|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 FORAGING STRATEGIES AND AGGRESSION PATTERNS OF NYLANDERIA FULVA (Mayr) (HYMENOPTERA: FORMICIDAE) IN NORTH CENTRAL FLORIDA By STEPHANIE KAY HILL A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 Stephanie Kay Hill
3 To my family and friends that shared in this journey with me
4 A CKNOWLEDGMENTS I express my deepest appreciation to all the folks who helped and guided me through my graduate studies. My m any thanks and gratitude to my maj or professor, Dr. Phil Koehler, for all of his encouragement, dedication, patience, and guidance. I would especially like the thank Dr. Roberto Pereira for his time and patie nce as we went through statistical processes and editing. I gratefully thank Dr. Grady Roberts for mentoring me through my agricultural education minor. Thanks to Dr. David Wil liams for all of his support and keeping me on the ball. Dr. Rebecca Baldwin, t hank you for being my mentor and allowing me to have so many teaching experiences A huge thank you and my upmost respect to Mark Mitola. I was lucky to have you assist me. We have many memories that I hope we can reminisce over years down the road. Thank you to the ever wonderful Liz Pereira and Tiny Willis. Without the two of you, the Urban Lab would fall apart. Thank you to my closest and dearest lab mates: Jodi Scott, Bennett Jordan, Joe DiClaro, Anda Chaskopoulou, Margie Lehnert, and Ricky Vazquez In particular, Jodi provided me with support and comic relief. Urban Lab. I w ould be amiss if I k the undergraduate students that assisted with field work and made day s go a little smoother. Thank you Holly Beard, Casey Parker, Ela Jaworski, Josh Weston, and Cory Goeltzenleuchter. Final thanks to my family. Thanks Mom for encouraging me and drivin g me to bug camp for ten years. A hug thanks to my husband. Thank you for listening to me rant and rave. Thank you for being my rock though every stage of this dissertation. I wish thank all those that provided funding for my research. Thank you to thos e who provided me scholarships throughout my graduate career.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 PRELUDE ................................ ................................ ................................ ............... 12 2 LITERATURE REVIEW ................................ ................................ .......................... 15 Classification ................................ ................................ ................................ ........... 15 Distribution ................................ ................................ ................................ .............. 16 Description ................................ ................................ ................................ .............. 17 Colony Structure ................................ ................................ ................................ ..... 18 Environmental Factors ................................ ................................ ............................ 20 Recognition ................................ ................................ ................................ ............. 20 F eeding Habits ................................ ................................ ................................ ........ 22 Relationship with Other Species ................................ ................................ ............. 23 Pest Status and Control ................................ ................................ .......................... 25 3 DISTRIBUTION AND DENSITY OF NYLANDERIA FULVA ON URBAN PLOTS AND THE EFFECT OF ABIOTIC FACTORS ................................ .......................... 29 Introduction ................................ ................................ ................................ ............. 29 Materials and Methods ................................ ................................ ............................ 31 Results ................................ ................................ ................................ .................... 34 Discussion ................................ ................................ ................................ .............. 38 4 AGGRESSIVE BEHAVIOR BETWEEN NYLANDERIA FULVA AND SOLENOPSIS INVICTA UNDER FIELD AND LABORATORY CONDITIONS ....... 57 Introduction ................................ ................................ ................................ ............. 57 Materials and Methods ................................ ................................ ............................ 58 Collection and Maintenance of Laboratory Colonies ................................ ........ 5 8 Aggression Test ................................ ................................ ............................... 59 Dyadic Interactions ................................ ................................ ........................... 59 Small Worker Group Interactions in a Neutral Arena ................................ ........ 60 Large Worker Group Interactions in a Neutral Arena ................................ ....... 61 Intruder into an Established Resident Territory ................................ ................ 61
6 Field Study ................................ ................................ ................................ ....... 62 Results ................................ ................................ ................................ .................... 62 Dyadic Interactions ................................ ................................ ........................... 62 Small Worker Group Interactions in a Neutral Arena ................................ ........ 63 Large Worker Group Interactions in a Neutral Arena ................................ ....... 63 Intruder into an Established Resident Territory ................................ ................ 64 Field Study ................................ ................................ ................................ ....... 65 Discussion ................................ ................................ ................................ .............. 66 5 FOOD PREFERENCE OF NYLANDERIA FULVA ................................ .................. 83 Introduction ................................ ................................ ................................ ............. 83 Materials and Methods ................................ ................................ ............................ 84 Results ................................ ................................ ................................ .................... 85 Discussion ................................ ................................ ................................ .............. 87 6 CONCLUSIONS ................................ ................................ ................................ ..... 92 Why Control of the Tawny Crazy Ant is Difficult ................................ ...................... 93 Integrated Pest Management Tactics for Tawny Crazy Ant Control ....................... 94 Identification ................................ ................................ ................................ ..... 95 Inspec tion ................................ ................................ ................................ ......... 95 Control Techniques ................................ ................................ .......................... 96 Sanitation ................................ ................................ ................................ ... 96 Exclusion ................................ ................................ ................................ .... 97 Chemical Control ................................ ................................ ....................... 98 Things to Remember ................................ ................................ .............................. 99 LIST OF REFERENCES ................................ ................................ ............................. 100 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 108
7 LIST OF TABLES Table page 3 1 Mean number of Nylanderia fulva for agers per sampling station on each type of ground cover in six sampling plots in North Central Florida. ........................... 43 3 2 Mean number of Nylanderia fulva forage rs per sampling station in each season at t he six sampling plots in North Central Florida. ................................ .. 44 4 1 Mean time and 95% confidence interval to 1, 10, 50, 90, and 99 percent mortality of Nylanderia fulva and Solenopsis invicta in small grou p interactions. ................................ ................................ ................................ ........ 72 4 2 Mean time and 95% confidence interval to 1, 10, 50, 90 of Nylanderia fulva and Solenopsis invicta in large group interactions ................................ .............. 73
8 LIST OF FIGURES Figure page 3 1 Mean number of foraging N. fulva on different gr ound covers through the seasons ................................ ................................ ................................ .............. 45 3 2 Aer ial view of the Gainesville sampling locations in relation to one anothe r ....... 46 3 3 Ground cover at each of th e sampling plots in Citra, FL. ................................ .... 47 3 4 Spatial distribution of average monthly foraging counts of N. fulva at Gaines ville plot 1 ................................ ................................ ................................ 48 3 5 Spatial distribution of average monthly foraging counts of N. fulva at Gainesv ille plot 2 ................................ ................................ ................................ 49 3 6 Spatial distribution of average monthly foraging counts of N. fulva at Gainesville plot 3 ................................ ................................ ................................ 50 3 7 Aerial view of the Citra sampling locations in relation to one another. ................ 51 3 8 Ground cover at each of the sampling plots in Citra, FL.. ................................ ... 52 3 9 Spatial distribution of average monthly foraging counts of N. fulva at Citra plot 1.. ................................ ................................ ................................ ................. 53 3 10 Spatial distribution of average monthly foraging counts of N. fulva at Citra plot 2.. ................................ ................................ ................................ ................. 54 3 11 Spatial distribution of average monthly foraging counts of N. fulva at Citra plot 3 ................................ ................................ ................................ ................... 55 3 12 Mean monthly density of foraging N. fulva and ave rage temperature in Gainesville and Citra plots during 2010 and 2011. ................................ ............. 56 4 1 Frequency of interaction types when N. fulva and S. invicta were the initiators of the interactions wh en introduced to ant of the opposite species in dyadic interactions. ................................ ................................ ................................ ........ 74 4 2 Mean number of N. fulva and S. invicta dead at sample times in small group interactions. ................................ ................................ ................................ ........ 75 4 3 Number of ants fighting at each sample time for small group interactions .......... 76 4 4 Mean number of N. fulva and S. invicta dead at each sample point fo r large group interactions. ................................ ................................ .............................. 77 4 5 Number of ants fighting at each sample time in large group interactions.. ........ 78
9 4 6 Frequency of interaction types when N. fulva and S. invicta were the initiators of the interactions after being introduced into resident colonies of opposite species. ................................ ................................ ................................ .............. 79 4 7 The mean number of N. fulva an d the summed other ants at each sampling plot for 2010 and 2011 ................................ ................................ ........................ 80 4 8 The ratio of ants found at Citra lot 2 in 2010 and 2011. ................................ ...... 81 4 9 The ratio of ants found at Citra lot 3 in 2010 and 2011. ................................ ...... 82 5 1 The percentage of N. fulva feeding on each of the food sources for the whole study, spring, summer, and fall ................................ ................................ ........... 90 5 2 Mean number of N. fulva collected in baited tubes containing a protein, lipid and a carbohydrate in six sampling sites in north c entral Florida from April to November 2011. ................................ ................................ ................................ 91
10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FORAGING STRATEGIES AND AGGRESSION PATTERNS OF Nylanderi a fulva (Mayr) (HYMENOPTERA: FORMICIDAE) IN NORTH CENTRAL FLORIDA By Stephanie Kay Hill May 2013 Chair: Phillip Koehler Major: Entomology and Nematology Native ecosystems are constantly at threat from invasive species. Better control and management t echniques can be designed with more information about an invasive species in its native and introduced range. The first step in this process is proper identification. U ntil 2012, an ant infesting the southern United States was misidentified as Nylanderia pubens when it really was Nylanderia fulva This misidentification le d to inefficient control techniques. The clarification of the species identification allowed research to be refocused on proper control strategies. The seasonal foraging activity of N. fulva was examined i n north central Florida. The foraging densities of N. fulva were at their lowest in the winter and early spring, when N. fulva foraged in small discrete locations most likely close to permanent nesting sites. Through the late sprin g and summer, the density and distri bution of the foraging increased later decreasing in the fall and winter. At certain times of the year, N. fulva foraged in discrete locations in north central Florida thus c ontrol measures need to be applied to discr ete locations because small populations are being targeted and not as much toxicant is needed, reducing the economic and environmental impact.
11 Laboratory and field investigations of the aggressive behavior of N. fulva towards S. invicta allowed identifi cation of the behavioral mechanisms responsible for resource competition. In laboratory one on one bioassays, N. fulva and S. invicta avoided conflict, most likely to decrease the risk of injury or death, and neither species experienced success in small g roup interactions. However, in large groups, N. fulva cooperated in defending nestmates. Field investigations showed th at this defensive behavior helped dictate the dominance of N. fulva over S. invicta These results explain the behavioral mechanisms re sponsible for N. fulva Food prefere nce and seasonal food change was also documented Results showed that N. fulva preferred protein as a food s ource year round; indicating that baits incorporating protein should be effective year r ound This research combined with current integrated pest management plans for N. fulva should help in managing this invasive species. Further research is need ed in assessing the ecological e ffect of N. fulva and in designing effective baits
12 CHAPTER 1 PRELUDE Ever since humans have been able to cross mountains, rivers, and oceans they have been transporting domestic animals, plants, and goods in order to survive. Unknowingly, non domestic species have travelled or been traded; whether it be on human s plants, or on the transportation device (Lockwood et al. 2007). Some of these exotic species have been able to colonize the new areas and have caused a great deal of problems. Exotic species are one of the greatest threats worldwide to native ecosystems and biodiversity (Vi tousek et al.1996, McGlynn 1999, Abbott 2006 Lockwood et al. 2007). Exotic species are also expensive, causing the United States alone approximately $137 billion dollars in damages each year (Leung et al. 2002). Human travel and tra de is on the increase (Hu and Reuscher 2004), thus the science of invasion ecology has an ever growing number of species in its repertoire. Invasive ants pose ecological and economical concerns especially in human developed ecosystems. The majority of i nvasive ants have a super colony structure (Heller 2005). Super colonies are polygynous, polydomous, and the workers do not show any aggression between nests (H lldobler and Wilson 1990). The high densities and lack of aggression between colonies affect the abundance and diversity of native ant species, other invertebrates, and some vertebrate species by becoming competitors for food re sources and territory (Porter and Savignano 1990, Human and Gordon 1996, Holway and Case 2001, Abbott 2005). They also b ecome economically expensive to manage and control (Abbott 2006, Silverman and Brightwell 2008). Nylanderia fulva are exotic ants that were introduced into Florida, Texas, and along the Gulf States (MacGown 2012). Nylanderia fulva form large super colonies and
13 take over vast expanses of land ( Zenner Polania 1990a, Fowler et al. 1994, Meyers 2008, Valles et al. 2012 ). Like other invasive ants, N. fulva causes dis ruption in natural ecosystems, a ffecting native an ts, other arthropods, and vertebrates (Zenner Polania 1990 a, Zenner Polania 1994, Aldana et al. 1995, LeBrun et al. 2012). The se ants have also been observed tending aphids, damaging electrical equipment, and causing home owners problems in the ir lawns ( Zenner Polania 1990a, 1990b, MacGown and Layton 2010, Calibeo and Oi 2010, 2011 ). The economic impact of N. fulva can only be speculated at this time, but it is expecte d to be substantial N ylanderia fulva have posed a great challenge to the pest control industry (Zenner Polania 1990b, Zenner Pol ania 1994 Pereira and Koehler 2007 ). Given the social behaviors of ants such as trophallaxis, shared resources, and grooming, active ingredients in pesticides should be passed around an ant colony. Other super colony ants, Linepithema humile (Mayr) and Anoplol epis gracil i pes (Jerdon) have been successfully controlled, despite their high densities, by taking advantage of their social behaviors (Kruchelnycky et al. 2004, Abbott and Green 2007). To manage high densities of N. fulva a broad spectrum bait i s normally applied to the entire treatable location (Culbert 2007, Pereira and Koehler 2007) with the idea that workers carry the bait back to t he nest, passing the toxicant to the other ants Bait applications have been unsuccessful in controlling N. ful va for long periods of time. Insecticidal sprays used for treatment on foraging trails and nesting areas also have been ineffective because N. fulva uses the accumulation of dead ants to bridge over the treatment (Drees et al. 2009, Nester and Raspberry 20 10).
14 This research was undertaken in an effort to answer questions about the invasion ecology N. fulva Spatial distribution and density studies were conducted on urban lots in North Central Florida. A thorough understanding or can help us understand it spread s and how to effect ively control its populations. The aggressive behavior between N. fulva and S. invicta was assessed to better understand the behaviors responsible for N. fulva Finally, seasonal food preference changes were determined to aid in bait attractant selection for a given time of year. Besides adding to the information we already have about this species this research will help in designing better control strategies and propel future r esearch.
15 CHAPTER 2 LITERATURE REVIEW Classification Nylanderia fulva (Mayr) is an ant species in the family Formicidae, subfamily Formicinae, and tribe Lasiini. The classification of this ant in the United States has been a source of constant confu sion. On record, large populations of these ants we re found in Florida in the 1990 s (Klotz et al. 1995) Originally N. fulva were identified as Paratrechina pubens (Forel) because of reports of P. pubens in south Florida in the 1950 s ( Trager 1984 ). L arge populations of these ants were discovered in Texas in 2002 ( Meyers 2008). The ants were classif ied as Paratrechina species near pubens because of minor morphological variations from P. pubens that were more similar to characteristics of Paratrechina fulva (Forel) ( Meyers 2008). The morphological similarities between P. pubens and P. fulva le d Creig hton (1950) and Bol ton et al. (2007) to synonymize the two species. Meyers but the research was inconclusive. In 2009, these ants were discovered in Mississippi an d then in 2010 in Louisiana and referred to as P. sp nr. pubens (Hooper Bui et al. 2010, MacGown and Layton 2010 ). In 2010, LaPolla et al. did a taxonomic revision, using molecular data, to the Prenolepis genus group in which P. pubens belonged. In the revision, Nylanderia was elevated from a subgenus to a genus, and 133 species, including P. pu bens, were moved into the genus Nylanderia from the genus Paratrechina (LaPolla et al. 2010). Z hao et al. (2012) compared molecular samples of Nylanderia pubens and Nylanderia sp. nr. pubens from Florida and Texas respectively. They determined that
16 the t wo groups were the same species. Unfortunately derived from few samples. With many areas offering abundant samples of this ant species, more accurate species identification and a better understanding of the species distribu tion could have be en obtained. Later in 2012, Gotzek et al. (2012) published a paper identifying N. sp. nr. pubens as Nylanderia fulva using morphometric and molecular data. The researchers had a representative sample of ants from the Caribbean Islands, Florida, Texas, Mississippi, and Louisiana (Gotzek et al. 2012). In this research, they also showed that N. pubens is restricted to the Caribbean islands, thus the ant populations introduced in the 2000s in Florida were most likely misidentified and the a ctual ant is N. fulva (Gotzek et al. 2012). Distribution Nylanderia fulva originates from Brazil ( Zenner Polania 1990a, Zenner Polania 1994 ). In the late 1960s and early 1970 s, N. fulva was introduced into Colombia as a biological control agent against leaf cutting ants ( Atta sp.) and poisonous snakes based on resear ch done in Brazil from the 1930 s and 1940 s (Zenner Polania 1990a Zenner Polania 1994 ). As of 2006, N. fulva can be found in all the regions of Colombia ( Germn et al. 2006, Zenner Polania 1990a). Nylanderia pubens has been documented from Anguilla, Argentina, Bermuda, Brazil, Colombia, Cuba, Guadeloup e, Martinique, Mexico, Panama, Puerto Rico, St. Croix, Lesser Antilles, and US Virgin Islands (Trager 1984, Wetterer and Keularts 2008, Pagad 201 1 MacGown 2012). Given possible misidentifications and the genetic evidence put forth, there is a possibility that some of the ants from those countries are actually N. pubens while others are N. fulva ( Gotzek et al. 2012).
