Skip to main content

Full text of "Distribution of the Argentine ant in the United States and suggestions for its control or eradication."

See other formats


ML. Brown, Jr. 


CIRCULAR No. 387 MAY 1936 






By M. K. Smith, associate entomologist, Bureau of Entomology and Plant 

Quarantine 1 

Mum l BRom 



Introduction ; 1 

How to recognize the Argentine ant 3 

Habits and economic importance 3 

Distribution and abundance 5 

Means of spread 5 

Response to ecological factors 7 

General distribution in the United States.- 9 
Occurrence and abundance in the infested 

States 11 

Control and eradication work in Mississippi.-- 24 

Surveying infested areas 25 

Estimating the cost of a campaign 26 

Supplies needed 26 

Organizing and conducting a campaign 30 

Control and eradication, etc.— Continued 

Best seasons for campaigns 32 

Reaction of ants to the poison ^ 33 

Determining the effectiveness of a cam- 
paign 33 

Eradication work 34 

Supplementary measures 34 

Success of control and eradication work in 

Mississippi 35 

Method recommended for the control or eradi- 
cation of the Argentine ant-_ 36 

Summary 37 

Literature cited 38 


Sometime previous to 1891 there entered the United States at 
New Orleans, La., a South American ant which was destined in less 
than 50 years to spread over not only a large part of the Southern 
States but a considerable area in California, as well as small areas 
in several other widely removed States. Within these limits the 
Argentine ant, Iridomyrmex hwnvilis (Mayr), as it is called, has 
proved to be probably the most annoying of the economic ants and 
a pest of no little importance. Although especially obnoxious to 
housekeepers, hotel managers, and cafe owners, it also causes serious 
losses to orchardists, planters, beekeepers, and others. 

According to Newell {10), 2 Edward Foster, of New Orleans, was 
the first person to observe this ant in the United States. From 
Foster's observations (3) it appears likely that the Argentine ant 

1 The work described in this circular was under the general direction of S. r A. Rohwer. 
The scouting and the evaluation of the Mississippi method of control and eradication were 
under the supervision of the author. The scouting was done by T. F. McGehee, L. C. 
Murphree, H. T. Vanderford, and D. B. Read. The States scouted by tbese men were as 
follows: Alabama, Mississippi, 'Arizona, New Mexico, Illinois (Murphree) ; Georgia, North 
Carolina, Virginia, West Virginia (Vanderford) ; Louisiana, Oklahoma (McGehee) ; 
Florida, South Carolina, Missouri, Arkansas (Read) ; Texas (Murphree, Vanderford, Reed, 
and McGehee) ; Kentucky and Tennessee (Vanderford and Murphree). Information on 
the status of the ant in California was furnished by Harry S. Smith, and in Maryland by 
B. N. Cory. Officials in entomological, plant quarantine, and extension work in the 
various States offered many helpful suggestions and aid. The data relative to the work 
in Mississippi were supplied by Clay Lyle and R. P. Colmer, of the State Plant Board of 
Mississippi. The photographs were made by Ross E'. Hutchins. 

s Italic numbers in parentheses refer to Literature Cited, p. 38. 

41271°— 36 1 


was first carried to New Orleans in coffee ships from Brazil. These 
ants were first found in California at Ontario in 1905 by E. S. G. 
Titus. In 1907 they were discovered at Berkeley by J. C. Bradley, 
and by 1910 C. W. Woodworth, of the University of California, 
reported that 5,000 acres in that State were infested with this pest. 
In 1915 Newell and Barber (11) listed 16 localities in California 
as being infested. Since that time the ants have continued to spread 
rapidly over new areas in the United States. 

As this insect increased in economic importance, many investiga- 
tions of its biology and control were undertaken. Newell and Bar- 
ber (11) have published a comprehensive account of the ant's life 
history and a description of its stages. Work by Horton (8, 9) 
in the citrus groves of Louisiana and by Woglum and Borden (17) 
in California deals with this ant in its relation to citrus, and the 
means of combating it. The development of a satisfactory poison 
bait by Barber (2) was the culmination of much work along this 
line by Barber, R. S. Woglum, and A. D. Borden, working for the 
Bureau of Entomology, United States Department of Agriculture, 
and Nickels (1£) , of the Department of Entomology of the Univer- 
sity of California. Not only has this bait given satisfactory results 
wherever it has been used in the South, but Ryan (13) has also in- 
dicated its successful use in citrus groves in California. Mississippi 
was the pioneer State in Argentine ant control and eradication 
work, regular control campaigns having been organized there as 
early as 1920. 

The act making appropriations to the Bureau of Entomology for 
the fiscal year 1931-32 provided funds for a further study of the 
Argentine ant. The work outlined was divided into two parts: (1) 
To determine the present distribution and relative abundance of 
the Argentine ant in the United States, and to map the infested 
areas as accurately as possible within a limited time and at a rea- 
sonable expense; and (2) to study the methods used by the State 
of Mississippi in controlling and eradicating the Argentine ant, and 
appraise the effectiveness of suppression campaigns conducted by that 

The survey to determine the distribution and abundance was begun 
in the fall of 1931 and completed late in the spring of 1933. During 
this time 4 men scouted for the ants in the principal towns and 
cities in 18 States. No effort was made to scout rural areas because 
of the great expense involved and the probability of finding only 
a small number of infestations. The States scouted most thoroughly 
were South Carolina, Florida, Georgia, Alabama, and Louisiana, 
one or more localities in every county being examined for ants. 
Scouting was also done in Oklahoma, North Carolina, Virginia, West 
Virginia, Kentucky, Tennessee, Mississippi, Illinois, Missouri, Arkan- 
sas, Texas, Arizona, and New Mexico. Since the status of the ant 
in California and Mississippi was rather well known, no work was 
done in California and only a few localities in Mississippi were 
scouted. The data collected in connection with this survey, which 
are summarized in this circular, are believed to be adequate for ap- 
praising the Argentine-ant situation as it existed in the United States 
at that time. 

During the summer of 1933 the author and L. C. Murphree in- 
spected certain areas in Mississippi believed to have been freed of 


Argentine ants as a result of campaigns directed by the State Plant 
Board. The methods of inspection and those used in the control and 
eradication work are described in this circular. 


Argentine ants are easily confused with other species, and in the 
course of the surveys native ants that resemble Argentine ants are 
occasionally encountered. The Argentine ant (fig. 1) can be dis- 
tinguished from these by the following characteristics : The worker 
is about one-eighth inch long and of an almost uniformly brown 
color. It is slender in 
form and moves 
rather quickly. The 
worker possesses a 
single joint (petiole) 
between the thorax 
and the gaster. When 
several workers are 
crushed, a character- 
istic musty or greasy 
odor (16) is given off. 
Other native ants 
either are odorless or 

they have Spicy or Figure l.— Argentine ant worker. Greatly enlarged. 

pungent odors, or a 
somewhat sweetish, nauseating odor similar to that given off by Tap- 
inoma ants. Some of the ants most closely resembling Argentine 
ants have the last-named odor. 


Although the Argentine ant affects our interests in many ways, it 
has made itself notorious primarily as a house-infesting insect. In 
the South this species invades the house almost continuously 
throughout the year, even in winter when temperatures are as low 
as 50° F. When a property is heavily infested, the ants may be 
found not only in pantries, dining rooms (fig. 2), and kitchens, but 
even in refrigerators, beds, etc. In certain sections dwellings have 
been vacated because of the ravages of this; pest. The presence of 
the ants in a house cannot be taken as reflection on the neatness of 
the housekeeper, as many houses kept spotlessly clean are invaded. 
Late in the summer and especially after heavy rains, perhaps be- 
cause the honeydew has been washed off the plants outside, the ants 
become unusually troublesome in the house. 

The Argentine ant is practically omnivorous in its feeding habits. 
If it has any preference for foods, this would seem to be sweets. 
A scum of dead ants from one-eighth to one-fourth inch deep has 
been seen in cans containing table sirup. Sugar, jelly, pies, and 
candies are consumed with relish. The ant will crawl on top of a 
block of ice to reach meat lying there. Occasionally it feeds even on 
corn meal. 

. Outdoors the ants make themselves noticeable by their numbers 
as well as by the damage they do. In some sections they are so 


abundant as to form numerous (rails up and down the trunks of 
trees, along streets, fences, the base of houses, and on the "round. 
In heavily infested areas it is not uncommon to see trails 5 or 6 
inches wide on the trunks of trees. By caring for and fostering 
mealybugs (Woglum and Neuls (25)) 'and aphids, the Argentine 
ant enables these insects to become so abundant that they do serious 
damage to plants. Such injury is especially noticeable in sugar- 
cane fields and citrus groves. In his scouting T. F. McGehee noted 
that ladybird beetles were not abundant in localities heavily infested 
with the Argentine ant. Horton (#) also has recorded the lack of 
predators and parasites of honeydew-producing insects in sections- 
infested with Argentine ants. This is no doubt due to the preda- 
ciousness of the ants as well as to the guard which they keep over 
the honeydew producers. The Argentine ants also affect plants 

Figure 2. — A scene familiar to housekeepers in ureas infested with the Argentine ant. 

