i-

DISTURBANCE OF COMMUNITY STRUCTURE IN

CENTRAL TEXAS BY THE RED IMPORTED

FIRE ANT, SOLENOPSIS INVICTA

by GERARDO RAFAEL CAMILO, B.S. A THESIS

IN ENTOMOLOGY

Submitted to the Graduate Faculty of Texas Tech University in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE

Approved

August, 1990 ^r^a ACKNOWLEDGEMENTS No C^) (* ^l express my deepest gratitude to Dr. Sherman A. Phillips, Jr., for giving me the opportunity to work under his direction. I am thankful also to Dr. Harlan G. Thorvilson and Dr. Michael R. Willig for serving on my thesis committee. Thanks are expressed to Drs. George C. and Jeanette N. Wheeler, Dr. James C. Trager, and Mr. James C. Cokendolpher for the identification of specimens. Thanks are also extended to Kathy Phillips and the graduate students of the Departments of Agronomy, Horticulture, and Entomology, and Biological Sciences, who provided help with the field work and helpful comments about my research. Many people helped me to reach this point. Dr. Bob Waide and Dr. Willig provided encouragement to continue graduate studies. Dr. Rip Phillips always knew when to pull the breaks on me, and other graduate students provided helpful comments on my research. I am deeply thankful to all of you. Finally, I want to thank my family, especially my mother, Josefina Rivera, for being so understanding, loving, and patient with me. They provided the extra encouragement needed. Thanks.

11 TABLE OF CONTENTS

ACKNOWLEDGEMENTS ii ABSTRACT iv LIST OF TABLES v LIST OF FIGURES vi CHAPTER I. INTRODUCTION 1 Community and Competition 1 Dominance and Diversity 2 Ant Communities 3 Competitive Replacement by the Red Imported Fire Ant 5 Statement of Study Objectives 8 II. MATERIALS AND METHODS 9 Study Area 9 Sampling Techniques 9 Cluster Analysis 11 Principal Components Analysis 11 m. RESULTS AND DISCUSSION 13 IV. CONCLUSIONS 25 LITERATURE CITED 2 8

111 ABSTRACT

The introduction of a competitor into an established community provides an excellent opportunity to study the factors that maintain the dominance, diversity, and interactions of that community. The introduction of the red imported fire ant, Solenopsis invicta Buren, into North America provides such an opportunity. A transect, with pitfall traps as sampling units, was established in central Texas, U.S.A., from areas uninfested to heavily infested by S.. invicta. Cluster analysis of the composition at particular localities depicts four distinct ant assemblages, two without and two with S.. invicta. A fifth aggregation of species was found and consisted almost entirely of S.. invicta. Principal components analysis reveals that disturbance of the habitat, mostly by humans, is the main factor controlling ant diversity. More specifically, increasing densities of the red imported fire ant were negatively correlated with species diversity. Conomyrma insana (McCook) can coexist with S.. invicta. probably because direct competitors or predators have been eliminated by S.. invicta. of the granivorous guild, mostly Pheidole spp., that occur with the red imported fire ant are being displaced, whereas more generalistic ants, like Monomorium minimum (Buckley) and Forelius pruinosus (Roger), coexist with S.. invicta.

IV LIST OF TABLES

1. Species assemblages in central Texas, U.S.A., as detected by cluster analysis 17

2. Principal components analysis scores of plots, percent of variation accounted for by each component (R^), and cumulative variation (IR2). 21

3. Correlation of species importance values to densities of Solenopsis invicta 24 LIST OF FIGURES

1. Systematic affiliation and sample sizes of ant species collected from August 1986 to October 1987 14

2. Dendrogram of ant species assemblages located in Kerr and Kimble Counties, Texas, U. S. A. Plots K and L were excluded from the analysis. 16

3. Principal components analysis of plots by species. Squares in the 1-2 axes plane are the geometrical projections of the plots in the ordination space. Axis 1 is correlated with disturbance, axis 2 with temperature, and axis 4 with density of S.. invicta. 20

VI CHAPTER I INTRODUCTION

Community and Competition The concept of community has generated a long series of arguments throughout the history of ecology. One school of thought is that the community is a "super-organism" in which the total is more than the sum of its parts (Clements 1916). Central to this theory is the view that the history of past interactions is the driving force that gives rise to the present communities (Odum 1971). The diversity of a community, that is, which species are present and in what numbers, is considered to be influenced mostly by competition. This community theory is based upon Cause's application of Malthusian population theory to interspecific ecological interactions (Hardin 1960). If two populations of different species, A and B, occupy the same niche and live in the same environment with limited resources, and A reproduces faster than B, then A will displace B, and B will become extinct (Miller 1967). This interpretation implies that interspecific competition in the past caused niche divergence, resulting in differences between species that are seen in contemporary communities (Schoener 1974). Yet other kinds of interactions, like predation and parasitism, were not considered as impo lant as competition in structuring communities (Seifert 1984). The other school of thought considers a community to be a group of species occupying a common area. The diversity of the community is a function of autecological rather than synecological processes (Strong et al. 1984). Such communities would not be holistic entities, but simply a collection of relatively autonomous populations occupying the same area at the same time. Community diversity is not affected just by competition but also by other biotic interactions and abiotic factors such as weather. Interactions like parasitism, predation and mutualism, can be the main factors responsible for the observed community structure (Strong 1984, Seifert 1984). The presumption that competition in the past is the general cause for species diversity and distribution in communities has been referred to as "the ghost of competition past" (Council 1980). Interspecific competition may not be common in communities because other natural factors (i.e., stochastic environmental fluctuations or parasites) frequently intervene to maintain densities below resource-defined equilibria (Strong 1984, Wiens 1977). Competition is more likely to affect species under moderately harsh environmental conditions, as opposed to either benign or extremely harsh environmental conditions (Council 1980). Under harsh conditions, such as droughts, populations of potential competitors may be reduced below the densities at which they would compete. Harshness, in this case, does not imply a reduction in resources. On the other hand, under benign conditions, predators, parasites, and pathogens may be more effective in maintaining population densities of prey or hosts below the level at which competition might occur. In addition, the intensity and duration of competition may not be enough as to be critical in the structuring of the community (Wiens 1977).

