The effect of wilting on the selection of leaves by the leaf-cutting ant Atta laevigata

Using field assays of leaf preference, we tested the hypothesis that wilting affects the selection of leaves by the leaf-cutting ant Atta laevigata (Fr. Smith). Detached leaves were left to air-dry until noticeably wilted. The area removed by the ants from wilted leaves was significantly greater than the area removed from fresh leaves, this effect being observed in several plant species, in leaves of different age, and in assays with different ant colonies. Leaves collected from water-stressed plants were also preferred to leaves from non-stressed plants. A. laevigata was found to employ a two-stage, size-related, strategy when cutting plants. Larger workers climbed the plant stem and dropped whole leaves to the ground by severing their petioles; smaller workers cut the lamina of the dropped leaves. The ants frequently left dropped leaves on the ground, until the next foraging day or even later, when they were harvested in a wilted condition in preference to newly-dropped leaves.It is possible that during wilting some repellent substances evaporate or become less effective, thus enhancing leaf palatability. Alternatively or in addition, changes in nutrient and water content may have rendered wilted leaves more palatable to leaf-cutting ants.     

Herbivory-induced extrafloral nectar increases native and invasive ant worker survival

Ascertaining the costs and benefits of mutualistic interactions is important for predicting their stability and effect on community dynamics. Despite widespread designation of the interaction between ants and extrafloral nectaries (EFNs) as a mutualism and over 100 years of studies on ant benefits to plants, the benefits to ants have never been experimentally quantified. The success of invasive ants is thought to be linked to the availability of carbohydrate-rich resources, though reports of invasive ant visits to EFNs are mixed. In two laboratory experiments, we compared worker survival of one native (Iridomyrmex chasei) and two invasive ant species (Linepithema humile and Pheidole megacephala) exposed to herbivorized or non-herbivorized EFN-bearing plants (Acacia saligna) or positive and negative controls. We found that non-herbivorized plants did not produce any measurable extrafloral nectar, and ants with access to non-herbivorized plants had the same survival as ants with access to an artificial plant and water (unfed ants). Ants given herbivorized plants had 7–11 times greater worker survival relative to unfed ants, but there were no differences in survival between native and invasive ants exposed to herbivorized plants. Our results reveal that ants cannot induce A. saligna extrafloral nectar production, but workers of both native and invasive ant species can benefit from extrafloral nectar as much as they benefit from sucrose.     

Small-scale disturbances spread along trophic chains: leaf-cutting ant nests, plants, aphids, and tending ants

Small-scale disturbances caused by animals often modify soil resource availability and may also affect plant attributes. Changes in the phenotype of plants growing on disturbed, nutrient-enriched microsites may influence the distribution and abundance of associated insects. We evaluated how the high nutrient availability generated by leaf-cutting ant nests in a Patagonian desert steppe may spread along the trophic chain, affecting the phenotype of two thistle species, the abundance of a specialist aphid and the composition of the associated assemblage of tending ants. Plants of the thistle species Carduus nutans and Onopordum acanthium growing in piles of waste material generated by leaf-cutting ant nests (i.e., refuse dumps) had more leaves, inflorescences and higher foliar nitrogen content than those in non-nest soils. Overall, plants in refuse dumps showed higher abundance of aphids than plants in non-nest soils, and aphid colonies were of greater size on O. acanthium plants than on C. nutans plants. However, only C. nutans plants showed an increase in aphid abundance when growing on refuse dumps. This resulted in a similar aphid load in both thistle species when growing on refuse dumps. Accordingly, only C. nutans showed an increase in the number of ant species attending aphids when growing on refuse dumps. The increase of soil fertility generated by leaf-cutting ant nests can affect aphid abundance and their tending ant assemblage through its effect on plant size and quality. However, the propagation of small-scale soil disturbances through the trophic chain may depend on the identity of the species involved.     

