UNIVERSITY OF CALIFORNIA PUBLICATIONS COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA THE CONTROL OF THE ARGENTINE ANT By C. W. WOODWORTH BULLETIN No. 207 Berkeley, Cal., October, 1910. BERKELEY THE UNIVERSITY PRESS 1910 Benjamin Ide Wheeler, Ph.D., LL.D., President of the University. EXPERIMENT STATION STAFF. E. J. WiCKSON, M.A., Director and Horticulturist. E. W. Hilgard, Ph.D., LL.D., Chemist (Emeritus). W. A. Setchell, Ph.D., Botanist. Elwood Mead, M.S., C.E., Irrigation Engineer. (Absent on leave.) Lerov Anderson, Ph.D., Dairy Industry and Superintendent University Farm Schools. M. E. Jaffa, M.S., Nutrition Expert, in charge of the Poultry Station. R. H. Loughridge, Ph.D., Soil Chemist and Physicist (Emeritus). C. W. Woodworth, M.S., Entomologist. G. W. Shaw, M.A., Ph.D., Experimental Agronomist and Agricultural Technologist, in charge of Cereal Stations. Ralph E. Smith, B.S., Plant Pathologist and Superintendent of Southern California Patho- logical Laboratory and Experiment Station. F. T. Bioletti, B.S., Viticulturist. E. W. Major, B.Agr., Animal Industry, Farm Manager, University Farm, Davis. George E. Colby, M.S., Chemist (Fruits, Waters and Insecticides), in charge of Chem. Lab. W. T. Clarke, B.S., Assistant Horticulturist and Superintendent of University Extension in Agriculture. H. M. Hall, Ph.D., Assistant Botanist. H. J. Quayle, A.B., Assistant Entomologist, Plant Disease Laboratory, Whittier. John S. Burd, B.S., Chemist, in charge of Fertilizer Control. C. M. Haring, D.V.M., Assistant Veterinarian and Bacteriologist. W. B. Herms, M.A., Assistant Entomologist. H. A. Hopper, M.S. A., Dairy Industry, University Farm, Davis. E. B. Babcock, B.S., Assistant Agricultural Education. W. T. Horne, B.S., Assistant Plant Pathologist. J. H. Norton, M.S., Assistant Chemist, in charge of Citrus Experiment Station, Riverside. J. E. Coit, Ph.D., Assistant Pomologist, Plant Disease Laboratory, Whittier. C. B. Lipman, Ph.D., Assistant in Soil Bacteriology. R. E. Mansell, Assistant in Horticulture, in charge of Central Station grounds. J. C. Bridwell, B.S., Assistant Entomologist. J. I. Thompson, B.S., Assistant in Animal Industry, Davis. E. H. Hagemann, Assistant in Dairying, Davis. R. M. Roberts, B.S.A., Field Assistant in Viticulture, University Farm, Davis. Roscoe Farrar, B.S., Assistant in Soils and Farm Crops, University Farm, Davis. B. S. Brown, B.S.A., Assistant in Horticulture, University Farm, Davis. Howard Phillips, B.S., Assistant in Animal Industry, University Farm, Davis. L. M. Davis, B.S., Assistant in Dairy Husbandry, University Farm, Davis. F. L. Yeaw, B.S., Assistant Plant Pathologist, Vacaville. F. D. Hawk, B.S.A., Assistant in Animal Industry. T. F. Hunt, B.S., Assistant Horticulturist. C. O. Smith, M.S., Assistant Plant Pathologist, Plant Disease Laboratory. Whittier. E. H. Smith, M.S., Assistant Plant Pathologist. C. H. McCharles, B.S., Assistant in Agricultural Chemical Laboratory. A. J. Gaumnitz, M.S., Assistant in Cereal Investigations, University Farm, Davis. S. S. Rogers, Assistant Plant Pathologist, Plant Disease Laboratory, Whittier. P. L. McCreary, B.S., Laboratory Assistant in Fertilizer Control. F. E. Johnson, B.S., Assistant in Soil Laboratory. M. E. Stover, B.S., Assistant in Agricultural Chemical Laboratory. Charles Fuchs, Curator Entomological Museum. P. L. Hibbard, B.S., Assistant Fertilizer Control Laboratory. \. I). [NGHAM, Assistant in Sylviculture, Santa Monica. B. A. Madson, B.S.A., Assistant in Cereal Laboratory. L. Bonnet, Assistant in Viticulture. Walter E. Packard, Field Assistant Imperial Valley Investigation, El Centre E. E. Thomas, B.S., Assistant Chemist, Plant Disease Laboratory, Whittier. Mrs. I). L. BUNNELL, Clerk to the Director. W. H. VOLCK, Field Assistant in Entomology, Watsonville. E. L. MORBIS, B.S., Field Assistant in PJntomology, San Jose. .1. S. Hr.NTKK, Field Assistant in Entomology, San Mateo. John T. BEAB8S, Foreman Kearney Park Station, Fresno. J. C. ROPES, Patron University Forestry Station, Chico. B. 0. Mi 1. 1. hi;, Foreman University Forestry Station, Chico. THE CONTROL OF THE ARGENTINE ANT By C. W. WOODWORTH. Two years ago we reported the presence of this insect in California in Circular No. 38 of this station. The exact localities then known to be infested were East Oakland, Alameda, San Francisco, San Jose, Cupertino, a point near Campbell, Los Angeles, a point near Azusa, and Upland. At the present time, through the discovery of new locali- ties and the natural spread of the insect, more than twice as much territory is known to be involved, amounting to about 1,000 acres in Southern California and in the neighborhood of 4,000 acres in the central portion of the State. THE PRESENT DISTRIBUTION. The East Oakland area includes also nearly all of Fruitvale and occupies all the territory between the bay and the hills as far west as Lake Merritt, and there are small outlying colonies scattered through Melrose on the one hand and the Piedmont section on the other, with one small colony in the center of Berkeley. Alameda is nearly all occupied and there are small, detached colonies on Bay Farm Island. The San Francisco colonies are all small and scattered, the largest being at the Presidio entrance. Small colonies are also found at Byron Hot Springs and at Stockton. The largest colony in the State is in San Jose and College Park, extending continuously in one direction for three and one-half miles, with many detached colonies within a radius of seven miles, including colonies at Santa Clara and those previously reported at Cupertino and near Campbell. Recently discovered colonies in and about Los Angeles places that district well toward the front as to the amount of infested territory, and five small colonies in the neighborhood of Riverside complete the enumeration of the points of distribution at present known and mapped. During the short time elapsing between the first discovery of the insect in the State and the publishing of the circular just alluded to, so many localities were found infested that we feared that it had 54 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. Fig. 1. — The egg of the Argentine ant. Very much enlarged. Fig. 2. — A very young larva of the Argentine ant. Very much enlarged. Fig. 3. — A full grown larva of the Argentine ant. Very much