded transects and

9 from roadsides. Overall, ants were collected in all but 2.3% of samples (Table 2.2).

All sampling methods combined on the nine wooded transects collected 62,434 ants representing 43 species while both of the collection methods (bait and quadrat) combined on the roadsides collected 2,986 ants representing 16 species (Table 2.2). Pitfall trapping collected 29 species, leaf litter extraction collected 31 species, dissection of coarse woody debris collected 28 species, active searching on tree trunks collected 32 species, Pecan Sandie baits collected 22 species, and active search in quadrats collected 14 species. Pecan Sandie baiting was conducted on both wooded transects and roadsides, giving it a larger potential number of species to collect.

Active search in quadrats was conducted only along roadsides, likely contributing to its low species catch when compared to the other methods.

Six species were collected by all sampling methods and were collected from both habitat types. These species were Aphaenogaster fulva Roger, Cyphomyrmex rimosus (Spinola),

Monomorium minimum (Buckley), N. fulva, Pheidole navigans Forel, and S. invicta.

Aphaenogaster fulva and M. minimum are native species while the remaining four are exotic.

Nylanderia fulva and S. invicta are invasive species and are known to disrupt ant communities where they occur (Porter et al. 1988, Porter and Savignano 1990, LeBrun et al. 2013).

Nylanderia fulva is reported here for the first time in Chatham County. This occurrence marks the northern most expansion of the species on the east coast of the United States and its presence on an international seaport poses the question of whether this species was transported to this location from another location in the U.S. or from elsewhere via the seaport. It also highlights the possibility of further transport aided by the large amount of traffic on the port both to and from domestic and foreign locations. The potential for human aided dispersal of this species is magnified with a growing population residing on Port of Savannah property.

This survey increases the ant species known in Georgia by two, to a total of 188, and the number of exotic ant species by two to a total of 26. These numbers are comparable to adjacent

44 states with a trend of decreasing species and decreasing exotic species with increasing latitude and as you move away from the coast. The large number of exotic species in Alabama and

Georgia may be explained by their position in relation to Florida, a state whose subtropical habitats are amenable to establishment by exotic species from tropical regions, and by the presence of large international maritime ports on their coasts.

In recent years there has been an increase in the number of studies aimed at identifying the role of ports in the introduction of exotic ant species. Several port cities in Japan have a high ratio of exotic to native ant species (23% and 64% of the ant fauna), comparable to that of the port of Savannah (29.8%) in relation to the state of Georgia (13.8%) (Table 2.3) (Harada et al.

2013, Harada et al. 2014). The abundance of exotic ant species in port cities can decrease the species richness of the ant community therein, paving the way for further invasion (Sunamura et al. 2007, LeBrun et al. 2013). The importance of monitoring these critical points of entry for invasive ant species has been highlighted by the repeated introduction of exotic species through these pathways (Suarez et al. 2005b, Stohlgren and Schnase 2006, Hulme 2009, Sakamoto et al.

2016). Increasing international trade brings with it the risk of invasion by exotic ant species, which has attracted interest in the potential for future invasions and their impact in uninvaded areas (Bertelsmeier and Courchamp 2014). The analysis of the trade pathways being exploited by exotic ants warrants further study with a focus on ports as a critical point for interception and eradication (Ward et al. 2006).

This survey provides a detailed look into the state of the ant community associated with an international seaport. It also provides a basic tool for informing interception and containment procedures previous to and following the discovery of a new invasive pest species.

45

References Cited

Agosti, D., and L. E. Alonso. 2000. The ALL protocol. Ants: standard methods for measuring

and monitoring biodiversity. Smithsonian Institution Press, Washington, DC 280: 204-

206.

Bertelsmeier, C., and F. Courchamp. 2014. Future ant invasions in France. Environ. Conserv.

41: 217-228.

Blonigen, B. A., and W. Wilson. 2006. New measures of port efficiency using international

trade data. National Bureau of Economic Research, Cambridge, MA.

Carbone, V., and M. D. Martino. 2003. The changing role of ports in supply-chain

management: an empirical analysis. Maritime Policy & Management 30: 305-320.

Harada, Y., D. Fukukura, R. Kurisu, and S. Yamane. 2013. Ants of ports-monitoring of alien

ant species. Bull. Biogeogr. Soc. Jpn. 68: 29-40.

Harada, Y., T. Yamaguchi, D. Fukukura, and H. Mizumata. 2014. Ants of ports on Amami

Islands-monitoring of alien ant species. Bull. Biogeogr. Soc. Jpn. 69: 83-90.

Hulme, P. E. 2009. Trade, transport and trouble: managing invasive species pathways in an era

of globalization. J. Appl. Ecol. 46: 10-18.

Ipser, R. M., M. A. Brinkman, W. A. Gardner, and H. B. Peeler. 2004. A survey of ground-

dwelling ants (Hymenoptera: Formicidae) in Georgia. Fla. Entomol. 87: 253-260.

LeBrun, E. G., J. Abbott, and L. E. Gilbert. 2013. Imported crazy ant displaces imported fire

ant, reduces and homogenizes grassland ant and arthropod assemblages. Biol. Invasions

15: 2429-2442.

MacGown, J. 2017. Ants (Formicidae) of the Southeastern United States.

http://mississippientomologicalmuseum.org.msstate.edu/Researchtaxapages/Formicidaep

ages/Identification.Keys.htm#.WbLR9sbZfVh.

Porter, S. D., and D. A. Savignano. 1990. Invasion of polygyne fire ants decimates native ants

and disrupts arthropod community. Ecology 71: 2095-2106.

46

Porter, S. D., B. Van Eimeren, and L. E. Gilbert. 1988. Invasion of red imported fire ants

(Hymenoptera: Formicidae): microgeography of competitive replacement. Ann. Entomol.

Soc. Am. 81: 913-918.

Sakamoto, Y., H. Mori, H. Ohnishi, H. Imai, T. Kishimoto, M. Toda, S. Kishi, and K. Goka.

2016. Surveys of the ant faunas at ports of Tokyo Bay and the Ogasawara Islands. Appl.

Entomol. Zool. 51: 661-667.

Stohlgren, T. J., and J. L. Schnase. 2006. Risk analysis for biological hazards: what we need to

know about invasive species. Risk analysis 26: 163-173.

Suarez, A. V., D. A. Holway, and P. S. Ward. 2005. The role of opportunity in the

unintentional introduction of nonnative ants. Proc. Natl. Acad. Sci. U. S. A. 102: 17032-

17035.

Sunamura, E., K. Nishisue, M. Terayama, and S. Tatsuki. 2007. Invasion of four Argentine

ant supercolonies into Kobe Port, Japan: their distributions and effects on indigenous ants

(Hymenoptera: Formicidae). Sociobiology 50: 659-674.

United Nations Conference on Trade and Development. 2015. Review of Maritime Transport.

UNCTAD, New York and Geneva.

Ward, D. F., J. R. Beggs, M. N. Clout, R. J. Harris, and S. O'Connor. 2006. The diversity and

origin of exotic ants arriving in New Zealand via human-mediated dispersal. Divers.

Distrib. 12: 601-609.

Wheeler, W. M. 1913. Ants collected in Georgia by Dr. JC Bradley and Mr. WT Davis. Psyche

20: 112-117.

47

Table 2.1. Ant species collected on and around the port of Savannah with collection method and habitat type noted. Wooded Species Pitfall Leaf litter Logs Trees Baits Quadrats transects Roadsides Aphaenogaster carolinensis Wheeler2 + + + + + - + -

Aphaenogaster fulva Roger2 + + + + + + + +

Brachymyrmex depilis Emery + + - + - + + +

Brachymyrmex obscurior Forel1 2 3 - - + - - - + -

Brachymyrmex patagonicus Mayr1 - - + + + + + +

Brachyponera chinensis (Emery)1 2 + + + + + - + -

Camponotus castaneus (Latreille) + - - + - - + -

Camponotus floridanus (Buckley) - - - + + - + -

Camponotus pennsylvanicus (De Geer)2 + - - + + + + +

Camponotus snellingi Bolton2 - - - + - - + -

Cardiocondyla venustula Wheeler1 2 3 - - - - + - - +

Cardiocondyla wroughtonii (Forel)1 2 - - - - - + - +

Colobopsis impressa Roger2 + + + + - - + -

Crematogaster ashmeadi Mayr + + + + - - + -

1 Denotes exotic species; 2 Species new to Chatham County; 3 Species new to Georgia. (+) Denotes a species collection by a sampling method or in a habitat. (-) Denotes a species absence from a sampling method or habitat type.

48

Table 2.1. (Continued) Ant species collected on and around the port of Savannah with collection method and habitat type noted. Wooded Species Pitfall Leaf litter Logs Trees Baits Quadrats transects Roadsides Crematogaster atkinsoni Wheeler2 - - + - - - + -

Crematogaster minutissima Mayr2 + + + + - - + -

Crematogaster pilosa Emery2 + + - + + - + -

Crematogaster pinicola Deyrup & Cover2 + - + + + - + -

Cyphomyrmex rimosus (Spinola)1 2 + + + + + + + +

Dorymyrmex bureni (Trager) 2 3 - - - - - + - +

Hypoponera opaciceps (Mayr)1 + + + + - - + -

Hypoponera opacior (Forel) + + + + - + + +

Lasius alienus (Foerster) + + + + + - + -

Linepithema humile Mayr1 2 - + - + - + + +

Monomorium minimum (Buclkey)2 + + + + + + + +

Myrmecina americana Emery2 - + - - - - + -

Nylanderia concinna (Trager)2 + + + + + - + -

Nylanderia faisonensis Forel + + + + + - + -

1 Denotes exotic species; 2 Species new to Chatham County; 3 Species new to Georgia. (+) Denotes a species collection by a sampling method or in a habitat. (-) Denotes a species absence from a sampling method or habitat type.

49

Table 2.1. (Continued) Ant species collected on and around the port of Savannah with collection method and habitat type noted. Wooded Species Pitfall Leaf litter Logs Trees Baits Quadrats transects Roadsides Nylanderia fulva Mayr1 2 + + + + + + + +

Pheidole bilimeki Mayr2 - - + - - - + -

Pheidole dentata Mayr + + + + + - + +

Pheidole dentigula Smith + + + + + - + -

Pheidole navigans Forel1 2 + + + + + + + +

Pheidole obscurithorax Naves1 2 - + - - + + + +

Pseudomyrmex ejectus (Smith)2 + - + + - - + -

Solenopsis abdita Thompson2 + + + + + - + -

Solenopsis carolinensis Forel - - - + - - + -

Solenopsis invicta Buren1 + + + + + + + +

Solenopsis picta Emery2 + + + + - - + -

Strumigenys louisianae Roger + + + + - - + -

Strumigenys membranifera Emery1 2 - + - - - - + -

1 Denotes exotic species; 2 Species new to Chatham County; 3 Species new to Georgia. (+) Denotes a species collection by a sampling method or in a habitat. (-) Denotes a species absence from a sampling method or habitat type.

50

Table 2.1. (Continued) Ant species collected on and around the port of Savannah with collection method and habitat type noted. Wooded Species Pitfall Leaf litter Logs Trees Baits Quadrats transects Roadsides Strumigenys ornata Mayr2 + + - - + - + -

Strumigenys rostrata Emery 2 + + + - - - + -

Strumigenys silvestrii Emery1 2 - + - - - - + -

Strumigenys talpa Weber2 - + - - - - + -

Temnothorax curvispinosus Mayr - + - - - - + -

1 Denotes exotic species; 2 Species new to Chatham County; 3 Species new to Georgia (+) Denotes a species collection by a sampling method or in a habitat. (-) Denotes a species absence from a sampling method or habitat type.

51

Table 2.2. Trapping statistics for the 2015 - 2016 survey at the Port of Savannah, Georgia.

Trap type Number of Ant species Number of Trapping Max. species in Total ants collected samples collected samples efficiency (%) one sample without ants Pitfall 180 29 6 96.67 10 47982

Leaf Litter 180 31 2 98.89 10 7902

Tree Trunk 180 33 1 99.44 6 2021

Logs 180 29 8 95.56 6 2721

Baiting 380 22 9 97.63 6 2984

Quadrat 200 14 4 98 6 1814

Total 1300 47 30 97.7 65420

52

Table 2.3. Number of ant species and exotic ant species in Georgia* and adjacent states*.

State/Port Number of species Number of exotic species Exotic ant fauna (%)

Port of Savannah 47 14 29.8

Alabama 173 28 16.2

Florida 230 65 28.3

Georgia 189 26 13.8

North Carolina 178 13 7.3

South Carolina 155 21 13.5

Tennessee 133 9 6.8

* Species numbers according to Joe MacGown’s “Ants (Formicidae) of the Southeastern United States”

53

Figure 2.1. Port of Savannah sampling locations. Yellow lines indicate wooded transects. White lines indicate roadsides sampled.

54

Figure 2.2. Pitfall sampling. (A) 9 dram vial containing propylene glycol before placement in the ground. (B) Rain guard suspended by nails above the installed pitfall trap.

55

Figure 2.3. Leaf litter collection. Leaves and soil were placed into a one gallon bag for transport to Berlese funnels. Samples were stored in a cooler with ice during transport. 56

Figure 2.4. Dissection of course woody debris. (A) Selected log placed in the dissection tray. (B)

Complete dissection of the log, at which point all ants found within were aspirated. Ants were immediately transferred to 75% ethanol. 57

Figure 2.5. Tree trunk sampling. (A) The litter surrounding the base of the tree was removed for easier access to the ants residing there. (B) The tree was searched for five minutes aided by a flashlight. Ants were aspirated and immediately placed into 75% ethanol. 58

Figure 2.6. Bait trap utilizing a nine dram vial and its cap along with crushed Pecan Sandie cookie placed in a small clearing on the forest floor. Ants found on the cookie, cap, or in the vial were aspirated and transferred to 75% ethanol. 59

Figure 2.7. Quadrat sampling. A 0.5 x 0.5 m quadrat placed on the margin of the pavement and roadside within which all ants were collected over a five minute period with the aid of a flashlight and then transferred to 75% ethanol. 60

60

50

40

Jacknife

30 Chao ICE Singleton

20

10

0 0 100 200 300 400 500 600 700 800

Figure 2.8. Rarefaction curves for all wooded transect samples combined. 61

CHAPTER 3

ECOLOGICAL IMPACTS OF THE TAWNY CRAZY ANT, NYLANDERIA FULVA

MAYR, ON THE ANT COMMUNITY AT THE PORT OF SAVANNAH, GEORGIA

Introduction

Invasive species can be ecologically and economically costly in their new range, disrupting native species and causing damage to crops and manufactured products (Pimentel et al.

