Title: Biology and ecology of ants of the genus Aphaenogaster in Florida
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098926/00001
 Material Information
Title: Biology and ecology of ants of the genus Aphaenogaster in Florida
Physical Description: viii, 177 leaves : ill. ; 28 cm.
Language: English
Creator: Carroll, John F., 1945-
Copyright Date: 1975
Subject: Ants   ( lcsh )
Entomology and Nematology thesis Ph. D
Dissertations, Academic -- Entomology and Nematology -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Statement of Responsibility: by John F. Carroll.
Thesis: Thesis (Ph. D.)--University of Florida, 1975.
Bibliography: Includes bibliographical references (leaves 169-176).
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00098926
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000425664
oclc - 38046238
notis - ACH4210


This item has the following downloads:

biologyecologyof00carr ( PDF )

Full Text








I am grateful for the assistance rendered me by several people

during this study. Particular thanks are due Dr. W. H. Whitcomb,

chairman of my supervisory committee, for his suggestions and encour-

agement, and to committee members Drs. W. F. Buren, T. C. Emmel, L.

C. Kuitert, and J. E. Lloyd for their assistance in preparing this


I wish to thank Dr. W. F. Buren for identifying numerous ant

specimens and for his aid in the systematics portion of my research.

Other individuals to whom I am indebted for identifications are:

H. A. Denmark, mites; Dr. T. J. Walker, myrmecophilous crickets;

G. B. Edwards, spiders; Dr. J. W. Kimbrough, fungi.

Additionally thanks are due Dr. A. P. Bhatkar, T. M. Neal, and

D. P. Wojcik for collecting specimens for me or loaning specimens to

me for examination.

I am also grateful to Dr. P. S. Callahan and J. C. E. Nickerson

for allowing me to examine the ants collected in their light trap





ACKNOWLEDGMENTS . . . . . . . . ... . . ii

ABSTRACT . . . . . . . . ... . . . . vi

INTRODUCTION . . . . . . . . . ... 1

REVIEW OF LITERATURE . . . . . . . . . ... 3

Distribution . . . . . . . . ... . . 3
Economic Importance . . . . . . . . . 3
Ecology and Biology . . . . . . . . . 5

MATERIALS AND METHODS ................... .. 15

Field Studies . . . . . . . . ... . . 15
Detection of Colonies . . . . . . ... 15
Collecting Trips . . . . . . . ... 17
Collection of Colonies . . . . . . ... 18
Dissection of Colonies . . . . . . ... 20
Field Observations . . . . . . . ... 21
Foraging Success . . . . . . . ... 22
Laboratory Studies . . . . . . . .... . 23
Maintenance of Colonies . . . . . . ... 23
Determination of Stadial Duration . . . ... 25

RESULTS . . . . . . . . ... . . . . . 27

General Appearance of Aphaenogaster Spp. . . . ... 27
Species of Aphaenogaster in Florida . . . . .. 28
Key to Workers of Florida Species . . . . ... 29
Appearance of Workers of Florida Species . . . ... 31
Taxonomy of rudis/texana Group . . . . . ... 36
Findings on A. ashmeadi . . . . . . . ... 46
Description of Female . . . . . . ... 46
Description of Male . . . . . . . ... 48
Distribution . . . . . . . .... . 49
Habitats . . . . . . . . ... . . 50
Nests . . . . . . . . ... . . . 51
Foraging . . . . . . . . ... . . 52






Feeding . . . .. . . . . .
Coexistence with other Ants . . . . .
Annual Cycles and Colony Size . . . . .
Development . . . . . . . . .
Wings on A. carolinensis . . . . . .
Description of Female . . . . ......
Description of Male . . . . . . .
Description of Worker . . . . . . .
Distribution . . . . . . . . .
Habitats . . . . . . . . .
Nests . . . . . . . . . . .
Foraging . . . . . . . . . .
Feeding . . . . . . . . . .
Coexistence of A. carolinensis with other Ants
Coexistence of A. carolinensis* with other Ants
Annual Cycles and Colony Size . . . . .
Development . . . . . . . ......
wings on A. flemini . . . . . . .
Description of Female . . . . . . .
Description of Male . . . . . . .
Distribution . . . . . . . . .
Habitats . . . . . . . . . .
Nests . . . . . . . . . . .
Foraging . . . . . . . . .
Feeding . . . . . . . . . .
Coexistence with other Ants . . . . .
Annual Cycles, Mating Flights, and Colony Size
Development . . . . . . . . .
ings on A. floridana . . . . . . .
DescripTion of Female . . . . . . .
Description of Male . . . . . . .
Distribution . . . . . . . . .
Habitats . . . . . . . . . .
Nests . . . . . . . . . . .
Foraging . . . . . . . . . .
Feeding . . . . . . . . . .
Coexistence with other Ants . . . . .
Annual Cycles and Colony Size . . . . .
Development . . . . . . . . .
ings on A. fulva . . . . . . . .
Distribution . . . . . . . .
Habitats . . . . . . . . . .
Nests . . . . . . . . . . .
Foraging . . . . . . . . .
Feeding . . . . . . . . . ..
Coexistence with other Ants . . . . .
Annual Cycles, Mating Flights, and Colony Size



Findings on A. lamellidens . . . . . . .. 110
Distribution . . . . . . . .... ... 110
Habitats . . . . . . . . ... . . 111
Nests ...... .. ........ .. .... 113
Foraging . . . . . . . .... ..... 113
Feeding . . . . . . . . ... ..... 114
Coexistence with other Ants . . . . . ... 115
Annual Cycles and Colony Size . . . . . .. 117
Development . . . . . . . . ... ... 118
Findings on A. mariae . . . . . . . ... 118
Description of Male . . . . . . . .. 118
Distribution . .. . . . . . ... 120
Habitats . . . . . . . . ... . . 120
Foraging . . . . . . . .... ..... 121
Annual Cycles . . . . . . . .... . 121
Findings on A. tennesseensis . . . . .. . . 121
Description of Male. . . . . . . . . 121
Distribution . . . . . .... .. . 123
Habitats . . . . . . . .. . . 123
Nests ................ ... .. . 124
Foraging . . . . . . . . ... . . 125
Coexistence with other Ants . . . . . ... 126
Annual Cycles and Colony Size . . . . ... 127
Development . . . . . . . . . . 128
Findings on A. treatae . . . . . . . .... 128
Distribution ... . . . . . . . . . 128
Habitats . . . . . . . . .. . . 129
Nests . . . . . . . . .. . . . 131
Foraging . . . . ... .. . . . . 131
Feeding . . . . . ... .. . . . . 132
Coexistence with other Ants . . . .... .. . 132
Annual Cycles and Colony Size . . . . ... 134
Development . . . . . . .... .. . . 135
Predators of AphaenoGaster Spp. . . . . . ... 135
Myrmecophiles Associated with Aphaenogaster Spp. .. .... 138

DISCUSSION . . . . . . . . . . . . . 140

Intrageneric Coexistence . . . . ... . . . 140
Consumption of Agaricales . . . . . . ... 144

TABLES . . . . . . . . . . . . . . 148

FIGURES . . . . . . . . .. . . . . 157

REFERENCES . . . . . . . .... . . . . 169

BIOGRAPHICAL SKETCH . . . . .... .. . . . . 177

Abstract of Thesis Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy



John F. Carroll

March, 1975

Chairman: Willard H. Whitcomb
Major Department: Entomology and Nematology

Investigations were conducted on the biology, ecology and sys-

tematics of the Floridian species of the genus Aphaenogaster (Hymen-

optera: Formicidae). Nine species of Aphaenogaster occur in Florida:

A. ashmeadi Emery, A. carolinensis Wheeler, A. flemingi M. R. Smith,

A. floridana M. R. Smith, A. lamellidens Mayr, A. mariae Forel, A.

tennesseensis (Mayr), and A. treatae Forel.

A complex of forms, whose constituents have been attributed by

various authors to A. rudis Emery, A. texana Emery, A. miamiana

Wheeler and varieties thereof, is synonymized under A. carolinensis

Wheeler. A. carolinensis was raised to species level from A. texana

carolinensis Wheeler on the basis of differences between Floridian

forms and Texas specimens of A. texana. Males of A. ashmeadi, A.

carolinensis, A. flemingi, A. floridana, A. mariae, and A. tennesseensis

females of A. ashmeadi, A. carolinensis, and A. flemingi and a worker

of A. carolinensis are described.

In Florida Aphaenogaster spp. occur primarily in wooded habi-

tats. Most species tend to be segregated into characteristic micro-

habitats. Soil dwelling species predominate in well-drained habitats,

while species nesting in rotten wood are most common in moist woodlands.

Nests are described and distributions of the species reported.

Aphaenogaster spp. forage chiefly on the ground, only occasion-

ally venturing more than three or four cm up herbaceous vegetation.

Species nesting arboreally also forage on the trunks and larger branches

of nest trees. The diet of Aphaenogaster spp. consists largely of arth-

ropods, which are captured or scavenged. Seeds and abscissed floral

parts are collected. Free growing fruiting bodies of mushrooms

(Agaricales) comprise a significant portion of the diet of some Aphaeno-

gaster species. Aphaenogaster spp. primarily consume Russula spp.

(Russulaceae) but also species of Tricholomataceae and Amanitaccae.

Aphaenogaster spp. are virtually free from Formicine competition for

this food source. Termites do not constitute a major food source.

Confined in the same containers, termite colonies coexisted for months

with Aphaenogaster spp. colonies in pieces of rotten wood.

Mating flights of A. fulva are described. A. fulva mating

flights are crepuscular, unlike those of A. treatae. Annual cycles of

eight Floridian species are reported. Eight species maintained in the

laboratory had similar developmental rates. In A. ashmeadi for example

the duration of the egg stage is 19 to 22 days, larval stage 14 to 22

days, and pupal stage ten to 13 days.

Toads, spiders, and other ants are important predators of

Aphaenogaster spp. workers. Myrmecophilous crickets, Myrmecophila

pergandei Bruner, were found in the nests of A. ashmeadi and A.

carolinensis. Ant species, nesting in or traversing the various

microhabitats utilized by each species of Aphaenogaster, are re-



Comparatively little public attention or scientific scrutiny has

been directed toward ants of the genus Aphaenogaster. These rather

generalized Myrmicine ants exhibit no bizarre or otherwise conspicuous

behavior patterns which might pique one's curiosity. They are, however,

neither small nor cryptic, nor solely nocturnal or hypogaeic foragers.

Instead these medium to large ants can be found, often abundantly, in

most natural terrestrial ecosystems in Florida. In North America

species of Aphaenogaster pose no obvious economic or health problems;

thus their anonymity.

Field observations and published accounts of Aphaenogaster spp.

being associated with or preying upon termites prompted this investi-

gation. Knowledge of the biology and ecology of these ants, it was

felt, might reveal them to be significant biological control agents,

limiting termite population levels.

In Florida few ant genera are represented by more species than

is Aphaenogaster. This fact raises certain questions regarding inter-

specific competition, and the specializations of the different species,

individually and collectively. How are so many closely related spe-

cies able to coexist in such a limited geographic region as north

Florida? Do different species feed on different kinds of food or

utilize different habitats? Are they different sizes? Hopefully this

investigation provides some answers to these questions.


A further indication of the need for studying the genus Aphaeno-

gaster was the number of taxonomic questions which arose at the outset

of this investigation and concerned the most common species of Aphaeno-

gaster in Florida.



The genus Aphaenogaster has a worldwide distribution. Wilson

(1971) and Wheeler (1922) indicated that species occur in all the major

faunal regions of the world except the Ethiopian. At least one species

of Aphaenogaster can be found in each of the 48 contiguous states

(Smith, 1947). Creighton (1950) provided distribution data for all the

then described species of Aphaenogaster in North America north of Mex-

ico. Subsequent work by other investigators has revealed the ranges

of several species to be greater than previously supposed. The distri-

bution of each Florida species is discussed individually in the results

section. The northward occurrence of species of Aphaenogaster was re-

ported by Gregg (1972). Wheeler (1910) conjectured that A. treatae

Forel, A. mariae Forel, and A. lamellidens Mayr had their evolutionary

origin in the southeastern United States, while A. fulva Roger was asso-

ciated with the northeast.

Economic Importance

North American species of Aphaenogaster are not significant pests.

Nevertheless, Smith (1965) reported A. fulva, A. lamellidens, A. rudis

Emery, and A. tennesseensis (Mayr) as occasionally infesting houses.

He had a single record of A. fulva infesting household foodstuffs, but

none for the other three species. None of the four species, he believed,

were pests of major importance in homes.

In Australia A. (Nystalomyrmex) pythia Forel and A. (Nystalo-

myrmex) longiceps F. Smith, known as funnel ants, cause serious damage

to commercial sugarcane and tobacco operations and to pastures. They

tunnel extensively in sugarcane fields and lacerate tender rootlets

(Wilson, G., 1969). Funnel ants harm tobacco seedlings by tunneling

in the soil and burying the small plants near their numerous nest

entrances with excavated soil (Smith and Atherton, 1944). Pastures

infested with funnel ants develop bare spots which in turn may be col-

onized by deep-rooted weeds rather than grasses. Heavy funnel ant

infestations are generally associated with the over-grazing of pas-

tures by cattle (Saunders, 1969). Chemicals are used to attempt con-

trol of these ants in pastures and sugarcane (Saunders, 1961, 1967,

1969 and Wilson, G., 1969). Smith and Atherton (1944) also report

that A. longiceps may tend aphids (Geoica sp.) on the roots of grasses.

As predators of other arthropods Aphaenogaster spp. may benefit

man as biological control agents. Jaynes and Marucci (1947) indicate

that A. rudis is an important predator of the codling moth, Carpo-

capsa pomonella (L.) in West Virginia apple orchards. A. rudis

attacks C. pomonella larvae which have dropped to the ground prior

to hibernation. It also attacks those in cocoons. A large propor-

tion of codling moth larvae released by Jaynes and Marucci near A.

rudis colonies were destroyed by the ants before they could make co-

coons. Over a three year period A. rudis was one of the most common

ants they found in their "biological control orchard." In nearby

commercial orchards where chemicals were heavily used, A. rudis popu-

lations were low.

Species of Aphaenogaster have long been reported as predators of

termites. This subject was treated in an economic vein in a few pub-

lications. In this paper, however, the Aphaenogaster-termite relation-

ship is discussed in the review of literature of Aphaenogaster biology.

All economic references are included in that section.

Whitcomb et al. (1973) found A. floridana preyed upon recently

mated queens of the economically important red imported fire ant,

Solenopsis invicta Buren.

Ecology and Biology

Much of the following portions of the literature review is

intended to acquaint the reader with the principal sources consulted

by the author. The contents of these and other publications are set

forth in some detail in the results and discussion sections.

Publications dealing specifically with the ecology or biology of

North American species of Aphaenogaster are few. Talbot (1951, 1954)

conducted population studies of A. rudis in Missouri and A. treatae

in Michigan. In a strip of field 100 by 200 ft she (1954) detected

63 colonies of A. treatae which she calculated to be one colony per

21 sq yds. She (1951) found from 34 to 3,445 individuals (all stages)

in 72 A. rudis colonies. She found multiple queens (two to 15) in

some colonies excavated in the spring. This led her to suggest that

colonies of A. rudis may mingle or coalesce in the spring just after

hibernation. Talbot (1953, 1957) studied populations of all ant

species in an old field in Michigan and in a Missouri woodland. In 25

selected plots (each ten sq m) in the Michigan field she found A.

treatae nesting in six plots and A. rudis in a single plot. A. treatae

colonies ranged in size from 191 to 3,221 individuals (all stages).

In 40 plots (each one sq m) in a Missouri woodland she found A. rudis

in abundance. Of 206 colonies representing 16 species, 62 colonies

were of A. rudis, the most numerous species. A single colony of A.

fulva was found. Headley (1949) censused the populations of 46 colo-

nies of A. rudis. They contained 58 to 1,440 individuals (all stages).

In the summer there were more brood than workers. He found eggs and

larvae in nests excavated during the winter. Only 11 of 36 colonies

sampled at the proper time had reproductive forms. Creighton (1951)

compared the ecological differences between A. huachucana Creighton

and A. texana Emery in Arizona. He found that A. texana occurs at

lower elevations, in habitats with a distinct Sonoran character and

nests in direct sunlight, while A. huachucana occurs at higher eleva-

tions, in habitats with Transitional or Canadian zone affinities and

nests in shade as well as sunlight.

Several authors performed ecological studies in which they asso-

ciated ant species inhabiting an area with particular plant communi-

ties and soil types in the area. These papers deal with all the ant

species in the areas studied, but they contain much valuable data

pertaining to Aphaenogaster spp. Among studies of this genre the most

pertinent are those of Van Pelt (1947, 1956, 1958) conducted in the

Gainesville area and at the Welaka Conservation Reserve, Putnam County,


The 1947 paper fails to include two species of Aphaenogaster

occurring in the Gainesville area, but was instrumental in helping me

find A. floridana. The later papers provide not only detailed data on

the habitats in which Aphaenogaster spp. occur at Welaka, but their

nesting habits and some biological information as well. Van Pelt (1966)

conducted a similar although less intensive study of the ants of the

Savannah River Plant, South Carolina. North Carolina has many of the

same species of Aphaenogaster as Florida. Carter (1962a, 1962p) car-

ried out ecological studies similar to Van Pelt's, but encompassing

the entire state of North Carolina. He reported on 12 recognizable

forms of Aphaenogaster and included data on activity and nesting habits.

