From The Ants Chapter 14:
The strongest evidence for ant-plant mutualism comes from the existence of domatia, or plant structures that serve no evident purpose other than to shelter ant colonies. Domatia increase the density of ants on the plant itself. It is easy to see how the phenomenon could arise in evolution. Ants are always quick to take advantage of whatever hollows and crevices plants have to offer. In most cases the shelters are not true domatia. They are adventitious, and the ants therefore live as parasites or commensals on the plants. These incidental nest sites can be divided into convenient categories as follows:
• Preformed cavities in live branches and stems, excavated by woodboring beetles and other insects and later occupied by ant colonies belonging to such diverse genera as Daceton, Podomyrma, Crematogaster, Azteca, Lasius, and Camponotus.
• Cavities in stems and branches that are naturally hollow or contain a pith soft enough to be easily excavated by ants. A legion of grasses, sedges, composites, and other herbaceous and shrubby plant forms provide this form of refuge and a great variety of ants occupy them. A typical field of sedges and weeds in Florida, for example, contains dense populations of one or more species of Pseudomyrmex, Colobopsis, Crematogaster, and Monomorium. The low trees and bushes of a Brazilian cerrado contain the same genera along with several species of Zacryptocerus.
• Natural or preformed cavities in bark. The bases of pine trees in the southern United States shelter an entire fauna of Hypoponera, Pheidole, Solenopsis, Crematogaster, Brachymyrmex, and other ant genera that nest adventitiously in the bark. Higher up can be found colonies of the small myrmicines Leptothorax bradleyi and Leptothorax wheeleri, which are specialized to this environment to some extent (Wilson, 1952). An ant restricted entirely to this microsite is Melissotarsus, which occurs in the bark cavities of living trees in tropical Africa and Madagascar. Its body form and locomotory behavior are modified for existence in tight spaces. The workers walk with their middle legs held upright and touching the roofs of the galleries. If placed outside, in the open, they are unable to move around in a normal manner (Delage-Darchen, 1972a).
• Roots of epiphytes. The tangled root systems of orchids, gesneriads, and other tropical epiphytes are ideal nest sites for ants. One of the most profitable ways to collect ants anywhere in the world is to hold an epiphyte over a pan or ground cloth and strike the root system several times with a trowel. In Central and South America, this technique yields large numbers of colonies of Hypoponera, Gnamptogenys, Strumigenys, Nesomyrmex, and other genera, many belonging to rare or previously undescribed species.
• Ants frequently nest in galls formed by cynipid wasp larvae; the phenomenon has been observed in Europe (Torossian, 1972; Espadaler and Nieves, 1983) and North America (Wheeler, 1910a). Leptothorax obturator of Texas appears to be a specialist on this nest site (Longino and Wheeler, 1987).
• A few ant species, such as the large ponerines Ectatomma tuberculatum and Paraponera clavata of the New World tropics, construct earthen or carton nests vertically against the sunken portions of tree trunks. Thus the tree provides a partial wall of solid wood that is virtually invulnerable. Paraponera clavata prefers trees of the abundant legume Pentaclethra macroloba, but the relationship is not obligatory. It is possible that the ants benefit from both the well-formed buttresses of the Pentaclethra and the extrafloral nectaries in the foliage of this species (Bennett and Breed, 1985).
None of these diverse structures appear to be “designed” to accommodate ant colonies. All are ordinary anatomical features of the plants that the ants exploit, apparently in a unilateral manner. In contrast, the domatia listed comprehensively in Table 14-1 do appear uniquely to serve as ant nests. They are characterized by cavities that form independently of the ants (even in greenhouses, where no ants are present), adventitious roots and tubercles that absorb nutrients from waste material carried onto the cavities, and even holes or thin windows of tissue through which ants can more conveniently enter and leave. Some of the most distinctive forms are illustrated in Figures 14-4 through 14-8. Domatia are almost always occupied by ant colonies in nature. Furthermore, plant species with the most complex domatia are typically occupied by only one or a small number of species specialized to live with them. Finally, species with domatia usually also manufacture food bodies, which are unique structures with no known function other than the feeding of ants (see also Table 14-2). In short, strong circumstantial evidence indicates that domatia are structures specialized in evolution to promote symbioses with ants. Benson (1985) has suggested that the ant domatia of many ant species evolved from the sheltered feeding sites used by mealy bugs and other homopterous insects. Because of the honeydew produced by the homopterans, ants were attracted to these sites, which were enlarged and otherwise structurally modified into ant domatia.