Dominance and Diversity The dominance-diversity relationships in communities have long been of interest to ecologists (Brown & Gibson 1983, Ricklefs 1987). A dominant species in a community is one that largely controls energy flow and strongly affects the environment of other species (Odum 1971). Dominance of a species in a community may be expressed by high density, high frequency of occurrence, or appreciable influence on the densities and occurrence of other species (Werner 1984). The concept of diversity involves several other ecological concepts. Two of these concepts are species richness, or the number of species present in the community, and equitability, or the distribution of relative importance values among the species in the community (Peet 1974). The two terms. 3 dominance and diversity, are complements of each other. Dominance relationships are manifestations of the "sorting out" of species in the community, as dominance develops, diversity is established (Whittaker 1965).

Ant Communities The high diversity and specialized behavior of ants make them important members of almost every terrestrial community (Carroll & Janzen 1973, Levings & Traniello 1981). Ants occupy all consumer trophic levels and exhibit all types of symbiotic relationships which have been demonstrated (Carroll & Janzen 1973). The structure of ant communities, as with any other community, depends on which ant species are present, in what numbers, and the type and intensity of the biotic interactions occurring at any given time (Levins et al. 1973, Heatwole & Levins 1972). However, the distribution and the number of species in a community are also a function of other factors. Species distributions have been found to be affected by climate, nutritional requirements, predators, and natural dispersal barriers (Culver 1974, Torres 1984a). Foraging ecology of ants, their highly aggressive behavior, and territoriality have a profound effect on the structure of ant communities. Levings and Traniello (1981) discussed the importance of spatial arrangement of ant nests and their territoriality in relation to the structure of ant communities. They found that ant species are regularly distributed with respect to conspecific and other sympatric species, and that these distributional phenomena are an outcome of high overlap of food utilization. In most of the species that they studied, ants with general food habits tended to have more efficient recruiting systems and most aggressive behavior. Departures from this pattern are suggested as a measure of interspecific competition. However, not enough evidence has been collected to test this hypothesis in a unequivocal way. The foraging habits of ant species that reach high population densities are also crucial factors affecting community structure. The foraging activities of the army ant, Eciton burchelli Mayr, have a significant effect on the structure of the community (Otis et al. 1986). This species is a swarm-raiding army ant, and because of its large daily food intake and large numbers (Wilson 1971), both arthropod abundance and arthropod taxa were reduced following the passage of foraging army ants (Otis et al. 1986). Factors other than those previously mentioned are influential in the shaping of ant communities. The occurrence of rare ant species and stochastic events influence the coexistence pattern in ant communities (Levins & Culver 1971). Microclimate variability, large overlap of food habits, and high levels of aggression are crucial factors in maintaining the diversity of ant communities in cultivated tropical areas and rain forests in Puerto Rico (Torres 1984b). The high number of coexisting species is attributed to non- equilibrial processes because the total ant densities are always below the environmental carrying capacity. Stochastic events, related to the random probability of finding food, are also major factors allowing ant species to coexist in tropical lowland forests (Torres 1984a). In woodlands and heath sites in Australia, the number of ant taxa is correlated positively with structural complexity of the vegetation (Andersen 1986a). The diversity of microclimates and variable weather conditions account for most variation in mesic ant communities. Competition is not considered a major influential factor, although some species compete for food resources. However, these events were few and of short duration. Arid area communities are more influenced by competition than those from mesic regions (Andersen 1986b). Xeric region communities are less diverse, and more tightly packed than mesic region communities, because resources, habitat, and food are less diverse and abundant. The fact that species in a community compete does not mean that the community is shaped by competition (Moulton & Pimm 1987). In ants, competition has been demonstrated to be both 5 intensive and extensive (Levings & Traniello 1981). Ant community succession in disturbed coastal heathland in Australia is determined by competition (Fox et al. 1985). As succession proceeds, a species of meat ant, Iridomyrmex sp., is replaced by another Iridomyrmex sp. as the dominant ant. These ants compete by defending their territories. Individuals of each species actively defend territorial space by constantly patrolling their territorial boundaries. These boundaries are breached when the defenders are eliminated. Fox et al. (1985) confirmed that interspecific competition plays an important role in maintaining the observed mosaic pattern of ant distribution, and thus the structure of the community. To summarize, a community is a group of species populations sharing a defined area in time and space. Diversity appears to be more related to each individual species interaction with abiotic factors (i.e., autecology) than to biotic relationships among the species (i.e., synecology). Only when the environment is stable enough to allow populations to reach densities defined by the carrying capacity (i.e., resource defined equilibria [Wiens 1977]), can competition take place on a scale influential enough to be crucial in the structuring of the community. This stability will take place in only moderately harsh environments, because in extremely harsh conditions populations will be depleted by direct action of the environment. Under extremely benign conditions, predators, parasites, and pathogens are more efficient and prevent population densities from reaching resource defined equilibria (Council 1980).

