Effects of ant exclusion during outbreaks of a defoliator and a sap-sucker on birch

1. The influence of a wood ant, Formica aquilonia, on the defoliation of the white birch, Betula pubescens, and on its invertebrate community was studied in ant-exclusion experiments during two outbreaks, the decline phase of the autumnal moth, Epirrita autumnata, and the peak year of the birch aphid, Euceraphis punctipennis, not tended by ants. 2. The numbers of the wood ant and Symydobius oblongus, a tended aphid, in birch foliage showed a strong positive correlation, and the former decreased rapidly when the distance from the ant mound increased, confirming that there was a distance-related gradient in arboreal ant predation. There may have been a parallel reduction in soil amelioration by ants through nest construction and food gathering. 3. The application of a glue ring around the trunk excluded ants totally from the canopy, inhibiting predation but not fertilization. 4. Ant-exclusion resulted in a 90–95% reduction in the growth of tended aphid colonies by mid-season. 5. Outbreak populations of the autumnal moth and the birch aphid were reduced by 45–67% and 77%, respectively, in control trees and correlated negatively with ant numbers. 6. The total percentage of leaf area damaged by moth larvae was 34% lower in ant-foraged than in unforaged trees. 7. Neither the distance from the ant mound nor its interaction with the glue treatment had any effect on the herbivores or folivory, indicating that the possible soil-ameliorating effect was weaker than predation, which reduced herbivore numbers at every distance studied (4–20 m). 8. Predation by ants also affected the abundance of syrphid larvae, predatory on both aphids, the percentage parasitism by a wasp, Aleiodes testaceus, on the autumnal moth, and thereby (or directly) its age (size) distribution. The presence of ants had no influence on spider abundance. 9. It is concluded that predation rather than soil amelioration is likely to be the reason why the degree of folivory and, during serious outbreaks, the mortality of trees are lower in the vicinity of wood-ant mounds.     

Sources of variation in the incidence of ant–aphid mutualism in boreal forests

• 1 The mutualism between wood ants of the Formica rufa group and aphids living in the canopy of trees is a widespread phenomenon in boreal forests, and it can affect tree growth. However, not all trees in the forest are involved in this interaction.
• 2 To assess the incidence of host trees involved in this ant–aphid mutualism and its spatial distribution in boreal forests, we inventoried sample plots with a radius of 10–15 m around wood ant mounds in 12 forest stands of two age classes (5–12‐year‐old sapling stands and 30–45‐year‐old pole stands) and two dominant tree species (Scots pine and silver birch) in Eastern Finland from 2007 to 2009.
• 3 The proportion of trees visited by ants out of all trees on the individual study plots were in the range 4–62%, and 1.5–39% of the trees on the plots were consistently visited by ants during all 3 years. The percentage of host trees increased with the ant mound base area on the plots. Trees visited by ants were larger and closer to the mound than trees not visited by ants. Within the group of visited trees, more ants were found on bigger trees and on trees close to the ant mounds.
• 4 Extrapolated from plot to stand level, we estimated that 0.5–6.6% of the trees were host trees in at least one of the three study years, and that only 0.01–2.3% of all the trees were consistently visited by ants during all 3 years. It is concluded that ant–aphid mutualism is a minor occurrence at the stand level.

     

Green islands — predation not nutrition

Summary The existence of green islands around wood-ant nests in otherwise damaged birch forests has been explained by two alternative hypotheses: (1) predation by ants protects the trees against defoliators, and (2) the ants ameliorate tree vigor by concentrating soil nutrients. The size of green islands and the nitrogen content of the soil and foliage do not support the nutrition hypothesis. These data and general knowledge about the foraging strategy of wood ants are consistent with the predation hypothesis.     

Interactive effects of soil-dwelling ants, ant mounds and simulated grazing on local plant community composition

Interactions between aboveground vertebrate herbivores and subterranean yellow meadow ants (Lasius flavus) can drive plant community patterns in grassland ecosystems. Here, we study the relative importance of the presence of ants (L. flavus) and ant mounds under different simulated grazing regimes for biomass production and species composition in plant communities. We set up a greenhouse experiment using intact soil cores with their associated vegetation.We found that plant biomass production in the short term was affected by an interaction between simulated grazing (clipping) and ant mound presence. Clipping homogenized production on and off mounds, while in unclipped situations production was higher off than on mounds. During the experiment, these differences in unclipped situations disappeared, because production on unclipped mounds increased. Plant species richness was on average higher in clipped treatments and patterns did not change significantly over the experimental period. Plant community composition was mainly affected by clipping, which increased the cover of grazing-tolerant plant species. The actual presence of yellow meadow ants did not affect plant community composition and production.We conclude that the interaction between ant mounds and clipping determined plant community composition and biomass production, while the actual presence of ants themselves was not important. Moreover, clipping can overrule effects of ant mounds on biomass production. Only shortly after the cessation of clipping biomass production was affected by ant mound presence, suggesting that only under low intensity clipping ant mounds may become important determining plant production. Therefore, under low intensity grazing ant mounds may drive the formation of small-scale plant patches.     