17 In the United States, N. f ulva has been documented in Florida, Texas, Mississippi, and Louisiana (McGlynn 1999, Meyers 2008, MacGown and Layton 2010, Warner and Scheffrehn 2010). Because of misidentification, it is not known how long N. fulva has been in the United States. In Flo rida, the first record of N. pubens was in Miami in 1953 (Trager 1984). The next documentation of this species was in 1990 in and around a hospital and several building s in Miami ( Klotz et al. 1995). Beginning in 2000, N. pubens were reported in Everglad es National Park, Fort Lauderdale, Port St. Lucie, Jacksonville, and Sarasota (Warner and Scheffrehn 2010). Most of these infestations are most likely N. fulva. Currently N. fulva can be found in 24 counties in Florida ( Alachua, Brevard, Broward, Clay, C ollier, DeSoto, Duval, Hardee, Hillsborough, Indian River, Lee, Manatee, Marion, Martin, Miami Dade, Orange, Palm Beach, Pasco, Pinellas, Polk, Saint Johns, Saint Lucie, and Sarasota) (MacGow n 2012). The first observations of N. fulva in Texas were in Har ris County in 2002 ( Meyers 2008). Gold (2012 ) list ed 24 counties with confirmed N. fulva ( Bexar, Brazoria, Brazos, Comal, Cameron, Fort Bend, Chambers, Galveston, Hardin, Harris, Hidalgo, Jefferson, Jim Hogg, Liberty, Matagorda, Montgomery, Nueces, Polk, Orange, Travis, Victoria, Walker, Wharton and Williamson). Along the Gulf Coast, there are three counties in Mississippi (Hancock, Harrison, and Jackson) and two parishes in Louisiana (Calcasieu and West Baton Rouge) with confirmed established colonies of N. fulva (Hoop er Bui et al. 2010, Morgan 2011 MacGown 2012 ). Description Nylanderia fulva is a one node, reddish brown ant species. The workers are monomorphic and can range from 2.0 2.4 mm in length. The males and queens are
18 slightly larger measuring 2.4 2.7 mm and 4.0 mm or larger respectively ( LaPolla et al. 2011, Gotzek et al. 2012, MacGown 2012 ). The head appears darker and shiny due to the lack of pube scence Characteristic to Nylanderia species, N. fulva has a 12 segmented antenna with no clu b. The scape is long, measuring two times the width of the head. The ants have no ocelli. The thorax and abdomen are smooth and shiny except for areas of dense yellow pubescence. The gaster, after feeding, is lighter in color and appear s striped becaus e it is stretched. The appendages are also covered with a yellow pubescence. An acidopore is present in the females ( Trager 1984, LaPolla et al. 20 11, Gotzek et al. 2012, MacGown 2012 ). In males, the defining characteristic from other species is a strong ly triangular (in profile) paramere that is lightly scleritized and has few macrosetae originating from the margin of the paramere, as opposed to the heavily scleritized paramere in N. pubens (Trager 1984, Gotzek et al. 2012 ). Colony Structure Most ant sp ecies live in colonies of cooperating specific individuals. Nylanderia fulva is a polygyne and polydomous; therefore they are considered a super colony ant species (Zenner Polania 1990a, M e yers 2008, Valles et al. 2012). Polygyne is defined as the presen ce of more than one queen in a nest and polydomy is defined as the use of two or more spatially separated nests by one ant colony (H lldobler and Wilson 1990). New nest s are formed in super colonies by budding; one or more queens will leave the original n est and for m a separate nest (H lldobler and Wilson 1990). Super colonies are then very large communities with multiple queens and many nests that allow workers to mix freely across the community. The lack of aggression allows resource allocation so reso urces can be used to expand the colony rather than be spent
19 on int raspecific aggression (Tsutsui et al. 2003). There is a complete loss of colony borders and thus, super colonies have the ability to take over a large expanse of land (H lldobler and Wilson 1990). Zenner Polania (1990a) characterized the two types of nests that N. fulva constructs depending on the abiotic conditions. The first is a permanent nest, which is always occupied by the colony and is found in protected areas. Permanent nests have been found occupying one square meter and containing immatures up to 40 cm below the ground surface. In the dry seasons or when the weather conditions are not favorable, only permanent nests are built. In the rainy season or when conditions are more favo rable, temporary nests are utilized. Temporary nests are found under leaf debris, in or under waste material, in soil crevices, and in abandoned arthropod habitats. When temporary nests are disturbed, the ants will move to another nest. Since temporary nesting sites do not buffer N. fulva from temperature and moisture variations, the ants can alter their conditions by moving nest locations or changing the nest size, which in turn affects the metabolic heat generated by the group (Seeley and Heinrich 1981 ). The number and type of caste and developmental forms differ between each nest type. In 80 permanent nests sampled in Columbia, the caste and developmental form number varied with 1 14 queens, 366 2302 workers, 214 1034 e ggs, 815 1314 larvae, and 5 65 4539 pupae (Zenner Polania 1990a). Temporary nest s are much smaller. The average number caste and developmental forms found in 556 seasonal nests were 46 workers, no reproductives 21 eggs, 84 larvae, and 49 pupae (Zenner Polania 1990a).
20 The number of permanent and temporary nests at a location vary depending on how infested the location is with N. fulva In areas that are completely infested, permanent nest s are more common than temporary nests. In areas that N. fulva have not completely infested the number of temporary nests outnumber the permanent nests by 88% (Zenner Polania 1990a). Zenner Polania (1990a) hypothesized that temporary nests are used to new areas. At a coffee plantation in Colombia, N. fulva was recorded advancing their territo ry at a rate of 100 m per month (Zenner Polania 1990a). Environmental Factors Environmental factors can encourage or suppress the local spread of invasive lldobler and Wilson 1990). The foraging workers are severely vulnerable to these two factors because they travel away from the nest (Krushelnycky et al. 2010). In Colombia, N. fulva have been found between 150 and 2600 m above sea level (Zenner Polania 1990a). In Puerto B oyaca, Colombia, the elevation is 150 m above sea level and the temperatures can reach up to 40C (Zenner Polania 1990a). At this elevation and temperature, the infestation of N. fulva was low as compared with higher elevations and lower temperatures (Zen ner Polania 1990a). Nylanderia fulva can develop within a wide range of average temperatures, ranging from 13C to 29C (Zenner Polania 1990a). In Colombia, N. fulva develops best at temperatures averaging 22C and at an altitude of 1500 m above sea leve l (Zenner Polania 1990a). Recognition The ability to distinguish between nestmates and non nestmates is critical in The ability to distinguish between nestmates and non nestmates prevents robbery of bro od, parasitism, and colony
21 With the exchange of queens, brood, and workers, the ability to distinguish nestmates is even more important in super colonies. The way insects discriminate nestmates and non nestmates is much debated ( Chen and Nonacs 2000, noir 2010 ). The three most generally excepted ways of insect recognition are prior association, phenotype matching, and recognition alleles Prior association is where indivi duals learn cues and become familiar with other individuals they encounter and treat them as Only those individuals encountered previously can be recognized as nestmates. Similar to prior association, in phenotypic matching, individuals have an internal template that allows them to recognize other individuals of the same nest and/or species Recognition alleles do not require learning cues; they are solely based on the recognition of alle les Lenoir 2010) There are other factors, such as diet, that have proven to play a role in insect recognition (Chen and Nonacs 2000, Say Piau and Chow Yang 200 3, Buczkowski and Silverman 2005 ). Colonies of Linepithema humile (Mayr), the Argentine ant, originating from the same polygyne colony, fed on two different protein diets, exhibited aggressive behavior when introduced to a common aren a (Buczkowski and Silverman 2005 ). There was a 100% attack rate between the two groups. This beha vior has also been witnessed in Paratrechina longicornis (Latreille) (Say Piau and Chow Yang 2003). This leads to the speculation that it is not only genetics that aid in nestmate recognition but also other factors such as diet Polygy ne colonies are a great example of this theory
22 because of the variety of genetic cues derived from having so many queens (Chen and Nonacs 2000). Giraud et al. (2002) argue that it is the genetic origin of a colony rather than diet that allows ants to discriminate nestmates from non nestmates. Aggression tests were performed on populations of Argentine ants along a 6000 km track around Spain, Portugal, France, and Italy. Two super colonies were distinguished, the main super colony and the Catalonian super colony. The tests showed, regardless of the distance between ants of the same super colony, there was no aggression. The genetic structure between the two super colonies was found to be largely different at eight microsatellite loci, and there was a great genetic diversit y within each super colony. They concluded that unfound recognition loci remained constant in the each super colony. T here is no literature about how N. fulva discriminate nestmates from non nestmates. Based on the information about L. humile and P. lo ngicornis, it can be hypothesized that both recognition loci and diet play a role in recognition. Feeding Habits Nylanderia fulva are opportunistic generalist s that tend aphids and consume the honeydew secretions, prey on live animals, and scavenge from d ead animals (Zenner Polania 1990a, Ulloa et al. 2000, Campos Farinha and Zorzenon 2005). In Brazil and Colombia, N. fulva workers have been observed in teracting with aphids and scale insects specifically tending them and collecting honeydew (Zenner Polan ia 1990a, Campos Farinha and Zorzenon 2005). In some cases, the ants transported the Hemipterans from infested plants to uninfested plants (Zenner Polania 1990a). In Colombia alone, N. fulva is associated with over 25 species of Hemipterans (Zenner
23 Polan ia 1990a). The association between the ants and Hemipterans is devastating to plant health. The plants have to combat the feeding Hemipterans as well as any diseases they may vector. Nylanderia fulva has been recorded feeding on different forms of anima l protein (Zenner Polania 1990a Zenner Polania and Ruiz Bolanos 1985). To capture small prey (other arthropods), workers may attack the prey and immobilize it (Zenner Polania 1990a). Then the ants will either suck the hemolymph from the prey or transpor t pieces of the prey back to the nest (Zenner Polania 1990a). Nylanderia fulva have also been documented to attack small animals such as rabbits, birds, lizards, and domestic animals (Zenner Polania 1990a). hyxia (Zenner Polania 1990a). The ants then suck the liquids from the prey or take pieces of it back to the nest. In excavation of N. fulva permanent nests, Lepidoptera, Coleoptera, Isoptera, Hymenoptera, Arachnid, bird, and lizard remains were found (Ze nner Polania 1990a). Cook et al. (2012 ) examined the macronutrient requirements for N. fulva They found that the ants, when given a ch oice, consumed more food with a 1:2 protein to carbohydrate ratio. In no choice tests, the ants regulated the amount of each nutrient that they took in to meet the requirements of the colony. In both experiments, there was a high rate of mortality. This may have been due to the use of solid food rather than liq uid food in the experiments and a low larva to worker ratio th at can lead to starvation due to the lack of enough larvae to break down the solid food to feed all of the workers. Relationship with Other Species The introduction of the N. fulva to a new region causes the reduction and elimination of native ants and ot her insects (Zenner Polania 1994, Aldana et al. 1995,
24 LeBrun et al. 2012). There have been direct observations, in Colombia, of competition between native ants and N. fulva (Zenner Polania 1994, Aldana et al. 1995). Zenner Polania (1994) compared the spe cies richness of native ants in areas free of N. fulva areas that border infestations of N. fulva and in N. fulva infested areas. In areas free of N. fulva the species diversity index was 0.77. The species index at the border of infestations dropped 8 8% to 0.09 compared to areas free of N. fulva In areas that were completely infested, the species diversity index was 0.0088. This is a 98.85% reduction from areas that were not infested. The only ant species found in the N. fulva infested areas were M onomorium floricola Forel and Dolichoderus diversus Emery, making up 5% of the individual ants sampled. The success of these two species was related to their size and nesting habits (Zenner Polania 1994) Aldana et al. (1995) sampled the species richness of ants in the Laguna de Sonso Natural Reserve, Colombia, and in adjacent areas from the reserve. Nylanderia fulva were first found in the reserve in 1990. The areas sampled adjacent to the reserve were not infested with N. fulva There were 28 ant spec ies found in the adjacent areas. In the infested reserve area, only 7 species of ants were found. This is a 75% reduction in the number of ant species. LeB run et al. (2012) examined the e ffect of N. fulva on native ant and arthropod diversity in Texas gr asslands. In N. fulva infested areas, the native ant densities were two orders of magnitude lower than the densities of N. fulva Curiously, in areas with high densities of N. fulva, the red imported fire ant, Solenopsis invicta Buren, were absent and abu ndance of non ant arthropods was greatly reduced.
25 Similarly, Zenner Polania (1990a) documented the reduction of other arthropods in areas infested with N. fulva There was a noticeable reduction in white grubs, leaf feeding Lepidoptera, and termite colo nies when compared to areas not infested with N. fulva. In Brazil, N. fulva was documented preying on the pupae of Lepidopteron pest, Brassolis sophorae (Linnaeus) (Campos Farinha and Zorzenon 2005). Besides affecting invertebrates, N. fulva has been do cumented attacking vertebrate fauna (Zenner Polania 1990a, Aldana et al. 1995). On farms in Colombia, N. fulva has been blamed for the deaths of small animals, chickens, and other birds (Zenner Polania 1990a). The death of these animals is believed to be asphyxia (Zenner Polania 1990a). Larger animals were attacked around their eyes, nose, and hoofs. In the Laguna de Sonso Natural Reserve, N. fulva blinded an iguana and caused irreversible damage to a large bird, Anhima cornutus ( Linnaeus) ( Aldana et al. 1995). More quanti tative research is needed to ass ess the effects of N. fulva on vertebrate fauna. Pest Status and Control Nylanderia fulva is considered a serious pest in its native and introduced regions. Native to Brazil, N. fulva is considered a tra mp species (Gusmo et al. 2008). It has been found infesting buildings, greenhouses, and tending honey dew producing insects (Campos Farinha and Zorzenon 2005, Gusmo et al. 2008). In Colombia, N. fulva was introduced into the coffee growing areas as a b iological control agent for leaf cutter ants ( Atta spp.) and snakes ( Zenner Polania 1990a). It soon turned into a serious pest problem by tending sap feeing insects, attacking domestic animals, entering structures, and displacing native ant species (Zenne r Polania 1990a Zenner Polania 1994 ). Nylanderia fulva has developed into a serious pest problem in the Caribbean and the
26 United States; especially in Florida and states along the Gulf of Mexico ( Warner and Scheffrahn 2010, Pereira and Koehler 2007, MacGo wn and Layton 2010, Calibeo and Oi 2011). In the United States, N. fulva is considered a nuisance pest because the high densities of workers can affect day to day activities, and in extreme cases, makes outside activities impossible (MacGown and Layton 20 10). In Texas, beekeepers have reported N. fulva killing honeybee larvae and taking over the hives to use as a nest once the bees had been decimated (Harmon 2009). Many ants are attracted to electricity, and N. fulva is no exception. At the Jacksonvill e Zoo in Jacksonville, Florida, N. fulva infested electrical switch boxes causing shortages (Calibe o and Oi 2010). Meyers (2008), Drees et al. (2009), Nester and Rasberry (2011) also reported evidence of the ants causing electrical boxes and switches to f ail. Nylanderia fulva is usually found in high numbers when a pest management professional is called to a location. To manage the high numbers of ants, broad spectrum bait is applied to the entire location (Culbert 2007, Pereira and Koehler 2007). If need ed, a residual insecticide is applied to the outside of structures to deter the ants (Culbert 2007, Pereira and Koehler 2007). Control is usually seen for a short period of time before the ant population resurges ( Gusmo et al. 2008, Drees et al. 2009, Ne ster and Rasberry 2011 ). The control of N. fulva is difficult due to their high numbers, weak foraging on most traditional ant baits, and the efficacy of pesticides. There have been a variety of pesticides tested to see the effects on N. fulva populations In Brazil, methoprene (insect growth regulator) granular bait reduced the number of foraging N. fulva around a hospital by more than half in an 18 week study
27 (Gusmo et al. 2008). Another insect growth regulator, pyriproxyfen, was tested in Texas for i t s efficacy on N. fulva (Nester and Rasberry 2011 ). Using a broadcast spray at the highest label rate, control was seen for 14 days then the population reb ounded (Nester and Rasberry 2011 ). In both of these instances, a reapplication of pesticide would h ave to be done for continued control. In Colombia, a bait mixture of corn bran, fish meal, 1% sugar solution, propionic acid, pork lard, and carbaryl 85 PM was applied after the rainy season and the ants were controlled for approximately two months (Zenn er Polania 1990b). The density of ants, in Columbia, decline at the end of the rainy season and into the dry season. Given the natural decline in ant densities during the application of the bait, full control measures may not have been observed. Meyers ( 2008) laboratory and field tested a variety of pesticides against N. fulva but no long term suppression of the ant populations was seen. In the laboratory, experiments were conducted on queenless colonies, making the data collected in those experiments q uestionable. Dinotefuran on corn bait matrix was field tested with an initial population reduction, but the population resurged within a week. The same population resurgence was seen after testing the effectiveness at the label rate of fipronil, chlorfen apyr, abamectin, acetamiprid, and bifenthrin. The overall conclusion of Meyers (2008) work was that more work needed to be done to find an effective chemistry t hat will control N. fulva for extended periods of time. Scott (2012) examined bait size and at tractants for N. fulva This author found that the ants preferred foods that were between 0.85 1.00 mm in size and contained a high protein bait matrix. In the laboratory, fipronil was added to a protein bait matrix and
28 was highly efficacious in eliminat ing N. fulva To fully test the effects of the formulated bait, field trials would need to be conducted. When dealing with an N. fulva infestation, a single control method cannot be used because of the super colony structure (Zenner Polania 1990b, Dre es et al. 2009, Calibeo and Oi 2010 ). An integrated pest management plan is needed to control N. fulva infestations. Landscape debris, lawn debris, and trash need to be removed as they provide a place for temporary nest. Trash cans need to be emptied re gularly and stored away from structures. To prevent entry into structures, landscape around the structure needs to be cut back so that no branches are touching the structure, and all cracks and crevices around the structure need to be sealed. Pest manage ment professionals need to apply pesticides at the right times of the year for the toxicants to be most efficient (Zenner Polania 1990b). More information about N. fulva and ecology is needed to design a better integrated pest management plan.