Competitive Replacement by the Red Imported Fire Ant Since the introduction of the red imported fire ant, Solenopsis invicta Buren, to the southern U.S.A., it has been considered a threat to the native ant fauna (Lofgren et al. 1975, Wilson et al. 1971). The rapid spread of this species (Lofgren et al. 1975), its high reproductive potential (Glancey et al. 1976), and its aggressive behavior (Bhatkar et al. 1972) are among the factors that 6 contribute to the successful establishment of S.. invicta. Foraging and recruitment abilities have also been noted as major factors in the success of this species (Jones 1985, Kidd et al. 1985, Phillips et al. 1986). An excellent opportunity for study is provided by the introduction of a competitor into an already established community (Schoener 1983). The red imported fire ant provides such an opportunity. An introduced species population would not be affected by parasitoids and pathogens as is the local fauna. Therefore, the neo-Malthusian forces associated with classical competition theory do not affect the North American populations of S.. invicta as effectively as they do populations in South America. Moreover, insufficient time has elapsed to evolve differences for reduction of competition. A review of many case histories of introduced species by Simberloff (1981) showed that little or no competitive effect was expressed by the native fauna. In contrast, the literature on fire ants indicate that the local extinction of native fire ants, as well as other ants, after the appearance of S.. invicta has occurred repeatedly. The occurrence of S.. invicta is associated with the extinction or displacement of other fire ants (Solenopsis spp.) in the United States. Wilson and Brown (1958) considered the substitution of the "dark phase" of the imported fire ant for a "light phase" as the evolution of character displacement caused by a pleiotropic polygene interaction in the founding population of imported fire ants in Alabama. This color phase change is now considered to represent the substitution of one species, S.. richteri Forel, by another, S. invicta (Buren et al. 1974). Prior to Buren's (1972) taxonomic revision, studies conducted on imported fire ants were probably inaccurate with respect to identification of the species, because the complexity of the S.. saevissima group (Baroni Urbani & Kannowski 1974). After the systematic revision of the imported fire ants, several papers dealing with competition and community disturbance by S_. invicta were published. Baroni Urbani and 7 Kannowski (1974) studied demographic parameters of S.. invicta and interspecific competition in a pasture in Louisiana and found that only two other species of ants were collected frequently, Monomorium minimum (Buckley) and Paratechina arenivaga (Wheeler). These ants coexisted with S.. invicta. although they used the same food resources. The other species of fire ant in the area, Solenopsis xyloni McCook, was found only in shaded areas beneath large trees. The behavior of this species precludes confrontation with S.. invicta which prefers unshaded areas. The replacement of ants by S.. invicta has been observed on several occasions. In Alabama, the once dominant Argentine ant, Iridomyrmex humilis (Mayr), was displaced as the dominant ant by S.. invicta (Glancey et al. 1976). Solenopsis xyloni and S.. geminata (F.) were not detected in the bait station survey, although they previously had been found in the study area. The authors suggested that S.. invicta probably displaced these species from the area. In Florida soybean fields, S.. geminata was one of the first species to disappear following S.. invicta invasion (Whitcomb et al. 1972). They hypothesized that this rapid disappearance occurred because S.. invicta "...is particularly aggressive to other ant species on its own trophic level." The outcome of competition between S.. invicta and S.. geminata is mediated by environmental disturbance (Tschinkel 1988). Tschinkel (1986) described S.. invicta as a weed species, because weeds are opportunists and are adapted for taking advantage of ecologically disturbed habitats. The characteristics attributed to S.. invicta for consideration as a weed species are its association with disturbed habitats, mostly of anthropogenic origin (Banks et al. 1985, Summerlin 1977, Tschinkel 1988), high reproductive rate (Tschinkel 1986), and effective means of dispersal and colonization. In northern Florida pinelands S.. invicta is found only in mowed road margins and around ponds, whereas S_. geminata replaces S.. invicta in more stable areas (Tschinkel 1988). Porter et al. (1988) quantified the competitive replacement of native fire ants by the red imported fire ant. The invasion of fields near Austin, Texas, by S.. invicta was monitored, and S.. invicta 8 replaced S.. geminata six-fold. In a period of three years following invasion, the number of colonies of S.. invicta was six times that of S.. geminata before the arrival of S.. invicta. These results suggest strong asymmetrical competition and a restructuring of the ant community following invasion by S.. invicta. The literature on competitive replacement after S.. invicta establishment suggests a restructuring of the ant community in specific, and the arthropod community in general. Unfortunately, very few studies are of sufficient duration, or were designed to quantify, rather than qualify, the effect of S.. invicta on the structure of the ant community.

Statement of Study Objectives Several investigators have recommended studies to determine the ecological relationships between S.. invicta and other ant species in North America (Maxwell et al. 1982). Among their recommendations are: 1) to determine the population dynamics of S.. invicta and other related ant species in disturbed and undisturbed habitats, and 2) to determine the extent of interspecific competition on S.- invicta populations found at the periphery of infested areas. Based on these suggestions, the objectives of the present study are: 1) To determine the composition of ant assemblages in areas heavily infested, moderately infested, and uninfested by S.. invicta. 2) to determine which factors may control the diversity of the ant community, and 3) to determine the effect of S.. invicta on the densities of other ant species with which it occurs. CHAPTER II MATERIALS AND METHODS

Study Area The study area is located on the Edwards Plateau in central Texas, within Kimble and Kerr counties. In general, the topography is rough, consists of eroded limestone with little top soil which provides good drainage. Elevation is ca. 300 m, and rainfall ranges from 355 mm to 863 mm annually. Temperatures range from 11°C in winter to 40° C in summer. The vegetation is dominated in the overstory by live oak, Ouercus virginiana Mill., shinnery oak, Q.. Harvardii Rydb., juniper, Juniperus Ashei Buchh., and mesquite, Prosopis glandulosa Torr. The understory consists of grasses (Correll & Johnston 1970), mainly switchgrass, Panicum virgatum L., several species of bluegrass, Poa spp., and grama, Bouteloua spp.