Dennis (1938) and Cole (1940) undertook investigations of this sort in

Tennessee, where several species of Aphaenogaster occur. Talbot (1934)

and Gregg (1944), working in the Chicago region studied ant distribu-

tion in relation to plant communities. Talbot delved more deeply into

the physical factors affecting ant distribution, finding for example

that A. rudis picea Emery requires a higher humidity for survival than

does A. rudis rudis. She (1953, 1951) conducted similar but less ex-

tensive investigations of an old field in Michigan and of a Missouri

woodland. In Ohio Headley (1952) studied the ecology of the ants

found in locust woods. In 27 sq m plots he found a total of seven

colonies of A. rudis in five plots. Three species had more colonies

in the study area than A. rudis. A more general work of Headley's

(1943) on the ants of Ashtabula County, Ohio, provides some brief

ecological and biological notes on Aphaenogaster spp. Saunders (1967)

studied some ecological aspects of the pestiferous funnel ants in

Australia, but his emphasis was on habitat data.

The most detailed data on the biology of a single species of

Aphaenogaster were reported by Fielde (1901, 1903, 1904a, 1904b, 1905)

regarding A. rudis. Although most of her work was carried out in the

laboratory and much of it only peripherally pertinent to this study,

it does provide basic information on development, annual cycles and

reproduction. In her 1901 paper she reported that A. rudis swarms in

September in the northeast and that larval development generally takes

18 to 21 days. Some of the experiments she performed on A. rudis were

of a rather unique nature and similar comparable data for other species

of ants are scant if not wanting altogether. For example, Fielde

(1904b) kept an A. rudis worker alive for over three years, a decapi-

tated worker ten days, and another missing its prothoracic legs for one

month. She kept a queen without its abdomen alive 14 days. Of 18

workers she submerged in water for four days 12 fully recovered. One

unfed worker survived 46 days. Regarding longevity, Haskins and

Enzmann (1960) report queens of A. rudis live an average of 8.7 years

under laboratory conditions. One survived 13 years.

The only study of actual mating flights of any North American

Aphaenogaster is that by Talbot (1966) on A. treatae. Funnel ants, A.

(Nystalomyrmex) spp. mate in the spring after a period of cold wea-

ther. Apparently copulation takes place in the air. Regarding mating

Fielde (1901) found that A. rudis workers carried into their nests and

cared for conspecific males from alien colonies. This led her to sug-

gest that "probably cross-fertilization is common if not universal (p. 131).

Ledoux (1971) reported that another form of colony propagation may

exist in the European A. senilis Mayr. In the fall all colonies seem

to have one dealate female, but in the spring a portion of the colonies

are queenless. In the laboratory only colonies from which queens were

removed produced alate females, apparently from eggs of the former

queens. Of the orphaned alate females only one survives to mate with

one or more of the males which have developed from worker-laid eggs.

The number of colonies seems to increase by the budding-off of groups

of workers who propagate functional colonies by the means described

above. Pleometrosis, the formation of new colonies by budding-off,

is known for other Myrmicines, but is often associated with multiple

queens. In light of Ledoux' work a statement by Fielde (1901) regard-

ing A. rudis is interesting:

Colonies captured and confined in my nests just before

swarming time, within a few days divided into as many

groups as there were queens, the queens disposing them-

selves as far apart as the limits of the nest permitted.

When a queen was then removed by me, the workers at once

carried the young and settled down by another queen.
(P. 429)

As mentioned previously, Talbot (1951) found multiple dealated females

in A. rudis nests in the spring. Headley (1949) also found multiple

dealate females in A. rudis nests. Crozier (1973) did not find multi-

ple dealate females in A. rudis colonies in Georgia. His isozyme

studies led him to believe that if colonies aggregate in early spring

as suggested by Talbot they must separate along original family lines

with high fidelity. Crozier (1974) on the basis of allozyme analysis

found that in most A. rudis colonies males arise predominantly or only

from queen-laid eggs. Colonies with worker-derived males he feels are

probably those whose queens have died. Bruniquel (1972b) describes

worker oviposition by the European A. subterranea Latreille. Alimentary

and male producing eggs are laid by workers. The alimentary eggs are

smaller than queen-laid eggs and male producing worker-laid eggs. Has-

kins and Enzmann (1945) report an instance of thelotoky in A. rudis.

Of 100 virgin females one produced worker offspring. Only 18 produced

any offspring at all. They also mention a possible instance of

thelotoky in A. lamellidens. At one Georgia locality Crozier (1972)

found that at a cathodal malate dehydrogenase locus A. rudis queens

were all heterozygous, while workers were in approximately equal numbers

homozygotes and heterozygotes. This genotypic difference suggested to

him that differential selection exists between castes and that A. rudis

colonies consist of a single once-mated queen and her worker progeny.

Several European authors have studied the biology of Old World

species of Aphaenogaster. Bruniquel (1970a, 1970b, 1972a, 1972b) pub-

lished on the nesting habits, reproduction, and social biology of A.

subterranea. He (1970a) found that groups of A. subterranea workers

when deprived of their queen had to be comprised of at least 75 to 100

individuals if they were to maintain the semblance of a colony. Over

a given period of time the mortality rate for groups of 75 to 100

queenless workers was very low compared to the mortality rate for

groups consisting of fewer workers. The mortality rate was inversely

proportional to the number of orphaned workers. In the laboratory

Bruniquel (1970b) found that when presented a humidity gradient of 35

to 100 percent at 220 C, 92 percent of queenless workers and 93.55 per-

cent of workers with queen and brood nested within the range of 90 to

100 percent relative humidity. This he compares with published data

for other ants. In the arboreal Dolichoderus quadripunctatus L., for


example, workers live in the range of 35 to 50 percent relative humid-

ity, while the brood needs a relative humidity of 60 to 75 percent.

Bruniquel (1972a) describes two types of nests constructed by A. sub-

terranea; those completely hypogeaic and those partly in rotten logs

and extending into the soil. Ledoux (1967) found that the optimal

temperature range at which the normal activity of A. senilis took

place was 20 to 280 C. This was the temperature within the nest.

Temperature outside the nest, with the exception of sub-lethal tempera-

tures, only influenced speed. Above ground activity occurred between

ten and 350 C. When confined in a temperature gradient apparatus, A.

senilis colonies dwelt in the portion of the gradient having tempera-

tures 23 to 250 C.

Buschinger (1973) noted workers of A. subterranea carrying their

larvae to prey which had been brought into the nest. The larvae fed

directly on the food item. Fielde (1910) and Wheeler (1901) also re-

port this behavior in A. rudis. Wheeler and Wheeler (1953) gave dead

arthropods directly to unattended A. rudis; larvae consumed the items.

Fielde (1901) also reports that A. rudis workers regurgitate food to

young larvae. Wheeler and Wheeler (1953) describe a generalized

Aphaenogaster (Attomyrma) larva.

Moderately stout; constricted slightly at the first

abdominal somite; thorax turgid; abdomen swollen; no neck.

Submature larva with its diameter greatest at abdominal

somites IV and V, diminishing anteriorly; anterior end

bent ventrally; no neck; posterior end broadly rounded.

Body hairs moderately numerous and rather short. Of two

types: (1) with long stout base and short dichoto-

mizing tip, on the thorax and abdominal somites I, IX,

and X; (2) with short base and long flexible dichoto-

mizing branches, on abdominal somites I-VIII; inter-

grades rare. Head hairs moderately numerous and

rather long. Of two types: (1) simple and (2) with

bifid tip. Posterior surface of labrum densely spinu-

lose, the spinules minute and in short arcuate rows

which tend to form a reticulate pattern. Apex of

mandible forming a rather slender tooth which is

slightly curved medially; two stout round-pointed

medial teeth; medial surface of basal half with sev-

eral short to very long spinules. Maxillae with the

apex spinulose; palp digitiform. Dorsal portion of

hypopharynx with sublongitudinal ridges; ventral

portion spinulose, the spinules minute and arranged

in rows which form a reticulate pattern.
(p. 54-56)

They also describe the larvae of A. fulva, A. rudis, A. rudis picea,

A. tennesseensis, A. texana, A. treatae, and A. treatae pluteicornis

G. C. and E. W. Wheeler. Biological notes are included for some spe-

cies. Male and female prepupae of A. flemingi are described by

Wheeler and Wheeler (1972).

Behavioral studies on the geotropic response of A. fulva workers

placed on an inclined plane were conducted by Barnes (1930) and Barnes

and Skinner (1930). They found the ants had a greater positive geo-

tropic response the longer they were isolated from their nestmates.

This change they believed was not due to increasing muscular weakness

from lack of food or to the posture the ants assumed in the isolation


Wheeler (1910) concluded that A. tennesseensis is a temporary

social parasite of A. fulva or A. rudis. He found mixed colonies of

these ants. This combined with the aberrant appearance of A. tennes-

seensis females led him to his opinion. He suspected that the rare A.

marine whose females resemble those of A. tennesseensis is also a so-

cial parasite.

Little has been reported regarding the predators of Aphaenogaster

spp. Edwards et al. (1974) indicate that the jumping spider, Stoidis

aurata (Hentz) feeds on species of Aphaenogaster as well as other ants.

Brown (1958) describes an unusual instance of a slave-making ant,

Formica subintegra Emery, raiding a colony of A. rudis, killing workers

and stealing brood. This is peculiar since the usual victims of F.

subintegra raids are species of Formica belonging to the fusca and

pallidefulva groups.

Aphaenogaster spp. have been generally considered to be largely

predaceous and facultative seed collectors (Forel 1901, Wheeler 1910,

Talbot 1954, Smith 1965, Wilson 1971). Wheeler (1910) and Hendrickson

(1930) describe the collection of seeds and plant parts by A. rudis.

Goetsch (1942) reports leaf-cutting and collection by A. testaceopilosa

(Lucas) in Spain.

Aphaenogaster spp. predation upon termites has been documented by

several authors dating back to Forel (1901). Other authors have re-

ported Aphaenogaster spp. nesting adjacent to termite colonies or

preying on them (King 1897, Adams 1915, Van Pelt 1958). Reports of

termite predation were generally instances of the author having broken

open a termite nest exposing the inhabitants to attack by Aphaenogaster

foragers (e.g. Forel, 1901). More recent papers by Smythe and Coppel

(1964, 1973) and Beard (1973) question the actual destructiveness of

Aphaenogaster predation to undisturbed termite nests. Their attacks on

termites are fortuitous. Referring to Reticulitermes flavipes (Kollar)

Beard states:

Compatability also seems a characteristic of Aphaeno-

gaster spp. in spite of previously reported predatory

activity. A. rudis is one of the most frequently en-

countered ants in termite habitats, and only fortuitous

predation has been observed. In laboratory situations,

A. rudis and R. flavipes can live in the same container

with nonapparent conflict.

(p. 398)


Field Studies

Detection of Colonies

I employed a variety of methods to detect field colonies of

Aphaenogaster spp. Three techniques were of particular value and fre-

quently used.

1. I broke open rotten logs, limbs, stumps, and cypress

knees with an entrenching tool. This was the quickest method

of locating nests of several species of Aphaenogaster. Ex-

perience soon dictated exclusion from examination dead wood

too dry for Aphaenogaster inhabitation.

2. The substrate was scanned for foraging Aphaenogaster

workers. Upon spotting a worker, pupae of the red flour

beetle, Tribolium castaneum (Herbst), or small particles of

old cheese were scattered in the immediate vicinity (one cm)

of the ant. Almost invariably the ant picked up the food

and returned directly to her nest. Except when she dis-

appeared beneath leaves, the ant was easy to follow as either

the dark ant or her light colored burden contrasted with any

background. Even if the returning ant was lost from sight,

the general direction of the nest was indicated by the

direction she was headed. Often no nest entrance is

discernible for soil dwelling species, so disappearance

of the ant might mean she entered the nest. In any case,

recruited workers soon emanated from the nest toward the

food source, thus providing opportunities to pinpoint

the nest.

Headley (1949) and Talbot (1954) used a similar tech-

nique for locating Aphaenogaster nests, however they

sprinkled bread or cake crumbs as bait. Cheese worked

well; apparently the ants could detect its odor a few mm

away. A major drawback of this method is that its suc-

cess depends upon the extra nidal activity of the ants.

For soil dwelling species, above ground activity was

greatly curtailed during cooler and drier months and

often during the heat of the day in warmer months. The

drawback was particularly serious since no comparably

rewarding technique was found for locating nests of

soil nesting species.

3. I inspected trunks of living trees for Aphaeno-

gaster workers. Smaller or younger trees generally did

not have rotten or termite riddled portions suitable

for Aphaenogaster nests. As a rule trees of less than

seven cm dbh were not examined by me unless they bore

some obvious indication of extensive internal decay.

Arboreal nests were discovered by observation of ants

carrying food items up the trunks to their nest entrances.

A technique I occasionally employed was overturning stones. Al-

though it may be a rewarding method elsewhere, looking under stones

was of little value in Florida. There is in Florida a scarcity of

rocks of suitable size and shape beneath which Aphaenogaster spp.

might nest. Wilson (1959) reports that in many tropical areas the ant

fauna beneath stones is comparatively depauperate. In a New Guinea

rainforest less than ten percent of the rocks he overturned had ant

nests or foragers beneath them.

Sometimes the nest of a soil dwelling species was detected by

my visually recognizing the nest entrance. Such discoveries were usu-

ally fortuitous, occurring while looking for foraging workers. How-

ever, I found colonies of A. floridana most often by searching for

their nest entrances. This species regularly nests in rather open

sandy areas where its distinctive thatch encircled nest entrances

were easy to see unless situated at the base of a clump of wire grass.

A. floridana nest entrances were not always thatched.

A technique I used infrequently was to scrape away surface leaf

litter to expose the upper portions of nests of soil dwelling species.

This technique was usually much less effective than following workers

carrying food. Furthermore, this method was unsatisfactory because it

made collection of complete colonies almost impossible. Many immatures

and workers were lost in the leaf litter.

Collecting Trips

I took collecting trips to nearly every county in peninsular Flor-

ida. The panhandle was visited less frequently. Collection and dis-

tribution data may be somewhat biased toward north central Florida,

because these areas are more accessible from Gainesville and were

examined more extensively and intensively.

Collection of Colonies

Soil-dwelling colonies with the exception of A. carolinensis were

collected without losing significant portions of the population through

escape. Having ascertained the main entrance as accurately as possible,

I scraped leaf litter away from the entrance for a radius of about 25


Nests of soil-dwelling species rarely extended laterally more

than 20 cm from an imaginary line perpendicular to the ground and pass-

ing through the main nest entrance. Since a large portion of a colony

usually resided within 13 cm of the surface, one could, by removing in

one piece a cylinder of soil, ten cm deep and 36 cm diam with the nest

entrance at the center, secure in toto those ants most apt to escape.

Such soil cylinders were removed by thrusting an entrenching tool per-

pendicularly into the soil along the perimeter of the 36 cm diam imag-

inary circle to a depth of about 20 cm. Soil cylinders were dropped

directly into plastic garbage bags. The remaining ants could not

readily escape from the hole produced by the removal of the soil cylin-

der. The rest of the nest was excavated one shovelful at a time. The

soil was deposited in an area cleared of leaf litter adjacent to the

excavation, and sifted by hand to detect ants. Portions of soil con-

taining large numbers of ants were deposited in a plastic bag. The

excavation was expanded downward and laterally until no further ants

were discovered for 20 cm in any direction. A. floridana nests were

excavated to a minimum depth of 80 cm.

Nests of A. carolinensis were confined largely to the leaf litter

stratum. No nest entrances could be discerned among the leaf litter.

After several returning foragers disappeared into the leaf litter

within a few cm of each other, it was assumed that the nest was di-

rectly below that spot. An entrenching tool was thrust into the leaf

litter and soil at an angle of about 200 from the horizontal and to a

depth of about seven cm. One hand was placed on top of this material,

pressing it tightly to the blade of the entrenching tool. The entire

shovelful was deposited in a plastic garbage bag. The material in

the bag was quickly examined for brood to determine if the major part

of the nest was where it was suspected to be. Leaf litter adjacent to

the excavation was pushed toward the excavated area and examined,

making certain no peripheral parts of the colony were missed. Excava-

tion of the remainder of the nest continued as for any soil-dwelling


Success collecting in toto Aphaenogaster colonies nesting in rot-

ten logs or stumps depended upon the size of the colony, stage of

decay of the log and whether the nest extended into the soil. A large

colony in a slightly rotten log was difficult to collect in toto, par-

ticularly if some of the nest was in the ground. Well-decayed logs

were easily and rapidly broken down by hand or using an entrenching

tool. Portions of a nest log containing the colony were dropped into

plastic bags. Adults and brood were brushed by hand from large in-

frangible sections of logs into collecting bags. Occasionally small

colonies and specimens hidden in infrangible parts of logs were aspi-

rated. Removal of colonies from cypress knees and logs was especially


Except in winter, as soon as a nest log was broken open, the

workers began to run about rapidly. Many ants ran or dropped from dis-

turbed logs. This posed a serious problem when collecting large colo-

nies. Some nests extended more than a meter along logs. This diffi-

culty was somewhat circumvented by spreading plastic bags along or

under nest logs or by clearing leaf litter away from nest logs. Es-

caping workers were exposed for a longer time and could be aspirated

or gathered up by hand.