29 CHAPTER 3 DISTRIBUTION AND DENSITY OF NYLANDERIA FULVA ON URBAN PLOTS AND THE EFFECT OF ABIOTIC FACTORS Introduction Nylanderia fulva (Mayr) is an invasive tramp ant that is polygynous (multiple queens per colony), polydomous (multiple nest s per colony), and unicolonial (shows no aggression between neighboring colonies) ( Zenner Polania 1990a Meyers 2008, Valles et al. 2012). Since ants from different nests and colonies are constantly interchanging members and resources, it may be more accurate to conside r them a super colony ant ( Hlldobler and Wilson 1990 Giraud et al. 2002 ) Super colonies are known to facilitate rapid population increases due to the high number of queens resulting in high brood production ( Hlldobler and Wilson 1990 ). Little researc h has been done with the ecology of N. fulva in the United States, and most of the research has focused on the problems that the N. fulva causes and possible control measures ( Meyers 2008, Valles et al. 2012). Nylanderia fulva is a serious pest species in both its native and introduced areas. It is native to Brazil where it is often considered a tramp species; found infesting buildings, greenhouses, and tending honey dew producing insects (Campos Farinha and Zorzenon 2005, Gusmo et al. 2008). Nylanderia fulva was introduced into Columbia, the Caribbean, and along the Gulf states in the United States ( Zenner Polania 1990a Pereira and Koehler 2007, MacGown and Layton 2010). It has developed into a serious pest problem by tending sap feeding insects, attac king domestic animals, displacing native ant species, entering structures, and causing shortages in electrica l boxes (Zenner Polania 1990a, Zenner Polania 1994, Pereira and Koehler 2007, Calibeo and Oi 2010, MacGown and Layton 2010).
30 C ontrol of N. fulva i s difficult and control is only temporary (Zenner Polania 1994, Gusmo et al. 2008, Drees et al. 2009, Neste r and Rasberry 2011 ). When a pest management professional is called to a location with a N. fulva infestation, the ants are usually in high numbers In order to manage the infestation, broad spectrum bait is normally applied to the entire location (Culbert 2007, Pereira and Koehler 2007). If the ants are entering a structure, a residual insecticide is normally applied to the outside of the structure (Pereira and Koehler 2007, Drees et al. 2009). Both of these treatments are only adequate to manage the ant population for a short amount of time and the ants usually resurge in high numbers resulting in new insecticide applications (Drees et al. 2009, N ester and Rasberry 2011 ). The cost of the pesticide for these treatments has the potential to exceed the monthly cost of a pest control contract, making this treatment strategy uneconomical for the pest control company as well as the consumer. The amount of toxicant applied into the environment with this control strategy is higher than normal, and the environmental implications need to be considered. The super colony nature and lack of successful management strategies have caused N. fulva to become a pes t of major concern for the consumer as well as the pest control industry. The Argentine ant, Linepithema humile (Mayr) is also a super colony ant. Extensive work has been done on the biology, ecology, and control of this ant and it is one of the best st udied examples of invasive super colony ants (Tsutsui and Suarez 2003). Similar to N. fulva the colonies contain high numbers of workers which support and care for many queens which can reproduce at high rates (Horton 1918). Also, Argentine ants have hi gh rate s of reproduction that allow them to expand over large
31 geographical area s ( Tsutsui et al. 2000 Tsutsui and Case 2001 Giraud et al. 2002 Heller 2005, Corin et al. 2007). Environmental factors also encourage or suppress the spread of Argentine ants ( Hlldobler and Wilson 1990, Krushelnycky et al. 2005, Schilman et al. 200 5, Heller et al. 2008 ). F oraging workers are severely vulnerable to desiccation and temperature stress because they are traveling away from the safety of the nest (H lldobler and Wils o n 1990, Krushelnycky et al. 2010). Rainfall and soil temperature have been shown to affect the spread of Argentine ants (Krushelnycky et al. 2005, Schilman et al. 2005, Heller et al. 2008). Spatial pattern research suggests that Argentine super colonies expand in the spring and summer with increased soil temperature and rai nfall, while in the fall and winter months the colonies contract (Krushelnycky et al. 2005, 2010, Heller et al. 2006, 2008). Based on known similarities between N. fulva and the Argentine ant infestations, a study was designed to examine the spatial distr ibution and density of foraging N. fulva on urban lots over two years. The objectives were to compare the populati on distribution over the seasons and ground cover, compare ant density among locations and seasons, and correlat e ant population with tempera ture and precipitation. Materials and Methods Six urban lots, three in Gainesville, FL and three in Citra, FL, were chosen for the study because of the known presence of N. fulva The locations in Gainesville were a recreational park and two urban house lots. The locations in Citra, a rural community, were three rural lots with houses. For all locations, sampling boundaries were set by using the portion of the lot with managed landscape; using the house as the center point and not extending over the pr operty lines. The sample areas were gridded in 4 x 4 m
32 squares (Brenner 1988, Abbott 2005). The markers, placed at each vertex, were made by bending the metal stakes of surveyor flag s into a U shape and moving the flag to the bend in the stake. The mark ers were pushed into the ground so the U in the middle of the stake was flush with the ground and only the flag was showing. At each vertex, the ground cover type was recorded as: sand, loose litter (mulch, leaf debris, and loose organic matter), grass, o r solid surfaces (asphalt, concrete, dirt roads). Ant foraging in the lots was sampled during 2010 and 2011 between April and November. Sampling was done only when the air temperature reached a minimum of 21C and was not over 32C based on observatio ns by Warner and Scheffrahn (2010) with Nylanderia pubens (Forel) in south Florida, which was believed to be the ant at the start of the study The sampling period was divided into spring (April and May), summer (Jun e, July, August), and fall ( September, October, November) for analysis purposes. Sampling was done by placing a 7.5 by 5.5 cm index card with approximately 5 ml of honey as flat as possible on the ground at each grid vertex. After 30 minutes, pictures were taken of the cards using a digital ca mera (Sony Cyber shot, 8.1 mega pixels). The sampling cards were collected and removed from the property after sampling was finished. The pictures were downloaded to a computer, and the ants were identified and counted on the pictures projected on the co mputer screen. This procedure allowed the pictures to be magnified as needed for better discrimination of the ant species and separate individuals. The average daily temperature and precipitation for both Gainesville and Citra were compiled. For Gainesv ille, data used was from the Gainesville Regional Airport, which is approximately 10 km from the sampling sites and was compiled by the Florida
33 Climate Center in Tallahassee, Florida. Citra weather data, collected at a station approximately 6.5 km from th e sampling sites, was compiled from the Florida Automated Weather Network maintained by the University of Florida. Data Analysis : Only N. fulva data was used in the analysis. Analysis was performed on square root transformed data (number of ants) for bo th distribution and density. The foraging distribution of N. fulva was analyzed using a mixed analysis of var iance (GLIMMIX procedure in SAS 10 software) (SAS Institute 2012 a ) with the mean number of N. fulva at each sampling point as a response variable and the ground cover, season, monthly temperature averages, and monthly precipitation averages as fixed effects. For mean comparisons, a P value of <0.05 was used to establish statistical differences. Contour maps of the ground cover at each sampling lot and the population d ensities were prepared using Proc Contour (SAS Institute 2012 a ). In the population density contour maps 0 1 represent low densities, 4 8 represent moderate densities, and 16 64 represent heavy densities. A regression analysis (SAS I nstitute 2012 b ) was used to examine the effects of location, year, and mo nth on foraging density with a P value of <0.05 to establish statistical differences. To determine which weather parameters (average temperature, minimum temperature, maximum tempera ture, and average precipitation) affected the foraging density, the parameters were examined using a screening platform designed to determine factors with the largest effect on the response variable by effect sparcity which assumes that only a small number of the effects are significant and the others are not. The parameters with a significant P value >0.05 were further analyzed using a regression analysis (SAS Institute 2012 b ).
34 Results Foraging Distribution: Results from the mixed model ANOVA showed that the ants were foraging more on sand than other ground cover types ( F = 77.71 ; df = 3 ; P < 0.0001) (Table 3 1). More N. fulva foraged in the summer months than in the spring and fall months ( F = 34.46 ; df = 2 ; P < 0.0001 ) (Table 3 2), but more ants foraged in the fall than in the spring. In the spring, the ants foraged mostly on sand, followed by loose litter, then grass and solid surfaces. In the summer and fall, the ants had no preference between foraging on sand or loose litter, but preferred these two ground covers over grass and solid surfaces ( Figure 3 1 ). Results from the mixed model ANOVA showed that temperature had an effect ( F = 10.75; df = 1 ; P = 0.002) on where the ants foraged, but precipitation did not ( F = 1.43 ; df = 1 ; P = 0.24). The Gain esville plots were an urban park and two urban residences. The relative location of the plots to each other is seen in Figure 3 2 and the ground covers are seen in Figure 3 3 In 2010, there were no other houses adjacent to Gainesville plot 1, an urban r esidence ( Figure 3 4 ), but in 2011, construction began on residences on both sides of the location. In April and May in both years, the an ts foraged in discrete areas on the plot, and ants were first seen foraging in the eastern side of the plot. In June and July, the ants foraged more broadly across the plot, but still in localized areas. In August, the ants reached the widest distribution. In September 2010, the ants forag ed in similar discrete areas as in July. In September 2011, the ants foraged in a locali zed area in the southeast corner of the plot. Ants foraged i n discrete areas in the spring and fall that had loose litter as a ground cover. This pattern was seen in 2010 and 2011, but there were fewer high density foraging areas in 2011 than in 2010.
35 Gainesville plot 2 ( Figure 3 5 ) was an urban park and the most natural sampling plot of all the plots included in this study. The north boundary of the sampling area had a creek running beside it and the south boundary was open grass The wooded area of the plot had loose leaf litter around the base of the trees. The ground cover away from the trees was variable including loose leaf litter, sand, and grass. A nts foraged more broad ly in 2011 than in 2010. In April of both years, the ants forage d in the area with trees close to the creek to the north. In July and August for both years the ants foraged broadly and heavily across the plot. The areas where the ants did not forage as readily were covered with grass. In November 2010 and 2011, the ants foraged in locations where they were first seen in the spring. Gainesville plot 3 ( Figure 3 6 ) was also an urban residence, but it was unoccupied during the 2010 sampling period. In January of 2011, occupants moved into the residence and started ir rigating the grass areas as well as landscaping around the house. During both sampling years, the house had mulch around its perimeter, but in 2011 more mulch w as added to the existing mulched areas. In April of 2010, the ants were first foraging in disc rete areas in the northeast corner of the plot. In May 2011, the ants foraged heavily across the north boundary of the plot and moderately on the south side of the plot. In July and August 2010, the ants foraged broadly and heavily across th e plot area, and foraged in similar areas in June and September. In October and November 2010, the ants foraged again in discrete areas on the plot with the northeast corner being the location with the highest foraging. In April 2011, the ants foraged in discrete are as similar to May 2010. Both areas of heavy foraging were covered in mulch and leaf litter. From June un til August 2011, the ants increasingly foraged more broadly
36 and heavily across the plot. Starting in September 2011 and going through November 2011, t he ants gradually reduced their foraging area and foraged in fewer numbers. The Citra plots were all rural residences. Each of the sample plots bordered the same orange grove, where N. fulva might have been introduced through transport of commercial oran ge containers from South Florida. The location of the plots in relation to each other is seen in Figure 3 7 and the ground cov er at each location is seen in F igure 3 8. The property of Citra plot 1 ( Figure 3 9 ) shared its eastern property line with the o range grove. Also, the orange grove continued across the road on the northern side of the property. In May 2010, the ants foraged on the western side of the plot. In June, July, and August 2010, the ants began foraging closer to the house and were distr ibuted in a wider area, with August being the month with the ants most widely distributed and having the heaviest foraging. In the fall months, the ants foraged in more localized areas and with a greater range of number of foragers. In April of 2011, the ants foraged on the east side of the plot and continued to forage across the southern boundary of the plot in the following month. In the spring of 2011, all of the landscaping around the house was removed and the trees and bushes in and around the plot area were severely cut back. In June and July of 2011, the ants foraged in more localized areas to the eastern side of the plot and not around the house as compared to the same months in 2010. In the fall of 2011, the ants foraged minimally in discrete ar eas on the eastern side of the plot. Citra plot 2 ( Figure 3 10 ) property shared its western boundary with the orange grove. In April, May, and June of 2010, the ants foraged in localized areas on the
37 northwestern corner and southern boundary of the plot. Both of these areas were covered in leaf litter. In July 2010, the foraging area became larger and in August, the ants foraged heavily on most of the property. Throughout September, October, and Novembe r, the ants foraged in localized areas tha t becam e smaller over time In 2011, the ants foraged i n smaller more localized areas than in 2010. Also, in the spring of 2011, this plot was infested with fleas, with the worst infestation between the house and the southern plot boundary line. The homeowners treated the pets as well as all of the plot area several times throughout the 2011 sampling period with unknown pesticides. The property of Citra plot 3 ( Figure 3 11 ) shared half of its eastern most boundary with the orange grove. There was a fence lined with trees separating the southern boundary from the adjacent property that backed up to the orange grove. In May and June 2010, the ants foraged on the southern side of the sampling plot. In July 2010, the forging area increased and the ants foraged al ong the southern plot boundary, on the west side of the property, and in a small wooded area directly behind the house. In August 2010, the foraging area expanded to cover most of the plot, with several areas of heavy foraging concentration. In September 2010, the areas of foraging reduced in comparison to the previous month, and this continued in October and November, with the ants foraging in more restricted areas. Throughout the 2011 sampling period, the ants foraged in smaller localized areas than 20 10. In 2011, the solid surface west of the northern most structure in this plot had heavy ant foraging where the homeowners fed their cats from June until August. The residents also had a fire pit where trash was accumulated and burned around which the a nts foraged from July to November in 2010 and June to October in 2011.