Sampling Techniques A transect was established parallel to interstate highway 10, from areas uninfested to heavily infested by S.. invicta. The transect consisted of 12 plots, each separated by ca. 4.8 km. A plot comprised 24 pitfall traps, in two parallel rows of 12, each row 1 m apart. All traps within a row were also 1 m apart. A single trap consisted of a plastic cup (440 ml) embedded in the soil with the lip of the cup flush with the ground surface. A second plastic cup was inserted into the first, and soil was piled around the cups to level the ground to the lip of the second cup. By so doing, cups were set in place without excavating a new hole each time sampling occurred, thus minimizing the "digging-in" effect (Greenslade 1973). Ethylene glycol was used as the killing and preservative agent (Adis 1979). To prevent larger, unwanted and vertebrates from becoming trapped, roofing shingles (ca. 10 x 10 cm) were placed over each trap with a space of approximately 1 cm between the trap and the cover. This procedure also prevented rain from flooding the traps. 10 Communities were monitored once each month, from August 1986 through November 1986, and again from March 1987 through October 1987. However, the samples from November 1986 and March 1987 were excluded from the analyses thus minimizing variation resulting from seasonal fluctuations (Samways 1983). Because two plots were almost exclusively infested with S.. invicta (plots K and L), they were excluded from all analyses. Voucher specimens were deposited in the Entomology Collection of the Department of Agronomy, Horticulture, and Entomology, Texas Tech University, Lubbock, Texas. Catalog numbers of the specimens are 7576 to 7583, and 7746 to 7767. Traps were in place for 72 hrs each time the transect was monitored. This time period facilitated maximum capture of ants (Jansen & Metz 1979), whereas the time between monthly sampling provided an adjustment period in which no ants were removed from the habitat (Samways 1983). The importance value (IVxik) of each species (x) was calculated for each plot (i), as:

IVxik = Xik (3N)-l(FN) ; where X is the number of individuals of species x collected in a specific plot, i, at a specific sampling period, k. The constant, 3, accounts for the number of days the traps were in place, so the IV is given in days; N is the number of traps in a plot; and F is the number of traps in which individuals of species x were found. Plot values were determined using ant species IVs, excluding S.. invicta. as the classificatory variables (Pielou 1984). The dissimilarity index used among the plots was the Euclidean distances in multi-species space (Pielou 1984).

Cluster Analysis A cluster analysis, based upon the centroid method, was performed to identify distinct ant assemblages (Pielou 1984). The total monthly data were pooled for this analysis. The statistical 11 analysis was performed using procedure CLUSTER (SAS Institute Inc. 1985). Results generated from the distance matrix were graphed as a dendrogram generated by procedure TREE (SAS Institute Inc. 1985). Because the definition of a cluster is subjective (Pielou 1984), I decided that a cluster included those plots that shared at least 80% of similarity, as expressed by the species IV. Plots that were not in the same cluster usually shared less than 10%. Because the distribution of ant species through space and time is not normally distributed (Culver 1974, Levings & Traniello 1981), the assumption of a simple correlation, that the two variables being correlated are bivariate normal, is not met (Sokal & Rohlf 1981). For this reason Kendall's rank correlation (Sokal & Rohlf 1981) was used to determine the degree of association between species richness and densities of S.. invicta.

Principal Components Analysis Principal components analysis (PCA) is an ordination method that produces a numerical and graphical representation of a unit response (either plots or species) in multiple dimensions and represents it in fewer dimensions. In this way, the principal gradients of variation may be delineated, as well as the response of the units to these components (Digby & Kempton 1987). The eigenvalue matrix for species and plots was generated by procedures PRINCOMP and CLUSTER (SAS Institute Inc. 1985). The densities of S.. invicta were excluded from the PCA to identify the effects of this species on the native ant fauna. The average score of each plot was used for constructing the ordination of the communities. Axes with eigenvalues less than 1 were considered unimportant. Species that were uncommon, trapped once, or with less than 10 individuals were excluded from the analysis, but were included in the species list of their respective assemblages. Construction of the axes was terminated during PCA when 99% of the original variation was included in the previously obtained ordination axes. 12 Disturbance of the habitats was quantified by assigning a ranked value depending on the level and frequency of the disturbance. If the habitat was mowed at least once a month it was assigned a disturbance rank of 10. If the habitat was not disturbed in the month between sampling periods, it was assigned a rank value of 5, whereas a habitat that was never mowed was assigned a value of 1. The association between PCA scores for plots and indices of disturbance were evaluated by Kendall's rank correlation (Sokal & Rohlf 1981), whereas the association between PCA scores of plots and temperature or densities of S.. invicta were each evaluated separately using Pearson's product-moment correlation (Sokal & Rohlf 1981). The mean of the maximum temperature for the three- day sampling period was used as the temperature index for the ordination. Importance values of each species occurring with S.. invicta were correlated (Pearson's product-moment correlation, Sokal & Rohlf 1981) through time to S.. invicta densities. This test identifies which ants are increasing or decreasing in importance in the community as the density of S.. invicta fluctuated. CHAPTER m RESULTS AND DISCUSSION

A total of 7,697 ant specimens was collected, representing five subfamilies and 35 species (Fig. 1). Solenopsis invicta accounted for ca. 42% of the total of specimens collected. Through cluster analysis, four distinct assemblages were identified (Fig. 2). The distance between clusters is expressed along the horizontal axis. The first bifurcation in the dendrogram separates sites occurring in undisturbed habitats from those occurring in disturbed habitats. The second division separates those sites that contain S.. invicta from those that do not. The final division separates sites that have high and low densities of S.. invicta. Density of S.. invicta is negatively correlated with species richness and thus diversity (Kendall's x=0.62; n=27; P<0.01). Twentyone species of ant occur in the undisturbed, uninfested plot (Table 1), and 14 in the disturbed, uninfested plots. Comparison of the previous assemblages with the assemblages containing S.. invicta (Table 1) suggests a pattern of species number reduction with increased densities of S.. invicta: nine species in the assemblage with low density of S.. invicta. and five in the high density assemblage. Species assemblage one (Table 1) occurs in an undisturbed habitat (not mowed or grazed for at least 8 years). It has the greatest diversity and contains ecologically specialized taxa, such as Pachycondvla harpax (F.), Labidus coccus (Latreille), Leptothorax terragina Wheeler, and Tetramorium spinosus (Wheeler). These species were found throughout most of the sampling periods (summer and fall 1986; spring, summer, and fall 1987). Species assemblages two, three, and four are in disturbed habitats (mowed or grazed several times each year). Species assemblage two (Table 1) is characterized by high diversity, and some specialized ants such as Leptogenvs elongata (Buckley) and Neivamvrmex nigriscens (Cresson). However, these species were found sporadically throughout the sampling periods. Differences 13 14