enlarged. Fig. 4. — The pupa of a queen of the common small black ant (Tapi- noma). Very much enlarged. Bulletin 207. THE CONTROL OF THE ARGENTINE ANT. 55 already become a generally distributed insect, but we are glad to be able to report that all recent additions to the list of known localities are of small extent and have increased the area of known infested territory by a very small fraction and with the natural increase have scarcely doubled the area. It is altogether likely that still other localities will come to light, but we do not anticipate that any con- siderable additions will be made. At the time of the publication of the circular the exact extent of none of the infested tracts had been accurately determined. Since then we have mapped all of the known areas and will present them when publishing a detailed account of our investigations. Fig. 5. — Head of a queen Argentine ant. The small eyes at the top of the head distinguish the queen from the worker. The arrangement of the teeth on the jaw distinguishes the Argentine ant from all other species. An interesting development of the situation has been the discovery of the insect in South Africa, where it occurs in several towns corresponding somewhat with its distribution in California. The previous record, besides the native countries Brazil and the Argentine Republic, were the Madera Islands, with an area of over 500 square miles completely infested as early as 1898, and the neighborhood of New Orleans. In the latter locality they were first noted in 1891, and in less than twenty years spread until, according to Professor Newell 's account, more than five thousand square miles were occupied, or over five hundred times as much territory as is now occupied in California. This estimate is probably too large since some of the territory included was not actually occupied by the invading ant, but only had scattered colonies of this species. Estimated on the same basis, the San Jose colonies could be given at from fifty to a hundred square miles instead of between three and four actually covered, but in a few years more they will undoubtedly occupy the entire district. 56 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. /#L i '^trfr \ \ Fig. 6. — The queen Argentine ant. The small figure below gives the natural size. This is fully twice as long as the worker. Bulletin 207. THE CONTROL OF THE ARGENTINE ANT. 57 NATURAL DISTRIBUTION SLOW. A reassuring fact made evident by these studies is that the insect spreads very slowly when not transported by human agencies. There is probably no group of insects less capable of rapid migration than ants as a class. There are a few species, including the Argentine ant, that have nesting habits rendering them liable to be transported through the channels of trade, but aside from this no ant is much of a migrant. Apparently the Argentine ant is unusually slow to occupy new territory. We have carefully watched the spread of this ant for two years and the average extension of the colony has not exceeded an eighth of a mile a year; the spread depends primarily upon the ant being carried by man. The first writers upon this ant have Fig. 7. — Side view of body of queen Argentine ant. The middle region of the body shows the stubs to which the wings were originally attached. expressed the opinion that the carrying of the workers into a new territory would be sufficient to establish the insect there. It is true that workers live a long while if in numbers. It is likely, therefore, that crackers, sugar, fruit, fresh or dried, or any other attractive substance will be a ready means of carrying the ants into a new district and that they may live in this new locality for considerable time. We have no evidence whatever that such a transportation will be sufficient to establish the insect in its new home. The establishment of a colony requires the presence of an insect capable of laying eggs, that is, a queen ant. To be sure there seems to be some evidence that the workers of some species of ants may sometimes produce eggs that develop into males, but this must be very rare and if common would be of no value in establishing a colony which is not complete till it has a fertile queen. We have kept colonies in the laboratory at East Oakland for months under all kinds of conditions without, in any 58 UNIVERSITY OF CALIFORNIA — EXPERIMENT STATION. case, obtaining any evidence of egg laying except by queens. We should judge that a queenless colony might live for months under favorable conditions, but ultimately its annihilation is a certainty. Could a colony be established by workers alone, nothing could have prevented the whole State becoming infested by this ant long ago. The moving of the queen is the essential condition without which the spread of the insect is impossible. METHODS OF TBANSPOKTATlON. The people of the State living beyond a few hundred feet from an infested territory are in no immediate danger of invasion of the ant except when at least one queen with attendant workers are transported by human agency. It will be well, therefore, to indicate the ways this may be brought about. There seems to be good evidence that in Fig. 8. — Face of the queen of a field ant (Formica). Notice the position of the large eyes. These insects do not follow trails. several cases the insects were brought in manure. It is a very common practice to haul manure to lawns, gardens, and orchards, and it is difficult to see how a load could be taken from an infested pile without transporting many queens. It would seem to be a good policy to prevent by stringent quarantine the moving of such material from infested to non-infested territory. A most careful inspection should also be made of nursery stock from infested nurseries, potted plants of freshly dug stock from infested land, either in the field or from the healing in ground, are possible means of transference. Even when there is no earth left on the roots the queen ant may have found temporary shelter in the packing material. Vegetables are another possible means of transportation. Ants are often particularly abundant in potato fields, and queens could very easily find shelter Bulletin 20' THE CONTROL OF THE ARGENTINE ANT. 59 in bags and boxes of potatoes. Household goods, food stuffs, or almost any commodity could easily afford temporary