Colonies nesting in living trees were exceedingly difficult to

collect in toto. Often nests penetrated deeply into the heartwood of

large trees. Some nests were located several meters above the ground.

A colony was collected from a storm-downed tree by cutting a section

from the trunk with a chain saw. Nest trees frequently were too large

for the use of a chain saw. Trees were chopped open with a hatchet or

an axe. Many ants were crushed or escaped during the chopping.

Dissection of Colonies

Nests of soil-dwelling species were dissected according to a

rather widely accepted procedure. A trench, about 90 cm long, 30 cm

wide, and 60 cm deep, was dug tangentially to a 45 cm imaginary radian

from the main entrance of the nest. The trench wall nearest the nest

was carefully sheared away in decreasing decrements, exposing nest

chambers and galleries. Contents of the chambers were aspirated or

plucked out with forceps. The initial trench was lengthened as needed.

The procedure was continued until after digging 30 cm in any direction

no more ants were found.

Nests in wood were broken open carefully and gradually using a

hatchet, entrenching tool or penknife as the situation warranted. The

ants nested in preformed cavities in the wood. Sometimes despite being

riddled with cavities the wood was extremely hard, making careful dis-

section difficult. In every case, once the ants were disturbed, they

fled, carrying brood, into any tunnels or cavities adjoining the nest

proper. The true dimensions of nests in wood were, therefore, always

difficult to discern, except for incipient colonies.

Field Observations

Field observations were made year round. Stations where several

congeners coexisted in rather close proximity were frequent observation

sites. At stations where several Aphaenogaster spp. occurred I could

easily check and compare the activities of the various species under

apparently similar temporal and climatic regimes. I observed foragers

to determine where and when they foraged and their interactions with

other organisms. To study interspecific competition and foraging

ranges pieces of cheese and dead insects were set out in natural situ-

ations and observed. The feeding habits of other species of ants were

noted to determine those competing with species of Aphaenogaster. Ob-

servations of competing ant species were corroborated with. collections.

Within a two m radius of certain Aphaenogaster nests all ants seen

foraging on the ground or leaf litter were collected. Collections

were limited to one hour. Small ants were collected in an aspirator;

large ants by hand or with forceps.

Night observations were made with the aid of a battery powered

headlamp. The light from the headlamp seemed to diminish the activity

of nocturnal foragers upon which it was shone. A transparent piece of


red plastic affixed to the front of the lamp did not markedly inter-

rupt the activities of the ants, but seriously impaired my ability to

see the ants.

To determine some of the vertebrate predators of Aphaenogaster

spp., whenever possible amphibians and reptiles were captured and kept

until they defecated. The feces were placed in a petri dish with iso-

propanol and examined microscopically for remains of Aphaenogaster spp.

Cabbage loopers of different instars were released in the field

near Aphaenogaster spp. so that attacks upon the larvae by the ants

might be observed.

Colonies of A. ashmeadi and A. fulva were watched daily for sev-

eral weeks in June and July to observe mating flight activity.

Foraging Success

I set out 25 dead house flies (Musca domestic L.) several dis-

tances from the entrance of an A. ashmeadi nest. Four baits were 15

cm (II), four 21.2 cm (III), four 30 cm (IV), eight 33.5 cm (V), and

four 42.4 cm (VI) from the entrance. The remaining bait (I) was

placed adjacent to the nest entrance. Each of the baits for distances

II, III, IV, VI was on a radian 90 from the nearest bait the same

distance from the entrance. Each bait at distance V was on a radian

30 and 600 from the nearest baits at distance V. The radii of baits

of II and IV were identical and offset from the radii of III and VI

(also identical). To 22.50 either side of radii II/VI were radii of

baits of distance V. One at a time each bait was placed down and ob-

served. The species of the first ant to find and begin to feed on or

remove the bait was recorded. I placed the baits in a sequence which

avoided the complications of possible recruitment resulting from pre-

viously placed baits.

Ten house flies were placed in the area described above. The

sequence of ant species removing or feeding on the baits was recorded.

Seven house fly heads were placed in the area and observed to see

which ants successfully removed to their nests.

Two Pheidole metallescens Emery colonies were in the area between

the A. ashmeadi nest and the outermost baits.

Laboratory Studies

Maintenance of Colonies

Colonies of nine species of Aphaenogaster were maintained in the

laboratory. When brought into the laboratory field collected colonies

were still in the soil or wood in which they nested in their natural

situations. This material and the ants within were emptied from plastic

collection bags into plastic tubs of various sizes. The sides of the

tubs were dusted with talcum powder, preventing the ants from escaping.

Some colonies were allowed to remain in the tubs with their natural

nesting substrate. Others were transferred to smaller containers per-

mitting better observation.

Plastic shell vials served as artificial nests. A moistened wad

of absorbent tissue paper was pushed inside a transparent shell vial,

filling about one-half its volume. A plastic snap cap with a hole

(five mm across) in the center was affixed to the open end of the vial.

The capped vial was placed in a tub containing a newly collected ant

colony still nesting in its natural substrate. The vial was buried in

the substrate leaving only the capped end exposed. The natural sub-

strate was allowed to dry for a few days. When the humidity in the

natural substrate became low, the colony moved into the more humid

vial. Often the vial was not placed in the tub until the substrate

had already begun drying for a few days.

Colonies were maintained in plastic vials or transferred into

plastic petri dish nests or Wilson cells. Petri dish nests had one

or more holes in the lids or sides. Damp tissue paper was placed in-

side petri dish nests and Wilson cells. Shell vial nests, petri dish

nests, and Wilson cells were placed in trays which served as foraging

arenas. Sides of the trays were dusted with talcum powder. Individual

queens or queens with incipient colonies of less than 20 workers were

maintained in closed petri dish or shell vial nests.

High humidity was maintained in laboratory nests. Water from a

squeeze bottle was squirted onto natural substrate daily or onto tis-

sue in artificial nests at weekly intervals. Water evaporated rather

slowly from the semi-enclosed artificial nests. However, the ants

continually tore off small pieces of tissue and carried them from

their nests. Therefore, new tissue had to be placed in the artifi-

cial nests every few weeks.

I fed the ants largely proteinaceous food principally in the form

of larvae and pupae of the red flour beetle, potato tuber moth

(Gnorimoschema operculella Zell.) and cabbage looper (Trichoplusia ni

Hubner). To a lesser extent a variety of other arthropods and meat

table scraps were fed to the ants. Ant media (Bhatkar and Whitcomb,

1970) was frequently fed to the ants.

The laboratory in which the ants were maintained was heated or

cooled by an airconditioner every few days during the course of the


Determination of Stadial Duration

To determine duration of the egg stage I placed viable queens and

four to ten of their workers in closed petri dish nests. The time

elapsed between the day the first egg (eggs) was laid and the day of

the eclosion of the first larva (larvae) was defined as the duration of

the egg stage.

The duration of the larval stage was defined as the time elapsed

between the eclosion and pupation of the first larva. Differential

feeding of larvae by workers varied the rates of larval development,

making it difficult to monitor the development of several larvae in a

single colony. The time elapsed between the day of pupation of the

first larva and the day of its eclosion was the duration of the pupal

stage. Durations of several subsequently pupating larvae were also


Durations of larval and pupal stages were determined by a similar

technique. Eggs were removed from queens and those laid within the next

24 hrs were placed in petri dish nests with four to six sibling workers.

The development of eggs could be monitored with more certainty than by

the method described above. Eggs were placed in petri dish nests with

sibling workers and the duration of larval and pupal stages determined.

Prepupae were placed in petri dish nests with sibling workers and the

duration of the pupal stage determined.

Development of colonies from founding queens collected in the

field were monitored in a similar fashion. Queens and their brood,


if any, were placed in petri dish or shell vial nests and growth of

the colonies recorded.


General Appearance of Aphaenogaster Spp.

Ants of the genus Aphaenogaster are morphologically rather gen-

eralized. They do not depart radically from the basic Myrmicine form.

The genus keys out in the final couplet of Creighton's (1950) key to

the Myrmicinae of North America. Smith (1947) provides a brief descrip-

tion of the genus.

Monomorphic. Medium-sized (3.25-7mm), slender

ants. Head usually distinctly longer than broad; in

some forms very noticeably narrowed posteriorly (when

the head is much narrowed posteriorly the antennae

are usually long and slender). Eye generally promi-

nent, usually not placed far from the middle of the

side of the head. Frontal carinae short, not distant

from each other. Anterior border of clypeus usually

with a distinct median emargination. Antenna 12-

segmented, the last 4 segments enlarged but not

forming a very definite club, the last 3 segments

shorter than the rest of the funiculus. Base of the

scape with a prominent lobe in some forms (treatae

Forel and its variants). Thorax usually with'a

distinct promesonotal suture. Anterior portion of

mesonotum sometimes protuberant. Posterior third

or more of mesonotum often very noticeably impressed.

Mesoepinotal suture distinct. Epinotum usually with

a pair of spines of variable length (in only a few

forms are the spines vestigial or almost absent).

Petiole generally pedunculate anteriorly, sometimes

very much so. Gaster oval, not truncate basally,

rather shining except for the sculpturing at the

base in a few forms.
(p. 555-556)

Brown (1974) synonymized the genus Novomessor, which occurs in the

western United States, with Aphaenogaster. Smith's description is

still functional for Florida species. See Figure six for line drawing

of Aphaenogaster worker.

Species of Aphaenogaster in Florida

Eight species of Aphaenogaster, all belonging to the subgenus

Attomyrma, were collected in this study.

A. ashmeadi Emery

A. carolinensis Wheeler

A. flemingi M. R. Smith

A. floridana M. R. Smith

A. fulva Roger

A. lamellidens Mayr

A. tennesseensis (Mayr)

A. treatae Forel

The rare A. mariae Forel has been reported from Florida (Creighton,

1950), but I never found this insect in the state. See section on

taxonomy for discussion of other species reportedly occurring in


Key to Workers of Florida Species

1. Antennal scape with rearward extending lobe along at least

basal fifth of scape . . . . . . . . ... . 2

SAntennal scape without conspicuous rearward extending lobe .. 3

2. Lobe of scape extending one fifth of scape length, lobe

thin and flattened; specimens from peninsula blackish .. ashmeadi

Lobe of scape extending one quarter of scape length,

lobe thick: specimens brown, gaster somewhat darker .treatae

3. Base of antennal scape with small forward projecting

lobe; in frontal view head two times longer than wide. . . 4

Antennal scape without forward projecting lobe; head in

frontal view less than two times longer than wide . . . 5

4. Propodeum with pair of long slender spines; in profile

petiolar node acute . . . . . . .... flemingi

Propodeum without spines; in profile petiolar node

low and rounded . . . . . . . .... floridana

5. Postpetiole with laterally compressed anteriorly

projecting flange along ventral midline; postpetiole

wider than long . . . . . . . .... .... 6

Postpetiole without forward projecting ventral

flange; postpetiole longer than wide . . . . . . 7

6. First gastric segment with striations spreading fan-like

from point of articulation of postpetiole and covering

basal half of dorsum of segment; dorsum of first gastric

segment with erect hairs. . . . . . . . mariae

First gastric segment without striations, or if

present then only extending over basal fifth of dorsum

of segment; dorsum of first gastric segment devoid of

erect hairs. . . . . . . . . ... tennesseensis

7. Outer face of frontal lobe with rearward projecting

tooth-like flange; specimens reddish with yellowish

gaster . . . . . . . . ... . . lamellidens

Outer face of frontal lobe without tooth-like flange;

specimens brown to black, or if reddish then gaster

dark brown . . . . . . . . ... . . ... 8

*8. Anterior border of mesonotum rising abruptly above

contiguous portion of pronotum, viewed anteriorly the

transverse welt thus formed distinctly depressed in

middle; propodeal spines often as long as declivious

face of propodeum; in frontal view head, excluding

mandibles often as broad as long . . . . .. fulva

Anterior border of mesonotum not rising abruptly above

contiguous portion of pronotum, or if it is higher then

anterior edge does not form transverse concave welt;

propodeal spines rarely as long as declivious face; in

frontal view head, excluding mandibles, not as broad

as long . . . . . . . . .. . carolinensis

*Spines sometimes shorter than declivious face in nanitics of fulva.

Appearance of Workers of Florida Species

See generic description for general appearance.

A. ashmeadi Emery- Length 6.0 7.5 mm. Head 1.4 mm long, 1.2 mm

wide.* Head narrowed posteriorly. Antennal scapes each with

flattened rearward projecting lobe. Lobe extending along basal

fifth of scape. Propodeal spines shorter than declivious face of

propodeum. Postpetiole large as in A. flemingi, A. floridana, and

A. treatae. Head with punctations and longitudinal rugae covering

frontal and lateral portions. Alitrunk, petiole and postpetiole

largely punctate; weak transverse rugae on pronotum and coarse

transverse rugae on dorsum of propodeum; some weak granulations

and rugae on petiole and postpetiole. Piceous; some specimens

from Florida panhandle not as dark. Most similar to A. treatae,

but differs in having smaller flatter basal antennal lobes,

darker color and fine punctations covering only about basal third

of first gastric segment.

A. carolinensis Wheeler- Length 3.7 6.9 mm. Head 1.2 mm long, 1.0 mm

wide. Head narrowed posteriorly less than in preceding species.

Antennal scapes simple. Propodeal spines rarely longer than

declivious face of propodeum. In large specimens spines somewhat

incurved, while in smaller individuals spines often diverge.

Postpetiole not as voluminous as in preceding species. Head with

punctations and longitudinal rugae covering frontal and lateral

*For each species head length measurement does not include mandibles,
but mandibles included in total length measurements. Head widths
measured in frontal view across widest part of heads. Total length
equals summation of length of tagma.

portions in larger specimens; in smaller specimens longitudinal

rugae confined to area of frontal longitudinal midline and later-

ally posterior and anterior to compound eyes. Alitrunk, petiole

and postpetiole largely punctate; pronotum rather shining in small

specimens. Transverse rugae on dorsum of propodeum and in larger

specimens on humeral areas as well. Not as rugose as A. fulva.

Head, alitrunk, petiole and postpetiole reddish or brown to black;

gaster darker in all specimens, sometimes yellowish brown distally.

Nanitics of A. fulva are similar; see description of A. fulva for

details. Differences between A. carolinensis and A. rudis and A.

texana are discussed in "rudis/texana group" taxonomy section.

A. flemingi M. R. Smith- Length 6.5 7.6 mm. Head 1.4 mm long, 1.2 mm

wide. Head narrowed posteriorly; among Floridian congeners only

A. floridana with head as narrowed. Small angular lobe projecting

forward from base of each antennal scape. Lobe less than two-

thirds the size of basal lobes on scapes of A. ashmeadi and A.

treatae. Propodeal spines about as long as basal face of propodeum.

Postpetiole voluminous as in A. ashmeadi, A. floridana, and A.

treatae. Head with punctations covering frontal and lateral por-

tions. Longitudinal rugae weaker than in A. ashmeadi and A.

treatae and largely confined to area of longitudinal midline and

laterally anterior and posterior to compound eyes. Alitrunk,

petiole and postpetiole largely punctate. Pronotum feebly punc-

tate and smoother; shining even in largest specimens. Alitrunk

lacking rugosity; among Floridian congeners only A. floridana

with alitrunk devoid of rugae. Castaneous with gaster slightly


A. floridana M. R. Smith- Length 6 7.2 mm. Head 1.4 mm long, 1.2 mm

wide. Head narrowed posteriorly, more so than in Floridian con-

geners. Small angular lobe projecting forward from base of each

antennal scape; lobes about the size of those on scapes of A.

flemingi. Only Floridian Aphaenogaster without propodeal spines.

In profile petiole with lower more rounded node than in Floridian

congeners. Postpetiole large as in A. ashmeadi, A. flemingi, and

A. treatae, but more elongate and with lower node. Head capsule

with punctations covering frontal and lateral portions except

occipital area which is nearly smooth. Longitudinal rugae largely

confined to area of head anterior to compound eyes. Pronotum

shining and nearly smooth; remainder of alitrunk moderately to

feebly punctate. Petiole and postpetiole punctate laterally;

smooth and shining above. Yellow to light orange brown; gaster


A. fulva Roger- Length 4.5 6.0 mm. Head 1.1 mm long, 1.1 mm wide.

Head excluding mandibles about as long as broad; in frontal view

occiput laterally roughly angularly convex. Antennal scapes

simple but thicker than in preceding species. Anterior border of

mesonotum rising abruptly above contiguous portion of pronotum;

viewed from anterior the transverse welt thus formed is distinctly

depressed in middle. Propodeal spines generally more upwardly

directed than in Floridian congeners; spines about as long or

longer than basal face of propodeum. Postpetiole scarcely more

voluminous than petiole; postpetiole not as elongate as in any

aforementioned species. Heavily sculptured. Punctations and

coarse longitudinal rugae covering frontal and lateral portions of

head; in small specimens longitudinal rugae confined to area of

longitudinal midline and laterally anterior and posterior to com-

pound eyes. Alitrunk, petiole and postpetiole with more rugosity

than in preceding species. Transverse rugae extending across

pronotum, particularly in larger.specimens. Dorsum of propodeum

with transverse rugae. Brown to piceous. Nanitics similar to A.

carolinensis and A. rudis; less heavily sculptured, heads narrower

and spines often shorter. However, mesonotal welt character seems

to hold even in nanitics. Reproductives easily distinguished.