(Newell and Barber.) 

directly. Gardeners frequently find their lettuce and other small 
seeds removed from the. seed beds. The ants will enter the calyx 
end of ripe figs and gnaw into the interior. Worker ants have also 
been noted to cut through the buds of carnation and orange in 
order to reach the nectaries, and they will feed on the sap of trees 
where the bark has been broken (Horton (9)). 

Poultry raisers in localities where Argentine ants are abundant 
have reported that the ants drive hens from the nests by crawling 
over them, especially if an egg is broken in the nest or when hatch- 
ing begins. The ants often kill newly hatched chicks to obtain their 
blood. According to Newell and Barber (//). many species of birds 
suffer similarly from the ravages of the ants. 

Beekeeping is especially hazardous in areas where the Argentine 
ant is abundant. The ants invade the hives for honey, and, since 
the bees cannot combat them effectively, this may lead to the destruc- 
tion of the colony. The author has seen a honeycomb in which 


every cell had been emptied of its honey content by Argentine ant 

As far as known, no one has proved that the Argentine ant 
transmits human diseases. The workers' habit of visiting garbage, 
refuse, feces, sputum — in fact, anything that might harbor disease 
germs — suggests that they might easily transmit such diseases as 
typhoid fever, diarrhea, and tuberculosis. Horton (8) has sug- 
gested that the workers may also transmit certain plant diseases 
through their habit of feeding on the sap flowing from the wounds 
of plants. 

Argentine ants differ from, native ants in that, once they have be- 
come established on a property, they will usually not tolerate other 
species. Therefore, instead of many different kinds of ants on a 
property, there is ordinarily an abundance of only one, the Argentine 
ant. The author has often witnessed combats in the field between 
native and Argentine ants. It is not unusual to see five or six Argen- 
tine ants attack a single native ant. Although some of the native ants 
have the advantage in size, or by the possession of a sting, they 
finally succumb to the attack of these South American invaders 
because they are outnumbered. 

The fact that the Argentine ant destroys practically all the native 
ants as it advances makes it comparatively easy to delimit an area 
infested by them. One must not conclude, however, that after 
Argentine ants are eradicated from an area, that area will be free 
of native ants also. Just as soon as the Argentine ants begin to 
disappear, native ants invade the territory, and within a very few 
years are apparently as plentiful as ever. In Mississippi the fire 
ant Solenopsis asyioni McCook, an arch enemy of the Argentine 
ant and often mistaken for it, is usually the most abundant native 
species following the eradication of the Argentine ant. 


The unusually rapid spread of the Argentine ant has been due 
to its ability to produce prodigious numbers, its habit of extermi- 
nating competitive species as it spreads, its lack of natural enemies, 
its omnivorous feeding habits, and especially to its ability to thrive 
in human habitations as well as in rural sections. 

Its distribution has been accomplished by both artificial and nat- 
ural means. Artificial spread has been by far the more important. 
Common carriers and merchandise have transported the ants great 
distances, thus establishing infestations in localities remote from 
one another. Along the important railroad lines in the Southern 
States, especially those in the Gulf coast section leading out of New 
Orleans, nearly every locality of any commercial importance is 
the seat of an Argentine ant infestation. Such cities as New Or- 
leans, Birmingham, Montgomery, and Atlanta have served as focal 
centers for the distribution of the ants in neighboring towns and 
rural areas. The ants can be carried in shipments of lumber, plants, 
dry goods, groceries, etc. The spread by means of boats and rafts 
has not been so great as by railroad, because there is less water 
transportation, but in areas where water transportation is common, 



accounted for many of t 1 

especially in southern Louisiana, il 


The natural spread of the Argentine ants is comparatively slow, 

because dispersion is dependent upon crawling rather than flight. 

Only the queens and males have wings, and the nuptial flight is so 

infrequent as to he negligible, mating usually taking place in the 

nest. The author re- 
calls only one instance 
in a period of 13 years' 
field work when the 
ants were actually 
seen in flight. The 
natural spread is 
effected for the most 
part in the search for 
food and nesting quar- 
ters, principally the 
former. Normally 
the advance is only 
a few hundred feet a 
year. They often 
spread in some direc- 
tions much more rap- 
idly than in others. 
Sometimes the ants 
will not cross a street 
on which there is 
heavy traffic for a 
number of years. To 
reach a source of food 
supply Argentine 
ants will cross water 
placed in saucers, but, 
they will not attempt 
to cross the water of 
streams of any depth 
or width, especially 
if there is a decided 

Floods occasionally 
carry the ants long 
distances. Colonies 
may be carried for 
miles on floating de- 
bris such as rotten 
logs. Often the ants 
the workers clustering 
mass. When such 


-Trees with rotten crevices, 
myriads of Argentine unts. 

which harbor 

will form a rotating ball in the water, 
on the outside of their brood in a spherica. 
balls of ants lodge in suitable locations, they can start new nests 
A scout once observed an ant colony on a wooden post that had 
been under water 3 days. After the water had receded, the colony 
was still present and apparently unharmed. Many infestations es- 
tablished by overflows have been found in southern 'Louisiana, where 


the ants exist in immense numbers in areas that are subject to flood- 
ing. Nesting in crevices and rotten sections of trees (fig. 3), or 
else in or behind vines twining around the trunks (fig. 4), the 
ants can easily withstand the highest waters. Some of the terri- 
tory in which they are found is either sparsely or not at all 
settled by humans. 


In the scouting 
work it has been 
noticed that Argen- 
tine ants are very 
susceptible to winds. 
When there is a 
strong cold w T ind. they 
attempt to protect 
themselves either by 
crawling into well- 
protected places or 
by almost completely 
ceasing activity. The 
writer has seen strong 
trusts of wind blow- 
workers several inches 
from their trails on 
concrete pavement. 
but apparently with- 
out harm. 

The ants are de- 
pendent on moisture 
for the success of their 
colonial activites. 
Hertzer (7) has 
shown that in the 
laboratory the ants 
prefer moist loca- 
tions, and she believes 
that moisture is es- 
pecially necessary for 
the transformation of 
the egg and larval 
stages. Lyle 3 noted 
that worker ants al- 
ways sought the moist 
spots in experimental 

cages, and when moisture was not available they died, apparently 
from no other cause than lack of it. There is, however, a limit to 
the quantity of moisture that they tolerate. The author has never 
seen them living in water-logged habitats. In swamps or low places 
they nest in logs or stumps above the water level (fig. 5) or seek the 

Fi<;i;re 4. — Vine-covered trees in southern Louisiana, 
which afford excellent nesting places for Argentine ants. 

8 Ltle. C. 
Thesis. Miss 


State Col. ) 

22 pp., illus. 1931. (Master's 



higher and better drained spots. Barber (/) claims that excessive 
water in their nests, especially in the winter, will reduce their popula- 
tion greatly. 

Although the Argentine ant is a tropical insect, and its northward 
spread is held in check by prolonged severe winters, the writer knows 
of but one other species of ant, Prenolepis imparls Say, in the South 

that is so tolerant of 
cold. Mention lias 
already been made of 
the Argentine ant's 
ravages in refrigera- 
tors. At air temper- 
atures of around 50° 
F., and sometimes 
below this, the ants, 
although active out- 
doors in places, move 
along their trails 
rather sluggishly. In 
protected places 
workers can occasion- 
ally be encountered 
moving at tempera- 
tures even as low as 
freezing, but the 
colder it is the less 
active a n d less in 
evidence are the ants. 
This makes scouting 
especially difficult 
in cold weather, and 
to be reasonably sure 
of an accurate survey 
it is best not to scout 
at temperatures lower 
than 65° to 70°. For 
short intervals the 
ants will resist tem- 
peratures much lower 
than this. The author 
once found in a rotten 
log, packets of work- 
ers that were thor- 
oughly encrusted with 
ice and were so stiff 
and motionless as to 
give the appearance of 
being dead. The temperature the night before had been 13°. When 
a mass of the workers was taken to a laboratory, they thawed out 
in an hour or so. without any apparent ill effect, and became as 
active as ever. Herbert {(>) states that Argentine ants can be lulled 
under cold-storage conditions by exposure to 31° for 24 hours, but 
the writer's field experience causes him to question these conclusions, 
which were based on a limited number of experiments. 

Figure 5. — Uprooted stump, surrounded by water below. 
in whicfi Argentine ants were found nesting. 


Argentine ants have been found at altitudes ranging from approxi- 
mately sea level (3 feet) to as high as 3,955 .feet. 

So far as could be observed, the ants are capable of living under 
a great variety of soil conditions. They have been found in soils hav- 
ing textures ranging from light sandy loams to clay loams, with 
poor to good drainage, and with varying amounts of organic matter, 
including the soils of the stream bottoms, of the prairies, and of the 
marshes, as well as those of the uplands. 