Further evidence of coevolution is provided by the legendary ferocity of many of the guest ants. The vast majority of Pseudomyrmex species not occupying domatia are timid and flee even when their nest is broken apart. In sharp contrast, Pseudomyrmex triplarinus, an obligate resident of Triplaris americana, falls upon any intruder touching the nest tree without hesitation or mercy. To be stung by several of these ants within a few seconds is like encountering a nettle--you pull back at once. Or conversely, if you want to locate Triplaris quickly in an Amazonian forest, shake one sapling after another until one produces a swarm of the stinging ants. The Pseudomyrmex also attack and remove intruding insects, and they are significantly more efficient than the Crematogaster that also occupy Triplaris (Oliveira et al., 1987a). The species of Camponotus offer a similar dichotomy. Most retreat or offer limited resistance when the nest is disturbed by a human being. Possibly the shyest ant species in the world is the Amazonian Camponotus paradoxus, whose workers disperse and hide so quickly that it is difficult to catch any specimens at all. At the opposite extreme is Camponotus femoratus, an obligatory resident of epiphytic ant gardens in South America. Diane Davidson (personal communication) describes its behavior as follows:
When I approached to within 1-2 m of their nests, workers of this species typically began to run back and forth and frequently jumped or fell onto me. Workers of all size classes of this polymorphic species attempted to bite, but usually only the major castes were capable of breaking the skin with their mandibles and causing a stinging sensation by simultaneously biting and spraying formic acid into the wound. In addition, these workers often exhibited a second type of apparently aggressive behavior, which I will term “coughing” behavior. With mandibles held wide open and the prothoracic legs upraised, they brought their legs down abruptly in a jerking movement that resembled a cough.
Tetraponera tessmanni, the obligate pseudomyrmecine tenant of the verbenaceous creeper Vitex staudtii in West Africa, was described by Bequaert (1922) as “exceedingly vicious and alert. When its host plant is ever so slightly disturbed, the workers rush out of the hollow stalks in large numbers and actively explore the plant. Their sting is extremely painful and sometimes produces vesicles on the skin.” Even more redoubtable is the African pseudomyrmecine Tetraponera aethiops, the obligate tenant of the small flacourtiaceous tree Barteria fistulosa: “As soon as any portion of their host plant is disturbed, they rush out in numbers and hastily explore the trunk, branches, and leaves. Some of the workers usually also run over the ground about the base of the tree and attack any nearby intruder, be it animal or man. All observers agree that the sting of the Tetraponera is exceedingly painful and is felt for several hours. Its effects can best be compared with those produced by female velvet ants.” The ant is feared by the natives of the Congo, who try to avoid the unpleasant task of cutting the small Barteria trees scattered through the forest. As a consequence individuals of Barteria fistulosa are often found standing by themselves in the center of clearings or near the sides of forest paths. The species is also abundant in secondary forest growth.
Not all myrmecophyte ants are this formidable. A few, such as the tiny Pheidole species that occupy Maieta and Piper, seem incapable of defeating most herbivores. Many authors have commented on the evident ineptness of these ants in the protection of their adopted plants. However, Letourneau (1983) points out that such ants can perform their service by removing the eggs and early developmental stages of the herbivores rather than face down the adults. She showed that Piper plants in Costa Rica occupied by Pheidole bicornis (now known to involve three species, Pheidole bicornis, Pheidole bicornisculpta, Pheidole passivaeferox (Longino, 2019)), their principal ant guest, suffered less damage than those deprived of the ants. The workers preferred to patrol new leaves, which are the most susceptible to insect damage. When Letourneau placed termite eggs on Piper bushes, the ants discovered more than 75 percent and dropped them off the plants within an hour. Risch et al. (1977) also found some evidence that the Pheidole chew through or push aside alien vines from their host plants, and they postulate that the ants also bring nutrients to the plant cavities as part of their nest material. No one has assayed the relative importance of these benefits to trees.