Diversity of ant-plant interactions: protective efficacy in Macaranga species with different degrees of ant association

The pioneer tree Macaranga in SE Asia has developed manyfold associations with ants. The genus comprises all stages of interaction with ants, from facultative relationships to obligate myrmecophytes. Only myrmecophytic Macaranga offer nesting space for ants and are associated with a specific ant partner. The nonmyrmecophytic species are visited by a variety of different ant species which are attracted by extrafloral nectaries (EFN) and food bodies. Transitional Macaranga species like M. hosei are colonized later in their development due to their stem structure. Before the colonization by their specific Crematogaster partner the young plants are visited by different ant species attracted by EFN. These nectaries are reduced and food body production starts as soon as colonization becomes possible. We demonstrated earlier that obligate ant partners can protect their Macaranga plants against herbivore damage and vine cover. In this study we focused on nonspecific interactions and studied M. tanarius and M. hosei, representing a non-myrmecophyte and a transitional species respectively. In ant exclusion experiments both M. tanarius and M. hosei suffered significantly higher mean leaf damage than controls, 37% versus 6% in M. hosei, 16% versus 7% in M. tanarius. M. tanarius offers both EFN and food bodies so that tests for different effects of these two food rewards could be conducted. Plants with food bodies removed but with EFN remaining had the lowest mean increase of herbivore damage of all experimental groups. Main herbivores on M. hosei were mites and caterpillars. Many M. tanarius plants were infested by a shootborer. Both Macaranga species were visited by various ant species, Crematogaster spp. being the most abundant. We found no evidence for any specific relationships. The results of this study strongly support the hypothesis that non-specific, facultative associations with ants can be advantageous for Macaranga plants. Food bodies appear to have lower attractive value for opportunistic ants than EFN and may require a specific dietary adaptation. This is also indicated by the fact that food body production in the transitional M. hosei does not start before stem structure allows a colonization by the obligate Crematogaster species. M. hosei thus benefits from facultative association with a variety of ants until it produces its first domatia and can be colonized by its obligate mutualist.     

Epiphyll deterrence to the leafcutter ant Atta cephalotes

Summary Epiphyll growth on leaves of the grapefruit Citrus paradisi and the understory cyclanth Cyclanthus bipartitus repelled the fungus-growing, leafcutter ant Atta cephalotes from harvesting leaves of these tropical plants. Experimental removal of epiphylls from leaves resulted in 2–3 times more herbivore damage by leafcutter ants as compared to matched leaves with epiphylls. Because of the protection from herbivore damage, host plants may derive a partial fitness benefit from association with epiphylls.     

Experimental evaluation of the impacts of two ant species on banana weevil in Uganda

The banana weevil, Cosmopolites sordidus (Germar), is an important pest of bananas. Predatory ants are increasingly being viewed as possible biological control agents of this pest because they are capable of entering banana plants and soil in search of prey. We studied ant predation on banana weevil in Uganda in crop residues and live plants in both laboratory and field experiments. Field studies with live plants used chemical ant exclusion in some plots and ant enhancement via colony transfer in others to measure effects of Pheidole sp. 2 and Odontomachus troglodytes Santschi on plant damage and densities of immature banana weevils.In crop residues, an important pest breeding site, twice as many larvae were removed from ant-enhanced plots as in control plots. In young (2 month) potted suckers held in shade houses, ant ability to reduce densities of banana weevil life stages varied with the weevil inoculation rate. At the lowest density (2 female weevils per pot), densities of eggs, larvae, and pupae were reduced by ants. At higher rates there was no effect. In older suckers (5–11 months) grown in larger containers, banana weevil densities were not affected by ants, but damage levels were reduced. In a field trial lasting a full crop cycle (30 months), we found that the ants tested reduced the density of banana weevil eggs in suckers during the crop, but did not affect larval densities in the sampled suckers. However, most larvae occur in the main banana plants, rather than associated suckers. Nevertheless, levels of damage in mature plants at harvest did not differ between Amdro-treated and ant-enhanced plots, suggesting the ant species studied were not able to provide economic control of banana weevil under our test conditions.     

Host-plant selection,diet diversity,and optimal foraging in a tropical leafcutting ant