Females of A. fulva with mesothoracic sternite and episternite

heavily rugose, while in A. carolinensis these sclerites are

mostly smooth and shining. Males of A. fulva with pair of thick

propodeal spines; males of A. carolinensis rarely with propodeal

spines, head narrow.

A. lamellidens Mayr- Length 6.0 7.7 mm. Head 1.5 mm long, 1.3 mm

wide. Head not as wide posteriorly as in A. fulva, but wider than

in A. carolinensis. Antennae simple. Outer face of frontal lobe

with posteriorly projecting tooth-like flange. Compound eyes

smaller than in A. carolinensis. Propodeal spines shorter than

basal face of propodeum. Postpetiole not as large as in A. ash-

meadi, A. flemingi, A. floridana, and A. treatae. Punctations and

longitudinal rugae covering frontal and lateral portions of head.

Alitrunk, petiole and postpetiole largely punctate. Transverse

rugae on dorsum of propodeum and pronotum. Brick red with dark

legs and reddish yellow gaster; gaster sometimes darker in pinned

specimens. In the field A. lamellidens superficially like A.

tennesseensis, but morphologically most like larger forms of A.

carolinensis. A. lamellidens distinguished from Floridian con-

geners by flange on frontal lobe.

A. mariae Forel- Length 5.3 mm. Head 1.1 mm long, 1.0 mm wide. Head

shape similar to that of A. lamellidens, but slightly narrower.

Antennae simple. Propodeal spines curved and longer than declivious

face of propodeum. Viewed dorsally postpetiole broader than long.

Postpetiole with laterally compressed anteriorly projecting flange

along ventral longitudinal midline; of Floridian congeners only A.

tennesseensis with similar flange. Very heavily sculptured. Pro-

nounced punctations and coarse reticulo-longitudinal rugae cover-

ing frontal and lateral portions of head. Dense punctations and

coarse mostly reticulate rugae over most of alitrunk. Petiole and

postpetiole largely punctate. First gastric segment with stria-

tions spreading fan-like from point of articulation of postpetiole

and dorsally covering basal fifth of segment. Only specimen exam-

ined reddish brown. Differs from A. tennesseensis in having erect

hairs on dorsum of first gastric segment and in having striations

on first gastric segment.

A. tennesseensis (Mayr)- Length 6.0 6.6 mm. Head 1.2 mm, 1.2 mm

wide. Head wide posteriorly; similar in shape to head of A. fulva.

Antennae simple. Propodeal spines curved and longer than declivi-

ous face of propodeum. Viewed dorsally postpetiole broader than

long; not as elongate or voluminous as in A. ashmeadi, A. flemingi,

A. floridana, or A. treatae. Postpetiole with ventral flange as

described for A. mariae, but slightly smaller. Heavily sculptured,

but less so than A. mariae. Pronounced punctations and reticulo-

longitudinal rugae covering frontal and lateral portions of head;

reticulations not as coarse as in A. mariae. Alitrunk except be-

tween propodeal spines largely covered with dense punctations and

coarse mostly reticulate rugae. Petiole and postpetiole largely

punctate. First gastric segment without striations covering basal

fifth as in A. maria. Unique among Floridian species of Aphaeno-

gaster in having dorsum of alitrunk, petiole, postpetiole and

gaster devoid of erect hair. Reddish with yellowish gaster.

A. treatae Forel- Length 6.4 7.6 mm. Head 1.4 mm long, 1.2 mm wide.

Head narrowed posteriorly; similar in shape to that of A. ashmeadi.

Antennal scapes each with thickened rearward projecting lobe.

Lobe extending along basal one-quarter of scape; lobe larger and

more three dimensional than in any Floridian congener. Propodeal

spines shorter than declivious face of propodeum. Postpetiole

large and elongate as in A. ashmeadi, A. flemingi, and A. floridana.

Punctations and longitudinal rugae covering frontal and lateral

portions of head. Alitrunk, petiole and postpetiole largely punc-

tate; transverse rugae on pronotum and coarser transverse rugae on

dorsum of propodeum. First gastric segment often with fine punc-

tations basal one-half of dorsum. Brown with darker gaster. A.

ashmeadi similar; see description of that species for differences

with A. treatae.

Taxonomy of rudis/texana Group

The genus Aphaenogaster is in dire need of taxonomic revision. A

complex of disputed forms contains some of the most common representa-

tives of the genus in Florida. Other species or forms seem to be


erroneously attributed to the state. Creighton (1950) recognized A.

rudis and A. texana as species. Both were originally described as

varieties of A. fulva and subsequently reported as occurring in Florida.

Under present taxonomy both names are referrable to several forms; some

of which may well be sibling species, while others may be geographic

subspecies or intergrades. A. rudis has a long history of nomenclatorial

confusion as well. This is not a taxonomic revision; few specimens from

outside Florida were examined and the types of A. rudis and A. texana

are not reposited in this country. In lieu of such a desperately

needed study, what follows is an attempt to briefly and pragmatically

deal with the multiplicity of names and forms involving Floridian rep-

resentatives of the "rudis/texana group."

For purposes of discussion first A. rudis and associated forms are

treated and then A. texana and its associated forms. However, in the

ensuing discussion it should become apparent that there does not seem

to be, following the current convention, a "rudis group" distinct from

a discrete texanaa group."

Creighton (1950) recounted the confused taxonomic history of A.

rudis and proposed certain changes of his own. In essence Creighton

discarded the long standing name aquia as representing an unrecognizable

form described by Buckley. According to Creighton aquia had been placed

in synonymy with A. fulva by Mayr, perhaps on the basis of type speci-

mens. However, Emery (1895) resurrected the name aquia based on speci-

mens Creighton believed were intergrades, perhaps between the forms

rudis and picea also described by Emery in his (1895) paper. All three

were considered by Emery as varieties of A. fulva. Creighton suggested

Emery's (1895) rudis be recognized as a species and picea as a northerly

and high altitude subspecies of rudis. In addition to the nomencla-

torial maze the actual complex of forms is a source of confusion.

Karyological and isozyme studies by Crozier (1970, 1973) led him to

suggest that at least three distinct forms exist in the eastern United

States. Two of these forms he believes may be sibling species. Assum-

ing the validity of Crozier's work and accepting Creighton's abandon-

ment of Buckley's aquia, what in fact is the rudis redescribed by Emery

and accepted by Creighton? To what does the name rudis apply? The

type locality of Emery's rudis was restricted by Creighton to Virginia.

Smith (1930) and Wheeler (1932) report examining specimens of A.

fulva aquia from Royal Palm Park and Dunedin. Smith (1965) stated that

A. rudis occurs in Florida. Van Pelt (1958) found specimens at Welaka

which he believed resembled A. rudis, but he reserved judgment on the

matter. Apart from the questionable application of the name aquia to

any form, these records seem dubious. Based on the examination of

specimens from Florida I believe that the name A. rudis is not justifi-

ably applicable, even in a rather broad sense, to any form I know to

occur in peninsular Florida. Perhaps some isolated enclaves exist near

the Georgia border, however, if A. rudis does occur in Florida at all

it would seem more likely to be found in the cool shady ravines in the

panhandle near the Alabama border, which harbor more northern ant spe-

cies. The Royal Palm Park and Dunedin records are so far south as to

be highly unlikely. Smith's concept of A. rudis must be very broad, be-

cause specimens (from Dade County) determined by Smith as A. rudis cor-

respond perfectly with Wheeler's description of A. miamiana. Smith's

identification was subsequent to Wheeler's description. Creighton

(1950) seemed to support Smith with a statement epitomizing the entire

rudis/texana dilemma. Creighton stated that A. miamiana, which was

described as a variety of A. texana, "seems rather closely related" to

A. rudis (p. 146). None of the Floridian forms that I have examined

have had the smaller compound eyes, thicker prothorax, or shorter post-

petiole associated with rudis-like forms from more northern states. Some

more northern specimens (Iowa) examined had heads of about the same narrow

proportions as A. miamiana. Nanitic specimens with narrower heads and

more slender alitrunks lend to the confusion. Carter (1962b) in North

Carolina could not determine with certainty some collections from the

piedmont and coastal plain forests as either A. rudis or A. texana. In

this paper the name rudis is not applied to any Floridian forms.

Emery (1895) described Stenamma (Aphaenogaster) fulvum var. texanum

from two workers, which he indicated were honey-yellow. Twenty years

later Wheeler redescribed texana, raised it to species level and de-

scribed two variants of it, furvescens and carolinensis. Menozzi

(1929) described A. (Deromyrma) silvestrii from a worker and dealate

female collected in Gainesville, Florida. Silvestrii was relegated to

the status of a texana variant in 1932 by Wheeler, who at the same time

noted Menozzi's error of placing silvestrii in the subgenus Deromyrma.

In the same paper Wheeler described two more variants of texana from

Florida. Based on specimens from Dade and Monroe Counties he described

the variant miamiana, while he described the variant nana from three

very small (2.5 mm long) workers from Gainesville. Smith (1934) de-

scribed A. texana macrospina from coastal South Carolina. In 1950

Creighton attempted to clarify the taxonomy of texana. He dismissed

Wheeler's nana and two other more northerly forms, pusilla Emery and

punctithorax Cole, as based on nanitics and "impossible of exact

recognition." Creighton acknowledged, in addition to the typical tex-

ana, the forms furvescens, carolinensis, and silvestrii. He believed

carolinensis exhibits sufficient geographic allopatry warranting sub-

specific designation, but he synonymized furvescens and silvestrii with

the typical texana. He raised macrospina and miamiana to species level.

Creighton (1950) and Van Pelt (1958) reported that A. macrospina occurs

in Florida. The description of macrospina was based on small specimens

from an immature colony, so in 1958 Smith synonymized macrospina with

flemingi. A. flemingi, although with definite affinities to the typical

texana, is a very distinctive species and concerns us no further in this


A. texana and its forms have been reported from Florida by several

authors. The type locality of Menozzi's silvestrii is Gainesville.

Smith (1930) and Wheeler (1932) recorded the variant furvescens from

Florida. Wheeler (1932) in addition to describing the variants nana

and miamiana from Florida, reported examining silvestrii-like specimens

from Gainesville. Creighton (1950) implied that not only the typical

texana, but also the variants furvescens and silvestrii occur in Flor-

ida. Van Pelt (1958) reported collecting A. texana at the Welaka Re-


Certain difficulties arise in the application of the name texana

to the Floridian forms of Aphaenogaster examined in this study. Even

in a broad sense, including the variants acknowledged by Creighton

(1950), silvestrii and furvescens, the name does not seem to refer to

Floridian forms.

No specimens of Aphaenogaster from Florida examined by me were con-

colorously honey-yellow and also bore propodeal spines as per Emery's

original description of texana. Specimens from Brownsville, Texas,

corresponded closely to Emery's description. They differed markedly

from any specimens from Florida. The Brownsville specimens, aside from

color differences, had heads distinctly more narrowed posteriorly, less

rugosity, and larger size. If texana is a monotypic species then I

know of no specimens of it from Florida.

Wheeler (1915), however, complicated matters. Making no mention

of a honey-yellow form, he redescribed the typical texana as uniformly

light ferruginous. Much of the description is in the form of a compar-

ison with the very different A. fulva and thus of limited value. Mor-

phologically the Brownsville specimens fit Wheeler's description.

According to Wheeler's description, the variant furvescens differs from

the typical texana only in its coloration. Furvescens is reddish brown

with a dark brown gaster. He fails to offer an explanation regarding

the disparity between the sizes he gives for females of the two forms;

texana 11 11.5 mm and furvescens 7.5 mm. Creighton (1951) inter-

preted this incongruity as that the texana measurement included the

wings, while the furvescens measurement was based on a dealated speci-

men. Judging from the size of the Brownsville workers, their queen

(dealated) would be surprisingly small if it were only 7.5 mm long. In

fact in a later paper Wheeler (1932) mentioned that the female of the

typical texana measured 8 8.5 mm. If furvescens differs from the

Brownsville specimens only in color, the author has seen no Floridian

forms referrable to furvescens. Some Floridian forms fit the descrip-

tion of the coloration of furvescens, but most of these conform closely

to Wheeler's (1932) description of miamiana. Wheeler, author of both

furvescens and miamiana maintained that both occur in south Florida.

Could he have mistaken some nanitics of miamiana for furvescens? He

(1932) stated regarding miamiana that it is "More robust and averaging

larger than other forms of texana: head broader and less narrowed be-

hind, ... antennal scapes stouter and slightly shorter; ... sculpture

decidedly coarser throughout (p. 5). The characters quoted all would

serve to separate workers from mature miamiana colonies from nanitics

of the same species. The unquoted characters are also rather variable

in miamiana-like specimens examined by the author and in many instances

would not permit separation of texana from miamiana. Wheeler (1932)

cited only a single record of furvescens from Florida, that being from

Royal Palm Park. One further anomaly, miamiana-like specimens examined

by the author tended to be smaller than the Brownsville specimens just

the opposite of Wheeler's description. It would seem that if furvescens

occurs in Florida it must be extremely rare.

Having disposed of the names texana and furvescens, there remains

the problem of the forms themselves. That is, what names do we apply

to them. A major difficulty is the question of sibling species in the

texanaa group." Reproductive isolation has not been demonstrated be-

tween any of the Floridian forms with affinities to texana. Based on

morphology and coloration there seems to be a gradation of forms with

the typical miamiana at one extreme. Creighton (1950, 1951) in his

treatment of the forms of texana neglected to mention the possibility of

sibling species. The fact that dark forms and light forms could be

found throughout most of the range of texana led him to conclude that

these forms, except for carolinensis, were merely variants. Because of

their sympatry they could not be considered subspecies. However,

carolinensis occurred in rather pure stands so he felt it was a true

geographic subspecies. In his 1951 paper Creighton re:itted the diverse

localities and habitats from which texana sensu latu has been reported;

abundant in Miami, Florida, damp woods at an elevation of 1800 feet in

the Great Smoky Mountains, at an elevation of 900 feet in the plains of

central Kansas, shady ravines in central Texas, Grand Canyon, Arizona,

and at the 6,000 foot level on sunny slopes in canyons of mountains in

southern Arizona. Responding to this data, Creighton (1951) stated:

All these stations lie south of the 38th parallel,

but I fail to see that there is any other feature which

they share that could be cited as a reason why they were

selected as nest sites by texana. Under such circum-

stances it appears hopeless to attempt a generalization

which will cover the behavior of texana over its entire

range. However the situation is by no means difficult

if field observations are limited to a particular part

of the range.
Cp. 90)

Such ecological data would seem to indicate the existence of, if not

more than one species, then at least of geographic subspecies. An ant

species with such a range would be rather unique among the North Ameri-

can fauna. In this regard the original appelation of furvescens to an

Arizonan form makes its application to a Floridian form slightly more

suspect. Until evidence of reproductive isolation between any of the

forms in the "texana/rudis complex" is demonstrated taxonomic judgment

on various forms is tentative at best.

At one extreme among the Floridian forms are specimens recognizable

as Wheeler's A. texana carolinensis. Some of these Floridian specimens

were compared by Dr. W. F. Buren with type material. This form was

characterized by its small size (all castes), minimal rugosity, faint

femoral punctation, slender alitrunk, rather short propodeal spines,

comparatively convex compound eyes. It differs most markedly from the

typical miamiana. See Figure six. Specimens examined included many

from mature colonies containing reproductive forms. This small form

diverges from the typical miamiana in the size of its colonies, which

are less populous. The carolinensis-like form regularly constructs

small rather superficial nests in leaf litter and soil. In north Flor-

ida it is commonly found in drier habitats in areas where miamiana-like

forms are largely restricted to river swamps or hydric hammocks. Mating

flights of the two extreme forms do not seem to be chronologically iso-

lated from one another. The smallest specimens of the carolinensis-like

form approach the measurements given by Wheeler (1932) for nana.

At the opposite extreme is the typical miamiana. Specimens from

south of Lake Okeechobee fit Wheeler's (1932) description very closely.

Figure one depicts the distribution of miamiana-like and more carolin-

ensis-like forms in Florida. These specimens are readily distinguish-

able from carolinensis-like forms by the characters given above. I

found only the miamiana-like form south of Lake Okeechobee. Forms less

typical of miamiana occur in central and north Florida. There speci-

mens are generally darker, with the head, alitrunk, petiole and post-

petiole brown to almost piceous. In north and central Florida those

specimens resembling the typical miamiana in color are more restricted

to drier habitats; however, they are morphologically more similar to

the carolinensis-like form. The small specimens of the northern forms

do not seem distinct from larger carolinensis-like specimens in morphol-

ogy or color. The description of Menozzi's A. silvestrii seems to be

of such intermediate forms.