shelter for a nest driven out of the ground by rain or irrigation, if not indeed affording a 3 # 4 Fig 9. — Worker of the smaller honey ant (Prenolepis) . Notice the shape of the abdomen. permanent resting place. The point of these remarks is that we now know enough of the habits of the Argentine ant to make evident that it is the wisest course to at once undertake rigid quarantine measures 60 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. to prevent undue spread of the pest. This is advisable whether eradication is considered possible or not. We should not allow ignorance or carelessness to spread the pest over the State any faster than it will go if left to itself. AEGENTINE ANTS DO NOT FLY. The establishment of new colonies, in the case of many species of ants, is preceded by the well-known flight of the young queens, each going off by herself and rearing a new family. Supposing that the same rule held in the case of the present species and that the extension 2S Fig. 10. — Head of the smaller honey ant. of the infested territory was brought about in this manner, we at once endeavored to follow very carefully the development of the young queens in the spring and to watch their flight so as to have a good idea of the range of spread by this means. To our surprise, however, while winged individuals were produced in great quantities, both in our laboratory and in the field, there was apparently no flight whatever. Once or twice the males came out of the nests in moderate numbers and fluttered their wings, but did not rise in the air, and the females remained within the nest until they had lost their wings and became egg producing. Possibly under different climatic con- ditions from that of East Oakland a flight might occur, but nothing of the kind was observed in San Jose by Mr. Morris, or as far as I am aware in any region by anyone. These observations make it very probable that the danger of spreading by the flight of the queen ants is practically nothing. This probably accounts for the very slow rate of spread of this species and is an extremely hopeful feature in case eradication work is attempted. Bulletin 207. THE CONTROL OF THE ARGENTINE ANT. 61 CONDITION OF MIGKATION. Another very hopeful and significant point in case eradication is attempted is the fact that the insect is not liable to occupy new territory except under the press of hunger. Under the promptings of hunger, however, it is very ready to move and the presence of a new source of supply may result in the shifting of the domicile within half an hour or so of the discovery of the food. On the other hand, when food is sufficient for the ants of the colony and no special attraction exists a road may furnish sufficient barrier to prevent the spread of the Fig. 11. — Face of Tapinoma sessile. This is the' ant most commonly confused with the Argentine species. Note particularly the arrangement of the teeth on the jaw. ant indefinitely. There are many cases where the ant has not crossed the road during the two years it has been under observation. It is very probable, therefore, that a single city block, surrounded by roads, may be treated as an independent colony as long as the ants are not more abundant in the adjacent blocks than the food available w r ill accommodate, that is, if poisons are used as described below in all the adjacent blocks to maintain the numbers sufficiently small, no fear need be entertained of a reinfestation. EEADICATION. This is a term that many entomologists believe should cease to be used because they consider it something that is impossible, and many have already expressed their conviction that this ant is here to stay no matter what may be done toward its destruction. The term has been so often used in this State for the partial destruction of an insect pest that perhaps many do not appreciate that it has never been accomplished anywhere in the world after an insect has secured the 62 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. foothold secured by this ant, and cannot be accomplished without more thorough work than has yet been given in this State or elsewhere. It is the easiest way, perhaps, to say that eradication is impossible and not try, or to reduce the numbers and call it eradication, but such a very small part of the State is now involved and the insect is capable of producing so much damage and annoyance it would be too bad not to make at least a serious effort toward eradication. Fig. 12. — Worker of Tapinoma sessile. Notice the top of the abdomen over hanging the ' ' hump. ' ; In some ways this insect is a particularly good one with which to work. It is so slow to spread, most of the individuals are incapable of reproduction. The queens live in definite families and are thus not liable to escape detection, but on the other hand present some problems of particular difficulty and require a most careful study of their life and habits to prevent one making fatal mistakes. The methods of eradication will not be essentially different from those available for control work, with the exception that some may be entirely proper for eradication but too expensive to be available for control. Thus defoliation of the kind used two years ago in Marys- ville for the white fly is justifiable only in an attempt at eradication, and such treatment every few years for control is out of the question. On the other hand, spraying or fumigation is available for control work, but hopeless for eradication for that same insect. In work directed against the Argentine ant we may find the same to hold or we may find that the most available control measure is at the same time the most promising treatment with eradication in view. It is BULLETIN 207. THE CONTROL OF THE ARGENTINE ANT. 63 probable that efficient work toward eradication will depend much more upon the organization and administration of the work rather than 'upon the means employed. No matter how effective a treatment may be, the result will be negative unless the campaign is so organized that the last individual queen insect is destroyed. The efficiency of the provisions for ensuring the death of this last individual is the index of the success of the effort. This fact cannot be too strongly emphasized. No matter how effective a treatment may be or how thoroughly it may be applied, it fails if the home of one queen is Fig. 13. — Head of Dorymyrmex pyramicus. This is the most abundant ant in the southern part of the State. The coarse teeth on the jaws distinguish it at once. overlooked. It is this fact alone that makes eradication difficult. We know many remedies that can be employed to destroy ants. We do not yet know what it is wisest to employ, and it is more than likely that we will find different treatments best under different conditions. I do not anticipate that this is a case where a specific will be found, but the work, like a battle, will require the direction of a resourceful commander. Moreover, it is not a work that can be done in a day, but should be undertaken with the idea that final success will only come by destroying the insect step by step. We now know more than forty separate colonies of the Argentine ant. If all were treated, and though failure resulted in the majority of cases, any success would promise final victory by continuing the fight along the same line. If the success of the work depended upon one general attack, I would have no confidence of victory, not because of the inefficiency of the remedy employed, but because of the difficulty of a perfect organization of the effort. ■>,,, l0 : 64 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. HOW IMPORTANT ANTS ARE. While it remains a question whether the Argentine ant can be eradicated in the State at large, there is but very little doubt that, taken in time, small colonies could be annihilated, and this raises the question at once as to the importance of destroying them. In the circular already referred to we have described very fully the serious- ness of this insect in Louisiana. Further study has shown that the danger due to the direct attack of this insect upon vegetation is not Fig. 14. — Worker of Dorymyrmex pyramicus. Notice the pyramid above the base of the middle and hind legs. Bulletin 207. THE CONTROL OF THE ARGENTINE ANT. 65 as large in this State as the experience in Louisiana led us to fear. In every other respect, however, the Argentine ant has proven itself quite as serious a pest as first indicated, and in addition its insidious attacks upon bee hives has, at least in one instance, put an amateur beekeeper out of business, and in two cases that have come to our attention have become an equal menace in aviaries by the attacks upon the nestling, and indeed there is considerable evidence that they will have an appreciable effect upon native wild birds in the same way. Fig. 15. — Head of Iridomyrmex analis. Argentine ant in California. This is the nearest relative of the As a household pest the Argentine ant certainly stands ahead of all other species, and many of the native species are serious enough. Whenever the Argentine ant gets established it destroys all other species of ants and becomes much more abundant than all the others combined. Unless one uses remedies they become almost unbearable in the kitchen, and it is probably very conservative to estimate that the presence of the ant reduces the commercial value of real estate for residence purposes from ten to twenty-five per cent. Many cases have come to our attention of people moving out of the infested territory on account of the ants and of sales of real estate falling through as soon as the prospective purchaser discovered the presence of the ant. NATIVE SPECIES. Often when the Argentine ant first enters the house it is not distinguished from the native species, the housewife only noticing that the ants have become particularly annoying and do not respond so readily to remedies. It will be very desirable, therefore, to point out the distinction between this species and the various native species with which it mav be confounded. 66 UNIVERSITY OF CALIFORNIA — EXPERIMENT STATION. There are some forty-six species of ants known to occur in Cali- fornia. These have been divided into four sub-families, three of which include species that give more or less annoyance in the house, the fourth including a single species of the legendary ants so abundant and troublesome in tropical America, but which is a very rare insect in California. /' Fig. 16.— Worker of Iridomyrmcx anaUs. BULLETIN 207. THE CONTROL (JF THE ARGENTINE ANT. 67 These sub-families are distinguished by the following key, the figures will illustrate the characters mentioned: Pedical one jointed. Labrum not reaching above the bases of the antennae, Formicinae, 15 species. Labrum extending up between the antennae, Dolichoderinae, 5 species. Pedical two jointed. Frontal ridges covering bases of antennae, Myrmecinae, 25 spec : es. Frontal ridges not covering bases of antennae, Dorylinae, 1 species, Echiton calif 'ornicus , Mayr. r^^rtK Fig. 17. — Head of a house ant (Monomorium) . This ant is an introduced species from Europe. Sub-family Formicinae. — This sub-family is often called Campono- tinae after its largest genus Camponotus. There are four genera in California, distinguishable by the following table : Clypeus scarcely any broader than high, Camponotus Distinctly wider than high. Clypeal fossa distinctly separated from antennal fossa, Prenolepis Confluent. Fourth joint of maxillary palpi not quite as long as fifth, Myrmecocystus A little longer, Formica The Carpenter Ants. Camponotus. — The largest ants in California are known as the Carpenter ants and have sometimes been called "wood ants." They are among the most abundant ants of the moun- tains and foothills wherever the country is wooded. There are five different species, mostly black in color and of large size. They live in rotten logs and stumps and are occasionally brought into town in fuel and may establish themselves in the woodwork of houses. There are two or three such colonies in Berkeley, for instance. It is only rarely that they become very annoying and usually they die out in a few years 68 UNIVERSITY OP CALIFORNIA EXPERIMENT STATION. after they are first introduced. I have never observed a conflict between these ants and the Argentine species, but there is no doubt that the Argentines would destroy them as they have the other species, even though they are between fifty and a hundred times as large. Wherever this ant becomes troublesome there is no doubt that the remedies suggested for the Argentine would be even more effective. Fig. 18. — Worker of the fungus growing ant (Cyphomyrmex). The most peculiar shaped ant in California. Field Ants. Formica. — I am using this term to designate the com- monest ants found in the gardens and fields in the Coast region of California. These ants also occur all over the world, though in the desert regions they are nowhere near as abundant as the harvest ants referred to below. Some five species have been recorded in California, ranging in color from black to a mixture of red and black. These ants are not infrequent visitors in the house, but always have their nests in the ground, where they often remain during the hotter parts of the day though they are active enough mornings and evenings. These ants resemble the Carpenter ants in not having any definite pathways; they seem to be guided entirely by sight in their movements and have the very peculiar habit of running a short distance and then stopping suddenly and after a moment they rush forward again. BULLETIN 207. THE CONTROL OF THE ARGENTINE ANT. 69 These sudden stops and starts make it quite difficult to follow one with the eye. They ascend trees and are perhaps the most constant visitors of aphis colonies. Their nest in the ground often has many openings and it is rather difficult to locate them all in the case of plowed ground. Where the ground has not been disturbed, however, it is very easy to destroy the nest with carbon bi-sulphide, pouring in an ounce or two and closing up all the openings with earth. In San Jose the Argentine apparently drives these ants ahead of them, because all Fig. 19. — Side view of body of the fungus growing ant around the territory infested by Argentines the black field ants are unusually numerous. Mr. Morris, the Horticultural Commissioner of Santa Clara County presented in a paper at one of the department, conferences, a very entertaining account of battles between these two species. Honey Ants. Prenolepis and Myrmecocystus. — The honey ants are among the most interesting members of this family, but are not very commonly troublesome in the house. The characteristic feature of these species is the fact that certain individuals are converted into living honey-pots, the abdomen swelling into a balloon-like structure for the storage of the secretions of plant lice which are gathered by these insects. Prenolepis imparls occurs all over the State, but is nowhere exceedingly abundant. These insects often close up their burrows' and remain underground for months at a time, usually coming out in the spring and taking advantage of the great abund- ance of aphids that usually occur during this season. A case came to my attention in East Oakland near the ant laboratory where there was no evidence of the honey ant during the fall and winter, but early in the spring a very populous colony opened up several entrances and were active several weeks before the Argentines finally annihilated them, occupying their burrows. I have found no other species of ant so far inside the Argentine territory, and it is probable that when attacked the year before they simply closed up their burrows and lived on their stock of food. 70 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. The following list includes the species of Formicinae known in California belonging to the above mentioned genera : 1. Prenolepis imparis Say. 2. Formica rufa L. subsp. obscuriventris Mayr, Var. integroides Emery. 3. Formica subpolita Mayr. 4. Formica tejonia Buck.?. 5. Formica pilicornis Emery. 6. Formica rufiventris Emery. 7. Myrmecocystus melliger Forel, subsp. mimicus Wheeler var. depilis Forel. Myrmecocystus melliger Forel, subsp. semirufus Emery, var. testaceus. 8. Myrmecocystus mexicanus Wesm. var. horti-deorum McCook. 9. Myrmecocystus lugubris Wheeler. 10. Camponotus levigatus F. Smith. 11. Camponotus maculatus Fabr., subsp. vicinus Mayr. Camponotus maculatus Fabr., subsp. vicinus var. semitestaceus Emery. Camponotus maculatus Fabr., subsp. ochreatus Emery. 12. Camponotus maccooki Forel. 13. Camponotus herculeanus L., subsp. pennsylvanicus DeGr., var. semi- punctatus Kirby. 14. Camponotus fallax Nyl., subsp. discolor Buckley var. cnemidatus Emery. 15. Camponotus hyatti Emery. Camponotus hyatti var. bakeri Wheeler. Sub-family Dolichoderinae. — This family also consists of four genera, as distinguished by the following table : Scale of peliole almost hidden by overhanging of abdomen, Tapinoma Freely exposed. Epinotum with a conical elevation, Dorymyrmex Not so elevated. Body densely pubescent, Liometopum Body moderately hairy, Iridomyrmex Each is represented by a single native species, and the latter includes the imported Argentine ant. The names are as follows : 1. Liometopum apiculatum Mayr var. occidentale Emery. 2. Tapinoma sessile Say. 3. Dorymyrmex pyramicus Eoger bicolor Wheeler 4. Iridomyrmex analis Ern. Andre. 5. Iridomyrmex humilus Mayr. Liometopum is a very rare insect, but all the other members of the sub-family are very common and very annoying household pests. Tapinoma sessile is the commonest little black ant all over the State and is one of the most abundant ants found in the gardens. The fact thai it is not found commonly far away from human habitations and is not credited to -the State in Dr. Wheeler's last list of North Bulletin 20' THE CONTROL OF THE ARGENTINE ANT. 71 American ants would seem to indicate that this insect is also an introduced species. The insect is about the same size and general appearance as the Argentine ant, but darker in color, not quite so active in its motion, and is most easily distinguished from the Argentine species by the fact that it possesses a very strong odor when crushed. This odor is produced by a liquid secretion which can be ejected from the abdomen as an appreciable drop and which is used in its contest with the Argentine species. As long as the supply of this secretion lasts the Tapinoma has no difficulty in keeping the Fig. 20. — Head of the fire ant (Solenopsis). This is the most troublesome species in the San Joaquin Valley. Argentine off, but after having put four or five Argentines out of the combat in this way finally the Tapinoma is put to rout and the Argentines are invariably victorious, because they always attack in sufficient numbers. We have observed many battles between these two species and the Tapinoma is always driven away from its feeding grounds and its home despoiled. Most of the ant remedies used along the Coast regions of Cali- fornia, both north and south, were devised to control this insect, and are quite sufficient but are not so satisfactory as the arsenic solution described below Avhich was devised for the Argentine species. Dorymyrmex pyramicns bicolor. — This ant is very slightly larger than the Argentine species, but resembles it in many ways and is readily distinguished by the fact that the abdomen is of a darker color than the body, on account of which fact the name bicolor was given it. This species occurs as the most abundant ant in the towns and road- sides in Southern California and is the species that the Argentine ant comes most commonly in conflict with in that region. One can 72 UNIVERSITY OP CALIFORNIA EXPERIMENT STATION. observe almost continuous fighting all along the boundary of an Argentine territory in that part of the State between these two species, and while the fight is always long continued the Argentine gradually extends its sway, annihilating the Dorymyrmex. The Dorymyrmex is one of the most annoying house ants of Southern California and that region has for years been a very good place for the sale of ant remedies, chiefly because of this species. The problem of its control is practically the same as of the Argentine species. Fig. 21. — Worker of the fire ant. The small figures show the difference in size between the largest and smallest workers of this ant. Tridomyrmex analis. — This ant is the nearest relative of the Argentine species, belonging in fact to the same genus. It is decidedly smaller than the Argentine ant and paler in color, in fact is almost yellow. Its colonies may be found intermixed with those of the previous species and its habits are not very different. In relative abundance it runs in the proportion of about one of Iridomyrmex analis to ten of the Dorymyrmex pyramicus, and were it not for the Dorymyrmex would probably become a very serious pest in the house. Iridomyrmex humilis. — This, the Argentine ant, which appears destined to become the only ant in California, is equally energetic as a garden and a house ant, and in the latter capacity particularly it has no equal. Every foot of space in the house from cellar to attic is Bulletin 20' THE CONTROL OP 1 THE ARGENTINE ANT. 78 patrolled almost hourly by specimens of this species, and when any- thing that is attractive to them as food is left exposed in an incredibly short time thousands of these insects will appear and occupy them- selves in conveying it to their nests. In doing this they walk over well defined paths as do all other members of this sub-family and in the case of this species they leave enough dirt or secretions from their bodies upon this path to make it distinctly visible wherever they continue to pass for some time. Their persistence renders them par- ticularly difficult to control, but the methods which were developed in our investigations have proven quite satisfactory and will be described below. Fig. 22. — Head of the acrobat ant {Cremastog aster) Family Myrmicinae. — This is the largest family of ants, the genera of which will be distinguished by the following key : Head with grooves behind the eyes to receive the antennae, Cyphomyrmez No antellal grooves. Antenna ten jointed, Solenopsis Eleven or twelve jointed. Petiole articulated to the dorsum of gaster, Cremastogaster Attached at the end. Club longer than the remainder of the funiculus, Pheidale Not as large. Mesoepinotal suture distinct. Posterior tibial spurs pectinate, Myomica Not pectinate. Eyes vestigial, Stenamma Eyes well developed, Messor Suture indistinct. Last three joints of antennae much shorter than the remainder of the funiculus, Pogonomyrmex Nearly as long. Mesoepinttum spined, Leptothorax Not spined, Monomorium 74 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. The species thus far credited to California are as follows : 1. Leptothorax nitens Emery var. heathi Wheeler. Leptothorax andrei Emery. Leptothorax pergandei Emery. Monomorium pharaonis L. Monomorium ergatogyna Wheeler. Pheidole barbata Wheeler. Pheidole hyatti Emery. Pheidole californica Mayr. Stenamma brevicorne Mayr. Stenamma nearcticum Mayr. Apheanogaster patruelis Forel subsp. bakeri Wheeler. Messor stodderdi Emery. Messor pergandei Er. Andre. 14. Messor andrei Mayr. 15. Myrmica bradleyi Wheeler. Pogonomyrmex barbatus F. Smith var. rugosus Emery. Pogonomyrmex occidentalis Cresson var. subnitidus Emery. Pogonomyrmex subdentatus Mayr. Pogonomyrmex californicus Buckley Pogonomyrmex californicus subsp. longinodis Emery. Cremastogaster lineolata Say subsp. leviscula Mayr var. californica Emery. Cremastogaster lineolata Say subsp. coarctata Mayr. Cremastogaster vermiculata Emery. Solenopsis aurea Wheeler Solenopsis texana Emery subsp. catalinae Wheeler. Solenopsis geminata Fabr. Cyphomyrmex wheeleri Forel. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Fig. 23. — Worker acrobat ant. often held up over the thorax. Notice the shape of the abdomen, which is Bulletin 207. THE CONTROL OF THE ARGENTINE ANT. To The little House Ants. Monomorium. — These are imported species found chiefly in towns but not particularly abundant. They disappear completely on the coming of the Argentine species, but the combat has not been observed. They will doubtless respond to the same treatment as for the Argentine ant. The Fungus-growing Ants. Cyphomyrmex. — But a single rather rare species inhabits California. These insects are very sluggish and do not enter dwellings since they live on a fungus that they cultivate on the droppings of caterpillars or similar substances. The Fire Ants. Solenopsis. — These ants resemble somewhat the Borymyrmex of Southern California and occupy economically some- what the same position in the San Joaquin. A striking difference is the presence of large headed soldier ants and a greater variation in size of the smaller ants of the colony and their stinging proclivities, which gives them the name of fire ants, makes them an added nuisance. The ant is found also in Southern California, but does not there become excessively abundant. We have not as yet seen this ant in conflict with the Argentine species. The ant usually has nests with a single, very definite opening and will, therefore, be very easily handled by the use of carbon bisulphide. Fig. 24. — Top view of the body of the acrobat ant. The Acrobat Ants. Cremastog aster. — A fairly common shining black ant about towns; was imported into this country from Asia by way of Europe. The above name is suggested because of their habit of walking about at times with their abdomen thrown up over their back. The colonies are seldom very large; they are rather sluggish creatures and not very commonly found in the house. They could undoubtedly be poisoned in the same way as suggested for the Argentine ant. The Harvesting Ants. — The remaining six genera are all harvest- ing ants, so called by the fact that they collect seeds of plants and 76 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. after sweating them in their nests remove the chaff and pile it around the entrance. They are the commonest ant of the plains and bare foothills. Their nests are most conspicuous objects and can be seen for miles. Some of them, particularly those living well up in the Sierras build enormous mounds. Pig. 25. — Worker of :i black harvester ant (Stcnamma andrei), Bulletin 207. THE control OP THE ARGENTINE ant. 77 The two commonest species are Messor andrei, a black species, most abundant in the foothills, and Pogonomyrmex calif ornicus, belonging chiefly to the plains. All these ants, living in their conspicuous nests in the ground, are in consequence very easily destroyed by the use of carbon bisulphide. The Argentine ant seems to have but little if any more difficulty in disposing of these large ants than of their smaller foes. Fig. 26. — Head of Stenamma andrei. CONTROL MEASURES. Ants abound where food and water are plentiful. The harvesting ants dig deep burrows to water and can live where there is no surface water obtainable, and nearly any species becomes independent of a natural water supply where there are nectar secreting insects, such as scales or plant lice, or where plants have glands that secrete liquid material. The Argentine ant is incapable of digging for water and will probably never thrive in the drier plains. Where there is no lawn or growing plants about a place a house may become quite like a desert and be apparently deserted by ants, but when either w r ater or food is about they come trooping back. BARRING ANTS OUT. Wherever ants are present it will be desirable to keep food material out of their reach, placing everything in cans, jars, or tight boxes as many campers do. Ants cannot burrow through the walls so it is a very simple matter to make closets or cupboards ant tight. The use of felt strips about door and window and the puttying up of the cracks about the floor or wall will accomplish this very effectively. Where ventilation is desirable the use of cloth instead of wire screens will admit of some air and is entirely ant tight. Small safes for meat can be made in this way and where out-of-door cloth closets are used in the hotter sections, consisting of a cloth covered space in which the 78 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. cloth is kept wet by a constant drip of water it is only necessary to see that the cloth is tacked tightly all around and that the door closes tight and the ants will be effecively barred out. WATEK BAEKIERS. Perhaps the best known method of barring ants out is the water barrier, such as is secured by setting the table legs in cups of water. This is effective against most ants, but the Argentine ant has no difficulty in crossing water. Oil they cannot pass, but it is objection- able. We have found that the addition of a very small amount of cresol, just enough to make the water milky, renders water effective. Fig. 27. — Body of Stenamma andrei from above. The odor of cresol disappears in a day or two and the water will continue effective indefinitely. The cresol has germicidal qualities that prevent the water from becoming foul. Indeed the cresol water makes possible the development of a system of ant-proofing that is thoroughly practical and efficient. ANT-PEOOFING. The rendering of a house ant proof may be accomplished, provid- ing there is a level all around the house where a trough can be fastened which can be kept filled with cresol water, and providing that the floor is made ant tight. It is even possible to allow for vines on the house, and still have it tight by allowing a space through which the vine may grow, but stuffing it with cotton to prevent the ants climbing up the stem. A pantry can be easily made ant proof by a trough just above the top of the doors and windows and the suspension of the shelves from supports above this line. We have tried these troughs for months under the most severe conditions and they have proven absolutely effective and they do not need attention more than two or three times a year. The writer will be very glad to advise with anyone contemplating the ant-proofing of houses or rooms. Bulletin 207. THE CONTROL OF THE ARGENTINE ANT. 79 METHODS OF KILLING ANTS. The methods thus far discussed have not been directed toward reducing the numbers, but toward protecting the food in the house from their attacks. The more direct methods of fighting ants are often very practical. Fig. 28. — Worker of the large red harvester ant (Pogonomyrmex calif ornicus) . These ants often coil themselves up as shown when picked up. DESTBUCTION OF NESTS. In the above description we have referred to the use of carbon bisulphide for the destruction of nests of ants. In all the species where there are large nests with a single opening this is by far the most satisfactory treatment. The plan is simply to pour down a few ounces of carbon bisulphide, either in the natural openings or in holes made by thrusting in a crowbar and covering everything w T ith earth. The gas formed by the evaporation of the carbon bisulphide effectu- ally destroys both young and old. This method can be applied to any species where the nest can be discovered, but in the case of the Argentine species it becomes the least valuable of any method, since the nests are usually scattered almost everywhere over the whole surface of the ground and the treatment to be effective would have to include the entire ground space for acres about the house one intends to protect. There are conditions also which render the carbon bisulphide treatment insufficient. In our desert regions the soil becomes so dry 80 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. and open that the gas formed by the evaporation of the carbon bisulphide dissipates itself so rapidly through the soil that it becomes practically impossible to destroy the whole nest. In these cases better results have been secured by the use of a solution of cyanide of potassium. This material has been used for several years in the parks at Denver as a secret preparation, and its efficiency was recently independently discovered by Mr. Woglum in Southern California. The method consists simply in making a rather strong solution of the salt in water and pouring it into the holes in the same way that carbon bisulphide is used, only very much more of the material is applied, the amount actually used of course depending on the size of the colony. Hydrocyanic gas is liberated in the soil by this means and operates very effectively against all ages. Another method which we found effective and cheap and free from the disadvantage of using so violent a poison as potassium cyanide was to apply distillate made into a miscible oil and diluted with perhaps twenty times its volume of water. To be thoroughly effective this must be applied very liberally, soaking the ground well with the mixture. When this is done it seems quite as effective as any other method. ANT POWDEES. Numerous ant powders have been from time to time recommended for ants, many of which are of but little value, such as borax, chalk and the like, which when placed across the path of the ant, may cause it to turn aside and in the case of some species apparently giving some temporary relief. These have absolutely no effect against & persistent ant like the Argentine. The most efficient powder is the common Persian insect powder (pyrethrum), often sold under various names as indicating its availability for use against ants. When this is applied liberally on the shelves in the pantry and around the edges of the room, it sometimes will keep the insects away for several days when it must be repeated. Some housewives have had very good results by putting out the powder just after sweeping in the morning and leaving it about until the next day, when the old powder is swept up and fresh material applied. It is rather an expensive manner of treating ants, because it must be kept up continuously, but were there no better methods available it would certainly be better than to allow the ants to have continued access to the premises, and there may be times when it may be worth while. It will be cheaper, how- ever, to buy the insect powder rather than the special preparations known as ant powders. BULLETIN 207. TH E CONTROL OP THE ARGENTINE ANT. 81 ANT FLUIDS. A very common form of ant remedy for sale in the drug stores consists in an alcoholic solution of corrosive sublimate, usually colored by some coloring matter to disguise the nature of the prepara- tion. Any preparation with a strong odor of alcohol is quite certain to be of this character. The method of applying is to paint the material about on the shelves; it immediately dries up but leaves the fine crystals of corrosive sublimate which seems quite deterent for most species of ants. The Argentine ant seems to be very little influenced by this poison, and in many experiments tried at the ant laboratory lines of ants were found re-established across the poison inside half an hour of the most liberal treatment. Very good results have followed the use of this material against Tapinoma, which is often mistaken for the Argentine ant, and even in the case of the Argentine sometimes temporary relief is secured where there is nothing particularly attractive to the ants beyond the barrier. ARSENICAL POISONING. We obtained by far the best results by the use of a very weak solution of arsenic and syrup. Most of the commercial ant poisons commonly known as ant pastes consist of arsenic and syrup, but are made very strong in arsenic. This kills the foraging ants almost immediately. We found by reducing the arsenic to between one-fourth and one-eighth of one per cent, they would take large quantities of the material to their nests and feed it to the young, and the whole nest would be killed by a slow poisoning. The most convenient way of exposing the poison to the ants is to use a large jar with a perforated cover and within it place a sponge saturated with the arsenic solution. The ants will enter through the perforations in the cover, fill themselves with the arsenic solution and carry it to their nests. The sponge will hold enough poison to require two or three weeks to empty it, and before that time the ants will almost entirely disappear. The number of jars to use will depend upon the abundance of ants. In the worst cases a half-dozen jars will serve for an ordinary sized house and lot, and if the ants are not very bad one jar may be enough, in such cases it is well to place it in the pantry or kitchen. The same remedy can be used for all the native species of ants and will be more effective against them 82 UNIVERSITY OF CALIFORNIA EXPERIMENT STATION. EESUME. The Argentine ant now occupies about 5,000 acres in California. There are more than forty separate colonies, varying from one acre to nearly 2,000 acres in extent. Except when carried by human agencies they spread at about the rate of an eighth of a mile a year. Manure and nursery stock are particularly liable to be the method of transportation, though they may go with all sorts of merchandise. Natural spread is not accomplished by flight. Migration normally occurs when the numbers exceed . the food supply. Eradication is a possibility and should be undertaken. Importance estimated as a lowering of the value of residence prop- erty between ten and twenty-five per cent. All native species are discussed and the treatment of each suggested. Control measures consist of barring ants out and methods of killing. The best barrier consists of water treated with cresol and makes possible a practical ant-proofing of rooms or houses. For many ants the nests may be destroyed by the use of carbon bisulphide, potassium cyanide, or oil. Ant powders and fluids are discussed, but a syrup containing a very small quantity of arsenic is recommended as the most available poison and the only really satisfactory method of killing the Argentine ant.