It was found that certain types of vegetation seem to attract the 
Argentine ant more than others. The presence of aphids, mealy- 
bugs, scale insects, and other honeydew-producing forms is undoubt- 
edly the reason for such preference. Trees that are especially 
frequented include oaks (chiefly water and live oaks), willow, hack- 
berry, pecan, sweetgum, pine, fig, maple, elm, and citrus. 


Argentine ant infestations have been found in nearly one-third of 
our States (fig. 6). The total area infested is approximately 4,000 
square miles. Practically all the infested areas, with the exception 
of those in California, are in the Southern States, chiefly in the Gulf 
coast section. The infestations in Arizona, Missouri, Illinois, and 
Maryland appear to be isolated, and of these only that in Arizona 
is typically an outdoor type, as in the other States the ants appear to 
survive the winter only in buildings, especially those that are heated. 
In a heavily populated urban area, where the houses are practically 
contiguous, particularly in northern locations, the ants might spread 
from house to house until they occupy all or most of the city, because 
of their ability to survive winter conditions in the interior of buildings. 

The Argentine ant has spread more rapidly eastward from New 
Orleans than it has westward or northward. Except for California 
and a single infestation in Arizona, it is not found farther west than 
the eastern half of Texas. In the northern parts of Alabama and 
Mississippi infested areas are few and small, and in northern Louis- 
iana there are practically none. 

Ealeigh, N. C., is the most northern locality in the Eastern States 
where the ant has been found to live outdoors and survive winter 
temperatures. The infestation here is at least 15 or 16 years old, 
having been found by the author (14-) in September 1919, and as yet 
it apparently extends only about 0.15 square mile. The slow spread 
of the ants in Ealeigh is an indication that they are very near the 
northern limit of distribution under outdoor conditions. An out- 
door infestation at Nashville, Tenn., found by Barber was exter- 
minated within a year, presumably by severe winter weather. Bar- 
ber also found the ants at Memphis several years ago, but neither 
the scouts of the Bureau of Entomology nor representatives of the 
office of the State entomologist of Tennessee have been able to find 
any there in recent years. 

In Mississippi, however, where there were only about a dozen 
known infestations in 1910 (Harned and Smith (5) ) , in 1934 there 
were at least 245, exclusive of the isolated areas within many of the 
infested towns. Inspectional work, of the State Plant Board of 
Mississippi has retarded the spread of the ant considerably, and 
were it not for such service the situation would be much worse. 

41271°— 36 2 



From what is known of the habits and distribution of the Argen- 
tine ant in Mississippi, it is probable that there are isolated rural 
infestations in many of the States, especially in the Gulf coast region. 
In Monroe County, Miss., 17 rural infestations are known. Of the 
known infestations in Mississippi, 26.7 percent are in rural com- 

Unless controlled, it seems likely that the ants will in time be gen- 
erally, if not continuously, distributed throughout those Southern 
States in which numerous scattered infestations now occur. In ad- 
dition they will probably occur in most of the southern half of 
North Carolina, the extreme southern part of Tennessee, the southern 
half of Arkansas, the extreme southern part of Oklahoma, the eastern 
half of Texas, as well as in a large part of California, particularly 
in the lower altitudes. 


The information on the occurrence of the Argentine ant obtained 
by scouting or through correspondence is summarized in table 1 and 
in the following paragraphs. Of the 18 States where scouting was 
conducted, 13 were found infested. 







Although the infestations in Alabama are rather generally dis- 
tributed over the State, they are more numerous in the central part 
and less prevalent in the northern part (fig. 7). Jefferson County, 
with 10 infested localities, and Wilcox County, with 7, are the two 
most heavily infested counties. Many of the infestations are ex- 
tremely small, indicating that they are recent. The largest infesta- 
tion is at Birmingham and embraces 124.02 square miles. Mont- 
gomery follows with 
31.39 square miles, 
and Mobile third with 
10.44 square miles. 

Practically all the 
infestations originated 
from articles shipped 
by railroad. A few 
towns, however, appar- 
ently became infested 
through truck or water 
shipments. Birming- 
ham is an excellent 
example of a large 
commercial city that 
has become a focal cen- 
ter for the spread of 
the ants to nearly all 
the small towns in its 

The scouting in Ari- 
zona was confined prin- 
cipally to the southern 
third of the State, but 
examinations were also 
made in two locali- 
ties north of Phoenix, 
Prescott and Kingman. 
The localities chosen 
were the most impor- 
tant commercial points 
along the main high- 
w a y s and railroad 
lines. They were as follows: Tombstone, Bisbee, Lowell, Douglas 
(Cochise County), Safford (Graham County), Phoenix, Tempe, 
Mesa, Glendale, Buckeye, Gila Bend, and AVickenburg (Maricopa 
County), Kingman (Mohave County), Tucson (Pima County), 
Nogales (Santa Cruz County), Prescott (Yavapai County), and 
Well ton and Yuma (Yuma County). 

The single infestation in the State is at Douglas, in Cochise 

Figure 7. — Map of Alabama showing number of localities 
scouted for Argentine ants and the infested localities 
(dots) in each county. 



In Arkansas the localities scouted were confined for the most part 
to the larger and more important railroad junctions or river towns 
in the eastern and southern sections of the State. The river towns 
lay along the banks of the White, Arkansas, and Mississippi Rivers. 
The localities scouted were as follows: Blytheville, Osceola, Wilson 
(Mississippi County), Marion (Crittenden County), Forrest City 
(St. Francis County), Marianna (Lee County) Helena and Barton 
(Phillips County), Clarendon (Monroe County), DeValls Bluff and 
Des Arc (Prairie County), Searcy (White County), Little Rock 
(Pulaski County), Benton (Saline County), Pine Bluff (Jefferson 
County), McGehee and Arkansas City (Desha County), Lake Village 
(Chicot County), El Dorado (Union County), Camden (Ouachita 
County), Arkadelphia (Clark County), Malvern (Hot Spring 
County), Hot Springs (Garland County), De Queen (Sevier 
County), Ashdown (Little River County), Texarkana (Miller 
County, Ark., and Bowie County, Tex.). 

. Of the two infestations found, that at Hot Springs is very small, 
only 0.04 square mile, while that at Texarkana is rather large, com- 
prising four separate areas which cover a total of 3.38 square miles. 
Both infestations probably originated from railroad shipments. 

Although the eastern and southern sections of the State border on 
Mississippi and Louisiana, respectively, the lack of infestations in 
these sections is probably due to -the fact that neither northern 
Louisiana nor northwestern Mississippi is as yet heavily infested. 


The distribution of the Argentine ant in California can only be 
summarized here in a general way, because the Bureau of Entomology 
did no scouting in this State, nor has the exact distribution been 
determined by the State authorities. The information given below 
and on the map (fig. 8) was kindly furnished by Harry S. Smith, 
of the Citrus Experiment Station at Riverside, Calif., and repre- 
sents a summary of reports supplied him by county agricultural 

The San Francisco Bay district including the Santa Clara Val- 
ley and the coastal plain to the south, the San Joaquin Valley south 
of and including Sacramento County, a considerable part of Los 
Angeles and Orange Counties, and western San Bernardino, River- 
side, and San Diego Counties may be said to be generally infested ; 
that is, colonies of ants were found in every residence and on a con- 
siderable portion of the orchard acreage. Dr. Smith is of the 
opinion that there are at least 2,000 square miles of infested territory 
in California, and that this area is far greater than in any other 
State. He says : 

The ant has been here for more than 30 years, conditions are ideal for 
its development, and the unusually extensive .distribution of balled ornamentals 
has seemed to disseminate it very widely. 

In Florida the known infestations are in the northeastern and 
northwestern sections of the State (fig. 9). As yet none of the infes- 



tations is large, that at Pensacola (Escambia County) being the 
largest, -with 3 separate areas totaling 0.20 square mile. Caryville 
(Washington County) has 0.11 square mile, and Palatka (Putnam 
County) and Jacksonville (Duval County) 0.06 square mile each. 
The history of these infestations is rather obscure, but it is probable 
that boats and railroads played almost equal parts in carrying the 
ant to this State. 

That Florida has so few infestations is surprising in view of the 
many factors that might aid the ant in becoming established there, 

«««,«* Ml 

t*KJ Generally distributed in settled portions 

Scarce or uncertain 
YYYYY.i Not known to be present 
I I No reports 

Figukb 8. — Map of California showing Argentine ant distribution. 

such as a semitropical or mild climate, a heavy tourist trade, and 
opportunities for the ants to come in by both boat and railroad. 


Georgia (fig. 10) is not so generally infested as Alabama, the in- 
festations being confined largely to the northwestern part, especially 
the area lying between Columbus and Atlanta. The extreme south- 
eastern, southwestern, and northern sections are practically free 
from infestations. The most heavily infested counties are Coweta, 
Troup, and Fulton. A large number of these infested areas lie along 



the Atlanta and West Point Railroad. Most of the infestations in 
Georgia, like those in Alabama, without doubt originated from ants 
infesting products shipped by railroad. 

