THE ANTS (HYMENOPTERA, FORMICIDAE) OF WESTERN TEXAS. PART II. SUBFAMILIES ECITONINAE, PONERINAE, PSEUDOMYRMECINAE, DOLICHODERINAE, AND FORMICINAE James C. Cokendolpher and Osczir F. Francke THE ANTS (HYMENOPTERA, FORMICIDAE) OF WESTERN TEXAS. PART III. ADDITIONS AND CORRECTIONS James C. Cokendolpher SPECIAL PUBLICATIONS, THE MUSEUM TEXAS TECH UNIVERSITY NUMBERS 30 AND 31 THE ANTS (HYMENOPTERA, FORMICIDAE) OF WESTERN TEXAS. PART II. SUBFAMILIES ECITONINAE, PONERINAE, PSEUDOMYRMECINAE, DOLICHODERINAE, AND FORMICINAE James C. Cokendolpher and Oscar F. Francke TEXAS TECH UNIVERSITY PRESS 1990 SPECIAL PUBLICATIONS, THE MUSEUM TEXAS TECH UNIVERSITY NUMBER SO Series Editor J. Knox Jones, Jr. Published 23 February 1990 Copyright 1990 Texas Tech University Press All rights reserved. No portion of this book may be reproduced in any form or by any means, including electronic storage and retrieval systems, except by explicit, prior written permission of the publisher. Special Publications of The Museum are numbered serially sind published on an irregular basis. Institutions interested in exchsmging publications should address the Exchange Librarian at Texas Tech University. ISSN 0149-1768 ISBN 0-89672-175-2 Texas Tech University Press Lubbock, Texas CONTENTS INTRODUCTION 5 MATERIALS AND METHODS 5 RESULTS AND DISCUSSION 8 KEY TO WORKERS OF THE NORTH AMERICAN SUBFAMILIES OF ANTS . . .. 8 SUBFAMILY ECITONINAE 8 GENERIC KEY TO WESTERN TEXAS ECITONINAE 9 CHECKLIST OF THE ECITONINAE OF WESTERN TEXAS 9 ACCOUNTS OF SPECIES 10 SUBFAMILY PONERINAE 13 GENERIC KEY TO WORKERS OF WESTERN TEXAS PONERINAE 13 CHECKLIST OF THE PONERINAE OF WESTERN TEXAS 14 ACCOUNTS OF SPECIES 14 SUBFAMILY PSEUDOMYRMECINAE 20 CHECKLIST OF THE PSEUDOMYRMECINAE OF WESTERN TEXAS 20 ACCOUNTS OF SPECIES 20 SUBFAMILY DOLICHODERINAE 21 GENERIC KEY TO WORKERS OF WESTERN TEXAS DOLICHODERINAE . .21 CHECKLIST OF THE DOLICHODERINAE OF WESTERN TEXAS 21 ACCOUNTS OF SPECIES 21 SUBFAMILY FORMICINAE 26 GENERIC KEY TO WORKERS OF WESTERN TEXAS FORMICINAE . . .. 26 CHECKLIST OF THE FORMICINAE OF WESTERN TEXAS 27 ACCOUNTS OF SPECIES 28 SUMMARY 44 ACKNOWLEDGMENTS 45 LITERATURE CITED 45 FIGURES 48 APPENDICES 71 INTRODUCTION This contribution is the second part of a study ofthe ants of western Texas. The first part (Moody and Francke, 1982) dealt with ants ofthe subfamily Myrmicinae; this contribution deals with the remaining five subfamilies. The third part, the final section, wUl contain additions and corrections to Part I. The objectives of this study are the same as in Part I of the series: to determine which ant species inhabit western Texas, to define geographic regions in which they occur, and to explore some of the abiotic factors correlated with their distribution. Ecological data gathered at each collecting locality were used to determine the preferred habitat of the different species. MATERIALS AND METHODS Study Area Texas can be partitioned into 10, major, vegetative zones (Correll and Johnson, 1970). These zones are useful in describing general environmental differences that often serve to limit biotic distributions. The study area for this research included all of Texas west of the 100th meridian as well as several localities from 10 to 60 kUometers east of that line; five of the vegetative regions are within this study area (Fig. 1). The Texas High Plains occupy approximately 81,000 square kilometers (Fig. 1) and receive from 43 to 58 centimeters of precipitation annually. The elevation of the High Plains ranges from 800 to 1300 meters. Much ofthe region is irrigated cropland, although large tracts also support rangeland. The Rolling Plains (Fig. 1) occupy approximately 77,000 square kilometers ofthe study area and have an average annual precipitation from 44 centimeters on the Texas-New Mexico border in Oldham County to 70 centimeters at the 100th meridian. Elevation ofthe Rolling Plains study area ranges from 450 to 800 meters. Almost all of the Rolling Plains within the study area is rangeland. Approximately 69 percent of the 97,000 square kilometers of the Edwards Plateau lies west ofthe 100th meridian (Fig. 1). The average annual precipitation of this area is 43 centimeters in the west and 71 centimeters along the 100th meridian. Ranging in elevation from 250 to 800 meters, the Edwards Plateau is primarily rangeland, with some cultivation practiced in valleys and areas with deeper soils. About 20,000 square kilometers ofthe Rio Grande Plains are within the boundary of the study area (Fig. 1). Within this area, the average annual rainfall is 45 centimeters, and the elevation ranges from 150 to 200 meters. With the exception of a few large farms, most of the land is used as rangeland. The Trans-Pecos region includes approximately 77,000 square kilometers in the area west of the Pecos River (Fig. 1). The average annual precipitation varies with elevation (450 to 2650 meters), ranging from less than 35 centimeters in some parts of the Chihuahuan desert to as much as 55 centimeters at higher elevations. SPECIAL PUBLICATIONS THE MUSEUM TEXAS TECH UNIVERSITY FIG. 1.