Ingestion of radiolabelled plant juices from cut surfaces of leaves during foraging and during substrate preparation for the fungus garden was demonstrated for laboratory colonies of the leaf-cutting ants Atta cephalotes (L.) and Acromyrmex octospinosus (Reich). Using P32 isotope, up to one-third of the radioactivity in the leaves was taken in directly by the colony as a whole, but this was much less when C14 was used. Additional plant material was taken in by the ants via the fungus garden. Large numbers of ants spend time in investigating cut surfaces of leaves and also in cutting and crimping leaves. Ingestion of plant material during these processes could play an important nutritional role in the colony and this may explain why many ants return from foraging apparently unladen. Large ants obtained most of their plant juice intake during foraging and the medium and small ants took much of theirs during substrate preparation. Inhibitory chemicals did not affect cutting but they reduced drinking and also reduced the intake of P32 from leaves during substrate preparation. The ants also ingested different amounts of P32 from leaves of different acceptability. It is suggested that leaf-cutting ants use fungus culture as a means of ‘sidestepping’ plant inhibitors. Nutrients are ingested directly from leaves when these are palatable, and the less palatable components are made available to the ants via the fungus.

In laboratory colonies of Atta cephalotes (L.) and Acromyrmex octospinosus (Reich) pick-up responses to paper discs impregnated with arrestant and inhibitory solutions and feeding responses to measured amounts of these solutions were progressively increased by increasing the concentration of arrestants and decreasing the concentration of inhibitor; the inhibitor had a much greater effect on feeding than on pick-up. Both species picked up paper discs impregnated with arrestants and Acromyrmex readily incorporated these into the fungus garden, while Atta did this less readily. Foraging of arrestive paper discs presented regularly for several days was maintained by Acromyrmex but declined in Atta. Both species were less selective during periods of high foraging activity and during a shortage of suitable substrate. Incoming foragers did not seem to influence the preferences of the outgoing ants, but different colonies showed different preferences. Initially little preference was shown for new materials but on subsequent encounters widely fluctuating preferences emerged, which stabilised within two weeks. Over longer time periods further changes in preferences occurred. The ants' pick-up, cutting or feeding response to a particular material is determined by its chemical and physical nature, the amount of chemical information which the ants obtain, and the threshold of their response. This threshold is modified by such factors as the genetic composition of the colony, the previous quantitative and qualitative availability of the substrate, and the general level of foraging activity. Practical implications of these findings for toxic bait production are discussed.

In laboratory colonies of Atta cephalotes (L.) and Acromyrmex octospinosus (Reich), discs cut from young leaves of several plant species were picked up more readily than those from old leaves. Pith sections impregnated with whole extracts of young leaves were consistently picked up in preference to others with extracts of old leaves, and non-lipid extracts were consistently preferred to extracts containing lipids. Arrestive properties were confined to the non-lipid extracts, and inhibitory properties were mainly found in the lipids, but removal of lipids from the whole extracts did not necessarily make them more arrestive. In most cases, dewaxing of leaves with chloroform made them more acceptable, and heating and freezing of leaves had a similar effect. Leaf waxes showed inhibitory properties. Wax extract of young leaves was often preferred to that of old leaves, and differences in preferences were observed for wax extracts from different plant species. The implications of these findings for the relationship between leaf-cutting ants and the vegetation that they cut are discussed.

The leaf-cutting ants Atta cephalotes (L.) and Acromyrmex octospinosus (Reich) were tested in a moving-air olfactometer for their responses to a variety of odours from leaves, fruit, flowers, essential oils, other ants and a fungus garden. The line of a foraging trail established in still air tended to loop downwind in the presence of a laminar air flow. The ants exhibited positive, negative and neutral responses to the test odours, confirming the existence of attractants and repellents in substrate materials. Neutral responses to the odours of several of the materials that were acceptable for cutting showed that arrestive materials were not necessarily attractive. Ants orientating towards the source of an odour often secreted a pheromone trail. Atta and Acromyrmex responded differently to several of the materials tested, and the responses to odours of young and old leaves were not totally consistent with the observed cutting preferences. Removal of wax from non-attractive leaves made them attractive. Both species responded negatively to lemon-oil odour, but some Acromyrmex workers showed a hostile response. A hostile response was also elicited by the odour from other ants. Fungus-garden odour elicited a positive investigatory response, and no response was shown to queen odour. Laden Atta workers did not respond to an odour that was attractive to unladen ants, and laden examples of Acromyrmex responded in small numbers to such odours. The addition of attractive chemicals to baits for the control of leaf-cutting ants would improve pick-up by making the bait particle easier to find.