38 Foraging Density: Results from the regression analysis showed that location ( F = 4.2 ; df = 1 ; P = 0.04), year ( F = 4.7 ; df = 1 ; P = 0.03) and month ( F = 3.9; df = 1 ; P = 0.001) had an effect on the density of ants. Gainesville plots had significantly more ants than Citra plots. In 2010, there were more ants foraging than in 2011. There were significantly fewer ants foraging at Citra in 2011 than any other location/year combination. In both April and November, for both 2010 and 2011, there were significantly fewer ants foraging than the rest of the months sampled. August had significantly more ants foraging than the rest of the months sampled ( Figure 3 12 ). The average temperature was the only weather parameter that had a significant effect on the density of ants ( t = 4.08 ; P = 0.002) with higher temperature associated with higher ant densities ( Figure 3 1 2 ). The minimum temperature, maximum temperature, and average precipi tation did not significant ly affect the density of ants. Discussion In sup er colonies of ants, a group of connected nests that share resources and reproduce together are considered a colony The foraging strategy of N. fulva is most likely based on their super colo ny structure. The distribution and density of foragers in this study suggest that N. fulva colonies expand and contract seasonally. The phenomena of expanding and contracting of ant colonies has been well documented in Argentine ants (Heller 2005, Heller and Gordon 2006, Heller et al. 2006, 2008). The ants aggregate d in connected nests in the winter and disperse d in the summer (Heller 2005). The climate affected the rate in which the ants disperse in the summer. Heavy rainfal l and high temperatures incr eased soil moisture and temperature which led to higher numbers of suitable nest locations (Heller 2005, Krushelnycky et al. 2005). This allow ed the ants to disperse across the landscape (Heller and Gordon
39 2006). In the winter months that we re drier and cooler, the ants contract ed back int o the aggregated nests that had a more stable microclimate (Heller 2005). Zenner Polania (1990a) documented the two types of nest s N. fulva inhabited, permanent and temporary. Permanent nest s were found in protected a reas with controlled microclimates and allow ed the ants t o become established in an area The temporary nests help ed advance the colo ny into new areas where they could forage for nutritiona l resources. Permanent nest s were inhabited year round but were essential in the late fall, winte r, and early spring, as they were constructed in protected areas. Temporary nest s were inhabited in late spring, summer, and ear ly fall when brood production was at its highest and there was a high demand for nutritional r esources. Temporary nest s are abandoned for the permanent nest if the abiotic conditions are not right or if disturbed. In the spring and summer mo nths, the colony network expanded to facilitate efficient collection of nutritional resources. Clustering i n permanent nest s in the fall and winter allowed for resources to be efficiently collected and utilized. Nylanderia fulva foraging was distinct ly se asonal in the sampled plots in north c entral Florida. Overall, in April and May, the foraging took place i n localized areas of the plots with a small number of foragers. The workers gathered nutritional resources in those localized areas to aid in brood production. As the densities of ants increased, the ants began to inhabit temporary nests and the foraging distribution increased. The foraging distribution and density peaked in August. In September, the foraging distributio n and density decrease d and w ith th is decrease, the foraging area s became more localized. This was most likely due to the temporary n ests being abandoned for the permanent nests. Brood production was reduced in the fall and the n eed for
40 nutritional resources was not as great (Zenner Polania 1990a). The permanent nests protect ed the colony from acute abiotic conditions. A seasonal pa ttern of high colony dispersal during the times of high reproduction followed by a contraction into few nests before winter is called seasonal polydomy and has been described for other super colony ant species (Cerda et al. 2002, Dillier and Wehner 2004, E lias et al. 2005). In Colombia, on a 390 2 m coffee plantation infested with N. fulva 80 permanent and 556 temporary nests were found (Zenner Polania 1990a). This helps explain why there were sev eral localized foraging areas in the spring months and even into the early summer months. The localized foraging areas most likely represent areas close to permanent nests. In the summer an d early fall, these were area s of heavy foraging. The areas around permanent and temporary nests were most likel y where the h eaviest foraging was taking place. In th e fall, foraging area s, again become localized. The ants were most likely foraging close to permanent nests, as observed in the spring. In the winter, N. fulva foraged in discrete localized areas near permanent ne sts to gather enough nutritional resources to survive the colder months. Knowing the ground cover that N. fulva foraged on given time of the year can aid in identifying where nest s are located. In the spring (April and May), the ants preferred to forage on sand, then loose litter. At this time, the ants are most likely inhabiting their permanent nests located in areas with sandy or loose litter ground cover. At the coffee plantation sampled in Colombia, 39 of the permanent nest were found at root edges and 26 at the base of trees (Zenner Polania 1990a). At the sampling plots in my study, the ground cover at the base of trees and at root edges, in the spring, was either sand or loose litter (mulch or leaf litter). In order to locate the permanent nest, it would be
41 advantageous to look around the base of trees and root edges in the spring. In the summer, there was no significant difference between sand, loose litter, and grass as ground cover. At the coffee plantation, Zenner Polania (1990a) found tempor ary nests in soil crevices, in dead leaves, at the base of coffee trees, in banana stems, rotten roots, grass roots, and in empty snail shells. These nests could have be en on or in a variety of ground covers. In the fall, sand and loose litter was again the most preferr ed ground covers, and, probably where permanent nests are found. Temperature was the weather factor that was shown to affect the distributions and density of the ants. Nylanderia fulva have been documented surviving and reproducing at tem peratures averaging between 13C and 29C (Zenner Polania 1990a). In this study, as temperatures increased, the foraging distribution and density increased. August is one of the hottest months and was when the foraging distribution was most widespread a nd the density was at its highest at all plots in both years. Although not significantly correlated to ant density, precipitation, which was lower in 2011 than 2010, and water availability may affect N. fulva densities. The only sampling plot that did no t see a decrease in ant density in 2011 was Gainesville plot 2. This is perhaps due to the presence of a creek along the north boundary of the sampling area. Precipitation was a good predictor of population size in Argentine ants (Heller et al. 2006, 2008 ). By knowing seasonal characteristics of nesting and foraging patterns, better management programs could be designed for N. fulva. Control measures need to be applied in the winter and early spring when the densities of N. fulva are low and the ants are for aging in small discrete areas where permanent nests with reproductives are
42 present. This would allow for the least amount of control effort and cost to be used, reducing the economic and environmental impact of controlling this ant. This strategy con trasts the control techniques used currently in which broad spectrum pesticides were used during the months when there was high ant densities (Culbert 2007, Pereira and Koehler 2007). Application of pesticide is usually dictated by costumer complaints. E ducation of customers about N. fulva would be crucial to implementing a successful management program. Nylanderia fulva was initially identified as Paratrechina pubens (Forel) before this th e species was identified as N. fulva Nylanderia pubens can forage at temperature no lower than 22C (Warner and Scheffrahn 2010). Nylanderia fulva can forage at temperatures as low as 13C (Zenner Polania 1990a), but because the ant was incorrectly iden tified s ample procedures may not have covered the foraging season for N. fulva In March 2011, one foraging sample was taken. It was not used in statistical analysis because of the lack of comparable data for 2010. Ant activity was seen during that samp ling time. Year round sampling may allow better identification of areas where permanent nests are present, which would allow better targeting for control measures.
43 Table 3 1. Mean number of Nylanderia fulva ea ch type of ground cover in six sampling plots in North Central Florida. Means followed by the same letter are no t signifi cantly different (Tukey Kramer P < 0.05). Ground Cover Mean Number of Ant Foragers SEM Sand 1.3 a 0.34 Loose Litter 1.0 b 0.27 Grass 0.7 c 0.19 Solid Surfaces 0.3 d 0.09
44 Table 3 2. Mean number of Nylanderia fulva foragers in each season at the six sampling plots in North Central Florida. Season Mean Number of Ant Foragers SEM S pring 0.3 a 0.08 Summer 1.7 b 0.50 Fall 0.9 c 0.27 Means followed by the same letter are not significantly different (Tukey Kramer P < 0.05)
45 Fig ure 3 1 Mean number of foraging N. fulva rs th r ough the seasons Within each season, means connected with the same letter are not significantly different by Tukey Kramer ( P < 0.05).
46 Figure 3 2 A e rial view of the Gainesville sampling locations in relation to one another. The white line s indicate property lines. Numbers correspond to the lots. The N. fulva population distribution is for August 2010 as shown in Figure 3 4,. 3 5, and 3 6 N N 50 m
47 Figure 3 3 Ground cover at eac h of th e sampling plots in Gainesville FL. Solid lines represent solid structures and circles represent trees and bushes. The size of the circle indicates the relative size of the tree/bush.
48 Figure 3 4 Spatial distribution of average monthly forag N. fulva at Gainesville plot 1. Dashed lines represent buildings solid lines represent driveways and roads and circles represent trees and bushes. The size of the circle indicates the relative size of the tree/bush.
49 Fig ure 3 5 N. fulva at Gainesville plot 2. Dashed lines represent buildings, solid lines represent driveways and roads, and circles represent trees and bushes. The size of the circle indicates the relative size of the tree/bush. A p r i l M a
50 Figure 3 6 N. fulva at Gainesville plot 3. Dashed lines represent buildings, solid lines represent driveways and roads, and cir cles represent trees and bushes. The size of the circle indicates the relative size of the tree/bush. A p r i l M a y
51 Figure 3 7 A e rial view of the Citra sampling locations in relation to one another. The white lines indicate property lines. Numbers correspond to the lots. The distribution is August 2010 as shown in Figure 3 9, 3 10, and 3 11 100 m
52 Figure 3 8 Ground cover at each of the sampling plots in Citra, FL. Solid lines represent solid structures and circles represent trees and bushes. T he size of the circle indicates the relative size of the tree/bush.
53 Figure 3 9 N. fulva at Citra plot 1. Dashed lines represent buildings, solid lines represent driveways and ro ads, and circles represent trees and bushes. The size of the circle indicates the relative size of the tree/bush.
54 Figure 3 10 N. fulva at Citra plot 2. Dashed lines represent buildings, solid lines represent driveways and roads, and circles represent trees and bushes. The size of the circle indicates the relative size of the tree/bush. A p r i l M a y J u n e J u l y A u g u s t S e p t e m b e r O c t o b e r N o v e m b e r
55 Fig ure 3 11 f N. fulva at Citra plot 3. Dashed lines represent buildings, solid lines represent driveways and roads, and circles represent trees and bushes. The size of the circle indicates the relative size of the tree/bush. pit. In Ju ne, July, and August 2011, a cat feeding station was located at the A p r i l M a y J u n e J u l y A u g u s t S e p t e m b e r O c t o b e r N o v e m b e r
56 Figure 3 12 Mean monthly density of foraging N. fulva ( and average temperature in Gainesville (A and B) and Citra (C and D) plots during 2010 (A and C) and 2011(B and D).
57 CHAPTER 4 AGGRESSIVE BEHAVIOR BETWEEN NYLANDERIA FULVA AND SOLENOPSIS INVICTA UNDER FIELD AND LABORATORY CONDITION S Introduction The factors that direct the spread and success of invasive species are also critical in attempts to reduce their impact ( Williamson 1996 ). There are numerous factors that allow inv asive species to become dominant such as competitive abilities, resource allocation, lack of natural enemies, or human interaction. However, the ir competitive abilities are thought to be the reason i nvasive species are able to dominate native species to acquire territory and resources (Sax et al. 2007). The Solenopsis invicta Buren and Nylanderia fulva (Mayr) are invasive species that were introduced from their native ra n ge of South A meric a to the southern United States (Vinson 1997 Meyers 2008 Ascunce et al. 2011 MacGow n 2012 ). Solenopsis invicta w as introduced in the 1930s and has become established in the United States It is not known when exactly N. fulva was introduced into t he United States, but became a Vinson 1997 MacGown 2010). Both species are known to displace native ants and other arthropod species as well as negatively affect vertebrate populations ( Zenner Polania 1990a, Zenner Polania 199 4 Aldana et al. 1995, Tschinkel 2006 Holway et al. 2002 LeBrun et al. 2012 ). In S. invicta and N. fulva of Brazil, their competitive behavior is not known. more individuals of the same species (intraspecies competition) or members of two or more speci es (interspecies competition) at the same trophic level for a common resource or omes about when there is a shortage or potential shortage of one or more resources. The
58 aggressive behavior between ant species is based on competition and is adaptive depending on the situation (Wilson 1970). In this study, the individual, small group, large group and colony aggressive interactions as well as numerical dominance between N. fulva and S. invicta were examined to better understand each species competitive strategies. Materials and Methods Collection and Maintenance of Laboratory C olonies Nylanderia fulva were collected from two different locations approximately 18 km apart, withi n Gainesville, FL: SW Depot Avenue (GNV1) and NW 50 th Avenue (GNV2). Solenopsis invicta were collected on the campus of the University of Florida at the Urban Entomology building in Gainesville, FL (UEL) and at the Vikal Buildin g Complex in Citra, FL (VBC). For N. fulva single stock colonies cons isting of 3,000 5,000 workers, 5 10 queens, and numerou s brood were established Single stock colonies o f S. invicta were established consisting of 3,000 10,000 workers, 1 queen, and numerous brood. Colonies of both species were maintained in Fluon (BioQuip Products, Rancho Dominguez, CA) coated trays containing nest cells composed of plaster filled (Dentsply, York, P A) plastic Petri dishes (11 x 15 mm). Red acetate paper covered the nest cells to filter light and make it more attractive for brood rearing. Both species were provided ad libitum with water, 10% sugar water, and a variety of food including freshly kille d American cockroaches ( Periplaneta ameri cana ), hardboiled egg, orange, and honey. All colonies were maintained at 27 1.5C, 40 7% RH, and a 12:12 L: D cycle.