Summary A month-long study was conducted on the comparative foraging behavior of 20 colonies of the leafcutting ant, Atta cephalotes L. in Santa Rosa National Park, Guanacaste Province, Costa Rica. The study was conducted during the middle of the wet season, when trees had mature foliage and the ants were maximally selective among species of potential host plants. The colonies always gathered leaves from more than a single tree species but on average one species constituted almost half the diet with the remaining species being of geometrically decreasing importance. Colonies exhibited greater diversity in their choice of leaves and lower constancy of foraging when the average quality of resource trees was lower, as predicted by elementary optimal foraging theory. Furthermore, the ants were more selective of the species they attacked at greater distances from the nest. However, the ants sometimes did not attack apparently palatable species, and often did not attack nearby individuals of species they were exploiting at greater distances.A classical explanation for why leafcutting ants exploit distant host trees when apparently equally good trees are nearer, is that the ants are pursuing a strategy of conserving resources to avoid long-term overgrazing pressure on nearby trees. We prefer a simpler hypothesis: (1) Trees of exploited species exhibit individual variation in the acceptability of their leaves to the ants. (2) The abundance of a species will generally increase with area and radial distance from the nest, so the probability that at least one tree of the species will be acceptable to the ants also increases with distance. (3) The ants forage using a system of trunk-trails cleared of leaf litter, which significantly reduces their travel time to previously discovered, high-quality resource trees (by a factor of 4- to 10-fold). (4) Foragers are unware of the total pool of resources available to the colony. Therefore once scouts have chanced upon a tree which is acceptable, the colony will concentrate on harvesting from that tree rather than searching for additional sources of leaves distant from the established trail.     

Effect of habitat complexity on biological control by the red imported fire ant (Hymenoptera: Formicidae) in collards

A potentially important and understudied biological control agent in US agroecosystems is the red imported fire ant, Solenopsis invicta Buren. Red imported fire ants may be particularly important biological control agents because we can manipulate their abundance with changes in habitat complexity. The effect of habitat complexity on biological control by fire ants was determined using plots of collards intercropped with white clover (complex habitat) and simple collard monocrops. The most economically significant pests of collards are larvae of the diamondback moth (DBM), Plutella xylostella (L.). Predation of DBM larvae by fire ants was more rapid and efficient in the intercrop than the monocrop. Red imported fire ants were 23% less abundant in the intercrop than the monocrop, however, suggesting that fire ants had a greater per capita effect on DBM survival in the complex habitat. Red imported fire ant predation of DBM larvae was significantly affected by larval density. Red imported fire ants also reduced the survival of leaf beetles, another economically significant pest taxa, by 45%. Furthermore, collard leaf damage tended to be inversely related to fire ant density and fire ants were more effective at reducing crop damage in the complex intercrop. Our study indicates the ability of red imported fire ants to be effective biological control agents and suggests that increasing habitat complexity can enhance red imported fire ant efficacy and herbivore control.     

Ant species confer different partner benefits on two neotropical myrmecophytes

The dynamics of mutualistic interactions involving more than a single pair of species depend on the relative costs and benefits of interaction among alternative partners. The neotropical myrmecophytes Cordia nodosa and Duroia hirsuta associate with several species of obligately symbiotic ants. I compared the ant partners of Cordia and Duroia with respect to two benefits known to be important in ant-myrmecophyte interactions: protection against herbivores provided by ants, and protection against encroaching vegetation provided by ants. Azteca spp., Myrmelachista schumanni, and Allomerus octoarticulatus demerarae ants all provide the leaves of Cordia and Duroia some protection against herbivores. However, Azteca and Allomerus provide more protection than does Myrmelachista to the leaves of their host plants. Although Allomerus protects the leaves of its hosts, plants occupied by Allomerus suffer more attacks by herbivores to their stems than do plants occupied by other ants. Relative to Azteca or Allomerus, Myrmelachista ants provide better protection against encroaching vegetation, increasing canopy openness over their host plants. These differences in benefits among the ant partners of Cordia and Duroia are reflected in the effect of each ant species on host plant size, growth rate, and reproduction. The results of this study show how mutualistic ant partners can differ with respect to both the magnitude and type of benefits they provide to the same species of myrmecophytic host.     

Induced biotic responses to herbivory and associated cues in the Amazonian ant-plant Maieta poeppigii

Ants inhabiting ant‐plants can respond to cues of herbivory, such as the presence of herbivores, leaf damage, and plant sap, but experimental attempts to quantify the dynamic nature of biotic defenses have been restricted to a few associations between plants and ants. We studied the relationship between certain features of the ant‐shrub Maieta poeppigii Cogn. (Melastomataceae) and the presence or absence of ant patrolling on the leaf surface in plants occupied by the ant Pheidole minutula Mayr (Hymenoptera: Formicidae). We also carried out field experiments to examine ant behavior following plant damage, and the potential cues that induce ant recruitment. These experiments included clipping of the leaf apex, as well as the presentation of a potential herbivore (live termite worker) and a foliar extract from Maieta on treatment leaves. The presence of ants patrolling the leaves of M. poeppigii is influenced by the number of domatia on the plant. Ant patrolling on the leaves of M. poeppigii was constant throughout a 24 h cycle, but the mean number of patrolling ants decreased from young to mature leaves, and from leaves with domatia to those without domatia. There was an overall increase in the number of ants on experimental leaves following all treatments, compared to control leaves. Visual and chemical cues associated with herbivory are involved in the induction of ant recruitment in the Maieta–Pheidole system. The continuous patrolling behavior of ants, associated with their ability to respond rapidly to foliar damage, may result in the detection and repellence/capture of most insect herbivores before they can inflict significant damage to the leaves.     