2005). The effects of climate change may be facilitating the success of new arrivals in temperate environments from more tropical environments (Walther et al. 2009). Ants are often particularly well suited to human mediated dispersal and can undergo genetic and behavioral changes upon arrival that allow them to become a dominant force in their new habitat (Tsutsui and Suarez

2003). These changes, caused by a reduction in genetic diversity, include an increase in the number and acceptance of queens in a colony and the shift from independent competing colonies to unicoloniality, often referred to as supercolony development (Tsutsui et al. 2000). These traits allow invasive ants to quickly outnumber and outcompete native species and to take over existing mutualisms for their own benefit (Holway et al. 2002).

Invasive ants reduce the number of coexisting, ecologically similar native ant species

(Wittman 2014). Ant communities under invasion may become disorganized when ecological pressures from previously co-occurring species are replaced by the invader (Sanders et al. 2003).

The phylogenetic structure of invaded ant communities may change from evenly disbursed and structured by competition to clustered communities containing fewer groups of closely related species (Lessard et al. 2009). The persistence or extirpation of a species in an invaded habitat may be attributed to its either complementary or similar resource utilization in comparison to the invader (Sarty et al. 2007). The size of workers may also allow species to withstand an invasion by avoidance because small ants can nest in areas too small for the invader (LeBrun et al. 2013). 62

This avoidance by size strategy may select for other invasive species, which tend to have smaller workers, to co-occur with a dominant invader (McGlynn 1999b).

In the case of the tawny crazy ant, Nylanderia fulva Mayr, we can see similar ecological characteristics and consequences of its invasion in the southern United States (Gotzek et al. 2012,

LeBrun et al. 2013). It has become a dominant, polygynous, supercolonial species in its invasive range while it is not so in its native range (McDonald 2012, Calcaterra et al. 2016). It forms mutualisms with other arthropod species where it occurs, keeping these resources from native species (Sharma et al. 2013). There is evidence of its ability to adapt to climatic variation in its invasive range, which would allow for a much broader potential distribution (Kumar et al. 2015).

Metrics for analyzing biodiversity, such as species richness and species accumulation, have been successfully used to estimate the incidence and abundance of ant species in a diversity of habitats, although their interpretation and application can be dubious in certain circumstances

(Gotelli and Colwell 2001). Along with this, the Jaccard similarity index can be used to compare the shared and unique species between habitats, but its appropriateness in habitats undergoing invasion has not been tested.

The purpose of this study was to monitor the effect of an invasion by N. fulva on the ant community at the Port of Savannah in Garden City, Georgia. Over the course of two field seasons we monitored the ant communities in wooded areas on the port property using multiple sampling methods. The sampled areas covered an area within the invasive population, along the invasion front and outside of the invaded area. We then assessed the usefulness of several diversity estimators along with the Jaccard index in analyzing the ant community, and changes therein, within and around an active and expanding invasive ant population.

Materials and Methods

Site Description

The Port of Savannah in Garden City, Georgia (32°08’00.00 N, 81°09’00.00 W) is located approximately 30 kilometers inland from the Atlantic shore, on the Savannah River. The 63 property consists of a 3 km docking area along the edge of the Savannah River. Immediately adjacent to the docking area are large paved areas for container storage. Beyond this are several forested/green areas fragmented by paved and dirt roads. Farthest from the river is the rail yard accompanied by more paved container storage (Figure 4.1).

Nine sites within forest fragments both on and surrounding the port property were chosen for sampling. Sites were selected based on a preliminary visual assessment of the N. fulva population in June 2015 (Figure 3.1). The selection of sites based on our previous survey was predicated on the N. fulva reproductive strategy of colony budding. Similar to the pharaoh,

Monomorium pharaonis (Linnaeus), and Argentine ant, Linepithema humile Mayr, budding colonies spread from a central location, remain contiguous with the original colony, share workers and queens, and spread across landscapes (Tsutsui and Suarez 2003, Buczkowski and

Bennett 2009).

Sites one and two were located near the center of the expanding N. fulva population.

These sites flanked an unpaved storage lot for damaged containers and trailers where N. fulva was initially discovered. These forest fragments were characterized by abundant canopy provided by mixed hardwoods and pines. Site one had a dense understory and was adjacent to a swamp. Site two was drier and had a thin understory composed mainly of oak saplings. The abundance of N. fulva was obvious in these two sites, with workers ubiquitous on the ground and foraging trails extending through the understory into the canopy.

Sites three through six had evidence of partial invasion in 2015. Site three was characterized by a heavy understory of oak saplings and a canopy of oaks and pines. Nylanderia fulva was present foraging on trees but not obvious on the ground. Site four had bare clay soil on the western end that remained consistently damp with an understory of palms and palmettos.

Eastward, this transitioned into an understory of oak saplings with a canopy of mixed hardwoods and abundant leaf litter groundcover. Site five had a moderate to thick understory of oak and hardwood saplings with a canopy of mostly hardwoods. Site six was located across a paved road

64 with heavy automobile traffic from dawn to dusk. This site had a sparse oak understory with a canopy of mostly pines with some oaks.

Sites seven had no N. fulva during sampling in 2015 although it appeared in 2016. Site seven had a moderate oak understory with an oak and pine canopy. Sites eight and nine had no detectable N. fulva either year. Site eight was similar to seven with a thicker understory. Site nine was mostly mesic with smaller hardwoods throughout and a muddy soil.

Sampling Methods

Forested fragments were sampled using one to three 90-m transects, depending on the fragment’s size and were sampled twice during June or July in 2015 and 2016. Samples were collected using five methods: pitfall trapping, leaf litter extraction, dissection of coarse woody debris, active searching on tree trunks, and baiting. Sampling was conducted over the course of three days to allow for 48 hours exposure to the pitfall traps (Agosti and Alonso 2000a) . Day one involved placement of pitfall traps along three transects and the sampling of coarse woody debris, tree trunks, and baits on one transect. On the second day, coarse woody debris, tree trunk, and bait samples were taken on the two remaining transects. The third day entailed collection of leaf litter samples and removal of pitfall traps from all transects. This three day protocol was repeated on all nine transects.

Single pitfall traps were placed along each transect at 10 m intervals, for a total of ten traps per transect. Pitfall traps were nine dram plastic vials lined with Fluon (Bioquip Products

Inc., Rancho Dominguez, CA) with propylene glycol added to 2-cm from the bottom (Figure 3.2

A). A small area was cleared of leaf litter and debris to expose the soil surface. Traps were installed in the ground with the open plastic vial flush with the soil surface. Traps were covered by a plastic plate, suspended approximately five centimeters from the ground by three nails, to prevent rainwater intrusion (Figure 3.2 B). Traps remained in the ground for 48 hours before they were collected. 65

Leaf litter samples were hand collected by filling a one gallon plastic bag with leaves and the top 5-mm of soil from a single location (Figure 3.3). Bags were placed in a cool environment until returned to the laboratory, where ants were extracted using Berlese funnels for one week.

A partially decayed dead branch or log (≈5-10-cm ID and no longer than 100-cm in length) was randomly selected to collect coarse woody debris (CWD). The CWD was dissected in a large, white, plastic pan. Ants were collected using aspirators and were transferred to vials of

75% ethanol (Figure 3.4).

Trees with trunks between 5-20-cm diameter at eye level were selected and litter, debris, and vegetation removed from the base of the tree (Figure 3.5 A). Each tree was visually searched for ants by two people for five minutes with the aid of flashlights (Figure 3.5 B). Trees were searched from the root zone to 3-m up the tree’s trunk and ants aspirated then transferred to vials of 75% ethanol.

Bait vials consisted of 1-cm of crushed Pecan Sandies cookies (Keebler Company, Battle

Creek, MI) in nine dram vials. A bait vial was upended allowing the cookie to fall onto the vial’s cap and the vial left on its side next to the cap on the ground (Figure 3.6). After one hour, ants on the bait were collected using aspirators and transferred to vials of 75% ethanol.

Analysis

Species richness and Sample based rarefaction. Species richness was compared among transects within and between years for each transect. Abundance of N. fulva in pitfall traps was used to assess the degree of invasion of each transect and to monitor the progression of invasion at each transect between years. Sample-based rarefaction curves were created for each transect by year using EstimateS software (EstimateS 9.1; Robert K. Colwell, 2016). Three rarefaction species richness estimators, Chao 2, Jackknife 1, and the Incidence based Coverage Estimator

(ICE), were used to assess species richness and sampling effort by transects (Longino 2000).

These estimators use the total number of species collected in a data set, the number of rare species

(uniques and duplicates), and the number of samples as parameters. Uniques are species that were 66 collected in only one sample at a site while duplicates were species collected in only two samples.

These parameters may be extrapolated into the sampling effort required to reach a maximum number of species for a sampling area (Gotelli and Colwell 2011).

The species accumulation graphs created using EstimateS show the estimated number of species present at each site with the slope at 50 samples indicating if additional sampling could be expected to collect new species from a transect (positive slope) or not (zero or negative slope).

Species collected only once (uniques) or twice (duplicates) were also plotted to show the potential for rare species at a site (Mao and Colwell 2005).

Jaccard similarity index. The Jaccard similarity index was used to examine pairwise shared and unshared species between transects. This similarity index is calculated by dividing the number of species shared by the total number of species from both transects providing a relative measure of similarity among two sites (Goodall 1966).

Results

Species Richness

Species richness ranged from 11 (transects one and two in 2016) to 27 (transect seven in

2016) with a trend of decreasing richness toward the center of the visually estimated N. fulva population (Table 3.1; Figure 3.1). Total species collected from both years ranged from 15

(transect one) to 31 (transect 5) and decreased with increasing proximity to the N. fulva population center (Figure 3.7). Differences in the number of species collected between years were negative for transects undergoing invasion and positive for transects where N. fulva was not present (Figure 3.7).

Abundance of N. fulva

The total number of N. fulva in pitfall traps from transects near the center of the of the population (transects one and two) numbered in the thousands and were an order of magnitude higher in 2015 and two orders of magnitude higher in 2016 compared to surrounding transects

67

(Figure 3.8). Areas with moderate abundance of N. fulva in 2015 and 2016 (transects three and four) had increases of about ten and five times respectively between years.

The number of N. fulva workers was at or near zero at the edge of and beyond the invasion front (transects five through nine) in 2015. The abundance of workers increased slightly in transects five and seven between years while transects eight and nine remained uninvaded.

Sample Based Rarefaction

Rarefaction curves for transects that were at the center of the N. fulva population

(transects one and two) leveled off by 50 samples and at that point the species richness estimates were near our observed species for those transects and decreased between years (Table 3.2;

Figure 3.9; Figure 3.10). Transects that were in areas with moderate abundance of N. fulva

(transects three and four) had rarefaction curves that, for the 2015 sampling, tended to increase steadily even at 50 samples and had a large number of unique species (Figure 3.11; Figure 3.12).

The species richness estimates for these transects in 2015 were much higher than our observed species and ranged from 32 to 47 for transect three and 33 to 44 for transect four. These estimates are in contrast to the 2016 rarefaction curves which leveled quickly, had species richness estimates similar to our observed species for these transects and had fewer unique species.

The rarefaction curve for transect five in 2015 was slightly increasing at 50 samples with estimates of species richness between 33 (Jackknife 1 and ICE) and 59 (Chao 2) and a high number of uniques. This differs from the 2016 rarefaction curve for this transect which had leveling and estimates between 26 and 28 (Figure 3.13).

In areas where N. fulva was not detected (transects six through nine) the rarefaction curves leveled and species richness estimates were similar to or slightly higher than our observed species from these transects (Figure 3.14-3.16). Transect seven is included in this group due to N. fulva occurring in only two samples from 2016 which were closest to a roadside inhabited by the advancing invasion front. Nylanderia fulva had likely not yet invaded this wooded area or had any measurable effect on the ant community therein.

68

Jaccard Similarity Index

The Jaccard similarity indices showed a pattern of higher similarity between transects that were either both inside the invaded area or both outside the invaded area. Similarity was lowest when comparing transects from the center of the N. fulva population with transects outside of the invaded area. The reduction in unshared species from the invaded transects was the main contributing factor in decreasing similarity between these groups.

The number of shared and unshared species for each of the pairwise comparisons was averaged for each transect in each year. Transects near the center of the N. fulva population had a decrease in both shared and unshared species between years (Table 3.3). Transect five had little variability in shared and unshared species between years. Transects outside the invaded area had increases in unshared species while shared species remained nearly unchanged.

Discussion

Species Richness

The number of ant species collected showed a pattern of reduction between years at transects affected by N. fulva and also in a single sampling year with greater proximity to the visually estimated center of the N. fulva population (Table 3.1; Figure 3.1). Transects one through five suffered a higher loss of species than those at the edge of the invasion between years (Figure

3.7). These transects also showed a reduction in the number of new species collected during 2016.

Transect five was partially invaded and suffered both a loss of species but also a gain of species in 2016. Transects six through nine had low species loss between years yet had substantial species gain between years.

The gain of species from 2015 to 2016 can likely be explained by sampling error. All potential species from a particular transect were not captured during the 2015 sampling year due to the randomness of sampling and the uneven spatial distribution of ants in the environment

(McGlynn et al. 2009). Thus, in 2016 by effectively doubling our sampling effort along a transect, we discovered additional species. This holds true for transects five through nine where 69

N. fulva had only partially invaded or had not invaded at all. At transects one through four species richness was reduced to such a degree that the sampling effort from the first year alone was sufficient to collect nearly all species from that location (Table 3.1).

Upon closer inspection of the species collected at affected transects, we observed non- random extirpation of species similar to that found by LeBrun et al. (2013). The species noticeably missing from transects with N. fulva tended to be large, ground foraging species such as Aphaenogaster spp., Camponotus spp., other Nylanderia species, Pheidole dentata, and

Solenopsis invicta. Even the arboreal Crematogaster spp. were missing from affected transects while remaining widespread at transects outside the invaded area. This is likely due to unavoidable interaction and competition for resources with N. fulva.

The ubiquitous nature of N. fulva in heavily invaded habitats makes avoidance impossible for some but not all co-occurring ant species. Certain species characteristics seem to allow for some species to survive the invasion, even along transects with an abundance of N. fulva. Very small, cavity dwelling species such as Monomorium minimum, Pheidole dentigula, Pheidole navigans, Solenopsis abdita, and several Strumigenys spp. persisted along transects regardless of the abundance of N. fulva. This persistence may be due to their ability to nest in spaces too small for N. fulva workers to enter, avoid aggressive interactions with N. fulva, or utilize food resources not dominated by N. fulva.

A similar persistence was observed in some larger species such as Colobopsis impressa and Pseudomyrmex ejectus. In the case of these species, their ability to nest in twigs and efficiently defend their entrances may be the key to their success. The twig nesting species were collected at all transects regardless of the presence or severity of invasion of N. fulva. In contrast,

Crematogaster spp. were readily removed from their environment despite being formidable competitors armed with chemical defenses (Daloze et al. 1991).

Additionally, the small arboreal species Solenopsis picta was able to withstand the N. fulva invasion, likely due to their small size. Also notably present throughout the invaded area

70 were Hypoponera opaciceps and Hypoponera opacior. These species are on the smaller side and tend to nest in logs, twigs, and moist microhabitats (Deyrup et al. 1988). These additional species have a combination of the previous characteristics that may allow them to persist in the presence of N. fulva.