59 Aggression Test The level of aggression between four colony pairs was assessed ( GNV1 UEL, GNV 1 VBC, GNV2 UEL, and GNV2 VBC ) using four differen t aggression bioassays similar to those designed by Bu c zkowski and Bennett ( 2008 ) : a) worker dyadic interactions in a neutral arena, b) small worker group interactions in a neutral arena c) large worker gr oup interactions in a neutral arena and d) intruder into an established resident territory Because S. invicta workers are polymorphic while N. fulva workers are monomorphic, S. invicta workers chosen were approximately the same size as the N. fulva work ers. An i ndividual ant w as not tested in more than one trial. Dyadic I nteractions Two workers of similar size one from each stock colony were selected at random on a wooden dowel ([ 152.5 mm x 2 mm ] Fisher Brand, Pittsburg, PA) and placed into a plasti c souffl cup ( 30 ml [ 1 oz Dart Manufacturing Company, Mason, MI]). Souffl cups were coated with Fluon within 1 cm of the bottom. This restricted the ants to a small area, maximizing the chance of the ants finding each other. The first interactions w ere scored on a 1 4 scale ( Suarez e t al 2002 Bu c zkowski and Bennett 2008) [1= igno re, 2= avoid, 3=aggression ( lunging, biting, and/or pulling physical ), 4= fighting (prolonged aggression and/or use of defensive compounds physical and/or chemical )]. On ly the first i nteraction was recorded to avoid compounding effects of more than one interaction (Buczkowski and Silverman 2005). The initiator was defined as the ant that made contact first. The interaction score was based on the initiator. T en rep licat es were performed ( Suarez e t al 2002 Bu c zkowski and Bennett 2008) for each colony pairing F or each replicate the first interaction initiator reaction or avoidance of the initiator, and the interaction score of the initiator were recorded
60 A Chi Squa re test was performed in the JMP statistical package (SAS Institute 2012 b ) to determine if N. fulva and S. invicta were distributed differently as first interaction initiator and if the initiator showed a reaction or avoidance The interaction score of th e initiator for N. fulva and S. invicta were compared using a one way ANOVA using JMP (SAS Institute 2012 b ) Small Worker Group Interactions in a Neutral A rena Twenty workers of each species of similar size were random ly selected and transferred to plastic Fluon coated vials (3 cm diameter, 8.5 cm high [Thornton Plastic Co. Salt Lake City, UT]). The ants were allowed to calm for 5 min, after which the two groups were combined by gently emptying the ants from the vials into a plastic Fluon coated arena ( 9 x 9 x 1.5 cm 3 high [Fisher Scientific, Pittsburgh, PA]). Three replicates for each of the four colony pairings were performed The number of ants fighting (aggression level 3 or higher) was recorded at 1, 5, 10 20, 40, 60 min and every hour until 8 h after the two species were combined. For each time point, the number of ants fighting, number of ants dead, and the ratio of workers of each species involved in fights were recorded To test for differences in survival between colony pairing and specie s, the univariate survival analysis from the JMP statistical package (SAS Institute 2012 b ) was used to fit the survival distribution to a linear model. A probit analysis was used to determine time to mortality for each species. The proportion of N. fulva and S. invicta dead at each sample time was compared using a one way ANOVA and means separated by Tukey test. A regression analysis ( SAS Institute 2012 b ) was used to examine the effects time and species on the number of ants fighting with a P value of <0 .05 to establish statistical differences The proportion of N. fulva and S. invicta
61 fighting at each sample time was compared using a one way ANOVA and means separated by Tukey test. Large Worker Group Interactions in a Neutral A rena One hundred and fifty workers of each species of similar size were randomly selected and t ransferred to plastic, Fluon coated vials (4.5 cm diameter, 7.6 cm high [Fisher Scientific, Pittsburgh, PA]). The ants were allowed to calm for 5 min, after which the two groups were co mbined by gently emptying the ants from the vials into a plastic Fluon coated arena (23 x 23 x 2 cm 3 high [Fisher Scientific, Pittsburgh, PA]). Three replicates for each of the four colony pairings were performed Data were collected and analyzed as des cribed for small group interactions. Intruder into an Established Resident T erritory Individual intruder workers were collected on a wooden dowel ([152.5 mm x 2 mm ] Fisher Brand, Pittsburg, PA) and introduced into the rearing trays (52 x 38 cm 2 ) containin g the stock resident colony of the opposite species The initiator was defined as the ant that made contact first. The interactions score was based on the initiator. The first interactions were scored on a 1 4 scale (as previously described). Only the first i nteraction was recorded to avoid compounding effects of more than one interaction (Buczkowski and Silverman 2005). The intruder was removed and discarded after each trial. Subsequent trials were p e r formed when the resident colony was no longer vis ibly agitated ( after approximately 10 min). Ten replicates with each of the four colony pairs were tested; five replicates with N. fulva workers acting as the intruder and five replicates with S. invicta workers acting as the intruder. For each replicate the first interaction initiator reaction or avoidance of the initiator, the interaction score for the
62 initiator, and the maximum number of workers collaborating to kill the intruder were recorded A Chi Square test was performed in the JMP statistical package (SAS Institute 2012 b ) to determine if N. fulva and S. invicta were d istributed differently as first interaction initiator when residents and if the initiator showed a reaction or avoidance. The interaction score of the initiator and the number of resident colony collaborators for N. fulva and S. invicta were compared using one way analysis of variance (ANOVA) using JMP (SAS Institute 2012 b ) Field Study Data used was collected from field populations of ants as described in chapter 3 The ants were identified as N. fulva, S. invicta, Pheidole spp., Dorymyrmex spp., or others counted and square root transformed for each sampling time in 2010 and 2011 for each lot sampled In order to determine if N. fulva was the dominant species at each of the lo ts all groups identified except N. fulva were summed for each year and the sums compared using one way ANOVA. For the locations and years that had significant statistical difference ( P < 0.05) between N. fulva and the summed other ants the percentage of each group was calc ulated to determine the dominant ant group Results Dyadic I nteractions Of the 40 interactions, N. fulva initiated 72 .5 % (29/40) interactions compared to 27.5 % (11/40) initiated by S. invicta 2 = 8.1; df = 1; P = 0.004) Of the interactions N. fulva initiated, a reaction was observed (scored 3 or 4) 2 = 2.8; df = 1; P = 0.09) while S. invicta was avoided 38% of the time When S. invicta was the in itiator, they reacted 64 2 = 0.8; df = 1; P = 0.37). There was no
63 statistical difference between the frequency of interaction types for the two species ( F = 1.6; d f = 1; P = 0.21) (Figure 4 1 ) with s mean interaction score at 3.2 0.17 and S. invicta a t 2.8 0.2 8 Small Worker Group Interactions in a Neutral A rena T here was no statistical difference in the number of N. fulva and S. invicta that 2 = 4.6; df = 1; P = 0.20) after small group interactions ( Figure 4 2 ). At 10 min there was a signifi cant difference ( F = 6.7; df = 1; P = 0.012) between the numbers of dead ants for N. fulva and S. invicta but there was no significant difference between the numbers of dead ants for each species at any of the other sample points. T here was no statistica l 2 = 0.6; df = 1; P = 0.43) in the time to mortality for N. fulva and S. invicta (Table 4 1). Only time ( F = 131.2; df = 12; P < 0.0001) had an e ffect on the number of ants fighting. There were significantly more N. fulva fighting than S. inv icta at 5 min ( F = 4.4; df = 1; P = 0.04 7) ( Figure 4 3 ), but all of the other sample points, there was no difference between the numbers of each species fighting. After 120 min, fighting ceased. There was no significant difference ( F = 0.3; df = 1; P = 0 .55) in the number of ants cooperating in fighting. On average, 2.3 0.28 N. fulva workers cooperated in attacking S. invicta workers while 2.0 0.28 S. invicta workers cooperated in attacking N. fulva workers. Large Worker Group Interactions in a Neu tral A rena More S invicta died than N. fulva 2 = 540.0; df = 1; P < 0.0001) in large group interactions ( Figure 4 4 ). At 1 ( F = 1.0 ; df = 1 ; P = 0.33 ) 5 ( F = 0.1; df = 1; P = 0.75) and 10 ( F = 2.1 ; df = 1; P = 0.16) min there were no significant diff erences between the numbers of ants dead for each species, but at all other times sampled, there were
64 s ignificant differences (20 min [ F = 5.7; df = 1; P = 0.03], 40 min [ F = 18.4; df = 1; P = 0.0003], 60 min [ F = 25.7; df = 1; P < 0.0001], 120 min [ F = 51 .4; df = 1; P < 0.0001], 180 min [ F = 63.6; df = 1; P < 0.0001], 240 min [ F = 73.6; df = 1; P < 0.0001], 300 min [ F = 87.8; df = 1; P < 0.0001], 360 min [ F = 70.7; df = 1; P < 0.0001], 420 min [ F = 67.7; df = 1; P < 0.0001], 480 min [ F = 68.7; df = 1; P < 0.0001]). ( Figure 4 5 ). Nylanderia fulva had a faster time to 1% mortality than S. invicta (Table 4 2). At all other mortality levels, S. invicta had a faster time to mortality than N. fulva Time ( F = 118.86; df = 12; P < 0.0001), ant species ( F = 28.8 4; df = 1; P < 0.0001), and the interaction (time x species) ( F = 4.43; df = 12; P < 0.0001) had significant effects on the number of ants fighting. There were significantly more N. fulva fighting than S. invicta at 1 ( F = 7.3; df = 1, P = 0.012) 5 ( F = 8.9; df = 1; P = 0.007) and 10 ( F = 9.8; df = 1; P = 0.005) min ( Figure 4 5 ) but not a t the other sample times After 240 min, there were no ants fighting. There was a significant difference ( F = 5.1; df = 1; P = 0.03) in the number of ants cooperating in fighting. On average, 9.6 0.88 N. fulva workers cooperated in attacking S. invicta workers while 6.8 0.88 S. invicta workers cooperated in attacking N. fulva workers. Intruder into an Established Resident T erritory When N. fulva was the resident they initiated the interaction and reacted 100% (20/20) of the time. When S. invicta was the resident, there was no observed statistical 2 = 1.8; df = 1; P = 0.18), and N. fulva initiated 65% (13/20) of the interactions whereas S. invicta initiated 35% (7/20) of the interactions. Of the 65% of the interactions N. fulva initiated when S. invicta was the resident, N. fulva reacted (scored 3 or 4) 92.3% (12/13) of the time and avoided (scored 1 or 2) 7. 7% (1/13) of the time. When S. invicta was the resident and initiator, they
65 always reacted (scored 3 or 4) There was no statistical difference between the form of interaction when N. fulva and S. invicta were compared ( F = 1.6; df = 1; P = 0.2) (Fig ure 4 6 ), with only a slight difference in mean interaction score between the species (3.7 0.34 for S inv icta and 3.2 0.16 for N. fulva ). Nylanderia fulva and S. invicta workers cooperated in fighting the opposing species ( F = 33.3, df = 1, P < 0.0001). Nylanderia fulva workers fought in groups in 100% (20/20) of the encounters, whereas S. invicta workers received little help from their nestmates and fought in groups only in 30% (6/20) of the encounters. On average, 6.7 0.59 N. fulva workers attacked a n intruding S. invicta worker (range 2 14 workers). In contrast, only 1.9 0.59 S. invicta workers attacked an intruding N. fulva worker (range 1 6 workers). Field Study Nylanderia fulva was clearly the numerically dominant ant species at 4 of the 6 lo ts sampled for both years ([Citra 1, 2010: F = 22.7; df = 1; P < 0.0001], [Citra 1, 2011: F = 22.9; df = 1; P < 0.0001], [Gainesville 1, 2010: F = 36.3; df = 1; P < 0.0001], [Gainesville 1, 2011: F = 18.0; df = 1; P = 0.0002], [Gainesville 2, 2010: F = 89. 4; df = 1; P < 0.0001], [Gainesville 2, 2011: F = 148.9; df = 1; P < 0.0001], [Gainesville 3, 2010: F = 98.8; df = 1; P 0.0001], [Gainesville 3, 2011: F = 55.7; df = 1; P < 0.0001] ( Figure 4 7 ). At Citra lot 2 in 2010, N. fulva was the dominant species compared to the summed other ants but not signif icantly ( F = 3.6; df = 1; P = 0.07) ( Figures 4 7 and 4 8 ) Through the all of the seasons in 2010, N. fulva was the numerically dominant species (Figure 4 8). In 2011 at Citra lot 2, the summed other ants were dominant ( F = 28.8; df = 1; P < 0.001) (Figures 4 7 and 4 8) Overall, t he dominant ant was S. invicta (33 .6 % ) followed by Pheidole spp. (31.0 %), Dorymyrmex spp. ( 17.6 %), N. fulva (15.5%) and
66 other s (2.3 %) (Figure 4 8) The numerical ly dominant spe cies changed every season (4 8). In the spring Pheidole spp. were dominant and then S. invicta in the summer. In the fall, Dorymyrmex spp. was dominant. In 2010 at Citra lot 3, the summed other ants were dominant, but not significantly ( F = 2.7; df = 1 ; P = 0.11) (Figures 4 7 and 4 9) Overall, S. invicta (43.8%) and N. fulva (41.8 %) were the numerically dominant species (Figure 4 9). In the spring, S. invicta was numerically dominant but the numerical dominance was reduced in the summer (Figure 4 9). In the fall, N. fulva was the numerically dominant species, followed by S. invicta Pheidole spp., Dorymyrmex spp. and others (Figure 4 9). At Citra lot 3 in 2011 the summed other ants were significantly dominant ( F = 62.9; df = 1; P < 0.0001) (Figures 4 7 and 4 9) Overall, t he dominant ant was S. invicta (82.8 % ) followed by Pheidole spp. (8.3%) N. fulva ( 7 9 %), others (0.6%) and Dorymyrmex spp. (0.4%) (Figure 4 9) Solenopsis invicta was numerically dominant in every season. In the spring, Pheidol e spp. was numerically dominant over N. fulva but the dominance switched in the summer and continued in the fall (Figure 4 9). Discussion In order to understand N. fulv a and S. invicta competitive ability, aggressive interactions were examined. The u se of aggression is dictated by competition for resources ( Wilson 1970 ). In controlled arena bioassays, the tendency of individuals to fight becomes more evident. Results from dyadic interactions sugges t that N. fulva workers have a higher tendency to fi ght even without the numerical advantage when compared to S. invicta. However workers from both species were observed trying to escape from the arena, suggesting that self preservation may represent a stronger impulse than aggressive behavior in unfamili ar territory
67 As group size increased, N. fulva tendency to fight increased, as predicted by the cost minimize r hypothesis ( Buczkowski and Bennett 2008 ) The presence of nestmates increases the aggressiveness of individuals because of shared risk ( Buczk owski and Bennett 2008 ). In small group interactions more N. fulva engaged in fighting than S. invicta It was not until 20 minutes into the interaction period that more S. invicta were fighting than N. fulva but this occurred when there were more N. f ulva dead than S. invicta Cooperative fighting was not significant in small group interactions, and a fter five hours all of the ants in both species were dead. These results suggest that individuals of both species were more motivated by self preservat ion than protection of small group s as observed in dyadic interactions. In large group interactions, cooperative fighting was a significant factor in determining the dominant species. There were more N. fulva workers fighting throughout the large group interaction study. Nylanderia fulva fought by projecting formic acid from the acidopore (MacGown 2012) and by pulling on S. invicta appendag es. The release of formic acid may also act as an alarm pheromone, eliciting defensive responses from other N. fulva workers Nylanderia fulva was able to create the numerical a dvantage by attacking in groups an d systematically killing S. invicta creating a numerical advantage This attack strategy is also consistent with the cost min imizer hypothesis and increas es the l ikelihood of winning the fight. Fighting in groups may be one of the tactics N. fulva uses to outcompete other ant species and invade new areas. How individual N. fulva assessed group size in order to decide to fight or not is unknown. Most ants use semiochemicals, and those semiochemicals increase as group size increases (Buczkowski and Silverman 2005, Buczkowski and
68 Bennett 2008). Cooperative fighting was not seen as readily in S. invicta, even though they can release an alarm pheromone from t heir mandibular glands that induces cooperative fighting ( Tschinkel 2006). Solenopsis invicta workers fought more individually, pulling on appendages and acting defensively with their stinger. Whole colony defensive ability was tested with the introduc tion of a worker to resident colonies of different species Nylanderia fulva workers were highly aggressive when S. invicta workers were introduced into the ir colony and always initiated the first encounter. This suggests that N. fulva has strong nestmat e cognition cues that allow them to quickly detect intruders Upon detecting the intruder, N. fulva workers would project formic acid and pull appendages. The cost minimizer hypothesis was supported in N. fulva reaction to intruders by cooperative figh ting and high aggression displayed toward in t r u ders When a N. fulva worker was introduced to a S. invicta colony, N. fulva initiated most of the interactions and reacted Even with the numerical disadvantage, N. fulva engaged in fighting suggesting a hi gh willingness to enter into aggression. When S. invicta was the initiator to the introduced N. fulva, S. invicta reacted to protect their colony from the intruder. The cost minimizer hypothesis was not supported by S. invic ta because cooperat ive fighting was not observed. Overall, both species would aggressively protect their colony in the case of an intruder, most likely killing the intruder and preventing recruitment workers of the opposite species The aggressive behavior in animals is o Theory of Combat ( Franks and Partridge 1993, Whitehouse and Jaff 1996 Buczkowski and Bennett 2008 ). The square law and linear law are two combat models presented in the Theory of Combat that are often used when describ ing ant aggression (Franks and
69 Partridge 1993, Whitehouse and Jaff 1996 Buczkowski and Bennett 2008 ). The square law states that the side with the most individual fighters wins the battle no matter the fighting ability of individual ants (Lanche ster 191 6, Whitehouse and Jaff 1996). By creating the numerical advantage in large group interactions, N. fulva followed the square law model. The linear law focuses more on the individual fighter; if a fight is a series of one on one battles, a few good fighting individuals is better than many poor fighting individuals (Lanchester 1916, Whitehouse and Jaff 1996 ). Solenopsis invicta fought individually more than in groups implying that individual power is favored over group size. Plowes and Adams (2005) analyze d the mortality rates of S. invicta fighting in different numerical ratios, and found that S. invicta follow the linear law Both Meyers (2008) and LeBru n et al. (2012) present primary evidence that N. fulva displaces S. invicta from overlapping invasive areas. Nylanderia fulva is a polygynous ant and c olonies can reach tremendous numbers (Zenner Polania 1990a, Meyers 2008, Valles et al. 2012) Given the high number of N. fulva workers, it is hypothesized that they follow the square law. Solenopsis invi cta has the ability to form monogyne or polygyne colonies (Vinson 1997). Depending on the type of colony, would dictate the combat strategy. Monogyne S. invicta colonies would most likely follow the linear law while polygyne colonies would most likely fo llow the square law. In the field study, at four of the six locations sampled N. fulva was the numerically dominant species implying that N. fulva were able to out compete the other ant species for resources because competition is density dependent (Wi lson 1970). The numerical dominance then suggest s the use of the square law of combat as seen in the large group interaction.