The largest infested area in the State is at Atlanta, comprising 
26.70 square miles. The next largest are Newnan, 13.68 square miles ; 
Moreland and St. Charles together, 12.48 square miles ; and Augusta, 
8.15 square miles. Atlanta, like Birmingham, has been the source 
of infestation of many of the surrounding small towns. 

>«i 3 

I 5 

1 \HOLM£3lJA(. 

5 \ 

3J\ 2 \ 

f^4» l-LA 

2 $]ir 3 Ji' 




<3/tf 3 \vAKULtJr\lii / 

yAYLon~)FY 2 1 S 2 

\ '/OiXf£ 3$ 



\/ * \ MARION ^<J-^_3J\ 




fmsco Lk^ 

J 2 (&** 


! 3 ( 

V3\H/USaOAM *-\ *" 

*t \ 

%ty 3 


< .3 

6 S^ 

<n \ 




1 s 3 T3\ 

S r 1 / Y3 A 





3 V, 


jT „U£ 

V 4 

^N3j-i l 2 



\3 l_ 4 — | 

\ DAO£ ffl 

\ 3 5 / 

FlGUKB 9- 

-Map of Florida showing number of localities scouted for Argentine ants and 
the infested localities (dots) in each county. 

In Illinois the following towns were surveyed : Cairo (Alexander 
County), Champaign and Tolono (Champaign County), Mattoon 
( Coles County), Neoga (Cumberland County), Areola and Tuscola 
(Douglas County), Edgewood and Effingham (Effingham County), 
Carbondale and De Soto (Jackson County), Vienna (Johnson 
County), Centralia, Odin, Salem, Alma, and Kinrmindy (Marion 
County), Metropolis (Massac County), DuQuoin and Tamaroa 
(Perry County), Mounds and Ullin (Pulaski County), Harrisburg 
(Saline County), Anna (Union County), Ashley (Washington 
County) , and Marion (Williamson County) . 

No Argentine ants were found as a result of the scouting, but the 
species is established in the Zoology Building at the University of 



Chicago. 4 This infestation supposedly originated in 1906 from 
ants introduced on plant material for use in connection with investi- 
gations on the Colorado potato beetle '(Leptinotarsa decemlineata 
Say). Colonies are located in the basement of the building and oc- 
casionally cause considerable annoyance in the laboratories. Al- 
though in warm weather the ants are found outdoors a number of 
feet from the building, they have not spread to other buildings or 

Figvjke 10. — Map of Georgia showing number of localities scouted for Argentine ants and 
the infested localities (dots) in each county. 

over the campus. Apparently they are able to survive the winter 
only indoors. 


Scouting in Kentucky was confined to the northern parts lying 
along the Ohio and Mississippi Rivers, and also along the Illinois 

4 Data on this infestation -were supplied by Mary Talbot, of the University of Chicago, 
and V. E. Shelford, of the University of Illinois. 

41271°— 36 3 


Central Railroad in the extreme western part of the State. The 
■localities examined were as follows : Wickliffe and East Cairo (Bal- 
lard County), Wellsburg (Bracken County), Petersburg (Boone 
County), Holt (Breckinridge County), Catlettsburg and Ashland 
(Boyd County), Bardwell (Carlisle County), Carrollton (Carroll 
:County), Weston (Crittenden County), Owensboro (Daviess 
County), Hickman and Fulton (Fulton County), Warsaw (Gallatin 
County), Greenup (Greenup County), Clinton and Columbus 
(Hickman County), Hawesville (Hancock County), Covington 
(Kenton County), Louisville (Jefferson County), Vanceburg 
(Lewis County), Smithland (Livingston County), Maysville (Ma- 
son County), Paducah (McCracken County), Brandenburg (Meade 
County), Westport (Oldham County), Henderson (Henderson 
County), and Uniontown (Union County). No Argentine ants were 
found. The known infestations nearest to Kentucky are in the 
extreme southern part cf Tennessee. 


In Louisiana the ant occurs not only in cities and towns, but in 
rural or sparsely settled areas, and in almost impenetrable swamps. 
The southern half of the State is not only generally but heavily 
infested, whereas in the northern part there are only a few widely 
separated infestations (fig. 11). Although the heavier infestation 
in the southern half of the State might be attributed to the milder 
climate and the earlier appearance of the ants, it is probably due 
largely to spread by floods and boats. 

The importance of railroad lines in the dissemination of the ants 
is clearly shown in Tangipahoa Parish, where 10 of 13 localities 
along the Illinois Central Railroad were found infested. There are 
more large infestations in Louisiana than in any other State. Such 
infestations range from 10 to 35 square miles on an average, and 
there are several as large as 120 to 135 square miles. Owing to their 
inaccessibility, many localities, especially in the southern part of the 
State, were not scouted. The infested area is therefore probably 
larger than that given in table 1. 


No scouting was carried on by the Bureau of Entomology in 
Maryland. The author first learned of the occurrence of the Argen- 
tine ant in this State when he received from E. N. Cory, State ento- 
mologist, specimens collected from the Clifton Park greenhouses in 
Baltimore. Later investigations by Dr. Cory and members of his staff 
revealed the presence of this ant in greenhouses at the following 
parks in Baltimore: Druid Hill, Carroll, Patterson, and Clifton. 

On October 7, 1931, the author and D. W. Hookom, of the Uni- 
versity of Maryland Department of Entomolgy, visited greenhouses 
in both Clifton and Carroll Parks where the Argentine ants -were 
established. At Carroll Park colonies were found nesting outdoors 
75 yards from the greenhouse, but at Clifton Park the ants were 
not more than 10 feet away. Little definite information regarding 
the history of the infestation could be obtained. Judging from the 
abundance of the ants at Carroll Park and from statements of green- 



house attendants, they undoubtedly have been present there for 
several years. 

As a result of an investigation during the winter of 1931, Mr. 
Hookom found that the ants lived outdoors in protected places, but 
the winter was comparatively mild. 


In Mississippi scouting for Argentine ants was confined to 24 
localities in the extreme northern part of the State, because the State 
Plant Board was rather well posted as to the general distribution 

Figuke 11. — Map of Louisiana showing the infested localities (dots) in each county. 

of the ant elsewhere. The summary of conditions in Mississippi is 
based on records of both the Mississippi State Plant Board and the 
Federal Bureau of Entomology. 

Although the ant is generally distributed throughout the State 
(fig. 12), it is most abundant in the southern and central sections, 
and least abundant in the northwestern portion. Railroads have 
been the principal means of spread. So clearly is thig indicated that 
one can locate the important railroad lines in the State by following 
on the map the dots representing the infested localities. The most 
heavily infested counties are Monroe, Hinds, Attala, and Holmes. 
Many of the infestations are small, but in gome of the towns the 
infestation covers the entire incorporated area and extends into the 
country for several miles. Those at Jackson, Meridian, and Kosci- 



usko undoubtedly cover from 12 to 15 square miles each. In a num- 
ber of towns there is more than one infested area; at Columbus, 
for example, 18 isolated infestations have been found. A large 
number of the infestations, particularly in the heavily infested 
counties, occur in rural areas, and these no doubt originated through 
introduction of groceries or other products from neighboring towns. 
Recent intensive' scouting shows that 39 of the 245 infested locali- 
ties have been freed of the pest, as a result of poisoning campaigns. 


The following towns 
were scouted in Mis- 
souri : Poplar Bluff 
(Butler County), Cape 
Girardeau and Delta 
(Cape Girardeau 
County), Jefferson City 
(Cole County), Boon- 
ville (Cooper County), 
Washington (Franklin 
County) , Hermann 
(Gasconade County ) , 
Charleston (Missis- 
sippi County), New 
Madrid (New Madrid 
County) , S e d a 1 i a 
(Pettis County), Ca- 
ruthersville (Pemiscot 
County), Ste. Gene- 
vieve (Ste. Genevieve 
County), and St. Louis 
(St. Louis County). 

Argentine ants were 
found in St. Louis, but 
only at the Missouri 
Botanical Gardens, and 
in a nursery nearby. 
From the information 
obtained it does not 
seem possible for the 
ants to survive outdoor 
winter conditions here. 
Within the greenhouses 
and other heated buildings the ants seem to thrive and are making 
themselves objectionable through their attendance on aphids and 
mealybugs. Only two infestations are known to occur farther north 
than the one in St. Louis, and both of these are also of the indoor 


The towns scouted in New Mexico lie along the railroad lines and 
are as follows : Las Cruces, Hatch, Rincon, and Mesilla Park (Dona 
Ana County), Silver City (Grant County), Lordsburg and Rodeo 

n Localities from which the ants 
hare been eradicated 

^Localities from tvh/ch the ants 
hare anparent/y been eradicated 

Figure 12.- 

-Map of Mississippi showing status of 
Argentine ant infestations. 