Rarefaction and Richness Estimators

Rarefaction species richness estimators tended to level at a smaller number of species for transects closer to the visually estimated center of the N. fulva population with some important exceptions. Transects one and two initially had species richness estimates of ~20 species in 2015 and 12 to 13 in 2016 (Figure 3.8, Figure 3.9). These estimates are consistent with the species we found with our sampling along these transects. The ant community has been severely reduced in the habitat surrounding these transects and the estimators imply that additional sampling is likely to find few if any additional species.

Transects three and four in 2015 had species richness estimates that would suggest that there was an abundance of species that were not collected by our sampling (Figure 3.10, Figure

3.11). The estimators indicated 30 or more species at 50 samples with all estimators having a strong positive slope implying that there would be many more species collected with additional sampling. These estimates contrast drastically with the results from our sampling along these same transects in 2016, in which the estimators leveled at 15 to 20 species. This discrepancy in species estimates can be reconciled by considering the ecological impact of N. fulva and the terms that the richness estimators use in their analysis of our sampling data.

The estimators use the number of samples taken, the number of species collected, and most importantly in this scenario the number of rare species. The number of rare species in a transect’s data augments the estimation of total species along that transect. Rare species require greater sampling effort to capture and therefore an abundance of rare species will increase the estimation of total species in that area as well as the number of samples needed to collect them.

The impact of N. fulva had a particularly disruptive effect on the relationship between rare species

71 and estimated potential new species in a particular habitat. In contrast to transects one and two, where species richness had already declined substantially, transects three and four in 2015 were still undergoing this reduction in species richness.

The extirpation of a species does not happen in a single instance but instead gradually, as the invading population develops (Mitchell et al. 2011). This results in normally common, abundant species becoming rare before complete extirpation. This artificial inflation of rare species along a transect invaded by N. fulva leads to species richness estimates far greater than any of our collected data would suggest. The behavior of the estimators in this type of scenario may have implications for future biodiversity estimates in studies where an invasion is not as obvious as one like that of N. fulva or in studies that do not have repeated sampling at study sites.

Transect five saw a similar exaggeration of the Chao2 estimator in 2015 that was lost in the 2016 estimate (Figure 3.12). The remaining estimators were only slightly elevated in 2015 compared to 2016 where they leveled at around 28 species. This transect was unique in the study due to its orientation being parallel to the direction of outward expansion of the N. fulva population. Thus, the transect had been affected by N. fulva in the portion of the transect nearest to the invasion front, while the smaller remaining portion, further from the invaded area, remained somewhat unaffected. This led to a loss of species from this transect in response to the pressures from the invasion, but also the presence of species normally extirpated by N. fulva at the far end of the transect.

At the time of sampling, the remaining transects, six through nine, had not yet been invaded by N. fulva. The species richness estimators had nearly leveled and ranged from about 20 to 30 species with little difference between years for each transect (Figure 3.13, Figure 3.14,

Figure 3.15, Figure 3.16). These transects demonstrate the effective use of these species richness estimators, giving us estimates near our combined transect species totals and can act as a control for the invaded areas in our study. The species most affected by the N.fulva invasion were still present along these transects, but with invasion likely, their survival is not guaranteed. 72

Jaccard Similarity Index

The Jaccard similarity index is commonly used to assess the relative diversity among and between sample sets and while this method can be applied to our data set, it may not be useful in the context of differentially invaded habitats. Because the index is relative to the species present along each transect compared to one another, a large reduction in species from one transect may not manifest a change in similarity between the transects and may in fact confound the use of such similarity indices.

The 36 pairwise comparisons from each sampling year showed no obvious relationship between the transects similarity. Transects at the center of the N. fulva population tended to have higher similarity with one another and the same was true of transects outside the limits of the population. A more revealing trend was observed in the individual components of the Jaccard index: shared species and unshared species. These components show us the absolute values used to calculate the indices without masking the data with ratios. The averaged shared and unshared species from each transect showed a clear relationship between the location of the transect relative to the N. fulva population and the average number of both shared and unshared species from each transect (Table 3.3).

Transects at the center of the invaded area had the least number of shared and unshared species in both years due to the severe reduction in species in these areas. The average unshared species from transects one through five decreased from 2015 to 2016 while the remaining transects had an increase in unshared species. Shared species from transects one through four also decreased from 2015 to 2016 while transects five through nine either stayed the same or increased. The decreasing unshared species from transects within the invaded area illustrates the homogenization observed by LeBrun et al. (2013). This resulted in an increase in average unshared species from other transects due to their absence in the invaded transects.

Overall, a correlation of decreasing presence of ant species with increasing presence of N. fulva workers was observed. Those most affected by N. fulva were large, ground nesting species

73 that were likely in direct competition for food, water, or nesting space with the invader. Species richness estimator aberration in habitats undergoing invasion may confound measures of richness for these habitats but may serve as an indicator of such disturbance with foreknowledge of this relationship. The results of the similarity index applied to this data show the inefficiency of such relative measures when comparing habitats experiencing different stages of invasion. Although, looking at the index’s component measures revealed a trend of increasing homogenization with increasing presence of N. fulva workers. These findings may assist in the detection of unknown or ongoing biological invasions and should facilitate the use of analytical tools used to describe the communities in invaded habitats. 74

References

Agosti, D., and L. E. Alonso. 2000. The ALL protocol, pp. 204-206. In D. Agosti (ed.), Standard

methods for measuring and monitoring biodiversity. Smithsonian Institution Press,

Washington and London.

Buczkowski, G., and G. Bennett. 2009. Colony budding and its effects on food allocation in the

highly polygynous ant, Monomorium pharaonis. Ethology 115: 1091-1099.

Calcaterra, L., S. Cabrera, and J. Briano. 2016. Local co-occurrence of several highly invasive

ants in their native range: are they all ecologically dominant species? Insectes Soc. 63:

407-419.

Daloze, D., M. Kaisin, C. Detrain, and J. M. Pasteels. 1991. Chemical defense in the three

European species of Crematogaster ants. Experientia 47: 1082-1089.

Deyrup, M. A., N. Carlin, J. Trager, and G. Umphrey. 1988. A review of the ants of the

Florida Keyes. Fla. Entomol. 71: 163-176.

Goodall, D. W. 1966. A new similarity index based on probability. Biometrics 22: 882-&.

Gotelli, N. J., and R. K. Colwell. 2001. Quantifying biodiversity: procedures and pitfalls in the

measurement and comparison of species richness. Ecol. Lett. 4: 379-391.

Gotelli, N. J., and R. K. Colwell. 2011. Estimating species richness, pp. 39-54. In A. E.

Magurran and B. J. McGill (eds.), Biological diversity: frontiers in measurement and

assessment, vol. 12. Oxford University Press, Oxford, New York.

Gotzek, D., S. G. Brady, R. J. Kallal, and J. S. LaPolla. 2012. The importance of using

multiple approaches for identifying emerging invasive species: the case of the Rasberry

Crazy ant in the United States. Plos One 7: 10.

Holway, D. A., L. Lach, A. V. Suarez, N. D. Tsutsui, and T. J. Case. 2002. The causes and

consequences of ant invasions. Annu. Rev. Ecol. Syst. 33: 181-233.

75

Kumar, S., E. G. LeBrun, T. J. Stohlgren, J. A. Stabach, D. L. McDonald, D. H. Oi, and J.

S. LaPolla. 2015. Evidence of niche shift and global invasion potential of the Tawny

Crazy ant, Nylanderia fulva. Ecology and Evolution 5: 4628-4641.

LeBrun, E. G., J. Abbott, and L. E. Gilbert. 2013. Imported crazy ant displaces imported fire

ant, reduces and homogenizes grassland ant and arthropod assemblages. Biol. Invasions

15: 2429-2442.

Lessard, J. P., J. A. Fordyce, N. J. Gotelli, and N. J. Sanders. 2009. Invasive ants alter the

phylogenetic structure of ant communities. Ecology 90: 2664-2669.

Longino, J. T. 2000. What to do with the data, pp. 186-203. In D. Agosti (ed.), Standard methods

for measuring and monitoring biodiversity. Smithsonian Institution Press, Washington

and London.

Mao, C. X., and R. K. Colwell. 2005. Estimation of species richness: mixture models, the role of

rare species, and inferential challenges. Ecology 86: 1143-1153.

McDonald, D. L. 2012. Investigation of an invasive ant species: Nylanderia fulva colony

extraction, management, diet preference, fecundity, and mechanical vector potential.

Texas A&M University.

McGlynn, T. P. 1999. Non-native ants are smaller than related native ants. Am. Nat. 154: 690-

699.

McGlynn, T. P., R. M. Fawcett, and D. A. Clark. 2009. Litter Biomass and Nutrient

Determinants of Ant Density, Nest Size, and Growth in a Costa Rican Tropical Wet

Forest. Biotropica 41: 234-240.

Mitchell, M. E., S. C. Lishawa, P. Geddes, D. J. Larkin, D. Treering, and N. C. Tuchman.

2011. Time-Dependent Impacts of Cattail Invasion in a Great Lakes Coastal Wetland

Complex. Wetlands 31: 1143-1149.

Pimentel, D., R. Zuniga, and D. Morrison. 2005. Update on the environmental and economic

costs associated with alien-invasive species in the United States. Ecol. Econ. 52: 273-288. 76

Sanders, N. J., N. J. Gotelli, N. E. Heller, and D. M. Gordon. 2003. Community disassembly

by an invasive species. Proc. Natl. Acad. Sci. U. S. A. 100: 2474-2477.

Sarty, M., K. L. Abbott, and P. J. Lester. 2007. Community level impacts of an ant invader and

food mediated coexistence. Insectes Soc. 54: 166-173.

Sharma, S., D. H. Oi, and E. A. Buss. 2013. Honeydew-producing hemipterans in Florida

associated with Nylanderia fulva (Hymenoptera: Formicidae), an invasive crazy ant. Fla.

Entomol. 96: 538-547.

Tsutsui, N. D., and A. V. Suarez. 2003. The colony structure and population biology of invasive

ants. Conserv. Biol. 17: 48-58.

Tsutsui, N. D., A. V. Suarez, D. A. Holway, and T. J. Case. 2000. Reduced genetic variation

and the success of an invasive species. Proc. Natl. Acad. Sci. U. S. A. 97: 5948-5953.

Walther, G. R., A. Roques, P. E. Hulme, M. T. Sykes, P. Pysek, I. Kuhn, M. Zobel, S.

Bacher, Z. Botta-Dukat, H. Bugmann, B. Czucz, J. Dauber, T. Hickler, V. Jarosik,

M. Kenis, S. Klotz, D. Minchin, M. Moora, W. Nentwig, J. Ott, V. E. Panov, B.

Reineking, C. Robinet, V. Semenchenko, W. Solarz, W. Thuiller, M. Vila, K.

Vohland, and J. Settele. 2009. Alien species in a warmer world: risks and opportunities.

Trends Ecol. Evol. 24: 686-693.

Wittman, S. E. 2014. Impacts of invasive ants on native ant communities (Hymenoptera:

Formicidae). Myrmecological News 19: 111-123. 77

Table 3.1. Species richness by year plus number of species collected for both years and net change in number of species from 2015-2016 by transect. # of Species Total Net Number 2015 2016 Species Change 1 15 11 15 -4 2 16 11 17 -5 3 22 15 23 -7 4 23 14 25 -9 5 24 23 31 -1 6 18 23 23 +5 7 22 27 27 +5 8 24 25 28 +1 9 18 21 24 +3

78

Table 3.2 Uniques and Duplicates collected and richness estimator values for each transect by year. Uniques Duplicates Chao 2 Jackknife 1 ICE Transect 2015 2016 2015 2016 2015 2016 2015 2016 2015 2016 1 5 2 2 2 21 12 20 13 19 12 2 3 0 3 3 18 12 20 12 19 12 3 10 4 2 1 47 23 32 19 39 17 4 9 2 2 2 44 15 33 16 34 15 5 9 6 0 4 59 26 33 28 33 28 6 2 4 3 5 19 25 20 27 19 26 7 2 6 3 3 23 33 24 33 23 31 8 6 3 1 4 37 22 25 24 23 22 9 7 4 1 6 34 26 31 29 30 27 79

Table 3.3 Shared and unshared species averages for each transect by year. Unshared Shared

Transect 2015 2016 2015 2016

1 3.75 2.75 10.13 8.25

2 4.5 2.13 12.25 8.88

3 6.25 4 14.88 11

4 8.88 3.75 15.13 10.25

5 9.13 8.88 14.88 14.13

6 4.88 8.5 13.13 14.5

7 7.5 12 14.5 15

8 9.13 10.13 14.88 14.88

9 5.87 7.63 13.13 13.33 80

Figure 3.1. Port of Savannah sampling locations. Yellow lines indicate wooded transects with numbers indicating the location of sample number one. Red and orange lines delimit the heavily invaded area and the limit of the advancing invasion front respectively. 81

Figure 3.2. Pitfall sampling. (A) 9 dram vial containing propylene glycol before placement in the ground. (B) Rain guard suspended by nails above the installed pitfall trap.

82

Figure 3.3. Leaf litter collection. Leaves and soil were placed into a one gallon bag for transport to Berlese funnels. Samples were stored in a cooler with ice during transport. 83

Figure 3.4. Dissection of course woody debris. (A) Selected log placed in the dissection tray. (B)

Complete dissection of the log, at which point all ants found within were aspirated. Ants were immediately transferred to 75% ethanol. 84

Figure 3.5. Tree trunk sampling. (A) The litter surrounding the base of the tree was removed for easier access to the ants residing there. (B) The tree was searched for five minutes aided by a flashlight. Ants were aspirated and immediately placed into 75% ethanol 85

Figure 3.6. Baiting. A bait trap utilizing a nine dram vial and its cap along with crushed Pecan

Sandie cookie placed in a small clearing on the forest floor. Ants found on the cookie, cap, or in the vial were aspirated and transferred to 75% ethanol. 86

35 30

25 20 15 10 Number Number species of 5 0 1 2 3 4 5 6 7 8 9 Transect

Figure 3.7. Number of species collected by transect. Blue bars indicate the total number of species collected from 2015 and 2016. Green and red bars show the number of species gained and lost, respectively, between years. 87

18000 15644 16000 14543 14000

12000

10000 2015 8000 7193 5855 2016

Number Number Workers of 6000

4000 1739 101 2000 641 138 173 110 3 0 0 0 0 0 0 0 0 1 2 3 4 5 6 7 8 9 Transect

Figure 3.8 Abundance of N. fulva workers in pitfall traps. Total number of workers in pitfall traps by transect and by year. Blue bars indicate abundances for 2015 and red bars indicate abundances for 2016. 88

Transect 1 2015 25

20

15

10

Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Transect 1 2016 25

20

15

10

Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Figure 3.9. Rarefaction curves for transect one by year. 89

Transect 2 2015 25

20

15

10

Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Transect 2 2016 25

20

15

10

Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Figure 3.10. Rarefaction curves for transect two by year. 90

Transect 3 2015 50 45

40 35 30 25 20 15 Number Number species of 10 5 0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Transect 3 2016 50 45

40 35 30 25 20 15 Number Number species of 10 5 0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Figure 3.11. Rarefaction curves for transect three by year.