Taxa total

PONERINAE Pachycondvla harpax (F.) 56 Leptogenvs elongata (Buckley) 14

ECrrONINAE Neivamvrmex nigrescens (Cresson) 485 Labidus coecus (Latreille) 312

MYRMICINAE Solenopsis invicta Buren 3,225 Pogonomvrmcx barbatus (Smith) 408 Monomorium minimum (Buckley) 398 Solenopsis xvloni McCook 389 Solenopsis geminata (F.) 265

Tetramorium spinosus (Wheeler) 173 Pheidole crassicornis tetra Wheeler 142 Pheidole tepicana Pergande 108 Monomorium pharaonis (L.) 76 Pheidole hvatti Emery 39

Solenopsis miplorhoptrum) sp. 2 23 Atta texana (Buckley) 19 Crematogaster laeviuscula Mayr 13 Pheidole sp. 1 12 Pogonomvrmex imherbiculus (Wheeler) 11

Solenopsis aurea Wheeler 9

FIGURE 1. Systematic affiliation and sample sizes of ant species collected from August 1986 to October 1987. 15 Taxa total

MYRMICINAE Crematogaster punctulata Emery 6 Pheidole sp. 2 3 Leptothorax sp. 2 Leptothorax terrigena Wheeler 1 Solenopsis (Diplorhoptrum) sp. 1 1

DOLYCHODERINAE Forelius pruinosus (Roger) 482 Conomvrma insana (Buckley) 163 Forelius foetidus (Buckley) 152 Tapinoma sesile (Say) 36 Conomyrma flava (McCook) 17

FORMICINAE Paratrechina terricola/vividula ^ 582 Camponotus discolor (Buckley) 9 Formica sp. 7 Brachvmyrmex depilis Emery 3 Camponotus sp. 2

^ Unable to differentiate between species.

FIGURE 1. Continued 16

Distance between cluster centroids WVn 1 1 1 1 1 1 1 1 1 1 species 322 100 90 80 70 60 50 40 30 20 10 0 assemblage nAA undisturbed A I one B

C withoutS. invicta

D two

E disturbed LAAA F J low density G three S. invicta I

high densitv H four

FIGURE 2. Dendrogram of ant species assemblages located in Kerr and Kimble counties, Texas, U. S. A. Plots K and L were excluded . from the analysis. 17 TABLE 1. Species assemblages in central Texas, U.S.A., as detected by cluster analysis. A plus (+) indicate the presence of a species within an assemblage.

Species Assemblage Species One Two Three Four

Pachycondvla harpax + Labidus coecus + Conomvrma flava + Brachvmvrmex depilis + Camponotus discolor +

Tapinoma sessile + Formica sp. + Crematogaster laeviuscula + Pogonomvrmex imherbiculus + Crematogaster punctulata +

Leptothorax terrigena + Solenopsis aurea + Leptothorax sp. + Solenopsis geminata + Solenopsis (Diplorhoptrum) sp. 1

Pheidole sp. 1 Tetramorium spinosus Pogonomvrmex barbatus + + + Forelius pruinosus + + + Pheidole tepicana + + +

Paratrechina terricola /vividula ''• + + + T.eptogenvs elongata + N^.ivamvrmex nigrescens + Camponotus sp. + 18 TABLE 1. Continued

Species Assemblage Species One Two Three Four

Forelius foetidus + Atta texana + Pheidole hvatti + Monomorium pharaonis + Solenopsis xvloni +

Monomorium minimum + + + Pheidole crassicornis tetra + + + Solenopsis (Diplorhoptrum^ sp. 2 + Pheidole sp. 2 + Solenopsis invicta + +

Conomvrma insana

a Unable to differentiate between species. 19 between assemblages one and two may have resulted from the latter being disturbed several times each year, contrasted with the lack of human disturbance to assemblage one. If the site in which assemblage one is found should be subsequently disturbed, it will probably become more similar to assemblage two. Assemblage two is maintained by human disturbance and can be considered a disclimactic community (Odum 1971). Assemblage one should not be considered a climactic community because the area was originally a grassland and the habitat is a patch. Assemblages three and four were found within the range of S.. invicta. Assemblage three (Table 1) has low densities of S.. invicta (~2 foragers per day per trap). The colonies of this ant examined near the plots suggest that they were monogynous because never more than one physogastric queen was found in each colony. Most species were collected sporadically, with the exception of M. minimum. In this assemblage, a subterranean species of thief ant, Solenopsis (Diplorhoptrum") sp. 2, was captured repeatedly. Tschinkel (1988) suggested that this subgenus may prevent the establishment of S.. invicta because of Diplorhoptrum's predacious habits on other ants. Assemblage four (Table 1) has low species richness and very high densities of S.. invicta (=200 foragers a day per trap). All the colonies examined around the plot were polygynous. No species of the subgenus Diplorhoptrum were collected in this assemblage. Although these observations do not refute Tschinkel's statement about the thief ants, they do not constitute any significant, quantitative support either. The location of each plot in the ordination is depicted in Figure 3; the PCA scores are provided in Table 2. Axis 1 accounts for the most variation in species composition among plots (Table 2), and is correlated to the disturbance of the habitat (Kendall's T=0.77; n=80; P<0.01). Plot A, corresponding to species assemblage one (Fig. 2), is isolated from the rest of the plots along this axis. Variation in axis 1 is mostly due to the large species turnover and variation in species densities, between species assemblage one and the rest of 20

100

A X I s

AXIS 2 AXIS 1

FIGURE 3. Principal components analysis of plots by species. Squares in the 1-2 axes plane are the geometrical projections of the plots in the ordination space. Axis 1 is correlated with disturbance, axis 2 with temperature, and axis 4 with density of S. invicta. 21 TABLE 2. Principal components analysis scores of plots, percent of variation accounted for by each component (R^), and cumulative variation (SR^).