The distribution of weaver ant pheromones on host trees

The visible anal spots deposited by Oecophylla smaragdina ants have been suggested to deter ant prey, affect interspecific competition and facilitate mutualists and parasites in tracking down Oecophylla ants. I measured the density of anal spots on host trees with and without ants and tested for correlations between spot density, ant activity and the likelihood of being detected by an ant. Spots were only found on trees with ants. On ant-trees, spots were distributed throughout the trees but with higher densities in areas with high ant activity and pheromone densities were higher on twigs compared to leaves. Also there was a positive correlation between spot density and the likelihood of being detected by ants. Anal spots may thus function as reliable cues to interacting species and be an important factor in shaping the community around Oecophylla colonies. Received 31 August 2006; revised 27 April 2007; accepted 2 May 2007.     

Ant-seed interactions: combined effects of ant and plant species on seed removal patterns

Seed dispersal by ants (i.e. myrmecochory) is usually considered as a mutualism: ants feed on nutritive bodies, called elaiosomes, before rejecting and dispersing seeds in their nest surroundings. While mechanisms of plant dispersal in the field are well documented, the behaviour of the ant partner was rarely investigated in details. Here, we compared in laboratory conditions the foraging behaviour of two ant species, the omnivorous Lasius niger and the insectivorous Myrmica rubra to which seeds of two European myrmecochorous plants (Chelidonium majus and Viola odorata) were given. Ant colonies were simultaneously presented three types of items: entire seeds with elaiosome (SE), seeds without elaiosome (S) and detached elaiosomes (E). The presence of elaiosomes on seeds did not attract workers from a distance since ants first contact equally each type of items. Although ants are mass-recruiting species, we never observed any recruitment nor trail-laying behaviour towards seeds. For ants having contacted seed items, their antennation, manipulation and seed retrieval behaviour strongly varied depending on the species of each partner. Antennation behaviour, followed by a loss of contact, was the most frequent ant-seed interaction and can be considered as a “hesitation” clue. For both plant species, insectivorous Myrmica ants removed items in larger number and at higher speed than Lasius. This fits with the hypothesis of a convergence between odours of elaiosomes and insect preys. For both ant species, the small Chelidonium seeds were retrieved in higher proportion than Viola ones, confirming the hypothesis that ants prefer the higher elaiosome/diaspore-ratio. Thus, in these crossed experiments, the ant-plant pair Myrmica/Chelidonium was the most effective as ants removed quickly almost all items after a few antennations. The presence of an elaiosome body increased the seed removal by ants excepting for Myrmica which retrieved all Chelidonium seeds, even those deprived of their elaiosome. After 24 h, all the retrieved seeds were rejected out of the nest to the refuse piles. In at least half of these rejected items, the elaiosome was discarded by ants. Species-specific patterns and behavioural differences in the dynamics of myrmecochory are discussed at the light of ant ecology. Received 10 September 2007; revised 5 February 2008; accepted 5 March 2008.     

Azteca protection of Cecropia: ant occupation benefits juvenile trees

Summary In this 15 month investigation I experimentally demonstrated that sapling Cecropia aff. obtusifolia in lowland western Ecuador grow more vigorously when occupied by the ant Azteca constructor than when the ants have been removed. Thus the interaction is directly beneficial to Cecropia juveniles. The difference in growth is associated with differences in herbivory and vine cover. Removal of ants significantly increases nocturnal Coleoptera herbivory on unoccupied plants. In contrast to the influence on beetle numbers, Azteca are ineffective against Homoptera and cecidomyiid gall flies. Although ant-occupied saplings had less chewing herbivore damage throughout the study, the ants were more effective protectors in the dry season than in the rainy season, when herbivore pressure increased. In addition to reducing herbivory, Azteca efficiently remove vines from occupied saplings.     