Competition for the ant plants is intense. Young plants are soon fully occupied by colony-founding queens of plant-ant species. It is common to find different internodes of very young Cecropia saplings occupied by one or more queens of Azteca. Sometimes the inhabitants belong to two species. As many as a dozen colonies are started in this manner in each tree. When the first workers emerge, they cut holes through the septa separating the internodes. Fighting and other forms of competition ensue, and all of the young colonies except one are either destroyed or perhaps even assimilated, so that in the end only one large colony and a single nest queen survive (Dan Perlman, personal communication). The same reductive sequence occurs in Pseudomyrmex ferrugineus, a resident of swollen-thorn acacias (Janzen, 1967), as well as Pseudomyrmex triplarinus, an obligate resident of Triplaris (Schremmer, 1984).
The pervasiveness of competition is indicated by the patchiness of distribution of the ants among plants of the same species. At the Manu National Park, Peru, Davidson et al. (1988b) found eight myrmecophyte species. Among 130 plants dissected, 127 contained ant colonies, and of these 126 belonged to one species only. In many instances the structure of the plant was found to bias which ant species can colonize it. Maieta guianensis, for example, has dense epidermal hairs (trichomes) covering the leaves and stems. Its usual guest ants, the tiny Pheidole minutula, are able to walk through the hairs without difficulty. A second and less common ant, a Crematogaster belonging to the victima group, apparently has to cut trails through the trichomes (Diane Davidson, personal communication). In addition, the narrow tunnel entrances leading into the pouch-like domatia are readily entered by the Pheidole queens during colony founding but not by the Crematogaster queens, which are forced to chew holes into the plant. Allomerus demararae enjoy a similar advantage in the occupancy of the myrmecophyte Cordia nodosa over species of Crematogaster and Azteca, which must cut trails through the trichomes just to move around (Davidson et al., 1988b). Put briefly, certain ant species occupy individual myrmecophytes predominantly and other species intrude occasionally, but only one species is found in each plant. The final result of this competitive pressure is that the myrmecophytes are saturated with ants.
There are many plants that have co-evolved with ants and now have structures and life histories that are intimately tied with ant partners. One example is the epiphyte Dischidia major where highly modified leaves (‘pitchers’) provide lodging for various ants, especially Philidris. In return, the plants obtain nutrients from organic debris brought in by the ants, through close contact with branching adventitious roots growing within inhabited pitchers.
A neotropical genus of fast growing pioneer trees. Ants of the genus Azteca are most commonly found living in and patrolling these plants. Neoponera luteola nests exclusively inside the stems of Cecropia tessmannii trees in Peru, feeding on glycogen-rich Müllerian bodies produced by the plant to guarantee the protection of ants against herbivores.
A complex of subspecies, this is an African forest understory plant with some forms having extrafloral nectaries and/or domatia. Two ant species Petalomyrmex phylax and Cataulacus mckeyi are known to be associated with these plants.
A southeastern Asian genus of pioneer tree species.
A neotropical plant genus. These dioecious species of trees are most often associated with Pseudomyrmex but other species will also inhabit their hollow stems.
- Janzen D.H. 1974. Epiphytic myrmecophytes in Sarawak: mutualism through the feeding of plants by ants. Biotropica 6: 237 – 259.
- Hölldobler B. and Wilson E.O. 1990. The Ants. Cambridge, Mass. Harvard University Press.
- Kaufmann E. and Maschwitz U. 2006. Ant-gardens of tropical Asian rainforests. Naturwissenschaften 93: 216 – 227.
- Peeters, C. & D. Wiwatwitaya 2014. Philidris ants living in Dischidia epiphytes from Thailand. Asian Myrmecology 6: 49-61.
- Weir JS and Kiew R. 1986. A reassessment of the relations in Malaysia between ants (Crematogaster) on trees (Leptospermum and Dacrydium) and epiphytes of the genus Dischidia (Asclepiadaceae) including ‘ant-plants’. Biological Journal of the Linnean Society 27: 113 – 132.