—The vegetational regions of western Texas (from Correll and Johnston, 1970): 1, High Plains; 2, Rolling Plains; 3, Edwards Plateau; 4, Trans-Pecos region; 5, Rio Grande Plains. Gray on insert map indicates study area. COKENDOLPHER AND FRANCKE—ANTS OF V^'ESTERN TEXAS. PART II 7 Collecting Procedures Ants were sampled at 691 collecting sites in 97 counties, covering an area of approximately 324,700 square kilometers. Collecting techniques used in this study were based on methods oudined by Wheeler and Wheeler (1963). During the day, aspirators were used to collect ants from nests found at each locality. At night, bait stations were set, at which peanut butter, hamburger meat, or grease were used to attract nocturnally foraging workers. In addition, ultraviolet lights (black lights) were used to collect night-flying males and females. Samples equalling one nest series usually consisted of several dozen workers, brood, and any reproductive forms observed. Data such as slope angle and exposure, nest description, and caste(s) collected were compiled for most nest series; the plant association for each collecting locality also was recorded. Ecological data for samples collected prior to March 1978 are unavailable, and their numbers are reflected under the "no data" column heading of Appendices 3 and 4. All ants collected were preserved in 80 percent ethyl alcohol and deposited in the Entomological Collection at Texas Tech University. The elevation of each collecting locality was determined using United States Geological Survey Maps with contour intervals of 30.5 meters (100 feet). Soil texture data for all localities were determined in the field by wetting a pinch of soil and rubbing it between the fingers. The accuracy of field determinations was verified for 355 of the collecting loc£tlities by comparing them to the soil texture class for each locality as given in the county soil surveys published by the Soil Conservation Service (United States Department of Agriculture, 1975). The remaining field determinations could not be verified because county soil surveys were not available for those localities. Nonetheless, the data were considered to be reliable and were used in analyses. Analytical Methods Elevation and soil texture data for each species taken at more than two localities were analyzed separately with Chi-Square tests. Elevation distribution data were partitioned into cells of 100-meter intervals for the analyses (Appendix 2). Eleven soil texture classes were used (Appendix 3). The distribution of those parameters among the 691 localities sampled were used to calculate the expected values of occurrence for each ant species, using the nuU hypothesis that each species is randomly distributed in western Texas with respect to both elevation and soil type. Following the methods of Snedecor and Cochran (1967), we assumed the smallest expected value for each ceU in each aneilysis was at least one; in reality, most were substantially larger. If, after combination of adjacent cells, the above criterion was not met or if the degrees of freedom were below seven, the analysis was not attempted. Slope angle and exposure data for each nest series were tabulated and analyzed. Ants were collected primarily from nests that were on level or slightly sloping surfaces, but a few species were found in areas with more severe slopes (Appendix 4). Nests located on slopes were not coUected in sufficient quantities to suggest any 8 SPECIAL PUBLICATIONS THE MUSEUM TEXAS TECH UNIVERSITY definite exposure trends or preferences for any one species. For this reason, the nest slope-exposure data are not included. Plant association data for each collecting locality did not show any relationship with individual ant species, except in those cases where an ant species was direcdy associated with certain plants. This relationship occurred for those ants found nesting in plants, such as arboreal nesting species of Camponotus and Colobopsis. In the Edwards Plateau area of western Texas, gaU-nesting ants are primarily associated with live oak trees {Quercus virginiana Mill). For convenient reference, all appendices and figures showing ant distributions in western Texas are grouped at the end of the text. RESULTS AND DISCUSSION The largest subfamily of the Formicidae, the Myrmicinae, was dealt with in the first part of this study (Moody and Francke, 1982). The remaining five subfamilies, with 24 genera present in westem Texas, are considered herein. The cerapachyines, regarded as a subfamily in the first part, are treated as members of the Ponerinae in this paper. The army ants, referred to the Dorylinae in the first part, are here considered to belong to the subfamily Ecitoninae. These changes are discussed under the appropriate subfamilial headings and are reflected in the key to the subfamilies below. Key to Workers ofthe North American Subfamilies of Ants (modified from Creighton, 1950) 1. Caster with a distinct constriction between the first and second segments, or, if this constriction is faint, the mandibles are linear and the petiole is produced into a conical dorsal spine Ponerinae Gaster without a constriction between the first and second segments 2 2. Abdominal pedicel consisting of two segments 3 Abdominal pedicel consisting of one segment 5 3. Frontal carinae narrow jmd not expanded laterally so that the antennal insertions are fully exposed when the head is viewed from above 4 Frontal carinae expanded laterally so that they partially or wholly cover the antennal insertions when the head is viewed from above Myrmicinae 4. Eyes large, suboval or reniform, and consisting of several hundred fine oramatidia Pseudomyrmecinae Eyes vestigial or absent, ifpresent consisting of a single ocelluslike structure Ecitoninae 5. Apex of gaster with a distincdy circular orifice, the acidopore, which usually is surrounded by a fringe of hairs Formicinae Apex of gaster lacking acidopre, but with a broad, slidike cloacal orifice; hairs, when present, not forming an encircling fringe Dolichoderinae Subfamily Ecitoninae The members of this subfamily commonly are referred to as army ants. They are predaceous in habits and stage massive, weU-organized raids, earning their common name. Army ants also are weU known for their cyclic nomadic behavior: the cycles are related to the reproductive condition of the colony, alternating between statary