(Hidalgo County), and Denting and Cambray (Luna County). 
Although the scouting was limited, it does not appear likely that the 
ants occur in this State. 


In North Carolina the known infestations are confined to the 
southern and central or east-central sections (fig. 13). None of the 
infested areas is very large, that at Wilmington, with two areas 
totaling 1.23 square miles, being the largest and apparently the 
oldest. The infestation at Goldsboro. covering 0.60 square mile, is 

Fiuuke 13. — Map of North Carolina showing number of localities scouted for Argentine 
ants and the infested localities (dots) in each county. 

next in size, while that at Raleigh is only 0.15 square mile. It is 
surprising that no infestation was found in the small towns near 


The following localities in southern Oklahoma were scouted : An- 
adarko (Caddo County), Durant (Bryan County), Ardmore (Carter 
County), Coalgate (Coal County), Lawton (Comanche County), 
Hugo (Chocktaw County), Clinton (Custer County), Pauls Valley 
and Lindsay (Garvin County), Chickasha (Grady County), Holden- 
ville (Hughes County), Altus (Jackson County), Mangum (Greer 
County), Hobart and Snyder (Kiowa County), Madill (Marshall 
County), Valliant and Idabel (McCurtain County), McAlester 
(Pittsburg County), Ada (Pontotoc County), and Frederick (Till- 
man County). That no ants could be found in southern Oklahoma 
is surprising, since a number of infested localities in northeastern 
Texas are not more than 50 to 75 miles away. A vigilant watch 
should be kept for the appearance of the ants in this section of the 


In South Carolina the ant was found to be fairly well scattered 
over the State, but was most prevalent in the southeastern, east- 
central, and northern sections (fig. 14). Spartanburg County, with 
five infestations, has the largest number ; Greenville County second, 
with four; and Charleston County third, with two. 

The largest infestation in the State, and undoubtedly the oldest, is 
that at Charleston, which was estimated to cover 6.29 square miles. 



The next largest infestation is at Spartanburg, which occupies 2.20 
square miles, and that at Gaffney third, with 1.12 square miles. In 
several towns there is more than 1 infestation, Spartanburg having 
12 and Greenville and Gaffney 3 each. 

The importance of the railroad in distributing the ant is clearly 
shown in Greenville and Spartanburg Counties, where nearly all the 
infestations lie along the main line of the Southern Railroad. Many 
of the towns in the vicinity of Charleston, where the ant has been 
known for at least 19 years, are apparently not infested. 

FlGUEB 14.- 

-Map of South Carolina showing 'number of localities scouted for Argentine 
ants and the infested localities (dots) in each county. 


In the southern part of Tennessee only the larger and more impor- 
tant commercial towns were scouted, and in the western section the 
work was confined to towns along the main line of the Illinois Cen- 
tral Railroad and the Mississippi River. Infestations were found 
in the extreme southern part of the State not far from the Alabama 
line, one at Pulaski in Giles County and another at Fayetteville in 
Lincoln County. It was surprising that no Argentine ants were 
found at Memphis or in any of the larger towns along the railroad - 
or river. 


The Argentine ant was found to be rather generally distributed 
over the eastern half of Texas, or that section lying east of Wichita 
Falls and Austin (fig. 15). This was to be expected in view of its 
nearness to Louisiana, its rather densely populated areas, and the 
more abundant vegetation here than in the western part of the State. 



The largest infestations in the State are as follows : Beaumont, 11.4 
square miles; Fort Worth, 11.37 square miles; Houston, 8.32 square 
miles; Dallas, 5.07 square miles. In spite of the large number of 
localities infested in Texas, the State is much less heavily infested 
than the States east of Louisiana. 

Apparently all but a very few of the infestations were started from 
produce shipped on railroads. The ant has not become established 
in many of the towns along the Gulf of Mexico, either because large 
boats do not come into these towns or because the boats are usually 
not from ports infested with Argentine ants. Frequent direct water 

FiGtniE 15. — Map of Texas showing number of localities scouted for Argentine ants and 
the infested localities (dots) in each county. 

communication between Louisiana and the southeastern section of 
Texas would no doubt hasten the establishment of the ants there. 


The following cities in Virginia were scouted: Lynchburg (Camp- 
bell County), Richmond (Henrico County), Petersburg (Dinwiddie 
County), Suffolk (Nansemond County), Norfolk (Norfolk County), 
Roanoke (Roanoke County), and Lexington (Rockbridge County). 
The occurrence of the Argentine ant at Norfolk had been reported 
in correspondence, but H. T. Vanderf ord was unable to find any ants 
after a 5-day search. The limited scouting failed to reveal any ants 
in the State. 



Less scouting was done in West Virginia than in any other State. 
Outdoor infestations so far north did not seem probable, especially 
since no typical ones had been found north of Ealeigh, K C. In 
passing through this State to Kentucky, the scouts spent 4 days in 
Huntington (Cabell County) and Charleston (Kanawha County), 
with negative results. 


According to Harned (-5) Argentine ants probably made their 
first appearance in Mississippi about the year 1900, and by 1909 
there were approximately a dozen known infestations in the State. 
By 1922 infestations in 40 towns had been recorded, and in 1934, 245 
infestations had been found. Since only about half a dozen new 
infestations had been discovered during the previous year, it was 
presumed that most of the infested areas in Mississippi had been 

As the number and severity of the infestations increased, ana 
property owners found that their own efforts to combat the pest with 
commercial poisons were either futile or at least only temporarily 
beneficial, they appealed to the State officials for help. By this time 
E. B. Barber, an investigator for the Federal Bureau of Entomology, 
had proved that the ants could not only be controlled but prac- 
tically eradicated in small areas. Control campaigns were thereforei 
inaugurated through a cooperative arrangement between the State 
Plant Board and the infested communities,, with the assistance of 
Mr. Barber in directing the work. The first of these campaigns were 
conducted in four towns — Durant, Laurel, Crystal Springs, and 
Woodville — during the fall of 1920, and they were so successful that 
others were undertaken in the following years. Since 1920 the plant 
board has supervised this work, while the municipalities and coun- 
ties have paid the expenses. This work reached its height in 1929, 
when 121 localities were poisoned for the ants, 1,136,028 cups of 
poison being used at a cost of approximately $35,000. Even in the 
depression years there has been a steady demand for control work, 
although the towns and counties have been handicapped by lack of 

The early campaigns were designated control campaigns, and the 
word "eradication" was not used. In 1924, however, an area em- 
bracing a block and a half at Fayette, Miss., was freed of the ants 
after two successive fall campaigns, at an exceedingly low' cost (15). 
So far as is known, this was the first town, not only in Mississippi 
but in the world, from which the ants had been eradicated. Later 
other towns began to eradicate the ants, and by 1928 six — Fayette, 
Shaw, Lyman, Landon, Moss Point, and State College — had been 
freed of them (J,). Since then State authorities have not only 
stressed the fact that it is possible to eradicate Argentine ants by 
timely and thorough campaigns repeated for several consecutive 
years, but they have encouraged the municipalities and counties to 
fight the ante in this manner. 

Since it has been definitely shown that both control and eradica- 
tion of Argentine ants can be achieved by the method used in Missis- 



sippi, it is desirable to describe- in detail how this is accomplished 
and to give some idea as to the cost of the work. 


Usually there is only one general infestation in a town, although 
in some towns as many as 15 to 20 isolated areas have been found. 
The ideal method of locating and delimiting infestations would be 
to scout the entire town, property by property, and block by block, 
but this is both costly and time-consuming. The plant board has 
therefore not attempted to scout entire towns with such thorough- 
ness unless the town authorities have indicated their intention of 


[]RES. |jjt;LARGE PH 




'liii m iiliii H i DDiiJDimi ^ liiilli lli l ili uuln liiii 

M. AND N. R. R. 


Im l H l H 


Res.D □ Ores. \J 

RES. RES. , 

Figure 16. — Map of an area infested with Argentine ants, showing the number of cups 
of poison required for the various blocks. 

continuing the work until eradication is accomplished. It is pos- 
sible, however, to determine the limits of an infested area without 
visiting every property or block. Since the Argentine ant destroys 
practically all the native ants, wherever native ants are found one 
can be reasonably sure that there are no Argentine ants. If the 
scout, therefore, proceeds from a given point of infestation and 
examines trees, fences, the ground, and the bases of pillars of build- 
ings, he can consider that he has reached the limit of the infested 
area when he first encounters native ants instead of Argentine ants. 
From this point he encircles the infested area and as he finds the 
division between the infested and the noninfested territory he in-i 
dicates it on a map (fig. 16). When the circle is completed he is in 
a position to estimate the cost of poisoning that area. 



The cost of a campaign is estimated on the basis of the number of 
cups required. In an ordinary city block this will range from 100 to 
200. One familiar with the work can estimate roughly the number 
needed after walking around the block, but a less experienced person 
should note the places where cups are needed, spacing them 20 to 25 
feet apart, regardless of the character of the block with respect to 
vegetation, buildings, fences, etc. In areas not bounded by streets 
the estimate should be made in such a way as to avoid confusion. 
The number of cups of poison required for each block in the infested 
area is indicated on the map (fig. 16). The total number of cups 
required for all the blocks and other areas is then computed. It is 
advisable to give the totals in even thousands, for manufacturers 
prefer to ship in such lots. 