Axis Plot 1 2 3 4 A 0.0 32.75 83.55 70.60 B 79.75 7.49 95.12 68.10 C 79.54 8.25 89.46 83.40 D 100.0 100.0 87.80 64.65 E 81.26 2.01 100.0 2.73 F 83.98 5.56 0.0 60.77 G 79.73 8.14 9.29 79.69 H 79.84 0.0 9.49 0.0 I 79.72 12.97 8.49 6.94 J 77.51 10.26 89.42 86.99 R2 0.8849 0.0477 0.0465 0.0130 IR2 0.8849 0.9326 0.9791 0.9921 22 the assemblages. Axis 2 is correlated to temperature fluctuations (r=0.79; v=80; P<0.01), even though measures were taken to reduce this variation. This variation is a within plot effect because a single reading was made for all plots. Species of army ants, Ecitoninae, usually appear in spring and early summer, whereas most myrmicines reach their peak activity in mid- and late summer. Dorylines and formicines were more active from mid-summer to early fall. The most active ant in middle to late fall was S.. invicta. Plot scores on axis 4 were highly correlated with the density of S.. invicta (r=0.63; v=32; P<0.01). The fact that this axis accounts for the least variation (Table 2) can be explained by the fact that six of the ten plots did not contain S.. invicta. A PCA of only the infested plots showed that 62.38% of the variation was contained in the first component, and is correlated with the density of S.. invicta (r=0.68; v=32; P<0.01). This result indicates that within areas infested with S.. invicta. this ant accounts for most of the variation in ant species diversity. Axis 3, even though accounting for more variation than the fourth, represents an effect that was not identified (i.e., herbicide or insecticide applications adjacent to the sampling plots). A fifth aggregation of ants found in the study area was excluded from the mathematical analyses because it consisted almost exclusively of S.. invicta. The habitat in which it occurred was located next to a recreational area, and the grounds were mowed several times each month. This area is located well within the range of S.. invicta. at the east end of the transect. The red imported fire ant has been established in this area for several years (Phillips, pers. comm.). Human disturbance in consort with S.. invicta produced a simplified community as observed by Summerlin et al. (1977). Densities of S.. invicta reached over 2,000 ants a day per trap, and all the colonies examined around the area were polygynous. On two occasions, individual foragers of M. minimum were observed foraging ca. 1200 CDST, with temperatures over 38° C; nonetheless, no individuals were found in the traps. Search of nearby shrubs rendered another species of ant. 23 Crematogaster laeviuscula Mayr, but again, no individuals were found in the traps. Of the nine species found in sympatry with S.. invicta (Table 3), two had significant positive correlations with densities of S.. invicta. and two were significantly negative. Monomorium minimum and Conomyrma insana (Buckley) were each significantly correlated to S.. invicta densities, whereas Pheidole tepicana and £. crassicornis tetra were negatively correlated with S.. invicta. 24 TABLE 3. Correlation of species importance values to densities of Solenopsis invicta.

Species Coefficient of Correlation (r)

Conomyrma insana -i- 0.588** Monomorium minimum -i- 0.489* Forelius pruinosus -i- 0.424 Paratrechina terricola/vividula -i- 0.079 Pogonomvrmex barbatus - 0.111 Pheidole sp. 2 - 0.159 Pheidole tepicana - 0.345* Pheidole crassicornis tetra - 0.584**

* P < 0.05. **P<0.01. CHAPTER IV CONCLUSIONS

Diversity among assemblages, mostly species turnover, is correlated to continuous disturbance of the habitat by cultural activities. Assemblage one might have differentiated from communities like that in assemblage two after a short ecological period, when mowing was stopped. Moreover, if the observed trend in importance values continues in the fire-ant-infested plots, then assemblage three might become more similar to assemblage four. The observed structure of assemblage four might be a transitory stage toward a simpler one, like the fifth aggregation observed in the recreational area. The consequences of this oversimplification of the ecosystem is a question that should be addressed in future studies. Tschinkel (1988) pointed out that the successful establishment of Solenopsis invicta is dependent upon natural or human-induced disturbance of the habitat. My results suggest the same relation. Moreover, the higher the degree or duration of the perturbation, the more successful the establishment of S.. invicta. In areas within the range of S.. invicta. the diversity of the ant assemblages was negatively affected by increasing densities of this species. Previous studies have indicated that Monomorium minimum and Forelius pruinosus coexist with S.. invicta because of temporal foraging isolation (Baroni Urbani and Kannowski 1974, Phillips et al. 1986). My results confirm those findings. On the other hand, Conomyrma insana seems to be able to coexist with S.. invicta. probably because direct competitors or predators have been eliminated by S.. invicta. which in turn allows C.. insana to increase in the community. Ants of the granivorous guild, such as Pheidole tepicana Pergande and P.. crassicornis tetra Wheeler, are being replaced from the assemblages containing S.. invicta. The granivorous ant, Pogonomvrmex barbatus (Smith) is not negatively affected, as are 25 9 f\ ants of the genus Pheidole. No significant correlation was found between P. barbatus and S.. invicta. suggesting that these two species are able to coexist. Porter (personal communication) has also found that P. barbatus is one ant species that is able to coexist with S.. invicta around Austin, Texas. Ants of the genus Pheidole exhibit more specialized food habits than do most of the other species that coexist with S.. invicta. They forage during the day in trails and recruit workers when a food source has been found. Competition for seeds could be a reason for replacement of these seed-harvesting ants by S.. invicta. Workers of S.. invicta are similar in size to those of Pheidole. but much smaller than Pogonomvrmex. The two species of Pheidole were probably able to coexist because populations of seed-harvesting ants remove less than 10% of the seeds from the environment, and these species are extremely selective (Whitford 1978). These results suggests a steady restructuring of the community as observed by Porter et al. (1988), even several years after the invasion by S.. invicta. The structure of the communities is dynamic. None of the areas studied represents a stable climactic community, although species assemblage one was, as compared to the other assemblages, the most diverse. Outside the range of S.. invicta. the most important factor in shaping ant communities is the disturbance of the habitat, which prevents populations from reaching resource- defined equilibria. Denslow (1985) reviewed the literature on the coexistence of species mediated by habitat disturbance, and concluded that diversity within a patch will be maximized when the present disturbance regime resembles the historical regime. Exotic disturbances, like the introduction of a new competitor, may shift the dominance and diversity relationships within the community. This shift usually results in the loss of species diversity. Within the range of S.. invicta. the eventual structure of the ant community is mediated mostly by competitive abilities within disturbed habitats. In addition, the loss of species richness follows the predictions of Denslow (1985), and the shifts observed in diversity, from 27 uninfested to heavily infested areas, may result in a new species equilibria in which S.. invicta is a permanent member. LITERATURE CITED