70 Studies on a similar super colony ant species, Linepithema humile (Mayr) provided evidence of decreased prey availability duri ng the course of an invasion (three years), suggesting that food availability is a limiting factor in ant density and L. humile reached the carrying capacity for the invaded area (Ingram 2002). Nylanderia fulva has the ability to maintain high worker den sities enabling them to be strong competitors and effective invaders, but high worker density can lead to quickly reaching the carrying capacity of an invaded area. At the four locations that N. fulva was numerically dominant, i t is likely that the carryi ng capacity ha d not been met. At Citra lot 2, N. fulva was the numerically dominant species in 2010, suggesting that N. fulva would be the numerically dominant species in 2011. From fall of 2010 to spring of 2011 the dominance changed from N. fulva to Pheidole spp The numerical dominance was dynamic over each season suggesting competition for resources Most likely, N. fulva had reached the carrying capacity at Citra lot 2. Depleting food re source availability and less than ideal weather conditions in 2011(chapter 3) allowed the other ant species to out compete N. fulva and become numerically dominant. The field dynamic between N. fulva and S. invicta was not evident at this lot. The field dynamic between N. fulva and S. invicta at Citra lot 3 is more evident. In the spring 2010, S. invicta was numerically dominant. Solenopsis invicta nest in more sunlight areas allowin g them to forage and begin reproducing earlier in the year Nylanderia fulva overwinters in less sunlight areas, and do not begi n to actively forage and begin reproducing until the temperatures increase in the late spring (chapter 3). In the summer, N. fulva can produce and maintain high worker densities, allowing them to out compete S. invicta because of numerical dominance (squar e law of combat) In the
71 fall this trend continued At Citra lot 3, N. fulva reached its c arrying capacity in fall 2010. The less than ideal weather conditions and with the low densities of N. fulva allowed S. invicta to out compete N. fulva for food re sources and territory Numerical strength usually dictates the winner of fights between groups of animals (Adams 1990, Traniello and Bechers 1991), and is supported by individual The evid ence that N. fulva displaces S. invicta from overlapping invasive areas is most likely in the beginning of an invasion by N. fulva. Nylanderia fulva increases in density and uses the large densities to numerically dominant over other ant species. The hi gh food resource availability in newly invaded areas allows the large densities of N. fulva to be sustained. As food resources begin to deplete, the large densities of N. fulva are not sustainable, thus other ant species are able to compet e for resources and territory. Both N. fulva and S. invicta are urban pest s of serious concern. The natural competition between these two species can be used in helping control them. A specific treatment, such as that suggested by Kabashima et al. (2007) can be appli ed first against S. invicta so N. fulva can help in eliminating the S. invicta which is a more dangerous species in interactions with humans. A second treatment applied a few days later can be more directly active against N. fulva so the second pest a nt species is eliminated. Such strategy would only be possible in locations where both species exist, and N. fulva has not yet become the dominant species. In locations where N. fulva is numerically dominant, a treatment specific for N. fulva would first be applied reducing N. fulva density would allow other ant species to begin to competing for food resources and territory maintain low N. fulva densities.
72 Table 4 1. Mean time and 95% confidence interval to 1, 10, 5 0, 90, and 99 percent mortality of Nylanderia fulva and Solenopsis invicta in small group (20 individuals) interactions. Mean Time in Min (CI 95) M ortality (%) N. fulva S. invicta 1 1.8 (1.08 3.09) 2.2 (1.68 2.94) 10 5.8 (4.08 8.28) 6.8 (5.62 8.18) 50 23.9 (18.47 31.03) 26.6 (23.18 30.54) 90 98.6 (67.12 144.81) 104.4 (84.79 128.58) 99 312.56 (178.22 548.19) 318.3 (234.64 431.84)
73 Table 4 2. Mean time and 95% confidence interval to 1, 10, 50, 90 of Nylanderia fulva and Solenopsis invict a in large group (150 individuals) interactions. The predictions to 90% mortality for N. fulva and 99% mortality for both species mortality were exponentially outside of the sampled times, thus a very weak representation of actual time to mortality. M ean Time in Min (CI 95) Mortality (%) N. fulva S. invicta 1 1.3 (1.06 1.71) 3.7 (3.32 4.07) 10 14.6 (12.59 16.97) 12.1 (11.28 13.00) 50 272.6 (242.0 307.16) 52.26 (49.58 55.09) 90 225.6 (210.07 242.18)
74 Figure 4 1 Frequency o f interaction types when N. fulva and S. invicta were the initiators of the interactions when introduced to ant of the opposite species in dyadic interactions
75 Figure 4 2 Mean number of N. fulva and S. invicta dead at sample times in small gro up (20 individuals of each species) interactions. Column p airs topped with an asterisk are significantly different ( F = 6.7; P = 0.012) from each other. Each error bar is represent s 1 standard error from the mean.
76 Figure 4 3 Number of ants f ighting at each sample time for small group (20 individuals) interactions. Column p airs topped with an asterisk are significantly different ( F = 4.4; P = 0.047) from each other. Each error bar is represents 1 standard error from the mean.
77 Figu re 4 4 Mean number of N. fulva and S. invicta dead at each sample point for large group (150 individuals) interactions. Column p airs topped with an asterisk are significantly different from each other (20 min [ F = 5.7; P = 0.03], 40 min [ F = 18.4; P = 0. 0003], 60 min [ F = 25.7; P < 0.0001], 120 min [ F = 51.4; P < 0.0001], 180 min [ F = 63.6; P < 0.0001], 240 min [ F = 73.6; P < 0.0001], 300 min [ F = 87.8; P < 0.0001], 360 min [ F = 70.7; P < 0.0001], 420 min [ F = 67.7 ; P < 0.0001] 480 min [ F = 68.7; P < 0.0 001]) Each error bar is represent s 1 standard error from the mean.
78 Figure 4 5 N umber of ants fighting at each sample time in large group (150 individuals) interactions Column p airs topped with an asterisk are significantly different from each other (1 min [ F = 7.3; P = 0.012], 5 min [ F = 8.9; P = 0.007], 10 min [ F = 9.8; P = 0.005]) Each error bar is represent 1 standard error from the mean.
79 Figure 4 6 F requency of interaction types when N. fulva and S. invicta were the initiators of the interactions after being introduced into resident colonies of opposite species.
80 Figure 4 7 N. fulva and the summed other ants at each sampling plot for 2010 and 2011 The asterisks represent a significant difference between N. fulva and all the other ant species collected ( F test, P < 0.05). Each error bar is repr esents 1 standard error from the mean.
81 Figure 4 8. The ratio of ants found at Citra lot 2 in 2010 and 2011.
82 Figure 4 9 The ratio of ants found at Citra lot 3 in 2010 and 2011.
83 CHAPTER 5 FOOD PREFERENCE OF NYLANDERIA FULVA Introduction Nylander ia fulva (Mayr) is a polygynous, polydomous, supercolonial, tramp ant species. It is established throughout Florida and along the Gulf Coast (Pereira and Koehler 2007 Meyers 2008 Calibeo and Oi 2011 MacGown 2012 ). Nylanderia fulva adversely affects th e fauna in ecosystems ( Meyers 2008), agricultural production ( Zenner Polania 1990a, Campos Farinha and Zorzenon 2005 Harmon 2009 ), and invades ya rds and occasionally structures ( Zenner Polania 1990a, C ampos Farinha and Zorzenon 2005, MacGown and Layton 20 10, Calibeo and Oi 2011). Extreme populations of N. fulva indicate copious food sources in the environment. Nylanderia fulva has been observed feeding on honeydew (Zenner Polania 1990a ), live and dead insects ( Zenner Polania 1990a, Meyers 2008, Warner and Scheffrahn 2004 ), as well as other dead animals ( Meyers 2008). In research studies with N. fulva, the ants have been baited with hot dogs ( Meyers 2008 Nester and Raspberry 2011 ) and 20% honey solution ( Meyers 2008). The observations and bait s used did not identify the food preference of these ants. In the field the ants get carboh ydrates from honeydew, but it is not known if the ants forage more on proteins or oils from animals. Cook et al. (2012) showed that laboratory colonies of N. fulv a preferred a 1:2 protein: carbohydrate ratio of dry artificial food. H igh rates of mortality in this study fail ed to indicate much relevance (Cook et al 2012). Opposingly, Scott (2012), showed that laboratory colonies of N. fulva preferred dry bait tha t was a 1.5:1 pr otein: carbohydrate ratio. T he addition of a liquid protein (ground crickets)
84 caused the ants to forage more readily on the bait. When the dry bait with added protein was field tested, N. fulva did not show a preference for either (Scott 2012). Seasonal s h ifts in food preference has been documented for Campontus pennsylvanious (DeGreer) (Tripp et al. 2000), Linepithema humile (Mayr) (Rust et al. 2000), and Solenops is invicta Buren (Sorenson et al. 1985). Knowing of food preference shifts helps in desi gning and using the proper bait attractants given the season. However, l ittle data exists regarding the season shifts in food preference for N. f ulva. The objective of this study was to determine the food preference over time of N. fulva in a field setting, and to determine if there is a seasonal shift in food preference. U will be helpful with designing m ore effective baits Materials and Methods Food preference studies with multi ple colonies of N. fulva were conducted on six urban lots; three in Gainesville, FL and three in Citra, FL. The locations in Gainesville were a recreational park and two urban house lots. The locations in Citra were three rural house lots. The sites wer e chosen because of the known existence of N. fulva and because these are representative of locations of where N. fulva can be found in Florida. The lots were sampled biweekly, beginning in April 2011 and ending in November 2011. Sampling was done only w hen the air temperature reached a minimum of 21C and was not over 32C based on observations by Warner and Scheffrahn (2004) The sampling period was divided into spring (April and May), summer (June, July, August, September), and fall (October, November ) for analysis purposes.
85 Three foods were used to determine food type preference; protein (tuna packed in water [StarKist, StarK ist Co., Pittsburg, PA]), lipids (unrefined soybean oil [Organic Soybean Oil, Eden Foods, Inc., Clinton, MI]), and carbohydrates (honey [Great Value, Wal Mart Stores, Inc,., Bentonville, AR]). Two grams of each food was weighed out in sample tubes (100 mL hinged lid containers from Fisher Scientific, Pittsburg, PA ) using a tabletop balance (Ohaus Cooperation, Parsippany, NJ). Th e tuna was well drained before weighing by pressing the water out of the can using the lid. T hree replications of the study were performed at each lot The replications were at least 5 meters apart and placed near different ant trails. The sample tubes were 10 to 15 cm from the trail. The sample tubes were laid on their side with the hinge cap holding the tube in place near a trail of foraging N. fulva. In each replicate, each food (protei n, lipid, and carbohydrate) was presented to the ants. The samp le tubes were set up in a triangle, approximately 15 cm apart with the openings facing inward. The ants in the sample tubes were collected after 30 minutes by closing the hinged lid trapping the ants inside. The number of N. fulva in each sample tube was counted to determine food preference. Th e data was normalized by square root transformation. An analysis of variance (ANOVA) ( SAS Institute 2012 b ) follow ed by a S determine which food was preferred by N. fulva (Hopper and Rust 1997). Percentage s ants foraging on each food source and for each season were calculated All tes ts of significance were evaluated at P = 0.05. Results All of the food choices were fed on by the ants during the study period. Ant s foraged signific antly more ( F = 41.4; df = 2; P < 0.0001) on the protein ( 5.8 0.23 ants)
86 than the carbohydrate ( 3.4 0.23 ants ) and the lipid ( 2.9 0.23 ants) ( Figure 5 1a ) There was no significant difference between the number of ants foraging on carbohydrates and lipids ( t = 1.96 ; P > 0.05). The same food preference was seen for the seasons as well (Figure 5 1 b d) In the spring the ants foraged significantly more ( F = 10.2; df = 2; P < 0.0001) on the p rotein (3.9 0.36 ants) than the carbohydrate ( 2.4 0.3 6 a nts) and the lipid ( 1.9 0.36 ants) (Figure 5 1b ) There was no significant difference between the number of ants foraging on carbohydrates and lipids ( t = 1.97 ; P > 0.05). In the summer, N. fulva foraged significantly more ( F = 29.1; df = 2; P < 0.0001 ) on the protein (7.4 0.33 ants) than the carbohydrate ( 4.4 0.33 ants ); and the lipid ( 4.1 0.33 ants ) (Figure 5 1c) There was no significant difference between the number of ants foraging on carbohydrates and lipids ( t = 1.95; P > 0.05). In the fa ll, N. fulva foraged significantly more ( F = 14.1; df = 2; P < 0.0001) on the protein ( 3.2 0. 29 ants) than the carbohydrate ( 1.6 0. 29 ants) and the lipid (1.1 0. 29 ants ) (Figure 5 1d) There was no significant difference between the number of ants f oraging on carbohydrates and lipids ( t = 1.9 7 ; P > 0.05). At Citra 1 there was no significant difference between food source preferences until July ( Figure 5 2 ). In July, t he ants preferred the carbohydrate source more than the other foods sources ( F = 7. 9 ; df = 2 ; P = 0.0 04 ). In the middle of Septem ber, the food preference changed from carbohydrates to protein, but there was no significant difference between the food preferences from the end of July until the end of the study. The protein source only be came significantly preferred at Citra 2 ( F = 4.1 ; df = 2 ; P = 0.03 ) in June. However, in July there was no significant difference between the food sources. In August, the protein started to be significantly preferred ( F = 4.2 ; df = 2 ;
87 P = 0.04 ) and this preference lasted until November when there was no significant difference in the food preference ( F = 0.6 ; df = 2 ; P = 0.56 ). June ( F = 18.4; df = 2 ; P = 0.0001) and July ( F = 7.5; df = 2 ; P = 0.0 0 5 ) are the only months when the protein source was signifi cantly preferred at Citra 3. In August, the protein sources was significantly preferred over lipid source, but neither the protein and carbohydrate or carbohydrate and lipid were significantly different from each other ( F = 4.2 ; df = 2 ; P = 0.04 ). At Gain esville 1, the ants significantly preferred protein in June ( F = 10.6 ; df = 2 ; P = 0.0 00 5), August ( F = 7.4 ; df = 2 ; P = 0. 0 0 6 ), September ( F = 9.2 ; df = 2 ; P = 0.0 02 ), and October ( F = 5.8 ; df = 2 ; P = 0.0 1) At Gainesville 3, the ants significantly pref erred protein in May ( F = 38.95; df = 2 ; P = 0.05), June ( F = 7.2 ; df = 2; P = 0.004 ), and August ( F = 12.9 ; df = 2 ; P = 0.0005 ), September ( F = 12.0 ; df = 2 ; P = 0. 00 0 8 ). At both locations there was no significant preference in July ([Gainesville 1 ; F = 1 .2 ; df = 2 ; P = 0.0] [Gainesville 3 : F = 3.3 ; df = 2 ; P = 0.0 7 ]). At Gainesville 2, the ants significantly preferred protein in May ( F = 4.6 ; df = 2 ; P = 0.03 ), July ( F = 5.0 ; df = 2 ; P = 0.02 ), and October ( F = 5.5 ; df = 2; P = 0.01 ). In August and Nove mber, the protein sources was significantly preferred ([August : F = 4.2; d f= 2 ; P = 0.03 ] [November : F = 3.9; df = 2; P = 0.04 ) over lipid source, but neither the protein and carbohydrate or carbohydrate and lipid were significantly different from each ot her. Discussion P rotein was highly preferred by N. fulva during the whole span of the study. The ants still foraged on the other two foods, but not as readily. Rust et al. (2000) found that Argentine ants, L. humile, also a polygynous, polydomous, superc olonial ant, preferred
88 proteins in the spring and the summer because of the need for protein for reproduction. They also found that carbohydrates were preferred in the fall and winter t o sustain the workers activity as described also by Markin (1970). Ar gentine ant colonies need more protein when they are producing more brood in the summer months (Markin 1970), and foraging i s demand for nutrition. Food preference of N. fulva may also reflect the nutritional needs of the colony throughout the year. Nylanderia fulva had a high demand for protein, especially during the colony growth phase in the spring In the summer and fall, the ants still foraged more on protein suggesting that the colony was still growing Because c arbohydra tes are used for energy and g iven the potentially high number of workers in the summer and early fall, it was expected that N. fulva workers would switch from preferring protein to preferring carbohydrates but this was not observed This may indicate th at N. fulva does not require as much carbohydrates to function or that there was a lack of the other proteins and lipids in the environment thus N. fulva was feeding on those sources from the experiment and they were collecting carbohydrates elsewhere L ipids are used to store energy and as expected, lipids were primarily collected in the fall in preparation for the winter months. At Citra 1 t he ants preferred carboh ydrates over the other two food sources. T he house was initially surrounded wi th thick vegetation, but in spring of 2011, the ow ners removed all the vegetation. The loss of vegetation may have caused a loss of sap feeding insects from which the ants collected honeydew causing a reduction in As this nutri ent is essential for survival, the ants foraged more on the carbohydrate source when presented in this experiment
89 In a previous study N. fulva demonstrated a change in food prefer ences (Meyers 2008). In a four week field study, N. fulva switched from preferring 20% sucrose solution to preferring protein and lipids in the form of hot dog s The study had to be stopped because the height of the grass made it impossible for the research to continue (Meyers 2008). The ants may have switched their food pre ference because of the availability of food in the environment as seen as Citra 1 Also, the short duration of the study does not provide information of a seasonal food preference In this study, N. fulva did not have a season food preference they prefer red protein at all times The odorous house ant, Tapinoma sessile Say also does not have seasonal f ood preference (Barbani 2003). Tapinoma sessile prefers carbohydrates year round but still feeds on proteins and lipids (Barbani 2003). Little is known ab out the bait preference of N. fulva. The fact that N. fulva does not show a seasonal food preference and that protein was the preferred food choice here and by Scott (2012) suggests that a program incorporating protein baits would be effective year round
90 Figure 5 1. The percentage of N. fulva feeding on each of the food sources for the whole study (a), spring (b), summer (c), and fall (d).