The cost of supplies and labor for a campaign that will require 
6,000 cups is estimated, on the basis of current prices, to be about 
as follows : 

6,000 paperoid cups, at $9.25 per 1,000 $55.50 

6,000 brackets, at $3.25 per 1,000___, 19. 50 

75 gallons of poison, at 60 cents per gallon 45. 00 

12 pounds of shingle nails, at 5 cents per pound . 60 

Incidentals (four 5-gallon kerosene cans, 1 bucket, 1 funnel, 4 hammers, 

4 grocery baskets) 6. 00 

Express or freight on supplies 2. 00 

Labor, 120 hours at 20 cents per hour 24. 00 

Total 152.60 

Cost per cup .0254 

In Mississippi many campaigns have been put on at an average of 
from 2.3 to 3 cents per cup. A well-conducted campaign should 
seldom cost more than 3 cents per cup. 


During the course of the control and eradication work in Missis- 
sippi three types of poison containers have been used— tin cans, 
aluminum cups, and paperoid cups (fig. 17). 

Tin cans with detachable lids, measuring 2% .by 4 inches with a 
capacity of approximately 10 fluid ounces, were first used. Two in- 
dentations were made at the top with pliers, which when the lids 
were placed on the cans formed holes through which the ants might 
pass. To attach a can to. a tree, post, or building, a six-penny nail 
was driven into the object and the can hung over it through one of 
the holes beneath the lid. Formerly a gill of poison and .a piece of 
sponge about the size of an egg were placed in each can, the sponge 
serving as a footing for the ants as they fed. It was later found that 
the sponge was unnecessary and that 1 to l x / 2 ounces of sirup per can 
was sufficient. 

The tin can was finally discarded because of its high cost. The cost 
of the early campaigns averaged from 6 to 7.5 cents per can, and 
from 2.2 to 2.5 cents of this was for the can. The tin can, however, 
is somewhat more substantial than the other containers and, because 



of its greater depth, when buried in the soil it is not so likely to be 
submerged by successive rains. 

The aluminum cups measure V/g by 2% inches and have a capacity 
of approximately 1 ounce. On the sides of the cups near the top are 
three small holes through which the ants can enter and leave. The 
cups are light and easily transported and can be salvaged and used 
again the following year, but they are not large enough to keep the 
sirup from drying out when exposed to much sunshine, and they 
are easily submerged when buried in the soil. Although they do 
not require brackets, as do paperoid cups, their cost is about one 
and one-half times that of paperoid cups without brackets. 

Fkiuke 17. — Poison containers that have been used in Argentine ant control and eradica- 
tion work in Mississippi: A, Tin can; B, paperoid cup; C, aluminum cup. 

The paperoid cups, which are made of heavy paraffined paper, 
resemble small drinking cups with tops. They measure 1% by 2% 
inches and hold approximately 1% fluid ounces. There are four small 
holes for the ants near the top but beneath the lid. To attach the 
cups to trees, buildings, or other objects, metal brackets are necessary 
(fig. 18) . These brackets are so constructed that the cups are suspended 
from them by fastening the lid over the top, and an emarginated area 
at the bottom holds the cup firm. The brackets are attached to 
objects by a small nail driven through a hole in their center. Brackets 
on buildings and fences will last for several years, but those on young 
trees sometimes sink so deeply into a tree after a year or so that they 
can no longer be used. The cups not only withstand severe storms 
when securely fastened to brackets, but they last much longer than 
they are needed, for it is seldom, if ever, necessary to refill them in 
Mississippi. Occasionally, especially where the cups are exposed 
to sunshine, the sirup may ooze through the paper, but never in 
quantities to be objectionable. For the first campaign a bracket will 
be required for each cup attached to an object. In succeeding cam- 
paigns, however, only about half as many brackets will be required, 
since most of the brackets first put up are still in place. 


CIRCULAR 3 8 7, l r 


The paperoid cups have the advantage of lightness and low cost, 
and when from half to two-thirds full they hold just about the 
quantity of sirup needed. They are not desirable for burying in the 
ground" because of their shallow depth and the fact that water can 
eventually soak through them. Their paper composition also makes 
them more easily destroyed by children. In spite of these objections, 
however, the paperoid cups have proved practical and inexpensive 
and are now being used in greater numbers than any other type of 
container. The first paperoid cups used in Mississippi were white. 

Figure IS. — Paperoid cup attached to bracket uu a tree. Note how the lid of the cup is 
suspended over the top of the bracket, and also how the cup rests in an emargination on 
the lower side of the bracket. 

and their conspicuous appearance attracted the attention of mis- 
chievous children, and so cups of a chocolate or deep brown color 
are now used. 


The poison used in the cups is a sirup containing sodium arsenite, 
made according to a formula devised by Barber (£) and now known 
as the ''standard Government-formula Argentine ant poison." There 
are several firms in Mississippi that manufacture this poison. Ac- 
cording to the State chemist of Mississippi, it should meet the follow- 
ing specifications : 

Sodium arsenite - 0. 16 percent. 

Acidity (as tartaric acid) 0. CM. i*>rceiit. 

Solids! 50.0° Brix. 

Since it is difficult always to obtain samples that meet these figures 
exactly, the manufacturers have been allowed the following range : 

Sodium arsenite 0. 12-0. 20 percent 

Acidity (as tartaric acid) 0. 00-0. 10 percent. 

Solids! 50- o - 50 - °° Bl ix 



Before being used, all poisons are analyzed from uniform samples 
taken from the containers by the plant board inspectors after they 
reach the, town or county where the campaign is to be conducted. If 
the poisons do not fall within the above ranges, they are rejected at 
the manufacturer's loss. Lyle, 5 who carried on a limited series of 
poison IrMs against Argentine ants both in artificial cages and under 
natural conditions, found that the ants will tolerate an even wider 
range in arsenic content as well as in acidity of the sirup. Until 
more work is done on this- subject, though, the above ranges should 
lie strictly adhered to. 

If the poisoned si rap is to be stored for several months, it is placed 
where it will not be exposed to exceedingly low or high temperatures, 
and it is well covered to prevent contamination by dust, water, etc. 
The poison should never be placed in any receptacle that has contained 
oil, kerosene, or grease, for these substances are very repellent to ants. 


Incidental supplies that are required include 5-gallon cans, funnels 
with strainers, hammers, and substantia] grocery baskets, as well as 
tanks for distributing the poison in the held (tig. 1!)). The plant 

Figube 111. — Equipment hsimI by :i Held 
temporary storage of poison, carpenter's 

tops, and hammer. 

ins crew — distributing tank, 
basket for unfilled paperoid c 

hoard men often bring these supplies with them. Five-gallon cans 
of the type usually sold for holding kerosene are, excellent for the 
temporary storage of the poison in the held. A can of this size will 
hold enough sirup to iill from 375 to 425 cups, which is about the 
number that a field crew can put out in half a day. 

For distributing the poison a rectangular galvanized-iron tank 
measuring 4 by 12 by 12 inches and having a capacity of 2i/o gallons 
is used. At the lower end of the tank is attached a half-inch hose 
with a faucet at the end. The tank is strapped to the back of the 
Held assistant. The cost of such a tank, including accessories, is 
about $5. It can be made by any competent tinner. 

! Lvi.k, C. See footnote .'!. 



After a locality has been surveyed, the plant board scuds an 
itemized estimate of the cost of a campaign to the local officials. If 
they decide to put on a campaign, they arc advised concerning the 

If «BH|i m 

^ ■ JH ■ • 1 

Hr^^ J^i fti 


Hfli '■■ 

-ia^^^l *■■■ f-T^Si^* flu B^^'i^^*" "^ 

_ _M.. ..<#,, 

.*r^rf-& ^ -^tf f-4frglTl t WTMfr 

■•'" ^ 

jliW" 11 ™ 1 ^,, 

. , > ^, j- - f > - 


~* ' '" :: • 

Fiquke 21). — Field crew distributing poison. Leader is nailing a bracket to a tree fur the 
reception of Hie cop of poison which his assistant is tilling. 

Figi-ke 21.— Placing cups of poison on pillar of house that is not enclosed at the base. 




supplies needed and where to purchase them, and are requested to 
notify the plant board when these materials have arrived. One or 
more plant board men are then sent to direct the campaign. These 
men purchase incidental materials and also help select the labor if it 
is not already provided. Supplies are charged to and the labor is 
paid for out of local funds. 

The most intelligent and industrious labor available is obtained. If 
the men have not had previous experience in Argentine ant control, 
they are thoroughly instructed before they begin work. 