Adis, J. 1979. Problems of interpreting arthropod sampling with pitfall traps. Zool. Anz. 202: 177-184.

Andersen, A. N. 1986a. Diversity, seasonality and community organization of ants at adjacent heath and woodland sites in south-eastern Australia. Australian J. Zool. 34: 53-64.

1986b. Patterns of ant community organization in mesic southeastern Australia. Australian J. Ecol. 11: 87-97.

Apperson, C. S. & E. P. Powell. 1984. Foraging activity of ants (: Formicidae) in a pasture inhabited by the red imported fire ant. Florida Entomol. 67: 383-392.

Banks, W. A., D. P. Jouvenaz, D. P. Wojcik & C. S. Lofgren. 1985. Observations on fire ants, Solenopsis spp., in Mato Grosso, Brazil. Sociobiology 11: 143-152.

Baroni Urbani, C. & P. B. Kannowski. 1974. Patterns in the red imported fire ant settlement of a Louisiana pasture: Some demographic parameters, interspecific competition and food sharing. Environ. Entomol. 3: 755-760.

Bhatkar, A., W. H. Whitcomb, W. F. Buren, P. Callahan & T. Carlysle. 1972. Confrontation behavior between Lasius neoniger (Hymenoptera: Formicidae) and the imported fire ant. Environ. Entomol. 1: 274-279.

Brown, J. H. & A. C. Gibson. 1983. Biogeography. C. V. Mosby, St. Louis.

Buren, W. F. 1972. Revisionary studies on the of the imported fire ants. J. Georgia Entomol. Soc. 7: 1-25.

Buren, W. F., G. E. Allen, W. H. Whitcomb, F. E. Lennartz & R. N. Williams. 1974. Zoogeography of the imported fire ants. J. New York Entomol. Soc. 82: 113-124.

28 29 Carroll, C. R. & D. H. Janzen. 1973. Ecology of foraging by ants. Ann. Rev. Ecol. Sys. 4: 231-257.

Clements, F. E. 1916. Plant succession, an analysis of the development of vegetation. Carnegie Inst. Wash. Publ. 242.

Council, J. H. 1980. Diversity and the coevolution of competitors, or the ghost of competition past. Oikos 35: 131-138.

Correll, D. S. & M. C. Johnston. 1970. Manual of the vascular plants of Texas. Texas Res. Found., Renner, Texas.

Culver, D. C. 1974. Species packing in Caribbean and north temperate ant communities. Ecology 55: 974-988.

Denslow, J. S. 1985. Disturbance-mediated coexistance of species. Pages 307-324 in P. S. White & S. T. A. Pickett (eds.). The ecology of natural disturbance and patch dynamics. Academic Press, New York.

Digby, P. G. N. & R. A. Kempton. 1987. Multivariate analysis of ecological communities. Chapman and Hall, London.

Fox, B. J., M. D. Fox & E. Archer. 1985. Experimental confirmation of competition between two dominant species of Iridomyrmex (Hymenoptera: Formicidae). Australian J. Ecol. 10: 105-110.

Glancey, B. M., D. P. Wojcik, C. H. Craig & J. A. Mitchell. 1976. Ants of Mobile county, AL, as monitored by bait transects. J. Georgia Entomol. Soc. 11: 191-197.

Greenslade, P. J. M. 1973. Sampling ants with pitfall traps: digging- in effects. Insectes Soc. 20: 343-353.

Hardin, G. 1960. The competitive exclusion principle. Science 131: 1292-1297.

Heatwole, H. & R. Levins. 1972. Trophic structure and faunal change during recolonization. Ecology 53: 531-534.

Jansen, M. J. W. & J. A. J. Metz. 1979. How many victims will a pitfall make? Acta Biotheor. 28: 98-122. on Jones, S. R. 1985. Foraging and recruitment abilities of Solenopsis invicta Buren, compared with other ant species indigenous to Texas. M. S. thesis, Texas Tech University.

Kidd, K. A., C. S. Apperson & L. A. Nelson. 1985. Recruitment of the red imported fire ant, Solenopsis invicta Buren. to soybean oil baits. Florida Entomol. 68: 253-261.

Levings, S. C. & J. F. A. Traniello. 1981. Territoriality, nest dispersion, and community structure in ants. Psyche 88: 265- 319.

Levins, R. & D. C. Culver. 1971. Regional coexistence of species and competition between rare species. Proc. Nat. Acad. Sci. 68: 1246-1248.

Levins, R., M. L. Pressik & H. Heatwole. 1973. Coexistence patterns in insular ants. Amer. Sci. 61: 463-472.

Lofgren, C. S., W. A. Banks & B. M. Glancey. 1975. Biology and control of imported fire ants. Ann. Rev. Entomol. 20: 1-30.

Majer, J. D. 1978. An improved pitfall trap for sampling ants and other epigaeic invertebrates. J. Australian Entomol. Soc. 17: 261-262.

Maxwell, F. G. et al. 1982. Panel III: Population dynamics of the imported fire ant. Proceedings of the Symposium on the Imported Fire Ant, Atlanta, Georgia.

Miller, R. S. 1967. Patterns and processes in competition. Adv. Ecol. Res. 4: 1-74.