The two extreme forms in the absence of demonstrable reproductive

isolation will have to be treated as a single species. To assign a name

to these forms one must go back to A. texana carolinensis (Wheeler,

1915), which preceded silvestrii (Menozzi, 1929) and miamiana (Wheeler,

1932). As discussed above Floridian forms seem quite distinct from

the typical texana and the variant furvescens as I interpret them. In

my opinion the Floridian forms represent a separate species from texana

and the name A. carolinensis is being applied to them. The only other

name possibly applicable is furvescens (Wheeler, 1915), and only if the

type material does conform to the description. Described from Tryon,

North Carolina, carolinensis has not been previously recorded from

Florida. Carter (1962) reported it from the piedmont and fall line

sand hill areas of North Carolina, but did not find the typical texana

in that state. If one only considers the populations in Florida, one

might interpret the frequent occurrence of small forms (typical caro-

linensis) in drier areas in north and central Florida as a response to

adverse conditions in such habitats. Females produced in such enclaves

might also tend to be small being reared under less than optimal condi-

tions. Congeneric competition might be important in this regard. In

south Florida miamiana-like forms occur in a wide variety of habitats

but have little or no congeneric competition. Judging from some col-

lections from central Florida one might also hypothesize a north-south

color cline exists in the complex. Considering the overall distribu-

tion of the typical carolinensis and the miamiana-like forms another

interpretation is possible. The two extreme forms may be geographic

subspecies with a zone of hybridization in north and central Florida.

In North Carolina carolinensis seems to be prevalent in the same gen-

eral habitats as in Florida. According to Carter (1962b), in North

Carolina carolinensis is common in upland forests and "bottomland col-

lections were generally made in well-drained, sunny and sandy loca-

tions." Perhaps, resolution of the problem rests in karyological and

isozyme studies like those of Crozier (1970, 1973) and Imai (1971),

laboratory mating tests or nutritional investigations.

Findings on A. ashmeadi

Description of Female

TL 9.3, HL 1.8, HW 1.5, SL 1.9, AL 3.1 mm.* General appearance as

in Figure three. Head shape as in Figure three; not as narrowed poste-

riorly as in workers. Conspicuous rearward projecting lobe extending

along basal one-fifth of antennal scape. Lobe flatter than in A.

treatae females. Lobe in A. treatae females longer, extending one-

quarter scape length. Scape bent at slight angle at distal end of lobe.

Anterior margin of clypeus somewhat more deeply incised than in A.

treatae females.

Alitrunk as in Figure three; propodeum with pair of short poste-

riorly directed spines, as in females of treatae and floridana. Spines

shorter than declivious face of propodeum.

*TL=total length; HL-head length; HW-head width; SL=scape length;
AL=alitrunk length.

Petiole and postpetiole as in Figure three; rather similar to

those of A. treatae females. Postpetiole distinctly more voluminous

than in females A. carolinensis, A. fulva, A. lamellidens, and A.


Mandibles with longitudinal striae which spread slightly as they

approach the masticatory border. Clypeus with slightly wavy longi-

tudinal rugae. Clypeal rugae distinctly less coarse and less reticu-

late than in A. treatae females. Coarse longitudinal rugae over front

and lateral portions of head. Some transverse connections between

rugae. Punctations interspersed between rugae. In occipital region

rugae weaker and reticulate to transverse. Pronotum with rugae paral-

lel to suture between pronotum and scutum. Scutum longitudinally

rugose; rugae somewhat fainter along longitudinal midline of scutum.

Longitudinal rugae continue across parapteron to scutellum, where

rugae are slightly more reticulate. Propodeum with transverse rugae;

area between spines with more feeble rugae. Mesothoracic sternite and

episternite with rather longitudinal rugae; rather smooth and shiny

ventrally. Coxae and femurs shining, but with some punctations. Coxae

with a few weak rugae. Petiole punctate with some dorsal and lateral

rugae. Postpetiole punctate with dorsal and lateral rugae tending to

be transverse. Basal half of first gastric segment finely punctate;

basal one quarter finely striate (Florida females of A. treatae have

punctations covering basal four fifths and striations basal one third

of first gastric segment). A. treatae females from more northern

states may have less gastric sculpture. Remainder of gaster rather


Erect hairs on head, alitrunk, petiole, postpetiole and gaster

arranged in typical Aphaenogaster fashion. Hairs on scutum slightly

shorter than those of A. floridana females. Piceous; some workers from

Florida panhandle are lighter brown, but no females from such enclaves

were found.

Description of Male

TL 5.6, HL 1.0, HW 0.8, SL 0.35, AL 2.2 mm. General appearance as

in Figure four. Head shape as in Figure four; more elongate than in

males of A. fulva, A. maria, and A. tennesseensis; in transverse cross

section head more rectangular, not distinctly wider ventrally than dor-

sally as in A. fulva, A. mariae, and A. tennesseensis males. Head wider

than in male of A. flemingi.

Alitrunk as Figure four. Dorsum of propodeum with pair of short

denticles. Metathorax with pair of latero-ventrad swellings, each nar-

rowing distally into a blunt point. Swellings not nearly as well devel-

oped as those of A. mariae and A. tennesseensis males.

Petiole and postpetiole as in Figure four. Petiolar node in pro-

file more acute than in A. floridana males. Viewed dorsally postpetiole

distinctly wider than petiole.

Less sculptured than males of A. treatae. Mandibles smooth,

shining, with feeble punctations and hint of longitudinal striations.

Clypeus shining and feebly punctate. Frontal and lateral portions of

head punctate with faint transverse rugae between ocelli; faint longi-

tudinal rugae directly anterior and posterior to compound eyes. Ali-

trunk largely shining and feebly punctate except scutum. Scutum

densely punctate with feeble rugae originating along longitudinal mid-

line on anterior half of scutum and running roughly longitudinally

the length of the scutum. Scutellum punctate and with faint rugae.

Dorsum of propodeum and lateral sclerites mostly glassy smooth and very

shiny in A. ashmeadi, but in A. treatae largely punctate. Punctations

and transverse rugae on declivious face of propodeum and laterally be-

tween spines and lateral protuberances. Some rugae between meso- and

metathoracic wing articulations. Petiole and postpetiole shining but

punctate particularly laterally and posteriorly. Gaster shining; weak

punctations at base of first segment.

Pilosity characteristic of genus. Few short hairs on dorsum of

alitrunk. On posterior of postpetiole six setae which overlap base of

first gastric segment. Piceous.


According to Creighton (1950), A. ashmeadi occurs in the eastern

Gulf Coast states. In the Atlantic Coastal Plain it has been found as

far north as New Bern, North Carolina (Carter, 1962b). Cole (1940) re-

ported A. ashmeadi from the Great Smoky Mountain National Park at an

altitude of about 500 m. Carter (1962a) did list this species among

the ants he collected in the North Carolina Piedmont.

In Florida, its type locality, A. ashmeadi occurs throughout the

panhandle and northern portion of the state. The extent of its south-

ward distribution in Florida is uncertain. The southernmost collection

was made near Trilby in northeastern Pasco County. Habitats ostensibly

similar to those in which A. ashmeadi regularly occurs in north central

Florida exist as far south as Highlands County. It was specifically

sought for in Highlands and southern Polk Counties. However, soil

nesting species such as this are usually difficult to find unless they

are active above ground. A. ashmeadi was found in the following coun-

ties: Alachua, Franklin, Hamilton, Hernando, Leon, Levy, Marion, Oka-

loosa, Pasco, Putnam, Sumter, and Union. A. ashmeadi was quite common

in the Gainesville region.


Among those species of Aphaenogaster consistently nesting in the

soil A. ashmeadi was found in the widest range of habitats. See Table

one. Although A. ashmeadi occurs in well-drained sand pine and turkey

oak-longleaf pine communities as well as mesic hammocks, it is most

abundant where the soil is moderately well-drained sand, as in scrub

oak communities and xeric hammocks. In Gainesville this insect is

particularly common in scrub oak woods and upland open mesic woodlands.

Although pine woods generally harbored rather sparse populations of A.

ashmeadi, at one pine woods site in Union County it was abundant. A.

ashmeadi often nests in shaded lawns and parklike areas. At Welaka

Van Pelt (1958) found it common in xeric hammocks, occasionally in blue

jack oak woods, mesic hammocks, Leon scrubby flatwoods and scrubs, and

rarely turkey oak woods and bayheads.

The microhabitat of A. ashmeadi nests in Florida was characterized

by moderately to well-drained sandy soil. Only once was a colony found

nesting in a log and then only partially. Leaf litter and midday shade

also characterize its microhabitats. However, leaf litter was not

always present, but was generally rather thick in natural A. ashmeadi

microhabitats. Generally herbaceous vegetation was rather sparse in A.

ashmeadi microhabitats, except in parklike situations where it is regu-

larly cut.


Nests of A. ashmeadi rarely extended below a depth of 40 cm. The

deepest nest I found extended down about 60 cm. A. ashmeadi nests have

one to six entrances, but usually only one or two. Generally there is

one larger primary entrance used by foragers returning with food. The

main entrance may change every few days or weeks. This main entrance

is often surmounted by a turret of short (less than eight mm) slender

twigs, bits of wood, caterpillar droppings, and other debris. Arthropod

remains are frequently included in turrets. Heads of Odontomachus

brunneus (Patton) and Camponotus floridanus (Buckley) were regular con-

stituents of turrets. One colony was found to have incorporated white

pin feathers in its turret, while the turret of another colony was

strewn with entire corpses of the bibionid, Plecia nearctica Hardy.

When nests were in heavy leaf litter distinct turrets were not always

visible. However, the same materials which comprised turrets were used

by the ants to form the walls of an entrance-way among the leaves. 'A

superficial chamber among the litter or between the leaf litter and the

soil was delimited by vertical walls of debris. Pupae and prepupae are

usually kept in the superficial chamber. Winged reproductive often

frequent the superficial chamber. The superficial chambers were vari-

ously shaped and up to three cm across and one to two cm deep. One to

three large tunnels up to a cm in diameter open into the superficial

chamber. The tunnels descend to large chambers three or four cm across

and one to two cm high. The chambers have rather flat floors and

curved ceilings. The uppermost chambers are only about two to 16 cm

below the surface. The queen was frequently found in a chamber about

25 cm below the surface. Brood may be in upper and lower chambers.

There are usually five to seven subsurface chambers. One chamber con-

tains refuse, mostly arthropod remains. Refuse is also brought to the

surface and deposited. Workers carrying refuse from the nest deposited

the refuse 30 to 80 cm from the nest entrance.


A. ashmeadi foraged both day and night during the warmer months.

Colonies in deeper shade tended to forage more in the mid-afternoon

than those exposed to more direct sunlight. All were somewhat more

active in the evening than midday. Foraging decreased considerably

during winter and was confined to the warmer hours of the day. As late

in the Spring as mid-March it was hard to find A. ashmeadi foragers in

the Gainesville region. Regarding A. ashmeadi, Carter (1962b) reported

that in North Carolina: "Activity was observed from June to August "(. 173).

Foraging of A. ashmeadi was typical of most species of Aphaeno-

gaster, particularly those nesting in the soil. Workers of A. ashmeadi

forage singly, recruiting co-workers when a large or appealing food

source is found. Foraging was almost completely restricted to the

ground and leaf litter strata. Foragers were rarely found on herbs,

grasses or trees. The maximum height at which an A. ashmeadi forager

was seen on a plant was approximately six mm. Foragers sometimes

searched fallen logs. They investigated beneath leaves and into

curled leaves. As the ants traveled slowly about, they moved their

antennae right and left, frequently touching them to the substrate or

raising them in the air. Foragers were sensitive to sudden vibrations

of the leaves. If a vibration was caused by dropping a pin-head-sized

pebble upon a dry leaf, an ant on or under the leaf would run to the

vicinity where the pebble hit and-search actively.

Foragers apparently found food by other modes of perception as

well. Items of food incapable of movement were detected by foragers.

Antennal contact was not necessary for some inert food items to be de-

tected. If an ant's antenna passed within less than two or three mm of

the food, the ant would turn toward the food and find it. Apparently

scent trails of other ant species or the food carried by those ants

also can be sensed by Aphaenogaster foragers. A. ashmeadi workers

crossing the path just traversed by a returning forager of another spe-

cies would stop and run rapidly in the direction taken by the burdened

ant even if she were a cm distant. The A. ashmeadi worker would some-

times lose the trail and run from side to side and backtrack to pick it

up again. Often the ant carrying the food outdistanced the A. ashmeadi

worker who would then lose the trail altogether.

A. ashmeadi foragers generally stayed in shaded leaf litter micro-

habitats. Rarely did this species venture into extensive areas of hot

bright sunlight. However, Whitcomb et al. (1972a)reported this species

foraged in Florida soybean fields.


The diet of A. ashmeadi consists mostly of other arthropods, which

are captured or scavenged. According to Carroll and Janzen (1973) very

few ant species are exclusively hunters, and may supplement their diets

by scavenging. Unless disabled or molting, large or quick arthropods

(e.g. locusts, many Diptera) escaped predation. Insect larvae were

more susceptible. Large lepidopterous larvae were attacked, whereas

adult beetles of similar size repelled A. ashmeadi foragers. A. ash-

meadi workers were able to capture apparently healthy noctuid larvae

up to four cm long. Capture of such large larvae required the combined

efforts of at least four or five A. ashmeadi workers. Aphaenogaster

workers appeared rather clumsy in cooperative attacks. They often

seemed to be working against one another, biting and seizing co-workers,

thereby frequently allowing the prey to escape. Jaynes and Marucci

(1947) described similar counter-productive behavior in attacks of A.

rudis upon codling moth larvae. Individual A. ashmeadi foragers were

able to capture naked lepidopterous larvae up to 1.3 cm long. No

field observations were made of successful A. ashmeadi predation upon

hairy lepidopterous larvae. In closed arenas in the laboratory the

ants subdued larger hairy larvae only after a prolonged struggle. Most

of the field observations of A. ashmeadi predation upon large lepi-

dopterous larvae were after heavy rainfalls, which might have dis-

lodged the larvae from vegetation, or of attacks upon prepupal larvae.

On several occasions while excavating A. ashmeadi nests, unharmed pre-

pupal notodontid larvae were discovered just below the soil's surface

within ten cm of the nests' entrances.

I released cabbage loopers near an A. ashmeadi nest, so that the

ants' predatory behavior might be observed in the field. One fourth

instar looper attacked by a worker managed to escape in a manner which

emphasizes the terrestrial nature of A. ashmeadi foraging. The ant

bit the larva several times. One time when the larva broke free it

climbed a small plant instead of running along the ground. The larva

climbed to a height of only about 2.5 cm, but the ant was unable to

find it. The ant was soon joined by two nestmates and all searched

unsuccessfully about the base of the plant for nearly five minutes,

never climbing it. Large lepidopterous larvae were capable of killing

Aphaenogaster workers by biting them.

Arthropods living on plants are virtually free from A. ashmeadi

predation as long as they do not venture to the ground. Some quite

desiccated arthropods brought in by foragers were obviously scavenged.

I assumed that any worker returning with a piece of a larger arthropod

had scavenged it unless other workers returned with pieces of the same

organism. Crushed or flattened arthropod remains brought in by for-

agers were assumed to have been accidently stepped or knelt upon by me.

Small spiders, Diptera, Orthoptera, ants, and various larvae brought

in by foragers generally appeared to have been captured. Food items

too large to be brought to the nest whole were dismembered and dis-

sected on the spot.

Other ants were regularly among the items brought in by A. ashmeadi

foragers. Most of the smaller ants appeared to have been attacked by

A. ashmeadi, while most larger ants appeared to have been scavenged.

Pheidole dentata Mayr and a small species of Paratrechina (Nylanderia)

predominated among the victims. At times individual A. ashmeadi

workers were observed carrying two of these smaller ants. The smaller

ants were gripping each other with their mandibles as if they were cap-

tured while fighting each other. Considering their abundance in A.

ashmeadi microhabitats and their numerous encounters with the larger

ant, relatively few P. dentata and P. (Nylanderia) sp. were captured

by it. Corpses of Odontomachus brunneus, Pseudomyrmex brunneus F.

Smith, Camponotus floridanus, and C. (Colobopsis) sp. were also brought

in by A. ashmeadi foragers. Van Pelt (1958) reported seeing A. ashmeadi

workers carrying dead bodies of 0. brunneus. An A. ashmeadi forager

rarely attacked larger ants, but was sometimes able to kill one when

attacked itself.

Termites seem to be of little significance in the diet of A.

ashmeadi. Once a termite worker was observed being brought into an

A. ashmeadi nest. An A. ashmeadi worker was seen carrying a nearly

dead alate Kalotermes approximatus Snyder. On another occasion I

saw an A. ashmeadi worker leaving her nest carrying a termite wing,

which she deposited on an external refuse area.

Small oligochaetes were preyed upon occasionally. A. ashmeadi

workers often scavenged larval exuviae of various arthropods. Pupal

skins were usually ignored. Lepidopterous feces were sometimes ob-

served being carried by A. ashmeadi workers. A bird dropping was

visited by several A. ashmeadi foragers, which were removing lepi-

dopterous eggs from the excrement. According to Van Pelt (1958), A.

ashmeadi is attracted to raw liver.