91 Figure 5 2 Mean number of N. fulva collected in baited tubes containing a protein (tuna fish), lipid (soybean oil), and a carbohydrate (honey) in six sampling sites in north c entral Florida from April to November 2011. Month Mont h
92 CHAPTER 6 CONCLUSIONS The tawny crazy ant, Nylanderia fulva (Mayr ) formerly misidentified as the Caribbean craz y ant, Nylanderia pubens is one of the more troublesome ants in Florida and throughout its introduced region in the southern United States The tawny crazy ant is a difficult ant to control, and th e discrepancy in proper identification over the last few years has c ontributed to failed control techniques along with t he super colony nature of the ant, and nesting strategies of the tawny crazy ant The tawny crazy ant does not sting and only rarely bite s when their nest s are disturbed. Distribution and N aming : The tawny crazy ant is originally from Brazil (Zenner Polania 1990a). It was introduced into Colombia as a biological control agent against pest s of the sugar cane but later became a serious pest itself. Currently the tawny crazy ant can be found in several countries in South American and the Caribbean islands and is known as a pest in many locations. It is not known when or exactly how the tawny crazy ant was introduced into the United States. The first reported identification of a similar ant was in the 1950s in Miami (Trager 1984). The ant was identified as Paratrechina pubens (Forel) which is now classified as Nylanderia pubens the Caribbean crazy ant. At that time, the ant was not considered an important pest. The next report of P. puben s was around a M iami hospital in the early 1990 s but the ant was still not considered an important pest. Beginning in 2000, a n ant with identical physical characteristics was observed in Fort Lauderdale, Port St. Lucie, Jacksonville, and Sarasota (Warner and Scheffrehn 2010) and quickly became a pest of concern Given the proximity of the infestations to port cities, it is thought that the ants ente red Florida through human trade, and that this
9 3 represented the same ant, N. pubens previously identified i n the Miami area. In 2002, an ant species we know to be the same one that became a pest in Florida was found in Texas (Meyers 2008). This ant was identified as a different species, Paratrechina species near pubens and known locally as the Rasberry crazy ant (Meyers 2008). In 2009, the ants were discovered in Mississippi and then in 2010 in Louisiana (Hooper Bui et al. 2010, MacGown and Layton 2010 ). T hese new infestations were first found also, in and around port cities. In 2010, both P. pubens and P. spp. nr. pubens were moved to the genus Nylanderia b ased on new genetic evidence showing differenc es between the genera Paratrechina and Nylanderia (LaPolla et al. 2010). The Caribbean crazy ant was then referred to as Nylanderia pubens. Genetic comparison of ants found in Florida and Texas led to the conclusion, i n the spring of 2012 that the ants in the two locations were the same species of ant (Zhao et al. 2010). Later in 2012, the ant was finally identified as Nylanderia fulva (May r). Nylanderia pubens and Nylanderia fulva are different species with differing biologies and ranges. The common n ame, tawny crazy ant was approved by the Entomolog ical Society of American in January of 2013 for N. fulva The initial misidentification of N. fulva as N. pubens has c ontributed to failed control actions and caused problems in research The ant misiden tified from 2000 is N. fulva and t he ant in Miami from the 19 50s supposedly was and is N. pubens. Why Control of the Tawny Crazy Ant is Difficult Tawny crazy ant colonies can have multiple queens and nests per colony. In the winter, the ants are usua lly found in only one or a few nests. These nests are referred to as permanent nests (Zenner Polania 1990a). Permanent nests are always occupied and are the nests the ants leave from to form temporary nests in the late spring and summer. Permanent nests are found in protected areas and have been found extending 40 cm below the grounds surface (Zenner Polania 1990a). In north central Florida, i n the spring and summer, the queens produce large amounts of brood causing the ant population to increase rapidl y leading to the formation of many temporary nests. These temporary nests allow
94 the ants to spread across the landscape to forage for resources. Temporary nests can be found in or under just about anything in the landscape. If temporary nests are distur bed, the ants will move to another nesting location. As the weather begins to cool in the late fall and winter, the ants may retreat back to the old permanent nest or may form new permanent nests (chapter 3) Having multiple queens per nest leads to a high reproductive potential. In permanent nests, up to 14 queens, 2300 workers, 1000 eggs, 1300, larvae, and 4500 pupae have been found (Zenner Polania 1990a). Temporary nests are much smaller than permanent nests with varying numbers of castes and broo d. Areas with a high amount to leaf debris and mulch are ideal places for tawny crazy ants to nest (chapter 3). These areas hold moisture that tawny crazy ants need to survive. Tawny crazy ants have also been found nesting in mulch piles, in rotting wo od, in and under pieces of garbage, and under flowerpots. The potential abundance and diversity of nesting sites can make tawny crazy ant nests hard to characterize, especially in the summer months. The high reproductive potential, nesting strategy and t he abundance of potential of temporary nesting locations allows tawny crazy ants to spread across a landscape. A single colony can cover acres of land, making them an area wide problem not just a single lot problem. Unlike most ant species, tawny crazy a nts from different colonies usually do not fight. Their non territorial nature is a factor contributing to large populations. However, the tawny crazy ant is highly aggressive against other ant species, often displacing other ants and other arthropod spe cies from areas (chapter 4). Tawny crazy ants prefer to feed on proteins year round compared to other food sources (chapter 5). Protein sources in a typical north central Florida lot could include: termites, springtails, caterpillars, grubs, spiders, and other ant species (Zenner Polania 1990a). Integrated Pest Management Tactics for Tawny Crazy Ant Control An integrated pest management (IPM) method offers a safer way to control tawny crazy ants and leads to a greater chance of success. An IPM method incorporates prevention and suppression of problem causing organisms through proper identification, inspection, and control techniques (sanitation, exclusion, and use of pesticides ).
95 Identification The first step in an IPM method is proper identification of the pest. Without proper identification, control methods can be useless and potentially dangerous. Tawny crazy ants are difficult to identify. At first glance, the tawny crazy ant looks like the red imported fire ant. Tawny crazy ants are reddish b rown in color and 2.0 2.5 mm or 1/8 th in length Inspection under the microscope will reveal 12 segmented antennae, one node, and an acidopore but t hese morphological characteristic s alone will not lead to correct identification of N. fulva For proper morphological identification, males are needed (Gotzek et al. 2012) and t his is the reason why mis identification s occurred The foraging trail s can aid in identifying tawny crazy ants. Tawny crazy ant foraging tra ils are wide (several ants ) and loose T he a nts move in a quick an d erratic crazy ant Inspection After proper identification as the tawny crazy ant inspections will help reveal where the ants are foraging and nesting. Look for areas where the ants w ould have access to food, water, and shelter. Check the base of landscape plants to see if the ants are trailing up and down to collect honey dew from sap feeding insects. Also, look around garbage containers, burn rings, and compost bins for foraging ant s. It may be helpful to put down a tablespoon of honey or tuna fish in several different places around the lot to see where the ants are foraging (chapter 5) Tawny crazy ant nests are not distinct like red imported fire ant or pyramid ant mounds Findi ng nests may require following foraging trails. D epending on the time of year, N. fulva nests and foraging can be in different locations (chapter 3) IN north central Florida, i n the late fall, winter and
96 early spring, the ants are foraging in discrete l ocation s most l ikely around permanent nests. P ermanent nests will be found in protected locations such as at the base of trees and at exposed root edges (Zenner Polania 1990a) Start by looking for workers foraging in and around mulch and leaf debris ar ound trees and structures. S huffle mulch and leaf debris to get the ants to come out of the nest. In t he late spring in early summer, the ants will spread out to temporary nests, increasing the area that must be inspected for nesting locations. Tawny cr azy ant populations are highest in the summer months when we can e xpect multiple foraging trails lead ing to different nests (chapter 3) Control Techniques All animals need food, water, and shelter to survive. By redu cing the amount of these three elem ents one can more easily reduce tawny crazy ant populations. Sanitation water, and shelter. If possible feed pets indoors. Besides feeding the ants, a nother possible side effect of feeding pets outside in heavily infested areas is inhalation of the ants by pets that could cause respiratory complications Clean up animal waste and other items that would provide food for the ants. Avoid leaving garbage in burn cans or fire rings. Regularly washing out garbage and recycling containers will also reduce feeding sources. Trimming the landscaping around structures will reduce the number of honeydew producing insects and help prevent the landscape plants being used as a bridge to enter structures.
97 Tawny crazy ants will form temporary nests in almost any lawn debris (fallen palm fronds rotting wood, leaf debris, etc.). By diligently removing this debris, possible nesting sites are reduced which will prevent extreme growth in tawny crazy ant populations Mulch and leaf debris provide food, water, and shelter to the tawny crazy ant The mulch and leaf debris hold in moisture that the taw ny crazy ant needs to survive. Proper drainage of mulched area will help in moisture r eduction. Also, check irrigation systems in mulched areas; a ny leak will contribute to tawny crazy ant survival Other arthropods and fungi in the mulch and leaf debris are a food source for the tawny crazy ant. In recent studies, mulch and leaf debris were the most likely place to find tawny crazy ants foraging year round (chapter 3) Move mulch back away from structures, keep it dry, and if possible avoid using mulch and leaf debris in landscaping. Tawny crazy ants usually enter structures to find fo od and water when populations are extremely large Good indoor sanitation will help reduce the presence of this ant as well as other pest species indoors Store food in tightly closed containers, dispose of food waste promptly, and clean up spills. Regul arly vacuuming will remove any food particles that maybe on floors or carpets and serve as food for ants Good sanitation practices by home owners/inhabitants in infested areas will help reduce tawny crazy ant populations. Also, g ood sanitation will all ow pest control specialist to more easily access the pest problem and effectively apply control measures Exclusion Exclusion techniques are designed to prevent pests from entering structures. Check that doorways and windows have tight fitting doors and screens with properly
98 installed weather stripping. Sealing cracks, crevices, and other entry points in the foundation and walls is essential in reducing tawny crazy ants and other pest s inside. Chemical Control Pesticides are the most common and recogni zed method of pest management. When handling any pesticide be sure to read and follow the label for maximum control and to reduce safety hazards. Currently, there are no over the counter pesticides that effectively control tawny crazy ants. It is best t o call a pest control specialist to help in managing tawny crazy ant populations. Before the pest control specialist arrives, it would be beneficial to follow the sanitation and exclusion methods provided here. Pest control specialist s have a limited amou nt of pesticides available and labeled for tawny crazy ant control making sanitation and exclusion methods even more important. Once a pest control specialist arrives, they will access the scope of the problem and design a management strategy. Some tech niques they may use, but are not limited to, include: Applying a residual pesticide, which the ants must come in contact with to be controlled, around structures to prevent tawny crazy ant entry. If the pest control specialist applies a residual pesticide be sure not to use water or a broom to remove the dead ants. Instead, use a leaf blower to ensure that the pesticide is not washed or scraped away and will continue to help protect the structure. Recent studies have shown that populations of the tawny crazy ant are the worst in the late spring, summer, and early fall (chapter 3). The sooner in the year a management plan is put into place, the better the chances are the control actions will produce desirable results Pesticide baits on which the tawny crazy ant feeds on, or a contact pesticide may be broadcasted across the landscape. These pesticides are available in granular and liquid form. The pest control specialist may decide to use bait stations to dispense baits to the ants. The goal with any behavior of sharing food throughout the colony via trophallaxis (oral exchange of nutrients between colony members). The foraging ants take the bait back to the
99 nest and share the bait throughout the colony. Liquid ba its are shared by adult workers, while solid bait granules are taken back to the nest for the larvae to digest and then workers feed to the colony the digested granules. This is because workers cannot digest solid food. If the pest control specialist no tices tawny crazy ants trailing up landscape plants, they may apply a systemic pesticide to the plants to control honeydew producing insects. This helps in eliminating the honeydew as a food source for the ants but does not affect nectar production Th ings to Remember Tawny crazy ants are an area wide problem, not individual lot problems. The reduction of a tawny crazy ant population on one lot does not guarantee that they the ants will not reinvade from adjacent areas. Tawny crazy ants affect entire communities. Follow the guideline s set forth by the pest control specialist. They can tell you how to maximize the potential of their management plan to control tawny crazy ants.
100 L IST OF REFERENCES Abbott, K. L. 2005. Supercolonies of the invasive y ellow crazy any, Anoplolepis gracilipes on an oceanic island: forager activity patterns, density and biomass. Insect Soc. 52: 266 273. Abbott, K. L. 2006. Spatial dynamics of supercolonies of the invasive yellow crazy ant, Anoplolepis gracilipes on Chris tmas Island, Indian Ocean. Divers Distrib. 12: 101 110. Abbott, K. L. and P. T. Green. 2007. Collapse of ant scale mutualism in a rainfo rest on Christmas Inland. Oikos 116: 283 291. Adams, E. S. 1990. Boundary disputes in the territorial ant, Azteca tri gona : effects of asymmetries in colony size. Anim Behav 39: 321 328 Aldana, R. C., M. L. Baena, and P. C Ulloa. 1995. Introduccin de la hormiga l oca ( Paratrechina fulva ) a la Reserva Natural Laguna de Sonso (Va lle del Cauca Colombia). B ull Museo Ento mol Univers. Valle 3: 15 28. Ascunce, M. S., C. C. Yang, J. Oakey, L. Calcaterra, W. J. Wu, C. J. Shih, J. Goudet, K. G. Ross, and D. Shoemaker. 2011. Global invasion history of the fire ant Solenopsis invicta Scie nce 331: 1066 1068. Barbani, L. E. 2003 Foraging activity and food preferences of the odorous house ants ( Tapinoma sessile Say) (Hymenoptera: Formicidae). M.S. thesis, Virginia Polytechnic Institute and State University, Blacksburg, VA. Bolton, B., G. Alpert, P. S. Ward, and P. Naskrecki. 200 7. Bolton's catalogue of the ants of the w orld. [ CD ROM ] Harvard University Press, Cambridge, MA. Brenner, R. J. 1988. Focality and mobility of some peridomestic cockroaches in Florida (Dictyop tera:Blattaria). Ann. Entomol. Soc. Am. 81: 581 592. Buczkows ki, G. and G. W. Bennett. 2008. A ggressive interaction between t h e introduced Argentine ant, Linepithema humile and the native house ant, Tapinoma sessile Biol. Inv asions 10: 1001 1011. Bucz kowski, G. and J. Silverman. 2005. Context dependent nest mate discrimination and the effect of action thresholds on exogenous cue recognition in the Argentine ant. Anim. Behav. 69: 741 749. Calibeo D. and F. Oi. 2010. Nylanderia fulva in Florida: seasonal population trends, distribution and o bservations pp. 11 1 3. In Proceedings 2010 Imported Fire Ant Conference, 19 22 April 2010, Little Rock, AR. ( http://www.extension.org/sites/default/files/w/c/c4/2010_Annual_IFA_Conferenc e.pdf ).
101 Calibeo, D. and F Oi. 2011. Integrated pest management (IPM) of the Caribbean c razy ant, Nylanderia (= Paratrechina ) fulva (Forel) ENY 2006. Gainesville, FL University of Florida. ( http://edis.ifas.ufl.edu/in889 ). Campos Farinha, A. E. C., and F. J. Zorzenon. 2005. Predatory and intera ction behaviors of Paratre china fulva (Mayr) (Hymenoptera: Formicidae) on phytophagous insects o n palm trees. Arq. Inst. Biol. 71: 143 144. Cerda X. A. Dahbi, and J. Retana. 2002. Spatial patterns, temporal variability, and the role of multi nest colonie s in monogynous Spanish desert ant. Ecol. Entomol. 27: 7 15. Chen, J. S. C. and P. Nonacs 2000. Nestmate recognition and intraspecific aggression based on environmental cues in Argentine ants ( Hymenoptera: Formicidae). Ann. Entomol. Soc. Am. 93: 1333 13 37. Cook, S. C., R. A. Wynalda, R. E.Gold, and S. T. Behmer. 2012. Macronutrient regulation in the rasberry crazy ant ( Nylanderia sp. nr. pubens ). Insect Soc. 59: 93 100. Corin, S. E., K. L. Abbott, P. A. Ritchie, and P. J. Leste. 2007. Large scale unic oloniality: the population and colony structure of the invasive Argentine ant ( Linepithema humile ) in New Zealand. Insect Soc. 54: 275 282. Creighton W. S. 1950. The ants of North America. Bull. Mus. Comp. Zool. Havard Univ. 104: 1 585. Culbert, D. 2007. Cra z y ants in our plants. ( http://okeechobee.ifas.ufl.edu/News%20columns/Crazy.Ants.htm ). d 'Ettorre, G. and A. Lenoir. 2010. Nestmate recognition pp. 194 209. In L. Lach C. L. Parr and K. L. Abbott ( ed s.), Ant ecology: Oxford Biology New York, NY Dillier, F. X. and R. Wehner. 2004. Spatio temporal patterns of colony distribution in mondomous and polydomous species in North African desert ants, genus Cataglyphis Insect Soc 51: 186 196. Drees, B. M., P. Nester, and R. Gold. 2009. Control of rasberry crazy ants in and around homes and structures. Texas A&M University Fact Sheet. College Station, TX. ( http://urbanentomology.tamu.edu/pdf/RCA_control_fact_sheet.pdf ). Elias, M., R. Rosengren, and L. Sundstorm. 2005. Seasonal polydomy and unicolonality in a polygynous population of the red wood ant Formica truncorum Trends Ecol. Evol. 14: 135 139.