The poison is dis- 
tributed by crews of 
t w o men each, a 
leader, who is respon- 
sible for the place- 
ment of cups, and his 
assistant, who carries 
the supplies (fig. 20). 

To oversee the work 
there are held super- 
visors who are respon- 
sible for from two 
to four crews each. 
These supervisors 
should have had pre- 
vious experience and 
also know how to 
handle labor. One 
man with a small 
truck is responsible 
for the transportation 
of the crews and their 
supplies in the field. 
In all but the very 
large campaigns this 
man can very well 
direct the entire work 
through the field su- 
pervisors. He sees 
that the poison and 
other supplies are kept 
handy for the field 
men at all times and placed" 

also that the crews 

are so placed that he can keep in contact with them with as little 
effort as possible. This means that only one crew works a block and 
this crew works the entire tier of blocks and in the same direction as 
the other crews. All crews are thus working in tiers of blocks not 
only adjacent to but parallel with each other. 

In working a block a crew covers it property by property, being 
sure that no part of it is overlooked. Cups of poison are distributed 
at intervals of about 20 to 25 feet. Under houses that are not en- 
closed (fig. 21). a cup is placed on every third or fourth pillar. If 
the base of the house is enclosed, cups are fastened to the sides of the 
house about 5 or 6 feet from the ground. On fences enclosing live- 

%*£•?$!;?&■ .$*• <'«£.• 

' ■■• ' '■ ' "•■ 


I'll i I RE 22. 

it poisoa ran often bo advantage 
clumps of weeds, bushes, etc. 



stock the cups are placed outside the fence, about 6 inches from the 
ground, where they will attract the least attention. Cups should be 
concealed as much as possible in order not to attract the notice of 
children. Where there are no trees, buildings, posts, or fences, the 
cups can be set in the crowns of dense plants (fig. 22) or buried in 
well-drained soil (fig. 23) and covered with clods of dirt or pieces 
of debris. It is always desirable to poison slightly beyond the limit 
of infestation in order to check any ants that might advance that far. 
Ordinarily from 50 to 75 feet beyond is sufficient. A crew can usually 
put out on an average 800 cups of poison a day ; an extra good crew, 
with the right type of territory in which to work, may average as 
many as 1.000 cups or even more. A map of the area should be kept 


Burying cups of poison in the ground. 

by the supervisor at all times, as this is the most satisfactory wa\ 
to record the work and the area worked by the various crews. 


The best time to put on a campaign is in the fall or winter. At 
these seasons little honeydew is being produced by scale insects, 
mealybugs, and aphids and there are few organisms for the ants to 
feed on ; consequently the ants eat the poison sirup greedily. Cool 
weather, with temperatures ranging from 50° to 60° F., usually 
stimulates feeding. Furthermore, there is more time available at 
this season of the vear. Campaigns have been put on in Mississippi 
when it was so cold that the men had to wear overcoats and then- 
hands became numb, with results just as satisfactory as those from 
campaigns put on during warmer weather, but the discomfort of the 
men slows up the work and there is danger of carelessness under 


such conditions. The cost of control work prevents most localities 
from attempting more than one campaign a year. Moreover, such 
splendid results have been obtained in Mississippi from fall cam- 
paigns that there has been no incentive for putting on spring 


All the cups of poison are not visited at once by the ants. When 
feeding does take place, the ants visit the poison rather regularly and 
in large numbers for from 6 weeks to 2 months, after which time 
they attend it at infrequent intervals and then only in small num- 
bers. Cups placed on trees, fences, and houses are visited most fre- 
quently. It has been noticed that when the sirup begins to thicken 
its palatability for the ants greatly decreases, and within a year 
it ceases to attract them, although many of the cups are not yet 
empty. It is generally believed that worker ants and their brood 
die within a few days after having fed on the poison. The greatest 
reduction in the number of ants evidently takes place during the 
first 2 months after the poison has been distributed. 


In the summer following a campaign a survey is made by plant 
board inspectors to ascertain its effectiveness. Inquiries are made 
of housekeepers, storekeepers, cafe owners, and hotel managers as 
to the situation on their property. This is rated as "good" if no 
ants have been seen in their buildings since the poisoning, "fair" 
if few ants have been in evidence, and "poor" if the ants have been 
as abundant as before the poisoning. The "good" and "fair" reports 
are considered, as satisfactory, the "poor" as unsatisfactory. Fol- 
lowing the first campaign in a locality, from 95 to 100 percent of the 
reports are usually satisfactory, and after successive campaigns 
have been put on 100 percent of the reports are often satisfactory. 
In obtaining these reports it is necessary to be cautious, especially 
near the edge of the infested area, lest the property owner mistake 
native ants for Argentine ants. Some of the native ants cannot 
be controlled by this poison. 

When the first campaign in a locality has been put on in a 
thorough manner, the abundance of the ants is usually so greatly 
reduced that few homes are troubled with them any more and the 
residents often conclude that they have been eradicated. This is 
rather unfortunate, for it is difficult for the plant board to induce 
that locality to put on a follow-up campaign. It is customary for 
the plant board to publish in the local paper a short account of the 
investigation, informing the public that, although the ants have 
been greatly reduced in numbers, they have not been eradicated. 

In putting on the second fall campaign, the same quantity of 
supplies as in the first campaign should be used. The summer after 
the second campaign the infested area can be scouted to determine 
what sections have been freed of the ants and to what extent the 
quantity of supplies can be reduced for the next campaign. 

Many towns will poison the ants one fall and then wait until 
they become annoying again before putting on another campaign. 


Such a course is always discouraged by the plant board, as the 
ants can reach their former numbers and cause serious annoyance 
within 2 or 3 years. 


The Argentine ant can be controlled by one thorough campaign, 
but these campaigns must be repeated if eradication is desired. 
Occasionally two campaigns are sufficient, but usually from three 
to five, and occasionally more, are necessary. To be successful, 
however, especial attention must be paid to the timeliness and thor- 
oughness of the work and to the protection of the cups. If eradica- 
tion has not been achieved after the second campaign, further treat- 
ment depends on the abundance of the ants. If they are still widely 
distributed and fairly abundant, a third general campaign with 
the same quantities of supplies as before is advisable. If, however, 
only a few colonies are found in a block, it is sufficient to repoison 
the infested spots, including an area within a radius of about 50 
feet. Fresh poison should be applied immediately and again in the 
fall. The intensive follow-up scouting should be continued for two 
summers after the area appears to be free from the ants. 


Although it has been shown that the ants can be eradicated from 
given areas by poisoning alone, it is frequently desirable to hasten 
eradication by supplementary measures such as burning and oiling. 
In woodland areas where the ants are colonized in stumps, logs, 
and rotton" branches of trees, many strong colonies can be destroyed 
by making fires in the woods and burning all the movable wood 
containing colonies. Stumps and immovable logs (fig. 24) can be 
burned by pouring on kerosene and setting them afire: 

The burning should be done when the wood is dry and there is 
no wind. The best time for this is probably late in the fall or in 
winter, when the temperature is 50° F. or lower and the ants in their 
nests are inactive. The fires should be carefully guarded to prevent 
them from getting out of control and starting woodland fires. With 
labor in the South ordinarily not costing more than 12.5 cents an 
hour, and kerosene selling at. 14 cents a gallon, the cost of treating 
is not more than $2 to $3 per acre. 

Digging colonies out of the soil and burning them with a blow- 
torch is such a laborious task that it is prohibitive except where the 
ante have been almost completely exterminated. It is doubtful if 
it would be practical where there are more than one or two colonies 
per block. This work requires a higher type of labor than some of 
the other control work. It sometimes takes half a day or longer to 
burn out a large colony, which has many ramifying galleries in the 
soil. As a worker can only burn, from one to three colonies a day 
by this method, the eradication of each colony would cost from 
$1.25 to $4. Waste oil, such as can be obtained from filling stations, 
is used to saturate the ground where the colony is dug up. From 3 
to 4 gallons will saturate the ordinary nest. 

The towns that have eradicated Argentine ants in Mississippi have 
used diverse methods in accomplishing this. In the 21 areas investi- 
gated by the Bureau of Entomolgy, the total cost of eradication per 
block has varied considerably. Since the adoption of the paper cups, 



however, and cheaper methods of combating the ants, the average 
yearly cost for eradication work has not been more than $3 to $C per 


In the summer of 1933 the writer and L. C. Murphree undertook 
an appraisal of the work done in Mississippi to control or eradicate 
the Argentine ant. For this study the State Plant Board kindly 
made available its records and offered both suggestions and advice. 
Seventeen towns, having a total of 21 infested areas, were selected 
in different sections of the State — i towns in the southern third, 11 
in the middle, and 2 in the northern third — where the infestations bad 

Fiodre ^4. — Preparing to set tire to a stuniy heavily infested with Argentine ants. 

been established for various periods of years, and the campaigns 
had differed both in measures and in the time required to destroy the 
ants. The infested areas ranged in size from 1 to 35 blocks. They 
also varied greatly in topography, quantity of vegetation, soil drain- 
age, and density of human population. It seemed that, if the ants 
could be eradicated from such diverse habitats and under such adverse 
conditions, the methods employed in Mississippi might be recom- 
mended for general use elsewhere. In each area search for Argentine 
ants was made by foot-by-foot and block-by-block scouting. This 
work required from one to two man-days per block, and the entire 
task accupied from May 15 to September 17. As a result, 13 of the 
21 areas, or 67.9 percent, were found to be free from the ants. 