Fruiting bodies of certain fungi comprise a significant part of

the diet of A. ashmeadi. Although during the course of a year the

greater part of the food of A. ashmeadi consisted of arthropod flesh,

there were times when mushroom tissue was brought into nests in

quantities surpassing that of arthropod flesh. A. ashmeadi most fre-

quently brought back pieces of Russula spp. and Marasmiellus sp.

These two genera are not closely related; the former belonging to the

family Russulaceae and the latter to the Tricholomataceae. Occasion-

ally taken was Amanita vaginata of the family Amanitaceae. A rotting

clump of Clitocybe tabescens, Tricholomataceae, was visited for at

least seven days by workers from a nest by A. ashmeadi colony.

All are members of the order Agaricales. Usually no more than two or

three mature Russula sp. grew within the foraging territory of an A.

ashmeadi colony.

Workers tore off chunks (two to four mm dian) of Russula ousilla.

(The gills and cap were the most frequent sites of attack, but pieces

were removed from the stalk or stipe as well. See Figure two.) A max-

imum of five or six, usually one to three, workers visited a mushroom

at one time. In comparison a large dead lepidopterous larva might

attract 2C .r more workers. A mushroom, however ephemeral, consti-

tutes a more permanent food source than a dead insect. Individual

Russula sp. were visited rather continuously by A. ashmeadi workers

over a period of three to five days, until the mushrooms decayed or

were consumed by other organisms. Pieces of mushroom were generally

carried nestward at a rather slow gait compared to the speed of a

forager returning with a dead insect of equivalent size. Perhaps

running back with a dead insect lessens the chances of the forager

being waylaid by a larger stronger species and having the food stolen.

There seems less need for an A. ashmeadi forager returning with a

piece of fungus to run fast, since no larger ant species foraging in

its microhabitats was known to feed on fungi and therefore apt to

steal its burden. General foraging continued while mushrooms were

being visited.

Marasmiellus sp. is a small mushroom. The ones harvested by A.

ashmeadi had not fully expanded their caps. They were about five mm

high. An A. ashmeadi worker would cut one off at the base and carry

it back to the nest. The stalk was grasped just below the cap so that

the cap was anterior to the ant's head and the stalk directly beneath

her body. Marasmiellus sp. were brought in while other workers contin-

ued general foraging. Several other genera of mushrooms, bracket and

other fungi were available in A. ashmeadi microhabitats but were not

seen being taken by the ants.

Occasionally unidentified seeds were collected by A. ashmeadi

workers. A. ashmeadi was never seen associated with honeydew excreting

Homoptera except in a predator prey relationship. Sucrose and water

solutions presented to colonies of this species in the laboratory were

imbibed by workers rather than pure water simultaneously presented

them. Sometimes workers filled the liquid with grains of sand or wood

chips. The particles were later carried into the nest.

The artificial diet (Bhatkar and Whitcomb, 1970) used to feed

laboratory colonies was taken readily by the ants at first. After

several weeks the quantities consumed by the colonies gradually de-


Coexistence with other Ants

Several species of ants traverse and nest in microhabitats in which

Aphaenogaster spp. occur. The term Aphaenogaster sp. microhabitat is

used in discussing coexistence of various Aphaenogaster species with

other ants. For our purposes an Aphaenogaster microhabitat is defined

as the portion of the habitat directly influencing or interacting with

an Aphaenogaster sp. colony. Interspecific competition among social

insects is exceedingly complex, involving many parameters beyond the

scope of this study; e.g. source of food items, criteria for selection

of food items, energy requirements of colonies, recruitment of workers.

I will make no attempt here to draw definite conclusions regarding

interspecific competition. The data in Tables three, four, and five

and the ensuing remarks may indicate areas of potential overlap in

food and nest site requirements of the more common ant species utiliz-

ing Aphaenogaster spp. microhabitats.

Table three shows the relative abundance of workers of various ant

species traversing the ground within a two m radius of A. ashmeadi

nests. Foragers of many species traversed to the very edge of A. ash-

meadi nest entrances and picked up baits in such proximity. The ubiq-

uity of foragers of other species about an A. ashmeadi nest is further

indicated in Table five. Also suggested in Table five are the capabili-

ties of different sized ants to successfully remove food items to their

nest without having the items taken by larger ants. Not shown in Table

five is the ability of certain species (e.g. Solenopsis spp.) to monop-

olize large immovable food items. Large items of animal tissue (dead

mammals) are infrequent food sources of Aphaenogaster spp.

Those ants commonly nesting in or traversing microhabitats of A.

ashmeadi are:

Ponerinae- Hypoponera opaciceps

Odontomachus brunneus

Myrmicinae- Aphaenogaster carolinensis

A. lamellidens

Pheidole dentata

P. floridana

P. metallescens

Crematogaster ashmeadi

C. clara

Solenopsis geminata

Solenopsis (Diplorhoptum) sp.

Leptothorax pergandei

Cyphomyrmex rimosus minutus

Trachymyrmex septentrionalis

Formicinae- Brachymyrmex depilis

Camponotus floridanus

C. castaneus

Paratrechina (Nylanderia) sp.

Prenolepis imparis

Formica pallidefulva

Remarks on individual species. Not listed above is the red im-

ported fire ant, Solenopsis invicta Buren. In northwest Florida S.

invicta is known to be a serious competitor of native predaceous ants

(Whitcomb et al., 1972). Bhatkar (1973) computed an index quantifying

aggression between the two species.

Although it forages on the ground like A. ashmeadi, Hypoponera

opaciceps (Mayr) is more specifically a subsurface forager in leaf lit-

ter. A. ashmeadi tends to forage on the surface of the leaf litter

layer. This indicates a partitioning of the leaf litter layer into

different foraging zones. I have observed Odontomachus brunneus prey-

ing upon and scavenging the same sort of arthropods in the same stratum

as A. ashmeadi. With their long powerful mandibles and potent sting 0.

brunneus workers are formidable predators. Although Table three indi-

cates few 0. brunneus workers were collected around A. ashmeadi nests,

the large 0. brunneus workers appeared to be able to search an area

rapidly. Recruitment does not seem important in the foraging strategy

of 0. brunneus. As in A. ashmeadi, workers of 0. brunneus take food

items from smaller ants; they also take food items from A. ashmeadi

workers. 0. brunneus workers regularly attack A. ashmeadi foragers,

but not vice versa unless within one cm of the latter's nest entrance.

Attacked A. ashmeadi workers were often killed or maimed, while only

once was an A. ashmeadi worker seen killing an attacking 0. brunneus

worker. I frequently saw 0. brunneus workers visiting mushrooms, but

never observed them removing or carrying pieces of the fungi. Occa-

sionally 0. brunneus workers were seen visiting extrafloral nectaries;

a dimension not known in the feeding habits of A. ashmeadi.

Interspecific coexistence among the various species of Aphaeno-

gaster is discussed in a subsequent section.

In all A. ashmeadi habitats examined at least one species of

Pheidole was common. See Table three. Table five shows the ubiquity

of Pheidole foragers about one A. ashmeadi nest. Fifteen of 25 dead

house flies placed around the nest were first attacked by a Pheidole.

Despite their ubiquity Pheidole spp. may lose many food items they find

to A. ashmeadi workers as suggested by the limited data in Table five.

Size of the food item may be a factor in the ability of Pheidole spp.

to collect it from an A. ashmeadi foraging area (Table five). Single

A. ashmeadi workers took cheese baits and dead house flies from as

many as three P. dentata Mayr workers and seven P. metallescens workers.

P. morrisi Forel and P. metallescens occur in well-drained rather open

A. ashmeadi microhabitats. I never found P. floridana Emery in abun-

dance in A. ashmeadi microhabitats. On one occasion a P. metallescens

worker was seen removing a piece of a mushroom, Russula sp.

Crematogaster ashmeadi Mayr and C. clara Mayr often nesc arboreally

or at least in wood rather than the soil. It is doubtful that they are

nest space competitors of A. ashmeadi. Solenopsis geminata (Fabricius)

is common in more open drier A. ashmeadi microhabitats. Hamburger

baits set on the ground in microhabitats where S. geminata was present

were in a few hrs covered with hundreds of S. geminata workers. A

hypogeaic species of Solenopsis (Diplorhoptum) sometimes nests near A.

ashmeadi nests. Some S. (Diplorhoptum) species are known to infiltrate

nests of larger ants and take food (Wilson, 1971). Leptothorax

pergandei Emery often occurs in A. ashmeadi microhabitats, but its col-

onies were small and few. Cyphomyrmex rimosus minutus Mayr usually

nests in decaying wood, but was sometimes found nesting in the soil.

In addition to plant material it collects insect remains to use in its

fungal substrate.

Brachymyrmex depilis Emery tends to forage more nocturnally.

According to Whitcomb et al. (1972b), sweet liquids are an important

part of the diet of Camponotus floridanus (Buckley). Although I have

found it nesting in the soil C. floridanus generally nests in wood.

Both C. floridanus and C. castaneus tend to forage more at night. The

Paratrechina (Nylanderia) group is in need of taxononic revision. At

least two forms occur in A. ashmeadi microhabitats; one form more prev-

alent in drier areas, the other common in more nesic habitats. I ob-

served A. ashmeadi workers taking small cheese baits and dead mosqui-

toes from P. (Nylanderia) sp. foragers. In the Gainesville area

Prenolepis imparis (Say) is very active above ground from November to

April, but is difficult to find during the warmer months when A. ash-

meadi is most active above ground. More needs to be known about what

P. imparis does during the summer in order to properly speculate on

the competitive relationship between P. imparis and A. ashmeadi.

Formica pallidefulva Mayr is common in more open or parklike habitats

which have thin leaf litter. It seems to nest in sunnier areas than A.


A. ashmeadi workers were seen carrying corpses of two arboreal ants,

Pseudomyrmex brunneus F. Smith and Camponotus (Colobopsis) sp.

Annual Cycles and Colony Size

Large female larvae were found in field colonies in mid-March in

the Gainesville region. Pupal females appear during early April.

Alates were found in mid-April. Males usually reach adulthood ahead

of females, but there is overlap in most colonies. Mating occurs in

July and August. I regularly observed reproductive at nest entrances

during late afternoon from early June to early July. Alates are present

in the superficial nest chambers of some colonies in late July. A male

was found at a light on 7 July. I observed a dealated female crawling

among leaf-litter on 22 July. On 8 August I discovered a founding

queen with a few eggs in a small cell (one cm across) just below the

loose leaf-litter at the base of an oak tree. Her original eggs were

lost when I collected her. She laid eggs her first day in captivity.

Her first larva closed on 28 August. On 10 September the colony con-

sisted of the queen, five eggs, seven larvae, and one pupa. Small

colonies with less than 30 small workers, which I discovered in the

spring, may have been founded by queens mated the previous summer.

The largest colony of A. ashmeadi I collected contained a queen,

423 workers, 46 male and 15 female pupae, 6 female and 53 male or

worker prepupae, 50 larvae, and at least 30 eggs. Most colonies pro-

ducing alates had only 100 to 250 workers and fewer males. Van Pelt

(1958) collected a large A. ashmeadi colony at the Welaka Reserve con-

taining a queen, 333 workers, 250 pupae, workers and eggs.


Duration of the e.. ,tage for workers of A. ashmeadi is 19 to 22

days. Larval development is more variable in duration; 14 to 22 days.

The shortest duration tor larvae, 14 days, occurred in incipient colo-

nies having no older larvae. The longest elapsed times were for larvae

in colonies where older sibling reproductive were present. The pupal

stage lasts tento 13 days. A few females remain in the pupal stage

eight to ten days.

Findings on A. carolinensis

Description of Female

TL 7.0, HL 1.4, HW 1.2, SL 1.4, AL 2.3 mm. for typical Floridian

carolinensis to TL 8.2, HL 1.6, HW 1.5, SL 1.6, AL 2.7 mm. for south

Florida specimens. General appearance as in Figure five. Head shape

as in Figure five; more convex posteriorly than females of ashmeadi,

flemingi, floridana, and treatae, but less so than fulva. Base of

antennal scape without lobe.

Alitrunk as in Figure five. Propodeum with pair of posteriorly

directed spines. Spines shorter than declivious face of propodeum.

Petiole and postpetiole as in Figure five. Postpetiole less volu-

minous than in ashmeadi, flemingi, floridana, and treatae, but more

voluminous and elongate than in fulva.

More sculptured than flemingi, but less than fulva. Mandibles

with longitudinal striae which spread as they approach the masticatory

border. Clypeus punctate and with slightly wavy longitudinal rugae.

Coarse longitudinal rugae over front and lateral portions of head; less

coarse near occipital border. Punctations between rugae. Pronotum

with rugae parallel to suture between pronotum and scutum. Rugae orig-

inating along longitudinal midline of scutum and running roughly longi-

tudinally the length of the scutum. Longitudinal rugae extend across

parapteron and scutellum. Propodeum with transverse rugae; area be-

tween spines with feebler rugae. Mesothoracic sternite and episternite

with rather longitudinal rugae, but smoother and shinier ventrad; large

specimens with more and coarser lateral rugae. Coxae and femurs

largely shining, but punctate, and in large specimens fore coxae with

some striations. Petiole punctate and longitudinally rugose, particu-

larly posteriorly. Postpetiole punctate and with rugae rather longi-

tudinal anteriorly, but transverse posteriorly. Gaster rather smooth

and shining and with only faint punctation.

Erect hairs on head, alitrunk, petiole and postpetiole arranged

characteristically for genus. Brown or reddish brown to piceous;

gaster dark brown to piceous in all specimens.

Description of Male

TL 4.5, HL 0.75, HW 0.75, SL 0.4, AL 1.7 mm. for typical carolinen-

sis to TL 5.0, HL 0.9, HW 0.75, SL 0.4, AL 2.0 mm. for south Florida

specimens. General appearance as in Figure four. Head shape as in

Figure four; narrower posteriorly than in fulva, mariae, and tennesseen-

sis, but less so than in flemingi. In transverse cross section head

rather rectangular instead of wider ventrally than dorsally as in

fulva, maria, and tennesseensis.

Alitrunk as Figure four. Dorsum of propodeum with pair of slight

protuberances, not quite denticles but more prominent than corresponding

raised areas in flemingi. In one specimen examined protuberances were

drawn out into denticles. It is not rare for Aphaenogaster males to

have one propodeal protuberance drawn out twice the length of the other.

Metathorax with pair of latero-ventrad swellings, each narrowing dis-

tally into a dull point. Swellings only slightly more developed than

in flemingi and less so than other Floridian congeners. Mid dorsum of

propodeum tends to collapse in some pinned specimens.

Petiole and postpetiole as in Figure four. Postpetiole not as

voluminous as in flemingi, but more elongate than in fulva.

Rather weakly sculptured. Mandibles smooth, shining (less so in

large south Florida specimens), and with feeble punctations. Clypeus

shining and with faint punctations. Frontal and lateral portions of

head punctate, most pronounced punctations between ocelli. Scutum

shining, with some faint punctation; humeral areas particularly smooth

and shining. Scutellum punctate laterally, smooth and shining along

longitudinal midline. Dorsum of propodeum largely smooth and shining,

punctate laterally and dorsally between protuberances. Mesothoracic

sternite and episternite somewhat punctate, but with rather smooth

shining appearance. Petiole shining, but with punctations. Post-

petiole shining, punctate, but smoother dorsally. Gaster smooth and


Pilosity generally characteristic of genus. A few hairs on

scutum and on protuberances of propodeum. Light brown with head


Description of Worker

TL 4.0, HL 1.0, HW 0.9, SL 1.3, AL 1.6 mm. for typical Floridian

carolinensis to TL 6.5, HL 1.3, HW 1.2, SL 1.6, AL 1.9 mm. for large

specimens particularly those from south Florida. General appearance as

in Figure six. Head shape as in Figure six; very similar to that of

lamellidens, particularly in larger carolinensis specimens. In frontal

view head more narrowed posteriorly than fulva and rudis, lacking

roughly angular convexity near occiput characteristic of non-nanitic

forms of those species. Head not nearly as narrowed as in flemingi

and typical texana. Base of antennal scape without lobe.

Alitrunk as in Figure six. Prothoracic region more slender later-

ally and dorsoventrally than in fulva and rudis; somewhat less slender

in largest carolinensis specimens. Propodeal spines shorter than

declivious face of propodeum; generally shorter and more divergent in

smaller specimens, longer and rather incurved in larger specimens.

Spines project posteriorly but at various angles above the horizontal;

rarely as upwardly directed as in fulva. Spines somewhat laterally


Petiole and postpetiole as in Figure six. Postpetiole less volu-

minous than in ashmeadi, flemingi, floridana, and treatae; more elon-

gate than in fulva.

Larger specimens with more and heavier sculpture. Mandibles with

faint longitudinal striations; more shining in smaller specimens.

Clypeus punctate and with rugae that tend to be longitudinal toward

middle of clypeus. Most of head capsule punctate. Longitudinal rugae

covering frontal and lateral portions of head in large specimens; in

smaller more typical carolinensis specimens longitudinal rugae more

confined to area of longitudinal midline between and posterior to

frontal lobes and laterally anterior and posterior to compound eyes.