Bigger ant colonies reduce herbivory and herbivore residence time on leaves of an ant-plant: Azteca muelleri vs. Coelomera ruficornis on Cecropia pachystachya

Summary The effect of defence force size in colonies of the ant Azteca muelleri on the time spent to localize, attack and expel the specialized herbivorous beetle Coelomera ruficornis from Cecropia pachystachya bushes was studied in an area of Atlantic forest in northeastern Brazil. Our results show that Azteca muelleri expel Coelomera ruficornis from Cecropia pachystachya and that the number of ants leaving a colony (defence force size) is negatively correlated with the residence time of an adult beetle on the plant. Colonies with larger defence forces recruited larger numbers of ants, resulting in faster herbivore discovery (r ²=0.80; n=17; P<0.001) and reduced herbivore residence time on a leaf (r ²=0.79 n=23; P<0.001) before being driven off by the ants. We also found a negative and significant relationship between herbivore damage on leaves and ant colony size (r ²=0.28; n=17; P<0.05). We conclude that larger colonies have more individuals available to patrol a plant and recruit defenders toward herbivores. This reduces the time spent to locate and expel susceptible herbivores from the plant. Since the plant probably benefits from reduced herbivory and the plant provides food for the ants, the association between Azteca muelleri and Cecropia pachystachya appears mutualistic.     

Density-dependent ant attendance and its effects on the parasitism of a honeydew-producing scale insect,Ceroplastes rubens

The intensity of attendance by a honeydew-foraging ant, Lasuis niger, on the red wax scale insect, Ceroplastes rubens, was estimated at different manipulated densities in the field. The time that individual ants were present and the total attendance time (seconds x number of ants) of ants on scale-infested twigs significantly increased as the density of C. rubens increased, i.e. ant attendance was density dependent. To determine the effects of density dependence of ant attendance on parasitism of C. rubens by Anicetus beneficus, we measured parasitism rates in the field at different density levels of C. rubens both with ant attendance and with ants excluded. Parasitism rates were higher when ants were excluded, at each density level. Although the parasitism rate significantly deceased as scale density increased, whether or not ants attended, the difference in parasitism rate between density levels was strikingly less without ant attendance. Therefore, the density-dependent decrease of parasitism rate was more pronounced with ant attendance. Mortality not due to parasitism showed density dependence in both conditions and did not change when ants were excluded. These results indicate that attending ants reduce parasitism and that, as a consequence of the density dependence of ant attendance, the efficiency of reduction of parasitism by ants is enhanced at higher densities of C. rubens.     

The urban fire ant paradox: native fire ants persist in an urban refuge while invasive fire ants dominate natural habitats

In contrast to the widespread extirpation of native fire ants (Solenopsis geminata) across southern US following the invasion by imported red fire ants (S. invicta), some residential areas of Austin form unexpected refuges for native fire ants. Ironically, these urban environments provide refuges for the native fire ants while adjacent natural habitats have been overrun by invasive fire ants. Resistance to invasive fire ants in these urban areas occurs mainly in older residential properties constructed prior to the S. invicta invasion, while more recent construction has allowed establishment by S. invicta. The invasive ability of S. invicta is often attributed to escape from parasitoids and efficient dispersal of polygyne multiple queen colonies. Here we also show the importance of landscape parameters in the invasion process, where low levels of disturbance and continuous plant cover in older residential areas form possible barriers to colonization. Dense leaf cover (high NDVI) was also found to be associated with native ant refuges. Long term residential land ownership may have resulted in lower recent disturbance levels and increased plant cover that support refuges of native fire ants.     

Honeydew collection by the invasive garden ant Lasius neglectus versus the native ant L. grandis

Honeydew collection performed by the invasive ant Lasius neglectus and by the native ant L. grandis was compared. The invasive ant collected 2.09 kg of honeydew per tree while the native ant collected 0.82 kg. The aphid Lachnus roboris was visited by both ant species. In holm oaks colonized by L. neglectus, aphid abundance tended to increase and its honeydew production increased twofold. The percentage of untended aphids was lower in holm trees occupied by L. neglectus. As tending ants also prey on insects, we estimated the percentage of carried insects. The native ant workers carried more insects than the invasive ant. Both ant species preyed mainly on Psocoptera and the rarely tended aphid, Hoplocallis picta. We conclude that the higher honeydew collection achieved by L. neglectus was the consequence of (1) its greater abundance, which enabled this ant to tend more Lachnus roboris and (2) its greater level of attention towards promoting an increase of honeydew production. Handling editor: Heikki Hokkanen     