The plant board believes that at least 39 infestations out of the 
245 known to have occurred in the State have been eradicated, and 
that this number would have been larger if more funds had been 
available for campaigns. 



The first step in the campaign is to delimit the infested area. 
Scouting for this purpose should be done late in the summer or early 
in the fall, and never when the temperature is below 65° to 70° F. 
In connection with the scouting a map of the infested area should 
be prepared which will not only be accurate, but clear to anyone using 
it. It should be drawn on such a scale that it can conveniently be 
used in the field. In addition to indicating the blocks and streets, 
it will prove helpful to note important landmarks, especially in 
those sections not represented by streets. The number of. cups of 
poison required for each block can be indicated on the map. 

The campaign should be put on in the fall or winter, when there 
is so little natural food available that the poison will prove attractive 
to the ants. For the comfort of the men it is best to do this in mild 

The poisoned bait recommended is the "standard Government- 
formula Argentine ant poison", which is a sirup containing sodium 
arsenite. Before being distributed it should be analyzed by a chem- 
ist, and if it does not meet the specifications it should be discarded. 

The most satisfactory type of container for the poison is the paper- 
oid cup. The most practical and economical method of filling the 
cups is in the field from tanks carried on the backs of the field crew. 
The cups should be placed from 20 to 25 feet apart, and they should 
be concealed as much as possible. On trees and posts they should be 
tacked as high as one can reach. On fences they are preferably 
placed low and, if the fence encloses domestic animals, outside the 
enclosure. Under houses that are open underneath, the cups can be 
put on the top of every third or fourth pillar, and under houses that 
are enclosed the cups can be fastened on the walls about 20 feet apart 
and as high as one can reach. Where there are no objects on which 
to tack the cups, they should bq buried in the ground or placed in 
clumps of grass or bushes. 

It is important to select the best type of labor available, as the 
success of the campaign is largely dependent on the thoroughness of 
the work. The men should be instructed before the work is begun 
and carefully supervised throughout the campaign. 

On the completion of the campaign, an article should be prepared 
for a local newspaper explaining the nature of the poison, how it 
affects the ants, and the results that may be expected. Protection 
of thei cups should be especially urged. An appeal to the children 
not to touch the cups can be directed through the superintendent of 
public schools. The aid of the Boy Scouts can also be enlisted. 

The following summer the effectiveness of the campaign should be 
determined by questioning property owners, and the results of this 
survey published at once in the local paper. Later the margins of 
the entire infested area should be scouted to ascertain if the ants 
have spread since the previous fall or if any infested areas have been, 
overlooked. The results of this scouting should be indicated on the 
same map that was used in the preliminary scouting. If the ants 
have spread, additional supplies should be included in the estimates 
for the second fall campaign; if not, the same quantity of supplies 


should be ordered. The second campaign should be handled in the 
same manner as the first. 

If eradication is to be attempted, the summer following the second 
fall campaign a thorough scouting of the entire infested area should 
be made to determine the exact status of the ants. If the ants are 
rather generally distributed over the entire area and there are more 
than one or two colonies per block, the infested area should be repoi- 
soned as in the previous fall. If, however, the ants have been nearly 
eradicated, it is only necessary to repoison within a radius of 50 feet 
from each colony located. This should be done immediately and 
again in the fall. 

The following summer, if possible, the entire infested area should 
be thoroughly scouted. If time or funds do not permit, then those 
areas where colonies were previously found should be carefully ex- 
amined. Often the ants have apparently been eradicated by the third 
summer, but in such a case the area should not be released as free until 
two additional annual scoutings have failed to disclose any ants. 

The use of supplementary measures is often practicable when the 
abundance of ants has been reduced to a colony or less per block. 
Such colonies can be dug out and the ants burned with a torch, 
after which the ground should be well saturated with waste oil. 
Even at the beginning of the work, in wooded areas it«is well to 
burn out as many colonies as possible by setting fire to logs and 
stumps and by gathering up all pieces of wood containing the ants 
and throwing them into a general fire. Every precaution should be 
taken to guard against forest fires. 


The Argentine ant, Iridomyrmex humdlis (Mayr), is a pest of no 
little importance throughout the Southern States and in California. 
It is especially troublesome as a house-infesting insect, being present 
almost continuously throughout the year. Out of doors the ants feed 
on the honeydew produced by scale insects, mealybugs, and aphids, 
and do indirect damage to vegetation by fostering these insects. The 
ants also steal seeds from seed beds, kill young poultry and birds in 
their nests, destroy colonies of honeybees, and feed on the sap or fruit 
juices from certain trees and plants, particularly citrus. 

The wide and rapid spread of this ant since its introduction some- 
time previous to 1891 has been accomplished chiefly by artificial 
means, the railroads being the principal means of dispersion. Nat- 
ural spread is due mainly to crawling, which varies from a few feet 
to several hundred feet a year, but heavy rains, floods, etc., also play 
an important role. Flight is a negligible factor in the spread of this 
• ant. 

Argentine ants have been found in all types of soil and at eleva- 
tions ranging from approximately sea level to nearly 4,000 feet. 
They seem to be affected by strong winds, and by moisture conditions 
in the soil. Although tropical insects, they can withstand more cold 
than most of our native ants. 

It is estimated that the Argentine ant occurs over an area of at least 
4,000 square miles in the United States. California and Louisiana are 
the most heavily infested States, and infestations also occur in Mis- 


sissippi, Alabama, Georgia, Texas, South Carolina, Arkansas, North 
Carolina, Florida, Maryland, Tennessee, Arizona, Missouri, and 
Illinois (States listed in order of decreasing size of infested area). 
The size of the infestations varies from a single colony to areas con- 
taining several hundred square miles. Although as yet the ants are 
sporadically distributed over the States mentioned, except in Mis- 
souri, Maryland, and Illinois, where the infestations are of the indoor 
type, there is nothing to prevent them from occupying the entire area 
within their present boundaries and also from spreading to additional 

Efforts to control and eradicate the Argentine ant in Mississippi by 
poisoning campaigns have resulted in freeing 39 out of 245 infested 
localities, and reducing the infestations in nearly all the others. The 
method used in Mississippi is therefore described in some detail in this 
circular. The ants can be controlled by one thorough campaign, and 
by repeating the campaign each fall it is possible to eradicate them 
in from 2 to 5 years. The method consists, in brief, in making careful 
surveys of infested areas and then placing cups of sirup containing 
sodium arsenite at proper intervals throughout these areas. Where 
eradication is attempted, supplementary measures, such as burning 
and oiling colonies, expedite the work, although these should not be 
resorted to until the numbers have been greatly reduced. The cost of 
eradicating ants should not be more than 3 cents per cup of poison, 
or $3 to $6 per block. 


(1) Barber, E. R. 


states. V. S. Dept. Agr. Bull. 377, 23 pp., illus. 



Farmers' Bull. 1101, 11 pp., illus. 

(3) Foster, E. 


ormans. Jour. Econ. Ent. 1 : 289-293. 

(4) Harked, It. W. 

1928. argentine ants eradicated at six PLACES. Miss. State Plant 
Bd. Quart. 8 (3) : 10-11. 

(5) and Smith, M. R. 


Ent. 15 : 261-264. 

(6) Herbert, F. B. 

1932. EFFECT of cold storage temperatures on the argentine ant. 
Jour. Econ. Ent. 25 : 832-833. 

(7) Hertzes, L. 


external conditions. Ann. Ent. Soc. Amer. 23: 597-600. 

(8) Horton, J. R. 


Bull. 647, 74 pp., illus. 



Farmers' Bull. 928, 20 pp., illus. 

(10) Newell, W. 


iridomyrmex humilis mayr. Jour. Econ. Ent. 1 : 21-34. 

(11) and Barber, T. C. 

1913. the argentine ant. TJ. S. Dept. Agr., Bur. Ent. Bull. 122, 98 pp., 


(12) Nickels, L. J. 


Ent. 4: [353]-358. 

(13) Ryan, H. J. 


Ent. 21: 682-690. 

(14) Smith, M. R. 


Jour. Econ. Ent. 12: 465. 



Mississippi. (Sci. Note) Jour. Econ. Ent. 17: 603-604. 



Ent. 20: 646-647. 

(17) Wogltjm, It. S., and Borden, A. D. 


. D. S. Dept. Agr. Bull. 965, 43 pp., illus. 

(18) and Neuls, J. D. 


Agr. Farmers' Bull. 862, 16 pp., illus. 


For sale by the Superintendent of Documents, Washington, D. C. ... - Price 5 cents