Moulton, M. P. & S. L. Pimm. 1987. Morphological assortment in introduced Hawaiian passerines. Evol. Ecol. 1: 113-124.

Odum, E. P. 1971. Fundamentals of Ecology. W. B. Saunders, Philadelphia.

Otis, G. W., E. Santana C, D. L. Crawford & M. L. Higgins. 1986. The effect of foraging army ants on leaf-litter arthropods. Biotropica 18: 56-61. 31 Peet, R. K. 1974. The measurement of species diversity. Ann. Rev. Ecol. Sys. 5: 285-307.

Phillips, S. A., S. R. Jones & D. M. Clabom. 1986. Temporal foraging patterns Solenopsis invicta and native ants of central Texas. Pages 114-122 in C. S. Lofgren & R. K. Vander Meer (eds.). Fire ants and leaf cutting ants: Biology and management. Westview, Boulder.

Pielou, E. C. 1984. The interpretation of ecological data, a primer on classification and ordination. John Wiley & Sons, New York.

Porter, S. D., B. Van Eimeren & L. E. Gilbert. 1988. Invasion of red imported fire ants (Hymenoptera: Formicidae): microgeography of competitive replacement. Ann. Entomol. Soc. Amer. 81: 913-918.

Ricklefs, R. E. 1987. Community diversity: Relative roles of local and regional processes. Science 235: 167-171.

Samways, M. J. 1983. Community structure of ants (Hymenoptera: Formicidae) in a series of habitats associated with citrus. J. Appl. Ecol. 20: 833-847.

SAS Institute Inc. 1985. SAS user's guide: Statistics, Version 5 Edition. Cary, N.C.

Schoener, T. W. 1974. Some methods for calculating competition coefficients from resource-utilization spectra. Amer. Nat. 108: 332-340.

1983. Guild experiments on interspecific competition. Amer. Nat. 122: 240-285.

Seifert, R. P. 1984. Does competition structure communities? Field studies on neotropical Heliconia communities. Pages 54-65 in D. R. Strong, D.Simberioff, L. G. Abele & A. B. Thistle (eds.). Ecological Communities: Conceptual issues and the evidence. Princeton University Press, Princeton. 32 Simberloff, D. 1981. Community effects of introduced species. Pages 53-83. In M. H. Nitecki (ed.), Biotic crises in ecological and evolutionary time. Academic Press, New York.

Sokal, R. R. & J. F. Rohlf. 1981. Biometry. W. H. Freeman, San Francisco.

Strong, Jr., D. R. 1984. Exorcising the ghost of competition past: Phytophagous . Pages 28-41 in D. R. Strong, D. Simberioff, L. G. Abele & A. B. Thistle (eds.). Ecological Communities: Conceptual issues and the evidence. Princeton University Press, Princeton.

Strong, Jr., D. R., D. Simberioff, L. G. Abele & A. B. Thistle (eds.) 1984. Ecological Communities: Conceptual issues and the evidence. Princeton University Press, Princeton.

Summerlin, J. W., A. C. F. Hung & S. B. Vinson. 1977. Residues in nontarget ants, species simplification and recovery of populations following aerial applications of Mirex. Environ. Entomol. 6: 193-197.

Torres, J. A. 1984a. Niches and coexistence of ant communities in Puerto Rico: Repeated patterns. Biotropica 16: 284-295.

1984b. Diversity and distribution of ant communities in Puerto Rico. Biotropica 16: 296-303.

Tschinkel, W. R. 1986. The ecological nature of the fire ant: some aspects of colony function and some unanswered questions. Pages 72-87 in C. S. Lofgren & R. K. Vander Meer (eds.). Fire ants and leaf cutting ants: Biology and management. Westview, Boulder.

1988. Distribution of the fire ant Solenopsis invicta and S.. geminata (Hymenoptera: Formicidae) in northern Florida in relation to habitat disturbance. Ann. Entomol. Soc. Amer. 81: 76-81. 33 Werner, E. E. 1984. The mechanisms of species interactions and community organization in fish. Pages 360-382 in D. R. Strong, D. Simberioff, L. G. Abele & A. B. Thistle (eds.). Ecological Communities: Conceptual issues and the evidence. Princeton University Press, Princeton.

Wheeler G. C. & J. Wheeler. 1985. A checklist of Texas ants. Prairie Nat. 17: 49-64.

Whitcomb, W. H., H. A. Denmark, A. P. Bhatkar & G. L. Green. 1972. Preliminary studies on the ants of Florida soybean fields. Florida Entomol. 55: 129-142.

Whitford, W. G. 1978. Foraging by seed-harvesting ants. Pages 107-110 in M. V. Brain (ed.). Production ecology of ants and termites. Cambridge University Press, Cambridge.

Whittaker, R. H. 1965. Dominance and diversity in land plant communities. Science 147: 250-260.

Wiens, J. A. 1977. On competition and variable environments. Amer. Sci. 65: 590-597.

Wilson, E. O. 1971. The insect societies. Belknap Press, Cambridge.

Wilson, E. O. & W. L. Brown. 1958. Recent changes in the introduced population of the fire ant Solenopsis saevissima (Fr. Smith). Evolution 12: 211-218.

Wilson, N. L., J. H. Dillier & G. P. Markin. 1971. Foraging territories of imported fire ants. Ann. Entomol. Soc. Amer. 64: 660-665. PERMISSION TO COPY

In presenting this thesis in partial fulfillment of the requirements for a master's degree at Texas Tech University, I agree that the Library and my major department shall make it freely avail­ able for research purposes. Permission to copy this thesis for scholarly purposes may be granted by the Director of the Library or my major professor. It is understood that any copying or publication of this thesis for financial gain shall not be allowed without my further written permission and that any user may be liable for copy­ right infringement.

Disagree (Permission not granted) Agree (Permission granted)

/Mfi^ (fo f(. C^:=^.c^',^ Student's signature /Student's signature

Y^C, /^fz? Date D/ke

".'^r