Head shinier near occiput; rugae and punctations weaker. Alitrunk

largely punctate; pronotum less so. Middle of pronotum smoother and

shinier in small specimens. Larger specimens with some transverse

rugae along humeral areas. Mesonotum and metanotum with somewhat

longitudinal rugae especially in larger specimens. Dorsum of pro-

podeum with reticulate and transverse rugae; rugae coarser and tend

to be more transverse in large specimens. Area between spines shiny

and with feeble transverse rugae. Coxae and femurs shining and

faintly punctate; punctations most noticeable on fore coxae, particu-

larly in large specimens. Petiole punctate, especially posteriorly

and laterally; shining dorsally in small specimens, some rugae in

large specimens. Postpetiole punctate, particularly laterally and

posteriorly; shining anteriorly, more so in small specimens. Dorsal

posterior of postpetiole of large specimens granulo-rugose. Gaster

shining. Head, alitrunk, petiole and postpetiole orange-red to brown

to piceous. Gaster dark brown to piceous; distally yellowish brown

in some specimens.


According to Creighton (1950), A. carolinensis occurs in the

Piedmont region from southwestern Virginia to northwestern Alabama.

Carter (1962b) collected A. carolinensis in the fail-line sandhill

region of North Carolina, but not in the Coastal Plain. Warren and

Rouse (1969) stated that A. carolinensis can be found as far west as

Oklahoma. In a foregoing section on the taxonomy of the species of

Aphaenogaster of Florida A. miamiana was synonymized with A. carolinen-

sis. Consequently the range of A. carolinensis is expanded to include

that of the former A. miamiana, which Creighton (1950) reported to be

Florida and northward and westward into Alabama and Mississippi. Smith

(1933) and Wheeler (1932) had previously reported the occurrence of A.

miamiana from Florida.

A. carolinensis has not previously been recorded from Florida.

Smith (1930, 1933), Wheeler (1932), and Van Pelt (1947, 1958) mentioned

the occurrence of forms of A. texana in Florida, some of which may be

attributable to the typical A. carolinensis. The typical form occurs

at least as far south as Marion County. No collections of the typical

A. carolinensis were made in the panhandle west of Leon County, although

it probably occurs there. It is quite abundant in Alachua and Marion


Atypical forms of A. carolinensis, i.e. miamiana-like and inter-

mediate forms here referred to as A. carolinensis*, occur throughout

the state of Florida. A. carolinensis* was collected in the panhandle

in Leon and Santa Rosa Counties. The southernmost collection site for

carolinensis* was Key Largo, Monroe County. Wilson (1965) also found

A. carolinensis* in upper Key Largo. A. carolinensis* was also col-

lected in the following counties: Alachua, Broward, Collier, Columbia,

Duval, Gadsden, Gilchrist, Hamilton, Levy, Marion, Nassau, Okeechobee,

Palm Beach, Pasco, Putnam, Sumter, and Volusia. I examined specimens

from Dade, Dixie, and Hendry Counties. Forms closely resembling

Wheeler's (1932) description of A. miamiana seem to be restricted to

south Florida. See Figure one. A. carolinensis* seemed to be common

except in the central portion of the peninsula. Several attempts to

find this ant in Polk County proved fruitless. Since A. carolinensis*

generally nests in rotten logs, its presence was readily surveyed. I

scanned tree trunks and the ground for strays if no colonies were found

in logs. Van Pelt (1958) did not list A. miariana among the ants he

collected at the Welaka Reserve. I found dark miamiana-like specimens

there in a mesic hammock. Van Pelt apparently identified specimens of

miamiana as A. fulva or A. texana. Furthermore, he mentioned some

Aphaenogaster specimens which seemed to key out to A. rudis in Creigh-

ton's (1950) key. Some of the specimens he believed might be A. fulva

he sent to M. R. Smith for examination.


In the Alachua, Marion County area the typical A. carolinensis is

most common in upland mesic hammocks, having rather open understories.

This ant is often associated with the presence of pines in such hammocks.

At one station it was abundant in a grove of loblolly pine, where little

other woody or herbaceous vegetation grew. At the Welaka Reserve typ-

ical A. carolinensis was numerous in a mixed slash pine-blue jack oak

woods. Specimens attributable to this species were collected in wet

mesic hammocks in Leon County. Van Pelt (1958) reported collecting A.

texana rarely in hydric hammocks at the Welaka Reserve.

Nests of A. carolinensis are often confined to the leaf litter-

humus layers and the upper ten cm of the soil. The leaf mold layer is

sometimes wetter than the soil immediately beneath it. A. carolinensis

Is not "restricted" to soil-nesting as A. ashmeadi, A. flemingi, A.

floridana, and A. treatae are in Florida. A. carolinensis nests are

frequently associated with buried pinecones or small branches the diam-

eter of one's finger. Varying portions of a colony reside within the

wood. Colonies collected in Leon County were usually nesting in moist

rotten stumps or logs. There was heavy leaf litter present, but the

site appeared subject to frequent saturation or flooding. Wheeler

(1915) and Carter (1962b) reported that A. carolinensis often nests be-

neath stones.

In the northern half of Florida A. carolinensis* is most commonly

found in mesic and hydric hammocks, bayheads, cypress domes, and river

swamps. It is not as prevalent in upland mesic hammocks as in the lower

more moist woodlands. I rarely found A. carolinensis* in xeric ham-

mocks in north Florida and never found it in fields or unwooded habi-

tats. It occasionally occurs in shaded lawn or parklike areas. In

south Florida A. carolinensis*, i.e. the typical miamiana, occupies a

much wider range of habitats. It not only occurs in the same type habi-

tats as in northern Florida, but also in pastures, sugar cane fields,

sand pine dune areas, willow and pond apple quasi-marsh land, cypress -

red maple strand forests, tropical hardwood hammocks and habitats dis-

turbed by man (i.e. Australian pine, Casuarina sp., groves).

A. carolinensis* nests in rotten logs, stumps, and cypress knees,

in rotten portions of live trees, and in the soil, sometimes under

stones. No arboreal nests were found in living pine or sweetgum trees.

One nest was found in the Australian pine. All arboreal nests of A.

carolinensis* I found were less than three m above ground level. Some

A. carolinensis* arboreal nests were higher than the highest A. fulva

arboreal nests. None were as high as most A. lamellidens and many A.

tennesseensis nests. Most A. carolinensis* arboreal nests are in

habitats subject to frequent yearly flooding. Few arboreal nests were

found in cypress domes. A. carolinensis* sometimes nests in smaller

trees than do A. fulva, A. lamellidens, and A. tennesseensis.

Epiphytes, mosses, lichens, and vines contribute to the micro-

habitat about arboreal nests. Most arboreal nests are in the core or

heartwood of trees. Stumps, logs, and cypress knees occupied by A.

carolinensis* colonies may be in various stages of decay, but least

often in early and advanced stages of decomposition. The wood was

moderately dry to saturated. Nests were rarely found in logs contain-

ing passalid beetles. In floodplain situations, however, nests are

occasionally immediately beneath the bark of slightly decayed logs.

Nests wholly in the soil are less common than nests in wood. Nests

frequently extend from logs or stumps into the soil. The soil in

which A. carolinensis* nests is highly organic in content in mesic

and swamp situations. The hummncks in which this species may nest in

hydric hammocks, bayheads, and cypress swamps consist almost entirely

of decayed vegetable matter. At the other extreme soil nests in sand

pine woods are partially in very well-drained sand. Nests in such

xeric situations occur at the shaded base of shrubs such as rosemary,

Ceratiola sp., where organic matter has accumulated. Nearly all the

nests of A. carolinensis* were shaded, but more so in north Florida.

The amount of leaf litter around A. carolinensis* nests in flood

plains varies depending on its deposition by water currents. There is

heavy leaf litter in mesic, hydric, and tropical hardwood hammocks.

Leaf litter is thin in sand pine woods except beneath shrubs. In

cypress domes, hydric hammocks, and river swamps flood waters often

isolate colonies on small hummocks connected only by the tree canopy.


Nests constructed in leaf mold consist of one to three chambers,

the -largest not more than three cm across. The pupae, prepupae and

some older larvae are usually in the uppermost portion of the nest

among newer drier leaf litter. Ants enter their nest by descending in-

to the leaf litter. No external sign of an entrance could be detected

for these nests; thus differing from the nests of A. ashmeadi, A.

flemingi, A. floridana, and A. treatae in which a turret of debris

usually surmounts at least one entrance. Chambers in newer leaf litter

are often cells between two superimposed rather horizontal leaves. The

lateral boundaries of the cell are walls of minute wood chips or where

the leaves are tightly appressed. Portions or all of some ground nests

are in buried small branches or pine cones. All the pine cones con-

taining A. carolinensis colonies appeared to have been gnawed by

squirrels and tunneled by insects other than ants. The cones and

branches were moderately well decomposed, breaking apart easily when

handled. The deepest any colony was found to penetrate the soil was

about 25 cm, and this was in well-drained sand in a bluejack oak-slash

pine woods. Nests in logs or stumps consist of one to five chambers.

Pupae are usually in the most exterior chamber, where it is drier.

It was difficult to determine the number of chambers used by

large carolinensis* colonies nesting in rotten wood. See explanation

in materials and methods section. Apparently no more than five cham-

bers usually contained brood. Nest chambers appeared to be preformed

by termites or other wood boring or tunneling insects. One colony

was found nesting adjacent to a colony of Reticulitermes virginicus in

a rotten stump. Pupae and prepupae are usually in the outermost cham-

bers. Often they are just beneath a layer of loosened bark rather than

in a distinct chamber. I rarely found the queen in outer chambers.

Winged reproductive forms tend to congregate in the outer chambers.

Eggs, younger larvae, and the queen are in inner and outer chambers.

Chambers opening to the outside or beneath bark partly enclosed by

walls constructed of wood chips and other debris. Nests in living

trees often open to the outside via a hole or crack in the tree trunk.

Such cracks are sealed with a wall of wood chips or carton-like material

leaving open only a small entrance hole. Workers returning to log

nests use one or sometimes two entrance holes. Soil nests may extend

as deep as 50 cm or more. In nests beneath stones the older brood is

just beneath the stone. When a nest is in both log and soil the older

brood is in the log except in winter.


When A. carolinensis nests in the ground it forages almost entirely

among leaf litter. Foragers were not observed climbing higher than five

cm on vegetation. Log nesting colonies also forage largely on the

ground. It does more subsurface foraging among leaf litter than do

other Aphaenogaster spp. Foraging continues through the day in the


A. carolinensis* forages primarily on the ground. Foragers were

occasionally seen climbing trees other than those in which they nested.

However, foragers were not observed on foliage of trees or higher than

five cm on herbs or grasses. Foragers from arboreal colonies generally

descend to the ground to forage. They search among leaf litter, rotten

logs and hummocks in river swamps. A. carolinensis* forages on mud

flats of flood plains. Carroll (1970) found strays of A. carolinensis*

in sugar cane fields in south Florida in January. Foraging continues

throughout the day. In north Florida foraging is reduced during winter

months, but in south Florida it is not so curtailed.


A. carolinensis workers brought into their nests the same type food

items as the preceding species. If the comparatively small size of A.

carolinensis workers is a factor in the size of the food items taken,

it was not discernible on the basis of observation alone. I saw A.

carolinensis foragers bringing in dead or dying Pheidole dentata minor

workers. Although A. carolinensis was not observed collecting seeds,

it was often seen visiting and removing pieces of mushrooms, particu-

larly Russula sp. and occasionally Amanita vaginata. Bhatkar and Whit-

comb (1975) found workers of this species associated with bracket fungi

and the mushroom, Agaricus sp. I never saw A. carolinensis tending

Homoptera, but Bhatkar and Whitcomb (1975) report that workers occa-

sionally visit the extrafloral nectaries of Cassia sp. In the labora-

tory workers imbibed solutions of sucrose and water rather than plain

water proffered them.

The feeding habits of A. carolinensis* are similar to those of the

typical A. carolinensis and A. fulva. In south Florida A. carolinensis*

feeds on a wider variety of species of organisms, because it occurs in

a broader range of habitats. It also feeds on tropical organisms whose

range is restricted to south Florida. In river swamps A. carolinensis*

feeds on mollusks and oligochaetes stranded by receding waters. I

infrequently observed termite predation in the field except when I

broke open termite nests. In the laboratory colonies of A. carolinen-

sis* and termites coexisted in the same logs for months. Instances of

field predation on termites consisted of only one or two termites being

brought in by A. carolinensis* foragers during a period of one or two

hr. A. carolinensis* workers were seen collecting seeds of sweet gum,

Liquidambar, and dogwood, Cornus, and fallen floral parts of elder-

berry, Sambucus. Foragers were observed collecting pieces of mushrooms,

Russula sp., in the field. I never saw workers tending Homoptera. In

the laboratory workers imbibed sucrose and water solutions proffered

them. Workers were seen visiting a mass of freshly oozing sap on a

winged elm, Ulmus alata, trunk.

Coexistence of A. carolinensis with other Ants

Several species of ants share nesting or foraging microhabitats

with A. carolinensis. Among the most common of those ants are:

Ponerinae- Hypoponera opaciceps

H. opacior

Odontomachus brunneus

Myrmicinae- Aphaenogaster ashmeadi

A. carolinensis*

A. fulva

A. lamellidens

Pheidole dentata

Crematogaster ashmeadi

C. clara

Solenopsis (Diplorhoptum) sp.

Leptothorax :er'z-

Cyphomyrmex rimosgs -ninutus

Trachymyrmex septe-crionalis

Formicinae- Camponotus floridanus

C. castaneus

Paratrechina (Nylanderia) sp.

Prenolepis imparis

Remarks on individual species. Most of the species occurring in

A. carolinensis microhabitats are also present in A. ashmeadi micro-

habitats. A. carolinensis is smaller than its Floridian congeners.

Although it remains to be demonstrated, A. carolinensis workers may not

be able to take food items as easily as its larger congeners do and its

workers may be more susceptible to having items taken from them.

Like A. carolinensis, Hypoponera opaciceps, and H. opacior (Forel)

frequently nest in leaf mold. I observed Odontomachus brunneus taking

food items from A. carolinensis foragers.

Coexistence among the Floridian species of Aphaenogaster is treated

in a subsequent section.

Pheidole dentata is abundant in A. carolinensis microhabitats. A.

carolinensis foragers were often seen carrying dead P. dentata minor

workers or with P. dentata workers clinging to their appendages. Al-

though Solenopsis (Diplorhoptum) sp. occurs in A. carolinensis micro-

habitats, I never found it nesting adjacent to A. carolinensis nests.

Leptothorax pergandei often occurs in A. carolinensis microhabitats, but

its colonies are small and few.

In A. carolinensis habitats Camponotus floridanus generally nests

in logs or stumps. On one occasion I saw a C. castaneus major worker

take a dead mosquito from an A. carolinensis forager. Paratrechina

(Nylanderia) sp., a dark species, is abundant in A. carolinensis micro-

habitats. See collections one and two under A. ashmeadi in Table three.

In upland mesic hammocks where A. carolinensis is common P. (Nylanderia)

sp. nests chiefly in leaf mold, while in wetter areas it often nests in

rotten logs or stumps.

Coexistence of A. carolinensis* with other Ants

The following ants were found nesting in or traversing microhabi-

tats used by A. carolinensis*.

Ponerinae- Platythyrea punctata

Hypoponera opaciceps

H. opacior

Leptogenys manni

Odontomachus brunneus

Pseudomyrmecinae- Pseudomyrmex brunneus

P. mexicanus

Myrmicinae- Aphaenogaster ashmeadi

A. carolinensis

A. fulva

A. lamellidens

A. tennesseensis

Pheidole dentata

P. floridana

Cardiocondyla spp.

Crematogaster ashmeadi

C. atkinsoni

C. clara

C. vermiculata

Monomorium floricola

Xenomyrmex stolli

Solenopsis geminata

S. invicta

Tetramorium guineense

Cyphomyrmex rimosus minutus

Dolichoderinae- Iridomyrmex pruinosus

Conomyrma sp.

Tapinoma melanocephalum

Formicinae- Camponotus floridanus

C. castaneus

C. (Myrmentoma) sp.

C. (Colobopsis) impressus

Paratrechina (Nylanderia) spp.

Lasius alienus

Remarks on individual species. The long list of species which co-

exist with A. carolinensis* reflects this ant's distribution from the

panhandle to subtropical south Florida.

Platythyrea punctata F. Smith is a south Florida species common in

tropical hardwood hammocks. I Erequently found Hypoponera spp. nesting

in logs containing A. carolinensis* nests. Leptogenys manni Wheeler

feeds primarily on Isopods (Wheeler, 1910), but will also scavenge other

dead arthropods.

Pseudomyrmex brunneus and P. mexicanus Roger forage chiefly on

foliage and slender twigs and stems. P. mexicanus, although expanding

its range, is not yet known from north Florida (Whitcomb et al., 1972).

Coexistence among Floridian species of Aphaenogaster is discussed

in a subsequent section.

Pheidole dentata colonies were often found in rotten logs and

stumps containing A. carolinensis* nests. However, in many river

swamps where P. dentata is common, A. carolinensis* is scarce or absent.

P. dentata often nests in drier logs than does A. carolinensis*. P.

floridana is more common in south Florida than northward. Four species

of Cardiocondyla occur in A. carolinensis* microhabitats. C. venustula

Wheeler is not known from north Florida and the other three species are

more common in south Florida. Crematogaster vermiculata Emery occurs

mainly in cypress and river swamps. Monomorium floricola (Jerdon) and

Xenomyrmex stolli Forel are primarily south Floridian. Tetramorium

guineense (Fabricius) is common in south Florida, but is scarce in the

northern part of the state.