102 G ermn, A. V., P. A. Daz, L. A. Lastra, N. C. Mesa, I. Zenner Polania, and L. A. Gomez. 2004. Reconocimiento de enemigos naturals de la homiga loca, Paratrichina fulva (Hymenoptera:Formicidae), en el municipio de El Colegio (Cundinamarca) y en el valle del ro Cauca. Rev. Colomb. Entomol. 30: 225 232. Giraud, T., J. S. Pedersen, and L. Keller. 2002. Evolution of supercolonies: t he Argentine ants of southern Europe. PNAS. 99: 6075 6079. Gold, R. 2011. Tawny (R asberry ) crazy a nt, Nylanderia fulva College S tation, TX. Texas A & M ( http://urbanentomology.tamu.edu/ants/rasberry.html ). Gotzek, D., S. G. Brady, R .J. Kallal, and J. S. LaPolla. 2012. The importance of using multiple approaches for identifying emerging inv asive species: the case of the rasberry crazy ant in the United States. PLos ONE 7 : e45314. Gusmo, F. A., N. Sibinel, D. Bajomi, and A. E. C. Campos Farinha. 2008. Tramp ant species control with methoprene baits, pp 71 75. In P roceedings, Proceedings of the Sixth International Conference on Urban Pest, 13 16 July 2008, Budapest, Hungary. OOK Predd Kft, Veszprem, Hungary. Harmon, K. 2009. Honeyb ees face new threat in Texas: "c razy" ants. (http://www.scientificamerican.com/blog/p ost.cfm?id=honeybees face new threat in texas 2009 08 07). Heller, N. E. 2005. Colony structure, climate, and spread in invasive A rgentine ant. Ph.D dissertation, Stanford University, Stanford. Heller, N. E. and D. M. Gordon. 2006. Seasonal spatial dyna mics and cause of nest movement in colonies of the invasive Argentine ant ( Linepithema humile ). Ecol. Entomol. 31: 499 510. Heller, N. E., N. J. Sanders, J. W. Shors, and D.M. Gordon. 2008. Rainfall facilitates the spread, and time alters the impact, of th e in vasive Argentine ant. Oecologia 155: 385 395. Hlldobler, B. and E. O. Wilson. 1990. The ants. Belknap Press of Harvard Univ ersity Press, Cambridge MA Holway, D. A. and T. J. Case. 2001. Effects of colony level variation on competitive ability in the invasive Argentine ant. Anim. Behav. 61: 1181 1192. 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 Hopper, L. M. and M. K. Rust. 1997. Food p reference and patterns of foraging activity of the southern fire ant (Hymenoptera: Formicidae). Ann. Entomol. Soc. Am. 90: 246 253.
103 Hooper Bui, L. M., R. Strecker, X. Chen, D. Aguillard, and A. Miller. 2010. Super colonies of crazy ants in Louisiana pp. 13 16. In Proceedings 2010 Imported Fire Ant Conference, 19 22 April 2010, Little Rock, AR. (http://www.extension.org/sites/default/files/w/c/c4/2010_Annual_IFA_Conferenc e.pdf ). Horton, J. R. 1918. The Argentine ant in relation to citrus groves. Bulletin 647 U.S. Department of Agriculture, Washington, D.C. Hu, P. S. and T. R. Reuscher. 2004. Summary of travel trends: 2001 national household travel s urvey. Dep t. of Energy Doc. ORNL/TM 2004/297. ( http://nhts.ornl.gov/2001/pub/stt.pdf ). Human, K. G. and D. M. Gordon. 1996. Exploitation and interference competition between the invasive Argentine ant, Linepithema humile an d native ant species. Oecologia 105: 405 412. Ingram, K.K. 2002. Plasticity in queen number and social structure in the invasive Argentine ant ( Linepithema humile ). Evolution 56: 2008 2016. Kabashima, J.N., L. Greenberg, M. K. Rust, and T. D. Paine. 2007. Aggressive interactions between Solenops is invicta and Linepithema humile (Hymenoptera: Formicidae) under laboratory conditions. J. Econ. Entomol. 100: 148 154. 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: Formicid ae) of peninsular Florida. Fla. Entomol. 78: 109 118. Krushelnycky, P. D., D. A. Holway, and E. G. LeBrun. 2010. Invasion processes and causes of success, pp 245 260. In L. Lach, C. L. Parr, and K. L. Abbott (eds.) Ant ecology. Oxford Biology New York, NY Krushelnycky, P. D., S. M. Joe, A. C. Medeiros, C. C. Daehler, and L. L. Loope. 2005. The role of abiotic conditions in shaping the long term patterns of a high elevation Argentine ant invasion. Divers. Distrib. 11: 319 331. Kruchelnycky, P. D., L. L Loope, and S. M. Joe. 2004. Limiting spread of unicolonial invasive insect and characterization of seasonal patterns of range expansion. Biol. Invas ions 6: 47 57. Lanchster, F. W. 1916. Aircraft in warfare. Appleton, New York, NY LaPolla J. S., S. G. B randy, and S. O. Shattuck. 2010. Phylogeny and taxonomy of the Prenolepis genus group of ants (Hymenoptera: Formicidae). Syst. Entomol. 35: 118 131.
104 LeBrun, E. G. J. C. Abbott, and L. E. Gilbert. 2012. Adding insult to injury in Texas grasslands: c razy a nt extirpates imported fire ant and further reduces native ant and arthropod diversity. In Program, 97 th Annual Ecological Society of America Meeting, 5 10 August 2012. ( http://eco.confex.com/eco/2012/webprogrampreliminary/Paper36549.html ) Leung, B., D. M. Lodge, D. Finnoff, J. F. Shogren, M.A. Lewis, and G. Lamberti. 2002. An ounce of prevention or a pound of cure: bioeconomic risk analysis of invasive species. Proc. Biol Sci. 269: 2407 2413. Ligon R. A., L. Siefferman, and G. E. Hill. 2011. Invasive fi re ants reduce reproductive success and alter the reproductive strategies of a nat ive vertebrate insectivore. PLos ONE 6: e22578. Lockwood, J. L., M. F. Hoopes, and M. P. Marchetti. 2007. Invasion Ecology. Bl ackwell Publishing, Malden, MA MacGown, J. A 2012. Ant s (Formicidae) of the southern United States ( http://www.mississippientomologicalmuseum.org.msstate.edu/R esearchtaxapag es/Formicidaepages/genericpages/Nylanderia_fulva.htm ). MacGown, J. A. and B. Layton. 2010. The invasive raspberry crazy ant, Nylanderia sp. near pubens (Hymenoptera: Formicidae), re ported in Mississippi. Midsouth Entomo l 3: 44 47. Markin G. P. 1970. The seasonal life cycle of the Argentine ant in a California citrus grove. J. Econ. Entomol. 63: 740 744. McGlynn, T.P. 1996. Non native ants are smaller than related native ants. Am. Nat. 154; 690 699. McGlynn, T. P. 1999. The worldwide transfer of ants: geographical distribution and ecological invasions. J. Biogeogr. 26: 535 548. Meyers J. M. 2008. Identification, distribution and control of an invasive pest ant, Paratrechina sp. (Hymenoptera: Formicidae), in Texas. Ph.D. dissertation, Texas A&M University, College Station. Morgan, J. W. 2011. Hairy crazy ants m ake L ouisiana their new h ome ( http://www.lsuagcenter.com/news_archive/2011/september/headline_news/Hair y crazy ants make La their new home.htm ) Nester, P. R. and T. Ras berry. 2011. Response of Nylnaderia sp. near fulva population to single application of Esteem 0.86% EC insect growth regulator, pp 103 106. In Proceedings 2011 Imported Fire Ant Conference, 4 7 April 2011, Galveston, TX. (http://www.extension.org/sites/default /files/2011_fa_conf_proceedings_sep21.pd f)
105 Pagad, S. 2011. Global invasive species d atabase: Nylanderia (= Paratrechina ) pubens (insect). ( http://issg.org/database/species/ecology.asp?si=1553&fr=1&sts=&lang=EN ) Pereira, R. M. and P. G. Koehler. 2007. Ca ribbean crazy ant s moving to Florida heartland. Florida PestPro. 3: 18 20. Plowes, N.J.R., and E.S. Adams. 2005. mortality during battles of the fire ant Solenopsis invicta Proc. R. Soc. B. 272: 1809 1814. Po rter, S. D. and D. A. Savignano. 1990. Invasion of polygyne fire ants decimates native ants and disru pts arthropod community. Ecolo gy 71: 2095 2106. Rust, M K. D. A. Reierson, E. Paine, and L. J. Blum. 2000 Seasonal activity and bait preferences of the Argentine ant (Hymenoptera: Formicidae). J Agric Urban Entomol 17: 201 2 12. Say Piau, L. and L. Chow Yang. 2003. Nestmate recognition and intercolonaial aggression in the crazy ant, Paratrechina longicornis (Latreille) (Hymenopte ra: Formicicidae). Soc iobiology 41: 295 305. SAS Institute. 2012 a SAS 10 software. SAS Institute Inc, Cary, NC. SAS Institute. 2012 b JMP statistical package SAS Institute Inc, Cary, NC. Sax, D. F., J. J. Stachowicz, J. H. Brown, J. F. Bruno, M. N. Dawso, S. D. Gaines R. K Grosber, A. Hastings, R. D. Holt, M. M. Mayfield, M. R. Rice. 2007. Ecological and evolutionary insights from s pecies invasions. Trends Ecol. Evol. 22 : 465 471. Schilman, P. E., L. R. B. Lighton, and D. A. Holway. 2005. Respiratory an d cuticular water loss in insects with continuous gas exchange: comparison across five ant species. J. Insect Physiol. 51: 1295 1305. Scott, J. M. 2012. Optimization of bait components for Nylanderia pubens (Forel). M.S. thesis, University of Florida, Gain esville. Seeley, T. D. and B. Heinrich. 1981. Regulation of temperature in nests of social insects, pp. 159 234. In B. Heinrich (ed.), Insect t hermoregulation. John Wiley and Sons, Inc., New York NY Silverman, J. and R. J. Brightwell. 2008. The Arge ntine ant: challenges in managing an invasive unicolonial pest. Annu. Rev. Entomol. 53: 231 252. Sorensen, A. A., T. M. Busch, and S. B. Vinson. 1985. Control of food influx by temporal subcastes in the fire ant, Solenopsis invicta Behav. Ecol. Sociobiol. 17: 191 198.
106 Stradling, D. J. 1978. Food and feeding habits of ants pp 81 106. In M. V. Brian (ed. ) Production ecology of ants and t ermites. Cambridge University Press, Cambridge MA Suarez, A. V., D. A. Holway, D. Liang, N. D. Tsutsui, and T. J. Case. 2002. Spatiotemporal patterns of intraspecific aggression in the invasive Argentine ant. Anim. Behav. 64. 697 708. Trager J. C. 1984. A revision of the genus Paratrechina (Hymenoptera: Formicidae) of the continen tal United States. Sociobiology 9: 51 162. Traniello, J. F. and S. N. Beshers. 1991. Maximization of foraging efficiency and resource defense by group retrieval in the ant Formica schaufussi Behav Ecol 29: 283 289. Tripp, J. M., D. R. Suiter, G. W. Bennett, J. H. Klotz, and B. L. Reid. 2000. Evaluation of control measures for black carpenter ant (Hymenoptera: Formicidae). J. Econ. Entomol. 93: 1493 1497. Tschinkel, W. R. 2006. The fire ants Harvard University Press, Cambridge, MA Tsutsui, N. D. and T. J. Case 2001 Population genet ics and colony structure of the Argentine ant ( Linepithema humile ) in its native and introduced ranges. Evolution 55 : 976 985 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 Relationships among native and introduced populations of the Argentine ant ( Linepithema humile ) and the sourc e of introduced populations. Mo l Ecol. 10: 2151 2161. Ulloa, P. C., J. Bustos, R. C. Aldana, M. L. Baena. 2000. Control de la hormiga loca, Paratrechina fulva (Hymenoptera: Formicidae), con cebos toxicos en la Reserva Natural Laguna de Son so (Valle, Colombia). Rev. Colomb. Entomol. 26: 151 156. Valles, S. M., D. H. Oi, F. Yu, X. X. Tan, and E. A. Buss. 2012. Metatranscription and pyrosequencing facilitate discovery of potential viral natural enemies of the invasive Caribbean crazy ant, Nyl anderia fulva PLo s ONE. 7: e31828. Vinson, S. B. 1997. Invasion of the red imported fire ant (Hymenoptera: Formicidae): spread, biology, and impact. Am. Entomol. 43 : 23 39. Vitousek, P. M., C. L. Loope, and R. Westbrooks. 1996. Biological invasions as global environmental change. Am. Sci. 84: 468 478.
107 Warner, J. and R. H. Scheffrahn. 2010. Caribbean crazy ant, Nylanderia pubens Forel. UF/IFAS Featured Creatures. ENY 284. Gainesville, Florida. University of Florida. ( http://entnemdept.ufl.edu/creatures/urban/ants/caribbean_crazy_ant.htm ) Weeks, R. D., L. T. Wilson, S B. Vinson, and W. D. Jones. 2004. Flow of carbohydrates, lipids, and protein among colo nies of polygyne red imported fire ants, Solenopsis invicta (Hymenoptera: Formicidae). Ann. Entomol. Soc. Am. 97: 105 110. Wetterer, J. K. and J. L. W. Keularts. 2008. Population explosion of the hairy crazy ant, Paratrechina pubans (Forel), on St. Croix U.S. Virgin Islands (Hymenoptera: Formicidae). Fla. Entomol. 91: 423 427. Williamson, M. 1996. Biological Invasions. Chapman & Hall, London England Wilson, E. O. 1970. Competitive and aggressive behavior. Soc. Sci. Inform. 9:123 154. Whitehouse, M E A., and K. Jaff. 1996. Ant wars: strategies, territory and nest defense in leaf cutting ant Atta laevigata Anim. Behav. 51: 1207 1217. Zenner Polania, I. 1990 a Paratrechina (Nylanderia) fulva (Mayr), in Colomb ia, pp. 290 297. In R. K. Vander Meer, K. Jaffe, and A. Cedeno (eds.), Applied myrmecology. Westview Press, Boulder, CO Zenner Polania, I. 1990 b Paratrechina (Nylanderia) fulva (Mayr) in Colom bia, pp. 701 70 7. In R. K. Vander Meer, K. Jaffe, and A. Cedeno (eds.), Applied myrmecology. Westview Press, Boulder, CO Zenner Polania, I. 1994. Impact of Paratrechina fulva on other ant species, pp. 120 131. In : D.F. Williams (ed.), Exotic ants. Biology, impact, and control of int roduced species Westview Press. Boulder, CO. Zenner Polania, I. and N. Ruiz Bolanos. 1985. Habotos alimenticos y relaciones hominga Nylanderia fulva con otros artropodos. Rev. Colomb. Entomol. 11: 3 10. Zhao, L., J. Chen, W. Jo nes, and D. Oi. 2012. Molecular comparisons suggest caribb ean crazy ant from Florida and r asberry crazy ant from Texas (Hymenoptera: Formicidae: Nylanderia) are the same species. Environ Entomol 41: 1008 1018.
108 BIOGRAPHICAL SKETCH Stephanie Hill, daugh ter of Paula L arrick, was born and raised in southeast Ohio. She graduated from Buckeye Trail High School in 2003 and went on to continue her studies at Hiram College where she earned a Bachelor of Arts in 2007. Stephanie entered the graduate program at t he Universit y of Florida in the entomology and nematology d epartment in the fall 2007. s student under the supervision of Dr. Roxanne Connelly. In 2009, she received her m entomology and n ematology. Stephanie stayed at the Uni versity of Florida, under the direction of Dr. Phil Koehler, to complete her doctorate in entomology and nematology