91

Transect 4 2015 45 40

35 30 25 20 15

Number Number species of 10 5 0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Transect 4 2016 45 40

35 30 25 20 15

Number Number species of 10 5 0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Figure 3.12. Rarefaction curves for transect four by year.

92

Transect 5 2015 60

50

40

30

20 Number Number species of 10

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Transect 5 2016 60

50

40

30

20 Number Number species of 10

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Figure 3.13. Rarefaction curves for transect five by year. 93

Transect 6 2015 35

30

25

20

15

10 Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Transect 6 2016 35

30

25

20

15

10 Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Figure 3.14. Rarefaction curves for transect six by year.

94

Transect 7 2015 40 35

30 25 20 15

Number Number species of 10 5 0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Transect 7 2016 40 35

30 25 20 15

Number Number species of 10 5 0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Figure 3.15. Rarefaction curves for transect seven by year. 95

Transect 8 2015 35

30

25

20

15

10 Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Transect 8 2016 35

30

25

20

15

10 Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Figure 3.16. Rarefaction curves for transect eight by year. 96

Transect 9 2015 35

30

25

20

15

10 Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Transect 9 2016 35

30

25

20

15

10 Number Number species of 5

0 1 50 Number of samples

Uniques Duplicates Classic Chao2 Jackknife1 ICE

Figure 3.17. Rarefaction curves for transect nine by year.

97

CHAPTER 4

USE OF ROADSIDE HABITATS AS DISPERSAL CORRIDORS BY THE INVASIVE

TAWNY CRAZY ANT, NYLANDERIA FULVA, AT THE PORT OF SAVANNAH,

GEORGIA

Introduction

Invasions by exotic species sometimes cause great ecological damage and economic loss, costing the United States almost 120 billion dollars annually (Pimentel et al. 2005). A biological invasion can be described as “a species acquiring a competitive advantage following the disappearance of natural obstacles to its proliferation, which allows it to spread rapidly and to conquer novel areas within recipient ecosystems in which it becomes a dominant population”

(Valery et al. 2008). The continental U.S. has over 2,000 introduced arthropod species and

Hawaii has approximately 4,500 with more than 95% of these species unintentionally transported by humans in water ballast from ships, in soil, or on plants (Pimentel et al. 2005).

The spread of invasive ant species occurs at both long distance and local scales. Long distance spread is achieved through jump dispersal events, where the candidate species is introduced to its new range and, through subsequent dispersal events, arrives in other habitats in its potential range (Suarez et al. 2001). On a local scale, the species becomes established through an increase in colony density and individual abundance. After establishment, the species begins to spread in a diffusion-like pattern as it exploits the resources in its new habitat (Suarez et al.

2010). Locally, the rate and pattern of spread is dependent upon whether the species reproduces via mating flights or colony budding (DeHeer et al. 1999).

Human-mediated jump dispersal events allow species to travel much farther than they would be able to on their own. These events have been the cause of multiple invasions of the

North American continent by exotic ant species (McGlynn 1999a). Two hundred and thirty two 98 ant species have been intercepted while being transported into the United States (Suarez et al.

2005b). Several successful ant invaders, including the red imported fire ant, Solenopsis invicta

Buren, were first found in cities with international seaports (Lofgren 1986, Storz and Tschinkel

2004).

The port of Savannah, Georgia is of particular concern for the transport of invasive species into the U.S. because of the number of containers that pass through the port, its plans for future growth, and the mild Georgia climate (warm temperatures and abundant rainfall). The

Garden City terminal resides on a 485.6 hectare (1,200 acre) property in Garden City, GA.

Between July and December 2016, 1.86 million 6.1 meter container units or TEU (twenty-foot equivalent unit) were moved through the port, in addition to 185,000 rail containers and 1.2 million truck containers moved (Gordon Hammer, Georgia Ports Authority, personal communication, February 23, 2017). In 2015 the port deepened and widened the Savannah River to accommodate vessels capable of carrying over 13,000 TEUs.

In June 2015, the highly invasive tawny crazy ant, Nylanderia fulva (Mayr), was collected for the first time in Chatham County, Georgia, on Port of Savannah property (Chapter

2). To date, this is the furthest north that this species has been detected along the Atlantic coast.

The tawny crazy ant is native to South America. The earliest records of the ant in North America are museum specimens from Texas in 1938 (Trager 1984). Active populations were rediscovered in 2002 near Houston, Texas (Meyers 2008). A large population of the ant was also found in

Florida, but confusion with a closely related species, Nylanderia pubens (Forel), obscured the details of its invasion history (Wetterer and Keularts 2008, Wetterer et al. 2014). The ant has since been found in all of the Gulf Coast states and is continuing to spread (Gotzek et al. 2012).

Despite the recent range expansion by the tawny crazy ant, only one published study has demonstrated the impacts this species has on native ant communities (LeBrun et al. 2013).

LeBrun et al. (2013) found that the tawny crazy ant reduced and homogenized grassland ant assemblages and displaced the red imported fire ant. The study also measured the rate of range 99 expansion in the grassland environment. However the study did not include areas of human disturbance, such as roadsides, that are known corridors of dispersal for exotic ants (Stiles and

Jones 1998).

The purpose of this study was to document the effect of a local scale invasion by the tawny crazy ant on ant communities in disturbed habitats on the Port of Savannah in Garden City,

Georgia (Chatham County). We also documented the displacement, based on resource dominance and competitive exclusion, of the red imported fire ant during this invasion. Finally we looked at the rate of expansion of the tawny crazy ant across a disturbed habitat over the course of a year.

Materials and Methods

Site Description

The Port of Savannah Garden City terminal in Garden City, GA is located approximately

30 kilometers inland from the Atlantic shore, on the Savannah River. The property consists of a three kilometer long docking area immediately adjacent to the Savannah River bordered by a large paved area (204.9 hectares) for container storage (Figure 4.1). Beyond this are several forested/green areas fragmented by paved and dirt roads. Furthest from the river is the rail yard accompanied by more paved container storage. Study sites were chosen based on a preliminary visual assessment of the port property to determine the extent and limit of the tawny crazy ant population.

Site one (Figure 4.2, Blue, 560 m) was located near the center and within the borders of the tawny crazy ant population. This site is composed of two large green areas adjacent to a container maintenance lot which was filled with damaged or stored trucks, trailers, and containers. This was also the location where the tawny crazy ant was first collected in June, 2015.

Sampling at this site was conducted in the unmaintained grassy transition areas between gravel or paved areas and wooded areas characterized by large hardwoods and abundant understory.

Site two (Figure 4.2, Green, 780 m) was a linear section of a paved street, Grange Road, on the north property line of the port. Upon initial investigation, tawny crazy ant workers could

100 be seen foraging along the roadside on the western half of the street only. The street was flanked on either side by water retention canals which were overgrown with grassy vegetation. A single linear transect, along the north side of the road, was used to sample this site.

Site three (Figure 4.2, Orange, 300 m) was located directly south of site one and was situated along the edges of a patch of mowed grass covered by a canopy of mature hardwood trees. This site, like site two, was located at the visual edge of the expanding tawny crazy ant population. Site three was bordered to the north by a 2.15 hectare gravel trailer storage lot and to the south by a traffic-intensive road connecting two areas of the port.

Site four (Figure 4.2, Purple, 670 m) was located to the south of site three and was just outside the known, visual perimeter of the tawny crazy ant population in June 2015. Sampling at this site surrounded a single large green area characterized by mature hardwoods and evergreens with a dense understory surrounded on all sides by both paved and gravel roads.

Collection Techniques

Ants were collected at each site using baiting and quadrat sampling to assess roadside habitats on and around the port property to investigate the movements of the tawny crazy ant outward, along disturbance corridors (roadsides), from their point of origin. At all sites, the sampling methods were alternated so that each sample was flanked by a pair of the other sample type. The samples were taken at least 10 meters, but no more than 25 m, from one another.

Baiting. Baits were made by placing crushed Pecan Sandies cookies (Keebler Company,

Battle Creek, MI) in 9 dram vials, and filling each vial to one centimeter from the bottom. A bait vial was upended allowing the cookie to fall onto the vial’s cap and the vial was left on its side next to the cap one to three meters beyond the gravel or paved areas adjacent to the wooded areas

(Figure 4.3A). After one hour, ants on the bait were aspirated and transferred to vials containing

75% ethanol.

Quadrat sampling. Quadrat sampling was conducted by placing a 0.5 x 0.5 m quadrat half on the paved/dirt road and half on the grassy area flanking the road (Figure 4.3B). A

101 flashlight was used to enhance the visualization of ants in and amongst the vegetation inside the quadrat. Vegetation and grass were gently moved and/or removed to highlight ant occurrence. All ants seen in the area of the quadrat during a five minute period were collected using an aspirator and then transferred to vials containing 75% ethanol.

Sampling was conducted along roadsides at and surrounding the Port of Savannah,

Georgia (Garden City, Georgia) which were selected for their orientation to what is believed to be the introduction site of the tawny crazy ant on the Port of Savannah property. Collections at sites two and three covered a transition (visually estimated) between invaded and non-invaded areas and were made in July and October 2015 and March and July 2016. Site one, adjacent to the center of the introduction site, and site four, beyond the visually invaded area, were sampled in

July 2015 and 2016.

Results

Species List

Combined sampling from all sites and dates resulted in the collection of 16 ant species representing 12 genera. Of the species collected, nine were exotic representing seven genera

(Table 4.1).

Co-occurrence of Ant Species

Site one. Five species were collected from site one (Table 4.2). Solenopsis invicta,

Brachymyrmex patagonicus Mayr, Monomorium minimum (Buckley), and Pheidole navigans

Forel were collected in quadrats along with N. fulva while only M. minimum was collected at a bait with N. fulva.

Co-occurrence of N. fulva and S. invicta in July 2015. Nylanderia fulva was present in all

20 bait samples and was the only species collected in those samples. Quadrat sampling resulted in the collection of N. fulva in all 20 samples, co-occurring with fire ants in samples 7, 8, 9, and 12

(Figure 4.4A) 102

Co-occurrence of N. fulva and S. invicta in July 2016. Baiting resulted in the collection of

N. fulva in samples 1-10 and 12-20. Solenopsis invicta was collected in the remaining sample, 11.

Quadrat sampling resulted in the collection of N. fulva in all 20 of the samples and it co-occurred with S. invicta in sample 8 (Figure 4.4B).

Site two. Thirteen species were collected from site two (Table 4.3). Aphaenogaster fulva

Roger, Brachymyrmex depilis Emery, B. patagonicus, Cyphomyrmex rimosus (Spinola),

Dorymyrmex bureni (Trager), Hypoponera opacior (Forel), Linepithema humile Mayr, M. minimum, P. navigans, Pheidole obscurithorax Naves and S. invicta were collected in quadrats along with N. fulva while only Cardiocondyla venustula Wheeler, C. rimosus, P. navigans and S. invicta were collected at a bait with N. fulva.

Co-occurrence of N. fulva and S. invicta in July 2015. Baiting resulted in the collection of

N. fulva in samples 1-10 and 12, while S. invicta was collected in samples 11-20. Quadrat sampling resulted in the collection of N. fulva in samples 1-12 and S. invicta in samples 2 and 9-

20 (Figure 4.5A).

Co-occurrence of N. fulva and S. invicta in October 2015. Baiting resulted in the collection of N. fulva in samples 1-7 and 10-15 while S. invicta was collected in samples 16, 17,

19, and 20. Quadrat sampling resulted in the collection of N. fulva in samples 1-16 and 19, and S. invicta in samples 12, 14, 15, 16, 18, and 20 (Figure 4.5B).

Co-occurrence of N. fulva and S. invicta in March 2016. Baiting resulted in the collection of N. fulva in samples 1-14 and S. invicta in samples 15-20. Quadrat sampling resulted in the collection of N. fulva in samples 1-19 and S. invicta in samples 17-20 (Figure 4.5C).

Co-occurrence of N. fulva and S. invicta in July 2016. Baiting resulted in the collection of

N. fulva in samples 1-20, while S. invicta was absent from all samples. Quadrat sampling resulted in the collection of N. fulva in samples 1-20 with S. invicta in samples 1, 16, and 20 (Figure

4.5D).

103

Site three. Eleven species were collected from site three (Table 4.4). Aphaenogaster fulva, B. patagonicus, Camponotus pennsylvanicus (De Geer), Cardiocondyla wroughtonii

(Forel), C. rimosus, L. humile, M. minimum, P. navigans, P. obscurithorax, and S. invicta were collected in quadrats with N. fulva while only P. navigans and S. invicta were collected at baits with N. fulva.

Co-occurrence of N. fulva and S. invicta in July 2015. Baiting resulted in the collection of

N. fulva in samples 1, 4, and 5. S. invicta was collected in samples 2-4 and 6-10. Quadrat sampling resulted in the collection of N. fulva in samples 1, 2, and 5. S. invicta in samples 3, 4, and 6-10 (Figure 4.6A).

Co-occurrence of N. fulva and S. invicta in October 2015. Baiting resulted in the collection of N. fulva in samples 1-10 and was the only species collected in those samples.

Quadrat sampling resulted in the collection of N. fulva in samples 1-10, co-occurring with S. invicta in samples 7, 8, and 9 (Figure 4.6B).

Co-occurrence of N. fulva and S. invicta in March 2016. Baiting resulted in the collection of N. fulva in samples 1-10, co-occurring with S. invicta in sample 10. Five additional baits were placed across the street to the south of site three (overlapping with site four). These baits collected N. fulva in samples 12-14 and S. invicta in samples 11 and 15. Quadrat sampling resulted in the collection of N. fulva in samples 1 and 5-9. Solenopsis invicta was collected in samples 4, 5, 6, 9, and 10 (Figure 4.6C).

Co-occurrence of N. fulva and S. invicta in July 2016. Baiting resulted in the collection of

N. fulva in samples 1-10 while S. invicta was absent from all bait samples. Quadrat sampling resulted in the collection of N. fulva in samples1-10, co-occurring with S. invicta in samples 3-5 and 8-10 (Figure 4.6D).

Site four. Nine species were collected from site four (Table 4.5). Brachymyrmex depilis,

B. patagonicus, P. navigans, P. obscurithorax and S. invicta were collected in quadrats along

104 with N. fulva while only B. patagonicus, P. navigans, and P. obscurithorax were collected at a bait with N. fulva.

Co-occurrence of N. fulva and S. invicta in July 2015. Baiting resulted in the collection of

S. invicta in samples 2-10 and 12-20, while N. fulva was absent from all bait samples. Quadrat sampling resulted in the collection of S. invicta in samples 1-13 and 15-20, while N. fulva was absent from all quadrat samples (Figure 4.7A).

Co-occurrence of N. fulva and S. invicta in July 2016. Baiting resulted in the collection of

N. fulva in samples 12-20 and S. invicta in samples 1-11. Quadrat sampling collected N. fulva in samples 4, 5, 12-15, 16, and 20. Solenopsis invicta was collected in samples 1-11 and 13-20

(Figure 4.7B).

Discussion

Co-occurring ant species

There were three species of ants whose numbers were unaffected by the invasion by the tawny crazy ant: Brachymyrmex patagonicus, Cyphomyrmex rimosus, and Pheidole navigans.

These species occurred on a relatively high number of our baits alongside S. invicta as well as N. fulva. All three are exotic themselves and have a head capsule size near or below the threshold

(0.6 mm) indicated by LeBrun et al. (2013) allowing for a decrease in competitive pressure from the tawny crazy ant. Also, noticeably absent from sandy roadside habitats invaded by N. fulva was D. bureni, a common species in human altered habitats with a conspicuous funnel shaped nest entrance, which at first co-occurred with S. invicta, but quickly disappeared as N. fulva expanded along the sandy roadside habitat.

Co-occurrence of N. fulva and S. invicta

Overall, the number of baits and quadrats containing N. fulva increased over the course of the year while the number containing S. invicta decreased (Figure 4.8). Although this trend was seen in both the baits and quadrats, the degree of turnover was greater in baits. Tawny crazy ants were collected at 49% of baits and in 50% of quadrats in July 2015 and were collected at 90% of

105 baits and in 83% of quadrats in July 2016. A corresponding decline was seen in the presence of S. invicta, having been collected at 51% of baits and in 61% of quadrats in July 2015 and at only

10% of baits and in 41% of quadrats in July 2016.

Seventy nine of 80 samples (bait and quadrat) taken from site one (Figure 4.2, blue) in

July 2015 and July 2016 contained N. fulva. Two years prior to our June 2015 discovery of N. fulva at this site, using the same baiting technique, only S. invicta was present on 20 Pecan

Sandies baits on this transect (Dan Suiter, personal observation). This location on the port is suspected to be the tawny crazy ant’s point of initial introduction. The near complete replacement of S. invicta by N. fulva at site one is not surprising given the high densities of N. fulva in the adjacent wooded areas (LeBrun et al. 2013).

North of the site one, site two (Figure 4.2, green) provided an opportunity to investigate the characteristics of the roadside ant communities within and beyond the invasion front as the population of N. fulva expanded over the year. Sampling in July 2015 showed N. fulva and S. invicta overlapped at about the halfway point along the road, with N. fulva moving eastward

(Figure 4.5A). By October, S. invicta was collected in six quadrat samples along the eastern half of the road and was at only four baits among the easternmost five samples, while N. fulva was collected in three quarters of the samples, nearly reaching the eastern end of site two. The quadrat samples detected N. fulva workers farther east and S. invicta workers farther west, with a fair amount of overlap, than did the bait samples which had no overlap in the two species, suggesting

N. fulva was outcompeting S. invicta at food resources as it moved along the road. The March

2016 samples also had overlap in the quadrat samples while the baits had none. The March sampling revealed that N. fulva had spread along the entire road in the quadrat samples with their presence at baits following slightly behind. The final sampling date in July 2016 returned only N. fulva in bait samples and only three instances of S. invicta in the quadrat samples.

To the south of site one, the site three samples (Figure 4.2, orange) documented the expansion of N. fulva along the edges of the grassy area and eventually across the road to the 106 south where it invaded site four. Initial samples at site three in July 2015 collected S. invicta nearly three times as often as N. fulva. By October, this relationship had reversed, with N. fulva in every sample and S. invicta in only three quadrat samples and on none of the baits. This trend was observed in the March 2016 samples as well, with the addition of five bait samples taken across the road to the south of site three. These five additional bait samples collected N. fulva three times and S. invicta two times. Across the multiple sampling dates it was not difficult to see the movement of N. fulva from north to south through the grassy area and eventually across the road, which is consistent with an outward expansion from their suspected area of introduction at site one. The final sampling event at site three in July 2016 collected N. fulva in all bait and quadrat samples while S. invicta was only in six quadrat samples and on no baits, showing the dominance of N. fulva over food resources in the presence of S. invicta.

South of site three, just across the street, is site four (Figure 4.2, purple) with samples encircling a wooded area surrounded by paved and dirt roads. The July 2015 samples collected no

N. fulva and had S. invicta in all but one quadrat and two bait. The samples from July 2016 had S. invicta in all but one quadrat again but only in about half the bait samples, with N. fulva dominating the remaining baits. The baits with N. fulva were located along the roadside across from site three. Although N. fulva surrounded the northern wooded area at site four, it was detected at a single point on a transect by four sampling techniques (pitfall, leaf litter, bait, and tree trunk sampling) from a related study (Chapter 3) of the wooded areas. The wooded area was sampled by collecting ten samples of the previously mentioned methods plus dissection of logs for a total of 50 samples which found 27 species in July 2016. Nylanderia fulva was detected less than 10 meters from the roadside, while remaining absent from the rest of the transect (Figure

4.9). This may suggest that N. fulva is able to spread faster along the disturbed margins of these wooded patches, surrounding them before moving into them.

The tawny crazy ant has several social and physiological characteristics that make it a formidable opponent of the red imported fire ant (LeBrun et al. 2013, LeBrun et al. 2014). The 107 tawny crazy ant’s extremely high abundance also allows it to overwhelm competing species by outnumbering them in aggressive encounters (Horn 2010). In addition to these factors, N. fulva is able to dominate in heavily wooded areas, unlike S. invicta, giving them access to resources unavailable to S. invicta and allowing them to greatly increase their numbers as they spread.

Several studies have revealed the repellant property of the defensive chemicals released by the tawny crazy ant (Chen et al. 2013, LeBrun et al. 2014, LeBrun et al. 2015). The venom of N. fulva can act as a toxic fumigant against the red imported fire ant as well as a detoxifying agent against the fire ant’s own venom (Wang and Henderson 2016). Zhang et al. (2015) showed a synergism between the tawny crazy ant’s venom and volatiles secreted from the Dufour’s gland, leading to increased attraction to the volatiles by workers. This increase in recruitment may lead to the dominance of tawny crazy ants through superior numbers as opposed to superior fighting ability (Horn et al. 2013). In addition to chemical advantages, the size of workers plays a role in the outcome of competitive interactions with larger species, likely because they cannot hide from the tawny crazy ant, leaving them more vulnerable to extirpation upon invasion (LeBrun et al.

2013).

Movement of N. fulva Along Roadside Habitats

The N. fulva population edge expanded about 400 meters along the linear road at site two over the course of the year with an increased rate of expansion occurring in the summer and fall months. There was a noticeable difference in the rate of expansion detected by quadrat and bait sampling. The baiting samples had a slight lag in the presence of N. fulva as well as in the loss of

S. invicta. This would suggest that S. invicta is able to dominate resources like the bait samples for some time after first contact with N. fulva although they are extirpated soon after. This also shows the ability of N. fulva to competitively exclude S. invicta from point sources of food, such as our bait samples, shortly after expansion into new territory. In numerous instances, active fire ant mounds were located within a meter of baits dominated by N. fulva. Combined with the site three samples, the site four samples show N. fulva moving about 250 meters over the course of 108 the year, traversing a heterogeneous environment and a heavily used road in the process. These rates of movement contrast with the findings of LeBrun et al. (2013) who measured the tawny crazy ant’s rate of expansion in a habitat described as coastal prairie. LeBrun et al. (2013) found the movement of two separate populations to be 175 m and 180 m over the course of the year, much slower than the rates recorded in this study through highly disturbed habitats.

It has been shown that anthropogenic disturbance corridors harbor and facilitate the growth and dispersal of both non-native plant and animal populations (Brown et al. 2006,

Merriam et al. 2006). Highly disturbed areas such as fire breaks, powerline cuts, dirt roads and roadsides are preferred by founding queens of both exotic and native species of ants (Stiles and

Jones 1998, King and Tschinkel 2007, 2016). This preference can lead these species to occupy disturbed habitats that extend deep into relatively undisturbed habitats which can act as refugia and harbor source populations for the species as they spread into the undisturbed areas (Stiles and

Jones 1998). The tendency of exotic ant species to occupy roadside habitats has led to the co- occurrence of dominant species like the red imported fire ant and the tawny crazy ant at the Port of Savannah, Georgia.

The regular movement outward of N. fulva from its initial point of invasion highlights the manifestation of its reproductive mode. Typical reproductive dispersal in ants involves a nuptial flight, where the reproductive castes take flight and mate, at which point the newly mated queens disperse via flight to found new colonies (Hart and Tschinkel 2012). The tawny crazy ant, along with some other invasive ant species, mates within the colony (Meyers 2008) resulting in a diffusion-like process called budding, where the colony spreads radially from its source as the ants are able to allocate resources and reproduce (Figure 4.10) (Suarez et al. 2010). This restricts the tawny crazy ant to terrestrial movements while dispersing into new habitats. This predictable mode of dispersal allows for easy monitoring of the colony’s expansion from its believed point of introduction (Figure 4.9). 109

References Cited

Brown, G. P., B. L. Phillips, J. K. Webb, and R. Shine. 2006. Toad on the road: use of roads as

dispersal corridors by cane toads (Bufo marinus) at an invasion front in tropical .

Biological conservation 133: 88-94.

Chen, J., T. Rashid, G. L. Feng, L. M. Zhao, D. Oi, and B. M. Drees. 2013. Defensive

chemicals of Tawny Crazy ants, Nylanderia fulva (Hymenoptera: Formicidae) and their

toxicity to Red Imported Fire ants, Solenopsis invicta (Hymenoptera: Formicidae).

Toxicon 76: 160-166.

DeHeer, C. J., M. A. D. Goodisman, and K. G. Ross. 1999. Queen dispersal strategies in the

multiple-queen form of the fire ant Solenopsis invicta. American Naturalist 153: 660-675.

Deyrup, M. 2003. An updated list of Florida ants (Hymenoptera: Formicidae). Florida

Entomologist 86: 43-48.

Gotzek, D., S. G. Brady, R. J. Kallal, and J. S. LaPolla. 2012. The importance of using

multiple approaches for identifying emerging invasive species: the case of the Rasberry

Crazy ant in the United States. Plos One 7: 10.

Hart, L. M., and W. R. Tschinkel. 2012. A seasonal natural history of the ant, Odontomachus

brunneus. Insectes Sociaux 59: 45-54.

Horn, K. 2010. Examining competitive interaction between Rasberry Crazy ants (Paratrechina

sp. nr. pubens) and Red Imported Fire ants (Solenopsis invicta) using laboratory and field

studies. Rice University.

Horn, K. C., M. D. Eubanks, and E. Siemann. 2013. The effect of diet and opponent size on

aggressive interactions involving Caribbean Crazy ants (Nylanderia fulva). Plos One 8: 7.

King, J. R., and W. R. Tschinkel. 2007. Range expansion and local population increase of the

exotic ant, Pheidole obscurithorax, in the southeastern United States (Hymenoptera:

Formicidae). Florida Entomologist 90: 435-439. 110

King, J. R., and W. R. Tschinkel. 2016. Experimental evidence that dispersal drives ant

community assembly in human-altered ecosystems. Ecology 97: 236-249.

LeBrun, E. G., J. Abbott, and L. E. Gilbert. 2013. Imported crazy ant displaces imported fire

ant, reduces and homogenizes grassland ant and arthropod assemblages. Biological

Invasions 15: 2429-2442.

LeBrun, E. G., N. T. Jones, and L. E. Gilbert. 2014. Chemical warfare among invaders: a

detoxification interaction facilitates an ant invasion. Science 343: 1014-1017.

LeBrun, E. G., P. J. Diebold, M. R. Orr, and L. E. Gilbert. 2015. Widespread chemical

detoxification of alkaloid venom by formicine ants. Journal of Chemical Ecology 41:

884-895.

Lofgren, C. S. 1986. The economic importance and control of imported fire ants in the United

States. Economic impact and control of social insects. Praeger, New York: 227-256.

McGlynn, T. P. 1999. The worldwide transfer of ants: geographical distribution and ecological

invasions. Journal of Biogeography 26: 535-548.

Merriam, K. E., J. E. Keeley, and J. L. Beyers. 2006. Fuel breaks affect nonnative species

abundance in Californian plant communities. Ecological Applications 16: 515-527.

Meyers, J. M. 2008. Identification, distribution and control of an invasive. Texas A&M

University.

Pimentel, D., R. Zuniga, and D. Morrison. 2005. Update on the environmental and economic

costs associated with alien-invasive species in the United States. Ecological Economics

52: 273-288.

Stiles, J. H., and R. H. Jones. 1998. Distribution of the Red Imported Fire ant, Solenopsis

invicta, in road and powerline habitats. Landscape Ecology 13: 335-346.

Storz, S. R., and W. R. Tschinkel. 2004. Distribution, spread, and ecological associations of the

introduced ant Pheidole obscurithorax in the Southeastern United States. Journal of

Insect Science 4.

111

Suarez, A. V., D. A. Holway, and T. J. Case. 2001. Patterns of spread in biological invasions

dominated by long-distance jump dispersal: insights from Argentine ants. Proceedings of

the National Academy of Sciences of the United States of America 98: 1095-1100.

Suarez, A. V., D. A. Holway, and P. S. Ward. 2005. The role of opportunity in the

unintentional introduction of nonnative ants. Proceedings of the National Academy of

Sciences of the United States of America 102: 17032-17035.

Suarez, A. V., T. P. McGlynn, and N. D. Tsutsui. 2010. Biogeographic and taxonomic patterns

of introduced ants, pp. 233-244. In L. Lach, C.L. Parr, and K.L. Abbott (eds.), Ant

Ecology. Oxford University Press, Oxford, NY.

Trager, J. C. 1984. A revision of the genus Paratrechina (Hymenoptera: Formicidae) of the

continental United States. Sociobiology 9: 51-162.

Valery, L., H. Fritz, J. C. Lefeuvre, and D. Simberloff. 2008. In search of a real definition of

the biological invasion phenomenon itself. Biological Invasions 10: 1345-1351.

Wang, C., and G. Henderson. 2016. Repellent effect of formic acid against the Red Imported

Fire ant (Hymenoptera: Formicidae): a field study. Journal of Economic Entomology

109: 779-784.

Wetterer, J. K., and J. L. Keularts. 2008. Population explosion of the Hairy Crazy ant,

Paratrechina pubens (Hymenoptera: Formicidae), on St. Croix, US Virgin Islands.

Florida Entomologist 91: 423-427.

Wetterer, J. K., O. Davis, and J. R. Williamson. 2014. Boom and bust of the Tawny Crazy ant,

Nylanderia fulva (Hymenoptera: Formicidae), on St. Croix, US Virgin Islands. Florida

Entomologist 97: 1099-1103.

Zhang, Q. H., D. L. McDonald, D. R. Hoover, J. R. Aldrich, and R. G. Schneidmiller. 2015.

North American invasion of the Tawny Crazy ant (Nylanderia fulva) is enabled by

pheromonal synergism from two separate glands. Journal of Chemical Ecology 41: 853-

858.

112

Table 4.1. Ant species collected from roadside habitats (all four sites combined) on the Port of Savannah in Garden City, Georgia during 2015 and 2016. Aphaenogaster fulva Roger Hypoponera opacior (Forel)

Brachymyrmex depilis Emery Linepithema humile Mayra

Brachymyrmex patagonicus Mayra Monomorium minimum (Buclkey)

Camponotus pennsylvanicus (De Geer) Nylanderia fulva Mayra

Cardiocondyla venustula Wheelera Pheidole dentata Mayr

Cardiocondyla wroughtonii (Forel)a Pheidole navigans Forela

Cyphomyrmex rimosus (Spinola)a Pheidole obscurithorax Navesa

Dorymyrmex bureni (Trager) Solenopsis invicta Burena a Exotic species according to Deyrup (2003).

113

Table 4.2. Site one species incidence. Percentage of traps (20 Pecan Sandies baits and 20 0.5 x 0.5 m quadrat samples) in which each of five ant species occurred at site one in July 2015 and July 2016. Species July 2015 July 2016

Bait Quadrat Bait Quadrat

Brachymyrmex patagonicusa 0 35 0 35

Monomorium minimum 0 5 10 10

Nylanderia fulvaa 100 100 95 100

Pheidole navigansa 0 0 0 5

Solenopsis invictaa 0 20 5 5 a Exotic species according to Deyrup (2003).

114

Table 4.3. Site two species incidence. Percentage of traps (20 Pecan Sandies baits and 20 0.5 x 0.5 m quadrat samples) in which each of 13 ant species occurred at site two in July and October 2015 and March and July 2016. Species July 2015 October 2015 March 2016 July 2016

Bait Quadrat Bait Quadrat Bait Quadrat Bait Quadrat Aphaenogaster fulva 0 10 0 0 0 0 0 0

Brachymyrmex depilis 0 0 0 0 0 0 0 10

Brachymyrmex patagonicusa 0 60 5 40 0 10 0 35

Cardiocondyla venustulaa 0 0 0 0 0 0 5 0

Cyphomyrmex rimosusa 0 30 5 15 0 0 0 20

Dorymyrmex bureni 0 5 0 0 0 0 0 0

Hypoponera opacior 0 0 0 5 0 0 0 0

Linepithema humilea 0 0 0 5 0 0 0 0

Monomorium minimum 0 0 0 0 0 0 0 10

Nylanderia fulvaa 55 60 65 85 70 95 100 100

Pheidole navigansa 0 35 15 10 0 0 0 35

Pheidole obscurithoraxa 0 10 0 5 0 0 0 0

Solenopsis invictaa 50 65 20 30 30 20 0 15 a Exotic species according to Deyrup (2003).

115

Table 4.4. Site three species incidence. Percentage of traps (10 Pecan Sandies baits and 10 0.5 x 0.5 m quadrat samples) in which each of 11 ant species occurred at site three in July and October 2015 and March and July 2016. Species July 2015 October 2015 March 2016 July 2016

Bait Quadrat Bait Quadrat Bait Quadrat Bait Quadrat Aphaenogaster fulva 0 0 0 10 0 0 0 0

Brachymyrmex patagonicus a 10 10 0 40 0 20 0 70

Camponotus pennsylvanicus 0 0 0 0 0 0 0 10

Cardiocondyla wroughtonii a 0 0 0 0 0 0 0 10

Cyphomyrmex rimosus a 0 10 0 10 0 30 0 20

Linepithema humile a 0 0 0 10 0 0 0 0

Monomorium minimum 0 0 0 10 0 0 0 0

Nylanderia fulva a 30 30 100 100 100 60 100 100

Pheidole navigans a 20 0 0 10 0 10 10 20

Pheidole obscurithorax a 0 0 0 0 0 10 0 0

Solenopsis invicta a 80 70 0 30 10 50 0 60 a Exotic species according to Deyrup (2003).

116

Table 4.5. Site four species incidence. Percentage of traps (20 Pecan Sandies baits and 20 0.5 x 0.5 m quadrat samples) in which each of nine ant species occurred at site four in July 2015 and July 2016. Species July 2015 July 2016 Bait Quadrat Bait Quadrat Brachymyrmex depilis 0 0 0 5

Brachymyrmex patagonicusa 5 55 5 75

Cyphomyrmex rimosusa 0 20 0 10

Dorymyrmex bureni 0 5 0 0

Nylanderia fulvaa 0 0 45 40

Pheidole dentata 5 0 0 0

Pheidole navigansa 0 30 5 25

Pheidole obscurithoraxa 0 10 5 15

Solenopsis invictaa 90 95 55 95 a Exotic species according to Deyrup (2003). 117

Figure 4.1. Port of Savannah property. The Port of Savannah is located in Garden City, Georgia

(Chatham County, Georgia). It is comprised of 485.6 hectares, much of which is container storage

(204.9 hectares), interspersed with green areas characterized by mature hardwood and evergreen trees with moderate to thick understory. The port is the fourth busiest in the U.S., and because it is in the southeastern U.S. is accompanied by mild temperatures and abundant rainfall, conditions conducive to the establishment of biological invaders. The yellow outline indicates the extent of the 485.6 hectare Port of Savannah property. 118

Figure 4.2. Location of bait and quadrat sampling sites. Blue lines indicate site one (560 m). The green line indicates site two (780 m). Orange lines indicate site three (300 m). Purple lines indicate site four (670 m). 119

Figure 4.3. Bait and Quadrat sampling. (A) A bait trap utilizing a nine dram vial and its cap along with crushed Pecan Sandies cookie. Ants found on the cookie, cap, or in the vial were aspirated and transferred to 75% ethanol. (B) A 0.5 x 0.5 m quadrat placed on the margin of the pavement and roadside within which all ants were collected over a five minute period with the aid of a flashlight and then transferred to 75% ethanol.

120

Figure 4.4. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits (circles) and in quadrats (squares) at site one during (A) July 2015 and (B) July 2016 at the Port of Savannah,

Garden City, Georgia.

121

Figure 4.5. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits (circles) and in quadrats (squares) at site two during (A) July

2015, (B) October 2015, (C) March 2016, and (D) July 2016 at the Port of Savannah, Garden City, Georgia. White indicates that neither species was collected.

122

Figure 4.6. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits (circles) and in quadrats (squares) at site three during (A) July

2015, (B) October 2015, (C) March 2016, and (D) July 2016 at the Port of Savannah, Garden City, Georgia. White indicates that neither species was collected. 123

Figure 4.7. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits (circles) and in quadrats (squares) at site four during (A) July 2015 and (B) July 2016 at the Port of Savannah,

Garden City, Georgia. White indicates that neither species was collected.

124

100 90 80

70

60 50

40 % % presence 30 20 10 0 2015 2016 2015 2016 Bait Quadrat S. invicta N. fulva

Figure 4.8. Incidence of S. invicta and N. fulva by year and sampling method. Percent of baits or quadrats containing either N. fulva or S. invicta during the July 2015 and July 2016 sampling dates at the port of Savannah.

125

Figure 4.9. Subsequent occurrence of Nylanderia fulva in four sample types (pitfall, leaf litter, bait, and tree trunk sampling) along a wooded transect following invasion of the adjacent site 4 roadside habitat in July 2016. 126

Figure 4.10. Occurrence of N. fulva (yellow) and S. invicta (red) collected on baits and in quadrats during (A) July 2015 and (B) July 2016 sampling events at the Port of Savannah, Garden

City, Georgia.

127

CHAPTER 5

POTENTIAL USE OF ARGENTINE ANT, LINEPITHEMA HUMILE MAYR,

SEMIOCHEMICALS FOR THE DELIVERY OF AN ACUTE TOXICANT

Introduction

The Argentine ant, Linepithema humile Mayr, is a nuisance pest found mainly in southern

California and the southeast United States (Vega and Rust 2001). It commonly manifests as a supercolony, able to dominate entire landscapes and become highly invasive in human altered habitats (Suarez et al. 2008). The supercolony contains many non-competing, reproductive queens that drive the abundance of this species where it occurs (Vasquez and Silverman 2008).

Contact insecticides are a reliable form of control for pest ant species such as the

Argentine ant (Wiltz et al. 2009). Current Argentine ant control methods utilize toxic baits and sprays in and around structures (Rust et al. 2003). Fipronil, a contact insecticide, is able to achieve high mortality with relatively low concentrations because it is readily transferred between worker ants through physical contact, its effect is delayed long enough for it to spread to critical colony members such as queens, and it kills the ants with efficiency (Hooper-Bui and Rust 2000,

Choe and Rust 2008, Hooper-Bui et al. 2015).

Delayed toxicity is a critical factor in the effectiveness of insecticidal agents against the

Argentine ant (Rust et al. 2004, Wiltz et al. 2009, Rust et al. 2017). Wiltz et al. (2009) conducted a series of tests to evaluate the effectiveness of bifenthrin, chlorfenapyr, thiamethoxam, and fipronil at killing Argentine ants in a laboratory colony. Workers were topically treated with the chemicals and mobility impairment as well as median lethal time was recorded. Time to mobility impairment was slowest in fipronil treated colonies although mortality was also highest in fipronil treated colonies, demonstrating fipronil’s ability to be spread effectively among the colony. 128

Choe and Rust (2008) tested eight insecticides for their ability to be spread by contact, and mortality in Argentine ants. After exposure to insecticide treated sand for one minute, ants were placed into a colony of untreated ants. Only fipronil, at rates of 0.002% and 0.004%, resulted in adequate contact transfer and subsequent high mortality rates within four days.

Mortality rates for the remaining insecticides were not significantly different from the control. A second experiment using fipronil treated ant corpses showed that necrophoresis played an important role in the contact transfer of the insecticide. Fipronil treated corpses placed closer to the nest resulted in higher mortality compared to corpses placed 30 cm away from the nest.

Costa and Rust (1999) tested the effects of fipronil and diazinon soil treatments on

Argentine ant foraging rates and mortality in potted plants. Twenty four laboratory colonies of

Argentine ants were each given one potted oleander plant in a 15 cm diameter pot. Colonies were given one of three rates, 5, 10, and 20 ppm, of a fipronil soil mixture consisting of granular fipronil mixed with moist soil or a broadcast treatment consisting of dry soil treated with 14 and

28 g of fipronil broadcast and added to the potting soil. After one week, foraging rates in all treatments but the control had dropped to zero. The soil mix treatment of fipronil had killed >90% of workers after one week, and had killed all queens after four weeks. Fipronil broadcast treatments killed <50% of workers at one week, and took eight weeks to kill all queens.

Vega and Rust (2003) marked foraging Argentine ants with a fluorescent brightener

(FB28) to determine the origin of resurgence after treatment with fipronil baits and sprays. Baits containing 0.0001% fipronil and sprays formulated at 0.06% fipronil achieved significant reduction in Argentine ant workers by four weeks. The percentage of marked ants in these treatments decreased throughout the experiment due to immigration from the surrounding areas, suggesting that to effectively control Argentine ants, a much larger treatment area may be necessary. Klotz et al. (2002) reported effective control of Argentine ant colonies around homes using a combination of baits or a barrier spray treatment. While both methods were effective at 129 controlling the ants, the baiting method required several reapplications to maintain adequate control while a single application of the spray equally effective.

Increased use of fipronil in residential areas is tied to the biological characteristics of the

Argentine ant in its invasive range (Gan et al. 2012). Spray applications of fipronil based insecticide formulations remain effective on treated surfaces for up to 20 days but can be found in runoff water from these same surfaces at 90 days post treatment (Jiang et al. 2014). The persistence of fipronil and its derivatives in the environment can lead to acute toxicity of non- target terrestrial and aquatic organisms in the affected areas, causing increased mortality and reduced reproductive potential (Overmyer et al. 2007, Pisa et al. 2015).

Greenberg et al. (2010) evaluated fipronil and bifenthrin levels in run-off and irrigation water directly following treatment at residential homes in 2007 and 2008. The 2007 treatment consisted of a spray application of the two insecticides using standard practices. The 2008 treatment applied product as a pin stream, utilized spray-free zones, and restricted the application of insecticides to the foundation of the homes. After one week the level of fipronil in the run-off water from the 2007 treatment was enough to cause acute aquatic toxicity to sensitive organisms.

At eight weeks post-treatment, the same was true for levels of bifenthrin in the run-off water. In contrast, the water samples from the 2008 treatment contained no detectable levels of fipronil at one week post-treatment and contained greatly reduced levels of bifenthrin when compared to the

2007 treatment.

Ecological concerns over the use of these insecticides have led to the investigation of alternative methods for Argentine ant control. Essential oils from plants can be a deterrent to ants accompanied by low mortality rates in field scenarios but the amounts needed to achieve comparable control levels to insecticidal treatments make the essential oils prohibitive to use

(Wiltz et al. 2007, Guerra et al. 2011, Scocco et al. 2012).

The use of Argentine ant trail pheromone mixed with a fipronil insecticide increased ant mortality compared to a fipronil treatment alone (Greenberg and Klotz 2000, Choe et al. 2014).

130

The efficacy of this method is dependent on the application method and ant trail density, with point source applications being more effective, and areas of heavy Argentine ant establishment being less affected (Suckling et al. 2008, Suckling et al. 2010, Suckling et al. 2011). These limitations lead to the need for spray applications for effective control of the Argentine ant when using trail pheromone, increasing the likelihood of groundwater contamination by the insecticide.

Efforts have been made to reduce the amount of fipronil needed to maintain acceptable control of Argentine ants by using granular formulations, instead of sprays, that are not as easily transferred to water sources (Klotz et al. 2010). The use of novel, species specific attractants such as fipronil laced termites for the Asian needle ant, a specialist predator on termites, allows for a targeted application of the insecticide to while reducing the amount of insecticide introduced into the environment (Buczkowski 2017). Additionally, Buczkowski and Silverman (2006) were able to alter the cuticular hydrocarbon profiles of the argentine ant by supplying some ants with a specific food source, affecting the behavior of a portion of the colony during interactions with conspecifics. Therefore, the ability to influence the behavior of the ants through their communication cues may be a powerful tool for their control.

Semiochemical based retrieval behaviors by the Argentine ant may present a partial solution to the issue of insecticide volume and application method. Evidence suggests that ant retrieval behavior, where ants are signaled to move something into or away from their nest, can be requisitioned by other insects and plants to aid in egg and seed dispersal (Hughes and Westoby

1992, Windsor et al. 1996). Elaiosomes on plant seeds and their mimics, the capitula on stick insects’ eggs, can cause certain species of ants to pick up and move these objects to a desired location for incubation. The stick insect’s egg capitulum has a large lipid component believed to be an incentive for retrieval, similar to the large lipid component of elaiosomes found on some plant seeds (Brew et al. 1989).

The chemistry behind necrophoresis, the transport of dead individuals away from the colony, was explored by Choe et al. (2009) and revealed that the molecules eliciting the response, 131 a combination of cuticular triglycerides, are always present whether ants are alive or dead. The signal is instead masked by a combination of other highly volatile compounds, dolichodial and iridomyrmecin, thus preventing conspecific necrophoretic behavior toward live workers. When an ant dies, the volatile compounds evaporate or degrade quickly, revealing the triglyceride signal and eliciting the associated retrieval behavior. Necrophoresis was determined to play an important role in the spread of fipronil among Argentine ant workers when worker corpses treated with the insecticide were placed near the nest (Choe and Rust 2008, Wiltz et al. 2010).

Fipronil baits and sprays are effective at controlling populations of the Argentine ant, but sprays have ecological drawbacks. Investigations intended to reduce the volume of insecticide needed to control this species are therefore warranted. The intent of this study was to explore the possibility of using ant semiochemicals as a means of delivering small volumes (ng) of contact insecticides. Extracts containing necrophoretic signal molecules from the Argentine ant cuticle were tested in combination with a fipronil suspension to assess the viability of this novel approach to ant management.

Materials and Methods

Study Organism

Leaf litter debris where Argentine ants were nesting was collected on the University of

Georgia Griffin campus in Griffin, GA in June and July of 2015 and 2016. The ants and debris were kept in Fluon-lined rearing bins (56 x 43 x 13 cm) and provided test tubes containing water or 20% sucrose water stoppered with a cotton ball. Freshly killed house crickets, Acheta domestica, were provided ad libitum as a protein source. Nesting chambers consisted of a 90 x 18 mm Petri dish filled three fourths full with Castone (Dentsply Sirona, York, PA). Castone is water absorbent and increases the humidity inside the nest chamber when wetted. Holes were drilled in either side of the Petri dish to serve as entrances to the nesting chamber. Ants moved, as the leaf litter dried, into the moistened, Castone-filled dishes allowing for easy collection of workers, queens, and brood. 132

Pupa Retrieval Time

Experimental colonies were created by removing a nesting chamber, containing multiple queens, thousands of workers, and grams of brood, from a rearing bin and placing it in the corner of a Pyrex glass tray (24.5 x 20 x 7.5 cm, Asahi Glass Company, Tokyo, Japan) with Fluon on the inner walls to prevent escape (Figure 5.1). Colonies were provided 20% sugar water and water, as described above, and allowed to acclimate for 24 hours before use in an experiment.

Retrieval of Intact Pupae

Nesting chambers containing queens, workers and brood were removed from rearing bins and placed in a Fluon lined plastic container (19.5 x 14 x 10 cm) and immobilized with carbon dioxide gas. Pupae were removed under a microscope using a small paintbrush. A single pupa was placed in the center of the open area of the dish (Figure 5.1D) containing the 24 hour old colony, and the time to retrieval recorded. Retrieval was defined by a worker picking up and bringing a removed pupa to a nest entrance. Ten experimental colonies were assayed three times for a total of 30 replicates. The experiment was repeated once per colony using pupae killed by freezing to determine if worker ants responded to pupa behavior.

Activity of Pupal Extracts

Cuticular extracts from ant pupae were obtained by placing two pupae on a piece of filter paper (2 x 3 mm, Whatman #5) and applying 10 µl of n-hexane or methanol in single drops directly onto the pupae. Pupae were discarded and the solvent allowed to evaporate for one hour before bioassay of the filter paper.

A 30 mm diameter white paper disc was placed under the glass tray in the center of the open area in front of the Castone nest (Figure 5.1D). A piece of filter paper, with or without the cuticular extract, was placed onto the glass bottom of an experimental arena in the center of the circular activity area. The number of ants surrounding and on top of the filter paper and inside the circle’s border was recorded every minute for twenty minutes. Treatments consisted of a no item control (nothing in the circle), an untreated filter paper, a methanol treated filter paper, and a filter

133 paper containing solvent-extracted cuticular compounds. Each treatment was replicated five times.

Activity of Fipronil-laced Ant Cadaver Extracts

Cadavers were obtained by placing live worker ants in a freezer for 24 hours after which they were allowed to air dry for one hour. Ant cadavers (≈ 600) were placed in a small beaker with 3 ml of methylene chloride that was gently swirled for two minutes. The ants were removed and the solvent added to a clean beaker containing 200 cylindrical paper wicks (MiTeGen, extra fine liquid wicks, 3 x <1 mm). Excess solvent was allowed to evaporate and the wicks were allowed to dry for an hour. The process was repeated without ants to obtain methylene chloride treated wicks (negative control). Fipronil (0.10%, 0.01%, 0.001%; Termidor SC, BASF Corp,

Research Triangle Park, NC, USA) was applied to one hour old ant corpses, methylene chloride treated wicks, and ant extract treated wicks with three pumps from a hand-held spray bottle. The fipronil treated items were allowed to dry for one hour and then 30 were gently placed into an experimental colony and replicated six times.

Experimental colonies were created by adding ~300 live worker ants (estimated gravimetrically) to a Fluon lined glass dish (24.5 x 19.8 x 7.5 cm; Pyrex, Corning Inc., Corning,

NY, USA) containing an empty, moistened nesting chamber, as described previously (Figure

5.2). Ants were allowed to acclimate for 24 h before experiments were initiated. Experiments consisted of placing 30 treatment items (ants or wicks) onto a flat, plastic platform made from one third of a weigh boat, which was then placed in the center of the open space in the glass dish

(Figure 5.2B).

The number of wicks or ants removed from the platform was recorded after one, two and

24 hours to obtain removal counts. Live ants and cadavers in each colony were removed after 72 hours and counted. Percent mortality was determined by dividing the number of dead ants by the number of dead plus the number alive.

134

Statistical Analysis

Activity of Pupal Extracts. Results from the activity experiment were count data. The number of ants within the border of the circle at each minute, summed for all 20 minutes, was analyzed using a negative binomial distribution by PROC GLIMMIX (SAS Institute Inc., Carey,

NC, USA) with colony as a random effect. Differences in least square means were determined by pairwise t-tests. Residual plots suggested an even distribution around zero, indicating the data were normally distributed (Figure 5.3A).

Activity of Fipronil Laced Ant Cadaver Extracts. The number of items removed by live ants at one, two, and 24 hours was analyzed by 1-way analysis of variance using PROC

GLIMMIX (SAS Institute Inc., Carey, NC, USA). Removal counts were modeled using a Poisson distribution with colony as a random effect. Residual plots suggested an even distribution around zero, indicating the data were normally distributed (Figure 5.3B). Percent mortality was modeled using a binomial distribution with replicate as a random effect and percent mortality was analyzed by 1-way analysis of variance with events by trials syntax by PROC GLIMMIX (SAS Institute

Inc., Carey, NC, USA). Residual plots suggested an even distribution around zero, indicating the data were normally distributed (Figure 5.3C).

Results

Pupa Retrieval Time

All 30 assayed pupae were moved into the Castone nest. Worker ants retrieved pupae with a mean time of 289.3 seconds (median = 246.5, range 56-688, n = 30). All ten of the dead pupae were brought back into the nest.

Activity of Pupal Extracts

Treatments were separated, for the activity, into three statistically distinct groups, corresponding to treatment type (F = 49.2; d.f. = 4,16; P < .0001; Table 5.1). The no item control treatment demonstrated the random occurrence of workers in the circular activity area without an introduced stimulus. This resulted in low activity in the circle, averaging 2.5 ants over the 135 duration of the experiment. The addition of an inert item, with or without exposure to just solvent, increased ant response by 14- (no-solvent blanks) to 17-fold (methanol only). There was no difference in ant response to wicks (no solvent) and wicks treated with methanol and allowed to dry. Treatments containing either pupal extract (hexane or methanol) were not significantly different and elicited activity that was more than three times that of any treatment not containing a pupal extract. The methanol extract elicited the greatest ant activity with an average of 169.4 ants counted in the activity area followed closely by the n-hexane extract with 143.6 ants.

Activity of Fipronil-laced Ant Cadaver Extracts

Ant cadavers and wicks containing the extract from ant cadavers were removed in significantly greater numbers (86.3% to 98.7% removal) than wicks not containing ant cadaver odors (2.3% to 5.7% removal) (Table 5.2, Hour 1; F = 23.0; d.f. = 9,45; P < .0001). Furthermore, wicks containing the extract from ant cadavers were removed at the same rate as ant cadavers, whether they were treated with insecticide or not, suggesting that the fipronil insecticide formulation was not repellent at the concentrations tested. Lastly, wicks not containing ant cadaver odors were ignored by the ants and were not removed (0.7 to 1.7 wicks removed at one hour). At two and 24 hours the patterns of item removal had not changed appreciably, either in number removed or statistical significance. At 24 hours, retrieval in treatments containing ant cadavers, or wicks with an ant cadaver extract, ranged from 90% to 100% while retrieval in treatments with wicks not containing the extract ranged from 16% to 45.7%.

Mortality of ants exposed to ant cadavers treated with fipronil or wicks containing an ant cadaver extract and treated with fipronil increased with fipronil concentration (Table 5.3; F =

425.7; d.f. = 9,50; P < .0001). Mortality in 0.001% fipronil treatment groups (14.0% or 19.3%) was similar to the untreated control group (15.1%). Mortality increased in the 0.01% fipronil treatments with no extract but remained near control levels (not significantly different) for the extract treated wicks (13.5% vs 15.1% mortality). At 0.10% fipronil, mortality was highest in the

136 treated ant corpses group (84.9% mortality). The fipronil treated wicks and extract treated wicks also had increased mortality (51.9% and 43.9% mortality).

Discussion

Pupa Retrieval Time

The pupa retrieval experiment demonstrated that Argentine ants have a strong retrieval behavior for immature stages; all pupae were retrieved in less than 12 minutes (mean and median retrieval times of 4.1 and 4.8 minutes). Dead pupae were also all readily returned to the nest. A number of published studies have demonstrated that cuticular hydrocarbons on the surface of ant eggs, larvae, and workers elicit specific behavioral responses from nestmates, including destruction of foreign eggs, recognition and care for larvae, and removal of nestmate corpses from the nest (Bonavitacougourdan et al. 1988, 1989, Choe et al. 2009, van Wilgenburg et al.

2010, Shimoji et al. 2012, Helantera et al. 2014).

Activity of Pupal Extracts

The results from the activity measure experiment give us an idea of the cues for pupa retrieval by giving the ants two forms of sensory cues, physical and chemical. The no item control demonstrated base line activity that would be expected from random movement of foragers in the arena environment. The blank filter paper and methanol treated filter paper increased the activity of the foragers in the sample area, providing evidence for the ants’ interest in new, inert objects in their foraging area. The methanol treated filter paper further demonstrated that methanol had no impact on the ants’ interest since interest in it was not significantly different from the blank control. When the solvent based cuticular extracts were added to the filter paper, the activity in the sampling area was enhanced four- (n-hexane) or five- (methanol) fold. The differences among the no item control, the inert objects, and the objects treated with a cuticular extract show an additive effect on interest and activity between the presence of a foreign object and the presence of a chemical cue on that object (Choe et al. 2014). Further experimentation using the pupal cuticular extracts as a means to elicit the ant’s retrieval behavior was hindered by damage to the

137 pupae resulting from exposure to the solvents used for extraction. Using the ant’s retrieval behavior to deliver fipronil laced objects into the heart of the nest may reduce the amount of insecticide needed for appreciable levels of control.

Activity of Fipronil-laced Ant Cadaver Extracts

The data in the previous experiment partially demonstrate Argentine ants’ use of cuticular compounds as recognition cues. According to van Wilgenburg et al. (2010), manipulation of cuticular hydrocarbon profiles of Argentine ants with homologous compounds elicited aggressive responses from conspecific workers. Krasnec and Breed (2013) showed that variation in the positional isomers of alkanes found on the cuticle of Formica argentea was sufficient for nestmate recognition. Sainz-Borgo et al. (2011) concluded that recognition cues are present over the whole body of the trap jaw ant, Odontomachus bauri. Wiltz et al. (2010) utilized a triglyceride, common on the cuticle of ants, to elicit interest by the Red Imported Fire ant,

Solenopsis invicta, in an inert carrier object treated with fipronil as a novel delivery method for the toxicant.

The removal experiment revealed a strong item retrieval response for all experimental treatment groups that were composed of an ant cadaver or cadaver extract. Removal of ant cadavers, ant cadavers treated with fipronil, and wicks treated with a cuticular extract and fipronil were statistically identical, demonstrating that the methylene chloride extract elicited a retrieval response from the ants similar to that of a recently killed nestmate and that the insecticide formulation was not repellent. The interaction between the ant’s response to the extract and the non-repellency of an insecticide might allow for a targeted application of insecticide, thereby helping to reduce insecticide run-off and ground water contamination (Mahler et al. 2009,

Greenberg et al. 2010, Gan et al. 2012, Jiang et al. 2014, Hayasaka et al. 2015).

Reconciling the removal data with its associated mortality was problematic due to the difference in trends between the two experiments. While removal was clearly associated with the presence of the extracted cuticular compounds, mortality generally increased with increasing rates

138 of fipronil, a trend found in previous studies (Choe and Rust 2008). Ant corpses treated with fipronil generated the greatest mortality at each fipronil rate, possibly due to a difference in the amount of the fipronil available for contact toxicity of worker ants compared to the treatments with wicks. Ant cadavers likely did not absorb liquid like the wicks and so more of the spray may have remained on the surface of the ant cuticle resulting in greater bioavailability of fipronil by contact with the cadavers than with wicks.

The near complete removal but lower mortality in the extract wick treatments when compared to the fipronil only wicks, which had low removal rates yet higher mortality at each fipronil rate, confounded the relationship between removal of wicks and mortality of ants handling them. This discrepancy may be explained by the extreme toxicity of fipronil to worker ants. In the activity measure experiment (Table 5.1, Blank control) the mere presence of a foreign object was enough to elicit a 14-fold increase of activity around the object. Interaction with wicks

(touching and crawling on and around), even without removal, may have provided ample opportunity for the ants to acquire lethal doses of fipronil via contact, especially at the high rates

(.01% and .10%). Additionally, compounds extracted from ant cuticle may have had a negative effect on wick absorbency, changing the bioavailability of the fipronil.

The novelty of this application method and its potential to reduce environmental deposition of insecticide warrants further exploration. Reducing the amount of insecticide needed to control this widespread pest species would have a significant and positive ecological impact and may lead to new control methods for other pest ant species.