Iridomyrmex pruinosus (Roger) and Conomyrma spp. occur in more

xeric or open A. carolinensis* microhabitats. TaDinoma melanocephalum

(Fabricius) occurs in a variety of habitats in south Florida. It uses

many different types of nest sites.

Twice Camponotus floridanus foragers entered the laboratory and

raided and destroyed A. carolinensis* colonies nesting in Wilson cells.

C. (Colobopsis) impressus (Roger) forages chiefly on foliage and slen-

der twigs and stems. I found at least one form (species) of Para-

trechina (Nlanderia) in every A. carolinensis% microhabitat I examined.

Lasius alienus (Foerster) occurs in the northern part of the state and

most abundantly in the northern tier of counties.

Annual Cycles and Colony Size

In Marion County alates of both sexes of A. carolinensis were found

in field colonies in May. Males and winged females occur in field colo-

nies in Alachua County at least as late as early July. No colony found

had more thanten females or 15 males, while some contained alates of

only one sex, usually male. Large colonies have 200-300 workers.

An A. carolinensis* colony collected on Key Largo in mid-March

contained female larvae and adult males. Most of the female larvae

pupated during the first two weeks of April. Prepupal females were

found in colonies from Sumter and Duval Counties in mid-April. I

found adult males in a colony in late April in Sumter County. Pupae

and larvae of females and male pupae and imagines were found in an A.

carolinensis* colony in Gainesville in late May. Alates were found in

a colony in Palm Beach County in early June. Alates of both sexes were

found in field colonies as late as mid-July in Gainesville. Founder

queens with brood were discovered in cells in wet rotten wood in late

June and early July in Gainesville. No flights of this species were

witnessed. In laboratory colonies sometimes females shed their wings.

Such dealates seemed to function as workers, while sibling alate fe-

males behaved and were treated by workers as sexual. These dealate

females foraged and defended the rest.

One colony with only about 50 workers contained six males and

seven alate females. Mature A. carolinensis* colonies contain an esti-

mated 800 or more workers. Small colonies with less than 30 workers

were found in the early spring and fall. These probably represent

first year colonies.


The duration of the egg stage of A. carolinensis and A. carolinen-

sis* is 16 to 22 days. The larval stage lasts 13 to 17 days. The

elapsed time of the pupal stage is nine to 11 days. One male pupa

closed after ten days. The only recorded duration of a female pupa

was 21 days.

Findings on A. flemingi

Description of Female

TL 9.5, HL 1.8, HW 1.5, SL 1.9, AL 3.0 mm. General appearance as

in Figure seven. Head shape as in Figure seven; narrower than in A.

floridana, but distinctly less narrowed posteriorly than heads of own

workers. Base of each antennal scape with small angular forward pro-

jecting lobe. Lobes approximating size of those of A. floridana rather

than much larger, rearward projecting, ones on scapes of antennae of A.

ashmeadi and A. treatae.

Alitrunk as in Figure seven; propodeum with pair of somewhat up-

wardly directed spines. Spines as long as declivious face of epinotum.

No females of similar Floridian species possess such long propodeal


Petiole and postpetiole as in Figure seven. In profile petiole

with rather acute node compared to females of A. ashmeadi, A. treatae,

and particularly A. floridana. Postpetiole proportionally more volumi-

nous than in females of A. carolinensis, A. fulva, A. lamellidens, and

A. tennesseensis.

Considerably more sculptured than workers. Mandibles with longi-

tudinal striae, which spread slightly as they approach the masticatory

margin. Clypeus with slightly wavy longitudinal rugae. Coarser longi-

tudinal rugae over front and lateral portions of head. Some transverse

connections between rugae. Punctations interspersed between rugae.

Posterior to ocelli, rugae weaker and reticulate to transverse. Prono-

tum with rugae parallel to suture between pronotum and scutum. Scutum

longitudinally rugose. Longitudinal rugae continue across parapteron

to scutellum where they are more reticulate. Propodeum with transverse

rugae; area between spines with more feeble rugae. Mesothoracic stern-

ite and episternite with rather longitudinal rugae, but smoother and

shinier ventrally. Coxae and femurs shining, but with some punctations.

Petiole punctate with some dorsal and lateral rugae. Postpetiole punc-

tate with dorsal and lateral rugae which tend to be transverse. Basal

one quarter to one third of first gastric segment finely punctate. In

A. ashmeadi, A. treatae, and A. floridana basal gastric punctation more

extensive and fine striations are more apparent. Remainder of gaster

of A. flemingi rather shining.

Hairs on scutum slightly longer, more slender and less numerous

than in A. floridana. Castaneous with gaster slightly darker.

Description of Male

TL 5.0, HL 0.85, HW 0.7, SL 0.35, AL 1.95 mm. General appearance

as in Figure eight. Head shape as Figure eight; narrower posterior to

eyes than Floridian congeners. Distinct depression between lateral

ocelli. In transverse cross section head rather rectangular. Antennal

scapes more slender than in A. floridana males.

Alitrunk as in Figure eight. Dorsum of propcdeum posteriorly with

pair of slight protuberances each surmounted by a "Y"-shaped carina.

Metathorax with pair of latero-ventrad swellings each narrowing distally

to rather blunt point. Swellings less well developed and less pointed

than in Floridian congeners except A. carolinensis.

Petiole and postpetiole as in Figure eight. Petiolar node in pro-

file less rounded than in A. floridana, but not as raised as in A. ash-

meadi. Viewed dorsally postpetiole nearly twice as wide as petiole.

Mandibles smooth, shining, and with feeble punctations and hint of

longitudinal striations. Clypeus shining and feebly punctate. Remainder

of head mostly punctate and devoid of rugosity except for a few feeble

transverse rugae between lateral ocelli. Alitrunk more feebly punctate

than in A. ashmeadi, A. floridana, and A. treatae. Scutum weakly punc-

tate and smooth at dorsal midline. Scutellum with denser punctations

and tending toward longitudinal rugae. Dorsum of propodeum glassy

smooth and shining. Mesothoracic sternite and episternite mostly glassy

and shining. Petiole and postpetiole shining, but largely punctate

particularly laterally and posteriorly. Gaster shining.

On posterior of postpetiole six setae overlap base of first gas-

tric segment. Brown with head darker brown.


As recently as 1950 Creighton reported that A. flemingi was known

only from its type locality in Mississippi. At that time, however, A.

macrospina had not yet been synonymized with A. flemingi. This ant has

since been found in the Atlantic Coastal Plain as far north as Pine

Bluff and New Bern, North Carolina (Carter, 1962b). I have seen

specimens from Louisiana.

This may be the least common of the soil-dwelling species of

Aphaenogaster in Florida. It is not nearly as abundant as A. ashmeadi

or A. floridana. In light of its known distribution in Florida, Missis-

sippi, and Louisiana its occurrence in the Florida panhandle west of

Leon County seems predictable. A. flemingi was not collected south of

Alachua and Putnam Counties in Florida. I collected this species sev-

eral times in the Gainesville area. Van Pelt (1958) found A. flemingi

occasionally at the Welaka Reserve. I have seen specimens from Colum-

bia County.


In the Gainesville area I found A. flemingi, like A. ashmeadi and

A. treatae, most commonly in xeric scrub oak woods. It was not found

in turkey oak woods or mesic hammocks, but occurs in pine woods where

A. ashmeadi and A. treatae are scarce or absent. At the Welaka Reserve

Van Pelt (1958) found A. flemingi in bluejack oak, longleaf pine and

Rutlege slash pine flatwoods. He believed the distribution of A.

flemingi on the reserve indicated a preference for pine growths. In

North Carolina Carter (1962b) found A. flemingi had a "distinct prefer-

ence for open, grassy sites of dry sandy soil "(p. 173). He too found

it in pine woods. I did not find A. flemingi in fields. Van Pelt

(1966) in his study of old field ants of the Savannah River Plant in

South Carolina did not find A. flemingi.

A. flemingi nests in only slight shade, often provided by herba-

ceous vegetation alone. Neither A. ashmeadi nor A. treatae nests are

exposed to as much direct midday solar radiation. The soil in which A.

flemingi nests is often not as well-drained as where A. floridana

nests. The soil in pine flatwoods is frequently near saturation and

temporary surface water collects in lower areas. Usually the nest is

at the base of a clump of grass or a herb. In Mississippi Smith (1928)

collected a colony of A. flemingi from the base of a rotten stump. I

found no nests associated with logs, stumps, or stones.


Nests of A. flemingi are of the same general structure as those of

A. ashmeadi. However, A. flemingi nests extend deeper and more later-

ally. There are one or two entrances, one of which is surmounted by a

thatched turret of small pieces of dried grass and plant parts. Turrets

of A. flemingi nests, which sometimes appear woven, contain arthropod

remains less frequently than do those of other Aphaenogaster species.

The turreted entrance serves as the passageway for foragers. Woody

roots usually obstructed excavation of A. flemingi nests. Brood was

often found directly beneath the clump of grass where the main nest

entrance was located.


A. flemingi forages mainly on the ground. Foragers were not ob-

served climbing higher than two cm on plants. Workers forage among

the herbaceous vegetation. I never saw foragers in the deep shade of

scrub oak thickets. Foragers wandered farther from their nests in open

areas than into nearer shady thickets. No foragers were seen more than

about 5.5 m from their nest. Foraging activity is greater in the even-

ing than at midday during the summer, particularly when the colony is

shaded only by herbaceous vegetation. On most summer days I could find

a few foragers in mid-afternoon. On summer afternoons colonies of A.

ashmeadi and A. treatae within 11 m of an A. flemingi foraged moder-

ately to actively, while very few workers were seen leaving or entering

the A. flemingi nest. Foraging activity decreases in winter. However,

on rather cool days in late November there is moderate foraging activ-

ity. Van Pelt (1958) called A. flemingi a moderately active forager.

In North Carolina Carter (1962b) found A. flemingi active from May to



The feeding habits of A. fleningi are like those of the preceding

species. However, foragers tend to bring in organisms associated with

A. flemingi nesting microhabitats. A. flemingi seems no more adept

than its congeners at cooperative attacks involving several workers

against rather large prey. A lepidopterous larva about two cm long,

starting to spin its cocoon between two overlapping leaves, was attacked

by A. flemingi workers. Ten workers surrounded the larva, yet it es-

caped. On only one occasion a forager was seen bringing in a termite.

In the laboratory A. flemingi workers readily attacked termites prof-

fered them. Smith (1928) reported breaking open a rotten stump con-

taining a termite colony and an A. flemingi colony. The ants did not

disturb the termites. However, this took place in Mississippi in Jan-

uary. I did not see A. flemingi collect seeds. It fed on pieces of

mushroom, Russula sp. in the laboratory and was seen visiting Russula

sp. in the field. A. flemingi was never seen tending Homoptera, but

in the laboratory workers drank sucrose and water solutions rather than

plain water proffered them. Van Pelt (1958) remarked that on several

occasions A. flemingi workers were attracted to molasses traps.

Coexistence with other Ants

The following species were found nesting in or traversing A.

flemingi microhabitats:






Odontomachus brunneus

Pseudomyrmex pallidus

Aphaenogaster ashmeadi

A. treatae

Pheidole dentata

P. metallescens

P. morrisi

Crematogaster ashmeadi

C. clara

Monomorium viridum

Solenopsis geminata

S. invicta

S. (Diplorhoptum) sp.

Leptothorax pergandei

L. texanus

Cyphomyrmex rimosus minutus

Trachymyrmex septentrionalis

Iridomyrmex pruinosus

Conomyrma sp.

Camponotus floridanus

Paratrechina (Nylanderia) sp.


Prenolepis imparis

Formica archboldi

F. pallidefulva

Remarks on individual species. Many of the species listed above

are discussed regarding A. ashmeadi. For special remarks about those

species listed but not mentioned below see corresponding sections under

A. ashmeadi or A. floridana.

In A. flemingi microhabitats Pseudomyrmex pallidus frequently

nests in hollow grass stems. It appears to forage very little on the


Coexistence among the various Aphaenogaster species is treated in

a subsequent section.

Pheidole metallescens was nesting or foraging within one m of

every A. flemingi nest I found in the Gainesville area. Except in some

pine flatwoods around Gainesville either Solenopsis geminata or S.

invicta was abundant wherever I found A. flemingi colonies. S. geminata

was seen raiding a small A. flemingi colony and routing the inhabitants

from their nest. In northwest Florida S. invicta has largely replaced

S. geminata in most habitats except mesic hammocks. Leptothorax tex-

anus Wheeler was found more regularly in A. flemingi microhabitats than

in those of other Aphaenogaster species.

Iridomyrmex pruinosus and Conomyrma sp. occur in the more xeric A.

flemingi microhabitats.

At least two forms (species ?) of Paratrechina (Nylanderia) occur

in A. flemingi microhabitats. A yellow nocturnal form lives in the

more xeric A. flemingi microhabitats, while at least one dark form

occurs where there is moist leaf litter. Formica archboldi M. R. Smith

was more often seen foraging in A. flemingi microhabitats than in those

of other Aphaenogaster species.

Annual Cycles, Mating Flights, and Colony Size

A colony excavated in Leon County on 28 April, 1973, contained

last instar female larvae. Males produced in a queenless laboratory

colony flew from their nest on several nights during May. These males

were active above ground or flew from 7:30 9:30 p.m. EDT. Another

laboratory colony contained both males and females, which intermittently

attempted nocturnal mating flights during August. The alates were

active from about nine p.m. to 12 a.m. The colony was placed in the

observation cage described in the materials and methods section. Eight

females and four males were present in the colony. No more than four

females flew on any one night. Alates flew no longer than one or two

seconds at a time. They alighted on the screened walls of the cage and

crawled about, usually upward. Encounters between opposite sexes were

frequent, but no copulations were observed. When a female encountered

a male she followed him for two or three cm. Females stroked males

with their forelegs and antennae, but the males moved away after a few

seconds or minutes. Flight activity would continue for two to 2.5 hrs.

Males appeared fatigued after 1.5 hr. They sometimes fell from the

walls, and landed upside down unable to right themselves.

Eight females which had flown in the cage were removed and placed

in separate shell vial nests. Two laid no eggs and died in early Sep-

tember. Six intermittently laid one or two eggs which they ate after a

few days. No larvae wereproduced. No copulations, therefore, seem to

have taken place at any time among this colony. The lights in the

laboratory were turned off for several minutes during flights. The

darkness elicited no observable change in their activities. Alates

still climbed toward the top of the cage and did not copulate. Perhaps

mating only takes place between non-siblings.

A few larvae and many general workers were found in an A. flemingi

colony excavated in late November. No colony excavated contained over

300 workers. The colony collected by Smith (1928) in mid-January in

Mississippi had 90 workers and "many larvae."


The duration of the pupal stage of A. flemingi is ten to 13 days.

Larvae are in the prepupal state three or four days.

Findings on A. floridana

Description of Female

TL 9.9, HL 1.9, HW 1.7, SL 2.1, AL 3.0 mm. General appearance as

in Figure nine. Head shape as in Figure nine; narrowed posteriorly,

but less so than in worker. Head wider than in female of A. flemingi.

Antennal scapes exceed occipital margin by at least one quarter their

length. Base of scape with small angular forward projecting lobe about

the same size as lobe on base of an A. floridana antenna. Occipital

margin with more developed rim than females of Floridian congeners.

Alitrunk as in Figure nine. Propodeum with pair of short posteri-

orly directed spines, like those of A. ashmeadi and A. treatae females,

but differing markedly from the unarmed propodeum of own worker. A.

flemingi female similar in many respects, but has long rather upwardly

directed propodeal spines.

In profile petiolar node slightly more rounded than in similar A.

ashmeadi, A. treatae, and A. flemingi; considerably more acute than in

own workers. Viewed dorsally postpetiole as broad as long; not as elon-

gate as in worker. In profile postpetiole rather evenly convex above,

more so than in Floridian congeners.

Far more heavily sculptured than workers. Mandibles with longi-

tudinal striae, which spread slightly as they approach masticatory

margin. Clypeus with slightly wavy longitudinal rugae. Coarser longi-

tudinal rugae over front and lateral portions of head. Some transverse

connections between rugae. Punctations interspersed between rugae.

Rugae weaker and reticulate to transverse posterior to ocelli. Prono-

tum with rugae parallel to suture between pronotum and scutum. Scutum

longitudinally rugose. Longitudinal rugae continue across parapteron

to scutellum, where they are more reticulate posteriorly. Propodeum

with transverse rugae; area between spines with more feeble rugae.

Mesothoracic sternite and episternite with rather longitudinal rugae,

but smoother and shinier ventrad. Coxae and femurs shining, but with

some faint punctations. Petiole punctate, with some dorsal and lateral

rugae or granulation. Postpetiole punctate, with dorsal and lateral

rugae tending to be transverse posteriorly. Basal segment of gaster

finely punctate.

Hairs on scutum longer than those of A. ashmeadi and thicker and

more numerous than those of A. flemingi. Yellowish to light-orange

brown, gaster slightly darker. Females of Floridian congeners darker.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs