26 PRINCIPES tVol. 23

Pfimipes,23(f), 1979, pp. 26-31 Notes on the Foraging Behavior of a Leal-cutting Ant on bacaba in the Northwest Arrrazon of

ilIrculrl J. Baucx BotanicalMuseum ol Haraard Uniuersity,Oxtord, Street, Cambridge, MA 02138

A group of ants whosepeculiar forag- cultivated , especially the coffee and ing behavior is of great concernto agri- orange.. . ." culturists and horticulturists is found in In contrast to the well-known above- many areas of the Neotropics and sub- ground activities of these ants, much tropics. These ants, variously known as less is known o{ their subterraneanac- leaf cutters, gardening ants, or parasol tivities in fungus gardening. At first it ants, clip sections {rom the leaves of was thought that the ants or their young herbs, shrubs, and trees,and carry these consumed the cut leaves. Thomas Belt, pieces to their nests, where they culture a nineteenth century naturalist, was the "fungus gardens" upon the lea{ residue. first to describe the true purpose of their According to l[eber (L972), of the ap- unique foraging behavior (Belt, 1874) 'othese proximately 10,000 known of ants are so common throughout ants, only lbO-ZOO species exhibit this tropical America, and have excited the behavior. The most conspicuousby their attention of nearly every traveller, there size, and the best known, are the genera still remains much doubt as to the use Acromyrnex and, Atta. to which the leaves are put. Some nat- Much attention has been devoted to uralists have supposedthat they use them the ants' habit of defoliating in directly as food; others that they roof forests and cultivated areas. Descriptive their underground nestswith them. I be- accounts of these ants are to be found lieve the real use they make of them is in many of the early books concerning as a manure, on which grows a minute natural history in the American tropics speciesof fungus, on which they feed- (Andre, 1904; Barbour, 1945; Bates, that they are, in reality, mushroom 1863; Belt, 1874; Ornton" lB70). Re- growers and eaters. . . ." In a letter to garding their activity in the Amazono Charles Darwin published in the June ooBut Ornton (1870) wrote the most Il, 1874, edition oI Nature, Fritz Muller prominentby their immense numbers are supported this theory, based on his in- tlre dreaded saubas (Atta sp.). Well vestigations of the stomach contents of beaten paths branch off in every direc- the leaf-cutting ants. There is a close tion through the forest, on which broad association between the ants and the columns may be seen marching to and fungus, in that neither can survive with- fro, each bearing vertically a circular out the other. Each colony cultures only piece of leaf-un{ortunately they prefer one species of fungus, and the workers OENOCARPUS 27 19791 BALICK: ANTS ON

anentire oenocarpusu"ii!::ir:u:1ilil,j::t"x;:'?i"rili"l'"""uuilffi: t',""Ti lv.Iitr1'ocolombia'

when in fact the presence or absence of of the colony insurt that no other con- leaf-cutting ants dictates the types of iaminating types of fungus or mold take crops that can be grown. Wilson (1971) over. the story of a foreigner living in Lea{-cutting ants are of economrcslg- relates British Honduras who planted all types ni{icance throughout their range' They of vegetables in his garden against the destroy great areas o{ cultivated plants advice of the local Indians, and who und g."rrlu.tds, rendering them useless awoke one morning to find his garden for a=griculture, and the ants actually "Into completely de{oliated, and a hole "o*pJ" with cattle {or {ood' Weber mound earth], ants, moving (Igi2) notes that ten adult colonies of in the [o{ in quickened steps, were carrying bits of )no iopisuoro use the same amount of toPs o{ the carrots, the gru., u" an entire cow. ManY agricul- our cabbage, beans-in fact our entire garden was iurists from north temperateregions who going down that hole." These insects visit the tropics for the first time and are play such a visible and important role unfamiliar wit[ the deslructivehabits of in the lives of the inhabitants of tropical rhese ants otte; {ail to understand the America, that colloquial folk names have reasonsfor the paucity o{ {ruit trees and been given to them throughout their vegetable crops cultivated by the -local including bachaco (Venezuela), people in otherwise lush tropical en- range, "experts" (), isau (ParaguaY), and viro.tments. These northern coqui (). often ascribe these cultivation practices sauba benefits derived from to the supposedlethargy of the people, The economic PRINCIPES [Vol. 23 ":4

'#':t , a;t'

2. Part of an upturned spathe of Oenocarpus with live ants at side collecting flowers and a section of rachilla, dead ants, and floating flowers in water.

these ants hardly balances their detri- be the pollinators in this species of mental actjvity: the anls { especially Oenocarpus. A{ter remaining open {or queens of the Atla\ are used by an undetermined length of time. the native peoples as a food source, and in staminateflowers fall {rom the rachillae. the United States they are the canned, Large soldier ants of the colony were precooked ants often found in gourmet carrying the fallen flowers back to their shops. nest,each holding a singleflower clasped I had an opportunity to observe leaf- strongly in its sharp cutting mandibles. cutter ants during field studies in the The freshly fallen spathe was upturned Northwest Amazon of Colombia, on a toward the sky and contained a reser- sunny August day in 1976. T}r'e locality voir of water from the daily afternoon was on the Vaupes River, across {rom rains. Many flowers had fallen into the the small village of Mitu, the capital of troughlike spathe, and dozens o{ ants the territory o{ Vaupes. A single flower- were swarming about, trying to collect ing specimen oI the blossoms. An ant would scurry into Magt. stood in the center of an open the spatheand carry off a flower. Climb- grassy field, with an underground nest ing out of the incurved spathe often of. Atta sexdens L. a few {eet {rom the proved an impossibletask for theseants. base of the . The flowering spathe With their load as large as their own of the palm, over one meter in length, bodies, they would slip into the pool of had fallen, the flowers had opened, and water after trying hopelessly to ascend were surrounded by small black stingless the slippery walls. Each ant thrashed bees (Meliponidae), that may possibly about in the water for a short time" and 1979J BALICK:.ANTS ON OENOCARPUS 29

-q rt

'1; ru !F] rt *k '*8".

3. An ant carrying a flower from the uptFned spathe_of Oenocarpus above, ants carrying flowers on a foraging traii below.

then drowned. The spathe was filled whether the ants were actually clipping with dead ants, many of which were of{ the blooms from the long "horse- still clutching their precious Oenocarpus tail" in{lorescenceof the palm. After flowers. As the lea{ cutters are known searching the flowers and leaves of the to be so destructive to cultivated plants tall palm with binoculars, I could find and flowers, I attempted to determine no ants engagedin this type of activity. 30 PRINCIPES lVol. 23

Several ants from the colony were mov- that it changes in rate of collection and ing up and down the thick trunk, per- type of material preferred, depending on haps scouting for a nocturnal assault. the seasonor time of day. He also men- In some areas these ants are known to tions that the ants avoid srrnny, heated be nocturnal, especiallyduring the sea- trails, which may accounf- for the ab- sons having hot and sunny days. sence of any ants seen collecting on the Observations of the returning ants at palm itself, especially in view of the the entrance to the colony tunnell indi- sunny condition when the was cated the ants in this group favored the studied. Do the ants return any bene{it palm flowers over all other material in to tke palm, either by protecting it or the area. In an attempt to get the in- perhaps by encouraging some sort of ooenthusiasm" sects to display their {or mycorrhizal association in the soil, or the flowers in a more dramatic manner, by nurturing it through the vast massof o'manure" I knocked the buds and open flowers underground produced by the down from the inflorescence, raining colony from the leaf residue? Weber flowers all over the grassy area sur- (L972) statedthat the fungus cultivated rounding the palm. Soon ants seemedto in leaf cutter nests exhibits biological come from everywhere to gather the buds activity against certain plant pathogens. and flowers, each picking one up and Are the ants then doing the palm a fur- following the trail back to the nest. By ther service through protection against positioning mysel{ at the entrance hole, attack by plant diseases? Are there I was able to collect herbarium material other palm flower-ant relationships and for botanical studies,simply by picking are they less evident in jungle areas off the flowers {rom the insects as they where the ants have a greater selection entered their underground nests. The of succulent plant parts to forage from? unrestrained preference shown by this Indeed further research and observation ant colony for the flowers ol Oenocarpus is needed to answer these and other re- bacaba, rqises several interesting bio- lated questions which arise from such logical questions. Are the palm flowers observations on Amazonian palms. While used for anything besidesthe construc- this single observation oI Atta sexd,ens tion o{ fungus gardens,or are they sim- is of no significance by itself, it is pre- ply easier to compost than the thicker, sented in an attempt to stimulate further more fibrous grassesthat surround the observations of this type by other stu- nest in great abundance? Do the ants dents of the Palmae,hoping that a por- expend energy by climbing the trees to tion of the time spent in the field will harvest flowers, or do they just wait un- be devoted to studies on such insect- til the blossoms fall to gather them? At palm relations as might be evident. what time o{ the day or night is their Those living within the natural distribu- peak of activity with this particular tion of the palms have a great advantage palm, and what other plant material is in this area of study in that yearlong ob- used when palm flowers are no longer servations can be made and careful field available? Observations by Rockwood experiments carried out, Future work (1975) in Costa Rica show that the for- will surely reveal much needed informa- tion on palm-animal interaction, along aging behavior of Atta is adaptive in with their possible coevolutionary impli- cations, vital topics in view of the cur- 'This also might have been one of the en- rent rate of the destruction of the rain trances to a series of underground trails char- acteristic of this genus and mentioned in forest and attempted elimination of what Fowler (1976). are presently thought to be insect pests. r9791 BALICK: ANTS ON OENOCARPUS 3t Acknowledgemenls Llrrnlruns Crrrn Axnn6, E. 1m4. A Naturalist in the Guianas. This study is part of an ongoing thesis Smith Elder and Company, London. investigation o{ the biology and system' Brnnoun, T. 1945. A Naturalist in Cuba' atics o{ the lessenia-Oenocorpus com' Little Brown and Company, Bciirton. 'W. plex. Past field work, during which the Bltrs, H. 1863. The Naturalist on the previous observations were made, was River Amazons. John Murray, London. supported in part by grants from Sigma Brr,t, T. 1874. The Naturalist in Nicaragua' Xi, the Anderson Fund of Harvard Uni' John Murray, London. Fowr-un, H. 1976. An unusual trail modifi' versity, and Centro de Desarrollo In- cation' observed in the leaf'cutting ant oolas Gaviotas" in Bogota, Co- tegrado Acromyrmex crassispinusForel. Biotrop' lombia. I wish to thank sincerely all ica B: 142. those persons and institutions who have Onr,rrorv,J. 1870. The Andes and Amazon: been so kind in assisting in this investi' or Acrossthe Continentof SouthAmerica. gation. Although too numerousto men- Harper and Brothers, New York. tion herein, for this particular topic the Rocrwoon, L. L, 1975. The effects of sea' interest and assistance of the following sonality on foraging in two species of (Atta) are gratefully acknowledged: Dr. Paulo leaf-cutting ants in Guanacaste Rica. Biotropica 7 z 176' Lugari C., Dr. T. Plowman, Dr. M. L. Province, Costa 193. Corne, Dr. E. O. Wilson, J.L. Zarucchi, Wenrn, N. A. 1972. Gardening Ants, the C. W. Greene, H. J. Hoyos, Jardin Attines. American Philosophical Society, "Joaquin in Botanico Antonio Uribe" Philadelphia. Medellin, and Instituto de Ciencias WIr-sorv, E. O. 1977. The Insect Societies' Naturales, Universidad Nacional de Belknap Press o{ Harvard University Colombia. Press,Cambridge.