139

References Cited

Bonavitacougourdan, A., J. L. Clement, and C. Lange. 1988. Larvae recognition in the

Camponotus vagus Scop ant - larvae phenotypes of cuticular hydrocarbon patterns.

Comptes Rendus De L Academie Des Sciences Serie Iii-Sciences De La Vie-Life

Sciences 306: 299-305.

Bonavitacougourdan, A., J. L. Clement, and C. Lange. 1989. The role of cuticular

hydrocarbons in recognition of larvae by workers of the ant Camponotus vagus - changes

in the chemical signature in response to social-environment (Hymenoptera: Formicidae).

Sociobiology 16: 49-74.

Brew, C. R., D. J. O'Dowd, and I. D. Rae. 1989. Seed dispersal by ants: behaviour-releasing

compounds in elaiosomes. Oecologia 80: 490-497.

Buczkowski, G. 2017. Prey-baiting as a conservation tool: selective control of invasive ants with

minimal non-target effects. Insect Conserv. Divers. 10: 302-309.

Buczkowski, G., and J. Silverman. 2006. Geographical variation in Argentine ant aggression

behaviour mediated by environmentally derived nestmate recognition cues. Anim. Behav.

71: 327-335.

Choe, D.-H., and M. K. Rust. 2008. Horizontal transfer of insecticides in laboratory colonies of

the Argentine ant (Hymenoptera: Formicidae). J. Econ. Entomol. 101: 1397-1405.

Choe, D.-H., J. G. Millar, and M. K. Rust. 2009. Chemical signals associated with life inhibit

necrophoresis in Argentine ants. Proc. Natl. Acad. Sci. U. S. A. 106: 8251-8255.

Choe, D. H., K. Tsai, C. M. Lopez, and K. Campbell. 2014. Pheromone-assisted techniques to

improve the efficacy of insecticide sprays against Linepithema humile (Hymenoptera:

Formicidae). J. Econ. Entomol. 107: 319-325.

Costa, H. S., and M. K. Rust. 1999. Mortality and foraging rates of Argentine ant

(Hymenoptera: Formicidae) colonies exposed to potted plants treated with fipronil. J.

Agric. Urban Entomol. 16: 37-48.

140

Gan, J., S. Bondarenko, L. Oki, D. Haver, and J. X. Li. 2012. Occurrence of fipronil and its

biologically active derivatives in urban residential runoff. Environ. Sci. Technol. 46:

1489-1495.

Greenberg, L., and J. H. Klotz. 2000. Argentine ant (Hymenoptera: Formicidae) trail

pheromone enhances consumption of liquid sucrose solution. J. Econ. Entomol. 93: 119-

122.

Greenberg, L., M. K. Rust, J. H. Klotz, D. Haver, J. N. Kabashima, S. Bondarenko, and J.

Gan. 2010. Impact of ant control technologies on insecticide runoff and efficacy. Pest

Manag. Sci. 66: 980-987.

Guerra, M. D., D. R. Suiter, and C. M. Scocco. 2011. Topical toxicity of nine essential oils to

Camponotus pennsylvanicus (Hymenoptera: Formicidae). Sociobiology 58: 419-426.

Hayasaka, D., N. Kuwayama, A. Takeo, T. Ishida, H. Mano, M. N. Inoue, T. Nagai, F.

Sanchez-Bayo, K. Goka, and T. Sawahata. 2015. Different acute toxicity of fipronil

baits on invasive Linepithema humile supercolonies and some non-target ground

arthropods. Ecotoxicology 24: 1221-1228.

Helantera, H., S. J. Martin, and F. L. W. Ratnieks. 2014. Recognition of nestmate eggs in the

ant Formica fusca is based on queen derived cues. Curr Zool 60: 131-136.

Hooper-Bui, L. M., and M. K. Rust. 2000. Oral toxicity of abamectin, boric acid, fipronil, and

hydramethylnon to laboratory colonies of Argentine ants (Hymenoptera: Formicidae). J.

Econ. Entomol. 93: 858-864.

Hooper-Bui, L. M., E. S. C. Kwok, B. A. Buchholz, M. K. Rust, D. A. Eastmond, and J. S.

Vogel. 2015. Insecticide transfer efficiency and lethal load in Argentine ants. Nuclear

Instruments & Methods in Physics Research Section B-Beam Interactions with Materials

and Atoms 361: 665-669.

Hughes, L., and M. Westoby. 1992. Capitula on stick insect eggs and elaiosomes on seeds:

convergent adaptations for burial by ants. Funct. Ecol. 6: 642-648.

141

Jiang, W. Y., A. Soeprono, M. K. Rust, and J. Gan. 2014. Ant control efficacy of pyrethroids

and fipronil on outdoor concrete surfaces. Pest Manag. Sci. 70: 271-277.

Klotz, J. H., M. K. Rust, L. Greenberg, and M. A. Robertson. 2010. Developing low risk

management strategies for Argentine ants (Hymenoptera: Formicidae). Sociobiology 55:

779-785.

Klotz, J. H., M. K. Rust, H. S. Costa, D. A. Reierson, and K. Kido. 2002. Strategies for

controlling Argentine ants (Hymenoptera: Formicidae) with sprays and baits. J. Agric.

Urban Entomol. 19: 85-94.

Krasnec, M. O., and M. D. Breed. 2013. Colony-specific cuticular hydrocarbon profile in

Formica argentea ants. J. Chem. Ecol. 39: 59-66.

Mahler, B. J., P. C. Van Metre, J. T. Wilson, M. Musgrove, S. D. Zaugg, and M. R.

Burkhardt. 2009. Fipronil and its degradates in indoor and outdoor dust. Environ. Sci.

Technol. 43: 5665-5670.

Overmyer, J. P., D. R. Rouse, J. K. Avants, A. W. Garrison, M. E. Delorenzo, K. W. Chung,

P. B. Key, W. A. Wilson, and M. C. Black. 2007. Toxicity of fipronil and its

enantiomers to marine and freshwater non-targets. J Environ Sci Heal B 42: 471-480.

Pisa, L. W., V. Amaral-Rogers, L. P. Belzunces, J. M. Bonmatin, C. A. Downs, D. Goulson,

D. P. Kreutzweiser, C. Krupke, M. Liess, M. McField, C. A. Morrissey, D. A.

Noome, J. Settele, N. Simon-Delso, J. D. Stark, J. P. Van der Sluijs, H. Van Dyck,

and M. Wiemers. 2015. Effects of neonicotinoids and fipronil on non-target

invertebrates. Environ Sci Pollut R 22: 68-102.

Rust, M., D. A Reierson, and J. H Klotz. 2004. Delayed Toxicity as a Critical Factor in the

Efficacy of Aqueous Baits for Controlling Argentine Ants (Hymenoptera: Formicidae),

vol. 97.

Rust, M., D. A. Reierson, and J. H. Klotz. 2017. Factors affecting the performance of bait

toxicants for argentine ants (Hymenoptera: Formicidae), vol.

142

Rust, M. K., D. A. Reierson, and J. H. Klotz. 2003. Pest management of Argentine ants

(Hymenoptera: Formicidae). J. Entomol. Sci. 38: 159-169.

Sainz-Borgo, C., A. Cabrera, and J. V. Hernandez. 2011. Nestmate recognition in the ant

Odontomachus bauri (Hymenoptera: Formicidae). Sociobiology 58: 701-718.

Scocco, C. M., D. R. Suiter, and W. A. Gardner. 2012. Repellency of five essential oils to

Linepithema humile (Hymenoptera: Formicidae). J. Entomol. Sci. 47: 150-159.

Shimoji, H., Y. Fujiki, R. Yamaoka, and K. Tsuji. 2012. Egg discrimination by workers in

Diacamma sp from Japan. Insectes Soc. 59: 201-206.

Suarez, A. V., D. A. Holway, and N. D. Tsutsui. 2008. Genetics and behavior of a colonizing

species: the invasive Argentine ant. Am. Nat. 172: S72-S84.

Suckling, D. M., L. D. Stringer, and J. E. Corn. 2011. Argentine ant trail pheromone disruption

is mediated by trail concentration. J. Chem. Ecol. 37: 1143-1149.

Suckling, D. M., R. W. Peck, L. D. Stringer, K. Snook, and P. C. Banko. 2010. Trail

pheromone disruption of Argentine ant trail formation and foraging. J. Chem. Ecol. 36:

122-128.

Suckling, D. M., R. W. Peck, L. M. Manning, L. D. Stringer, J. Cappadonna, and A. M. El-

Sayed. 2008. Pheromone disruption of Argentine ant trail integrity. J. Chem. Ecol. 34:

1602-1609. van Wilgenburg, E., R. Sulc, K. J. Shea, and N. D. Tsutsui. 2010. Deciphering the chemical

basis of nestmate recognition. J. Chem. Ecol. 36: 751-758.

Vasquez, G. M., and J. Silverman. 2008. Queen acceptance and the complexity of nestmate

discrimination in the Argentine ant. Behav. Ecol. Sociobiol. 62: 537-548.

Vega, S. J., and M. K. Rust. 2001. The Argentine ant - a significant invasive species in

agricultural, urban and natural environments. Sociobiology 37: 3-25.

143

Vega, S. Y., and M. K. Rust. 2003. Determining the foraging range and origin of resurgence

after treatment of Argentine ant (Hymenoptera: Formicidae) in urban areas. J. Econ.

Entomol. 96: 844-849.

Wiltz, B. A., D. R. Suiter, and W. A. Gardner. 2007. Deterrency and toxicity of essential oils to

Argentine and Red Imported Fire ants (Hymenoptera: Formicidae). J. Entomol. Sci. 42:

239-249.

Wiltz, B. A., D. R. Suiter, and W. A. Gardner. 2009. Activity of bifenthrin, chlorfenapyr,

fipronil, and thiamethoxam against Argentine ants (Hymenoptera: Formicidae). J. Econ.

Entomol. 102: 2279-2288.

Wiltz, B. A., D. Suiter, W. Gardner, and C. W. Berisford. 2010. A Novel Delivery Method for

Ant (Hymenoptera: Formicidae) Toxicants, vol. 3. Midsouth Entomologist.

Windsor, D. M., D. W. Trapnell, and G. Amat. 1996. The egg capitulum of a Neotropical

walkingstick, Calynda bicuspis, induces aboveground egg dispersal by the ponerine ant,

Ectatomma ruidum. J. Insect Behav. 9: 353-367.

144

Table 5.1. Mean number of ants inside the border of a 30-mm circle around paper wicks treated with methanol and n-hexane extracts containing cuticular compounds from live pupae. Treatment Mean (± SE) number of antsa Methanol cuticular extract 169.4 ± 35.9 a

Hexane cuticular extract 143.6 ± 30.5 a

Methanol solvent only 43.5 ± 9.6 b

Blank control 34.8 ± 7.8 b

No item control 2.5 ± 0.9 c

F = 49.2; d.f. = 4,16; P < .0001 a Mean of n = 5

145

Table 5.2. Average number of insecticide treated and untreated ant cadavers and paper wicks containing a methylene chloride extract of Argentine ant corpses removed at one, two, and 24 hours. Treatment Hour 1a Hour 2b Hour 24c

Ant Cadavers 26.1 ± 2.6 a 27.8 ± 2.2 a 30.0 ± 2.2 a

.001% fipronil + cadavers 26.8 ± 2.6 a 30.0 ± 2.3 a 30.0 ± 2.2 a

.01% fipronil + cadavers 26.8 ± 2.6 a 30.0 ± 2.3 a 30.0 ± 2.2 a

.10% fipronil + cadavers 25.9 ± 2.5 a 27.8 ± 2.2 a 30.0 ± 2.2 a

.001% fipronil + cadaver extract + wicks 27.7 ± 2.7 a 30.0 ± 2.3 a 30.0 ± 2.2 a

.01% fipronil + cadaver extract + wicks 29.6 ± 2.8 a 30.0 ± 2.3 a 30.0 ± 2.2 a

.10% fipronil + cadaver extract + wicks 25.9 ± 2.5 a 26.5 ± 2.2 a 27.0 ± 2.1 a

.001% fipronil + wicks 0.7 ± 0.3 b 0.8 ± 0.4 c 7.0 ± 1.1 c

.01% fipronil + wicks 1.5 ± 0.5 b 2.0 ± 0.6 bc 4.8 ± 0.9 c

.10% fipronil + wicks 1.7 ± 0.5 b 2.7 ± 0.7 b 13.7 ± 1.5 b a F = 23.0; d.f. = 9,45; P < .0001 b F = 26.2; d.f.= 9,45; P < .0001 c F = 24.0; d.f. = 9,45; P < .0001 146

Table 5.3. Mortality of Argentine ants three days following exposure to insecticide-treated and untreated ant cadavers and paper wicks containing a methylene chloride extract of ant cadavers. Treatment Percent Mortalitya

Ant Cadavers 15.1 ± 4.3 g

.001% fipronil + cadavers 19.3 ± 1.2 f

.01% fipronil + cadavers 58.9 ± 7.3 b

.10% fipronil + cadavers 84.9 ± 4 a

.001% fipronil + cadaver extract + wicks 14.0 ± 2.1 g

.01% fipronil + cadaver extract + wicks 13.5 ± 1.5 g

.10% fipronil + cadaver extract + wicks 43.9 ± 7.9 d

.001% fipronil + wicks 18.3 ± 6.2 f

.01% fipronil + wicks 25.9 ± 5.1 e

.10% fipronil + wicks 51.9 ± 5.9 c

F = 425.7; d.f. = 9,50; P < .0001 a Means ± SE based on n = 6.

147

B C

D A

Figure 5.1. Test arena for pupa retrieval and activity measure experiments. Harborages (A) along with water and sugar solution tubes (B and C) were provided. Experimental items (intact live and dead pupae or paper wicks containing a methanol or hexane extract) were placed in the center of the circle (D) and time to retrieval was measured for pupae or ant numbers counted surrounding paper wicks inside the circle’s border with or without pupal extracts.

A, Castone nesting chamber.

B, 20% sucrose solution filled test tube with cotton stopper.

C, Water filled test tube with cotton stopper.

D, Experimental item placement area.

148

B A

Figure 5.2. Test arena for item removal and mortality experiments. Harborage (A) containing ~

300 worker ants was provided and the weighboat (B) was added when the treatment items were placed in the arena.

A, Castone nesting chamber.

B, Weighboat containing 30 experimental items.

149

Figure 5.3. (A) Residual plot for the activity measure experiment. (B) Residual plot for the removal experiment at hour one. (C) Residual plot for the mortality experiment.

150

CHAPTER 6

CONCLUSIONS

These studies have shown that the impact of invasive ants is great, and will be a force to consider with globalization and connectivity defining the modern era. Understanding new threats to our remaining biodiversity is the first step to ensuring its longevity. Sophisticated measures to tackle the problem of invasive species will need to be considered more thoroughly in order to maintain practices that are both effective and sustainable. While this research has begun to investigate the possibilities of species specific control strategies, more effort will be needed to see these methods become safe, reliable, and cost effective.