The fitness consequences of bearing domatia and having the right ant partner: experiments with protective and non-protective ants in a semi-myrmecophyte

The fitness advantage provided by caulinary domatia to myrmecophytes has never been directly demonstrated because most myrmecophytic species do not present any individual variation in the presence of domatia and the removal of domatia from entire plants is a destructive process. The semi-myrmecophytic tree, Humboldtia brunonis (Fabaceae: Caesalpinioideae), is an ideal species to investigate the selective advantage conferred by domatia because within the same population, some plants are devoid of domatia while others bear them. Several ant species patrol the plant for extra-floral nectar. Fruit production was found to be enhanced in domatia-bearing trees compared to trees devoid of domatia independent of the ant associate. However, this domatium effect was most conspicuous for trees associated with the populous and nomadic ant, Technomyrmex albipes. This species is a frequent associate of H. brunonis, inhabiting its domatia or building carton nests on it. Ant exclusion experiments revealed that T. albipes was the only ant to provide efficient anti-herbivore protection to the leaves of its host tree. Measures of ant activity as well as experiments using caterpillars revealed that the higher efficiency of T. albipes was due to its greater patrolling density and consequent shorter lag time in attacking the larvae. T. albipes also provided efficient anti-herbivore protection to flowers since fruit initiation was greater on ant-patrolled inflorescences than on those from which ants were excluded. We therefore demonstrated that caulinary domatia provide a selective advantage to their host-plant and that biotic defence is potentially the main fitness benefit mediated by domatia. However, it is not the sole advantage. The general positive effect of domatia on fruit set in this ant–plant could reflect other benefits conferred by domatia-inhabitants, which are not restricted to ants in this myrmecophyte, but comprise a large diversity of other invertebrates. Our results indicate that mutualisms enhance the evolution of myrmecophytism.     

Foliar nectar production and ant activity on a neotropical tree,Ochroma pyramidale

Summary In second growth forest in lowland Costa Rica, ants forage at the foliar nectaries of juvenile Ochroma pyramidale. The relationship between leaf development, foliar nectar production and ant visitation indicates that nectar secretion and ant maintenance are greatest following rapid leaf expansion. Nectar measurements in the glasshouse corroborate field measurements showing that nectar production on a sapling is continuous through time and correlated with distribution and abundance of ants within a sapling. The presence of two nectary types, leaf vein and petiolar, on the leaves of O. pyramidale results in the continual maintenance of ants on the leaf undersurface. Nectar production of a sapling increases with increasing leaf area resulting in greater number of ants per sapling. Energetic costs of nectar production and ant maintenance appear low, representing about one per cent of the total energy invested in leaves.Spatial and diurnal patterns of ant activity changed very little over the study period. Removal and exclusion of ants from saplings results in the utilization of foliar nectar by trigonid bees. A significant difference in leaf damage between ant-visited and unvisited saplings, coupled with ant behavioral characteristics, is consistent with the hypothesis that ants act as antiherbivore agents on Ochroma.     

When are ant-attractant devices a worthwhile investment? Vicia faba extrafloral nectaries and Lasius niger ants

Most studies aiming to determine the beneficial effect of ants on plants simply consider the effects of the presence or exclusion of ants on plant yield. This approach is often inadequate, however, as ants interact with both non-tended herbivores and tended Homoptera. Moreover, the interaction with these groups of organisms is dependent on ant density, and these functional relationships are likely to be non-linear. A model is presented here that segregates plant herbivores into two categories depending on the sign of their numerical response to ants (myrmecophiles increase with ants, non-tended herbivores decline). The changes in these two components of herbivores with increasing ant density and the resulting implications for ant-plant mutualisms are considered. It emerges that a wide range of ant densities needs to be considered as the interaction sign (mutualism or parasitism) and strength is likely to change with ant density. The model is used to interpret the results of an experimental study that varied levels of Aphis fabae infestation and Lasius niger ant attendance on Vicia faba bean plants. Increasing ant density consistently reduced plant fitness and thus, in this location, the interaction between the ants and the plant can be considered parasitic. In the Vicia faba system, these costs of ants are unlikely to be offset by other beneficial agents (e.g., parasitoids), which also visit extrafloral nectaries.