Colony success of the submissive ant Formica fusca within territories of the dominant Formica polyctena

Abstract. 1. Twenty-three nests of the submissive ant Formica fusca L. were sampled in two adjacent territories of the dominant wood ant Formica polyctena Först. The nests were dug up at different distances from the wood-ant mounds. Distance is assumed to be inversely related to the extent of disturbance of F. fusca by F. polyctena.
2. Colony success of F. fusca was assayed by counting the numbers of workers (colony size), and worker and sexual offspring in a nest. For individual size, the head width and dry weight of fifty workers per nest were measured.
3. Colony size correlated positively with the numbers of worker and sexual offspring in one territory, but only with worker offspring in the other.
4. Distance from the wood-ant mound correlated positively with colony size and numbers of worker and female offspring in one territory. In the other territory distance correlated positively with number of females, but negatively with number of worker offspring. No significant correlations between distance and the size measurements were obtained.
5. Because of its nonaggressive behaviour F. fusca may nest fairly close to a wood-ant mound but is likely to show reduced abundances in terms of nest density and forager number, and, ultimately, lowered fitness.     

Nest mounds of red wood ants (Formicaaquilonia): hot spots for litter-dwelling earthworms

A previously undocumented association between earthworms and red wood ants (Formicaaquilonia Yarr.) was found during an investigation of the influence of wood ants on the distribution and abundance of soil animals in boreal forest soil. Ant nest mounds and the surrounding soil of the ant territories were sampled. The ant nest mound surface (the uppermost 5-cm layer) harboured a much more abundant earthworm community than the surrounding soil; the biomass of the earthworms was about 7 times higher in the nests than in the soil. Dendrodrilusrubidus dominated the earthworm community in the nests, while in soils Dendrobaenaoctaedra was more abundant. Favorable temperature, moisture and pH (Ca content), together with abundant food supply (microbes and decomposing litter) are likely to make a nest mound a preferred habitat for earthworms, provided that they are not preyed upon by the ants. We also conducted laboratory experiments to study antipredation mechanisms of earthworms against ants. The experiments showed that earthworms do not escape predation by avoiding contact with ants in their nests. The earthworm mucus repelled the ants, suggesting a chemical defence against predation. Earthworms probably prevent the nest mounds from becoming overgrown by moulds and fungi, indicating possible mutualistic relationships between the earthworms and the ants. Received: 21 November 1996 / Accepted: 3 April 1997     

Contributions of female oviposition patterns and larval behavior to group defense in conifer sawflies (hymenoptera: diprionidae)

We studied the effects of adult oviposition and larval interactions on the defensive potential of gregarious behavior in conifer sawflies. Aggregation size and distribution initially reflected adult host plant selection and oviposition behavior. The contagious distribution of egg clusters resulted in part from the utilization of individual trees by multiple females, and of single host shoots by several females. Trees with the greates degree of prior defoliation received the most eggs, even though the potential for larval crowding made resource depletion possible. Foliar monoterpene and nitrogen contents explained only a small proportion of variability in female host utilization. Conifer needle architecture restricted the size of larval subgroups within aggregations, and limited the degree of defensive cohesiveness between subgroups. Subgroup turnover was frequent and independent of local food depletion. Risk of predation from wood ants varied with larval aggregation size and predator foraging level. When ant activity was high, large aggregations suffered greater numerical losses, but showed lower per capita predatory risk, than small groups. Results suggest that female oviposition patterns are influenced in part by the defensive benefits gained by offspring in large aggregations. Against ants, dilution effects and defensive synchrony due to gregariousness contribute to the overall antipredator strategy of sawfly larvae.     

三种植物源药剂对红火蚁的防治效果评估及其对本地蚂蚁群落影响

红火蚁Solenopsis invicta Buren是一种危险性入侵害虫,可对人类健康、农林业生产、公共安全和生态系统产生严重的负面影响。目前,化学农药是防治红火蚁的主要方法。然而,大多数化学农药并不适用于有机农场、水产养殖区等生态敏感区。为筛选出用于生态敏感区红火蚁防治的药剂,本文结合室内和田间试验评估了植物源杀虫剂苦参碱Matrine、鱼藤酮Rotenone和除虫菊素Pyrethrin对红火蚁的防治效果。室内试验结果表明,苦参碱、鱼藤酮和除虫菊素对红火蚁均具有较高的毒性。田间试验表明,25.00 mg/L和12.50 mg/L苦参碱、250.00 mg/L 和125 mg/L鱼藤酮、75.00 mg/L和37.50 mg/L除虫菊素灌巢处理,红火蚁蚁巢减退率分别达到100%、81.25%、87.50%、79.17%、71.73%和66.67%。然而,除25.00 mg/L苦参碱外,其他处理均导致在被灌巢的蚁巢周围短期内出现新蚁巢。灌巢处理后,处理区的蚂蚁多样性指数和均匀度指数升高,优势度指数下降。综上所述,苦参碱、鱼藤酮和除虫菊素适合灌巢处理防治红火蚁,其中0.5%苦参碱可溶性液剂能有效降低红火蚁种群数量,使得蚁巢减退率达100%,同时对本地蚂蚁群落没有负面影响,因此该商品制剂在红火蚁防治方面具有较好的开发潜力。     

Host tree architecture mediates the effect of predators on herbivore survival

Abstract.  1. Vegetation structural complexity is an important factor influencing ecological interactions between different trophic levels. In order to investigate relationships between the architecture of trees, the presence of arthropod predators, and survival and parasitism of the autumnal moth Epirrita autumnata Borkhausen, two sets of experiments were conducted.
2. In one experiment, the architectural complexity of mountain birch was manipulated to separate the effects of plant structure and age. In the other experiment the trees were left intact, but chosen to represent varying degrees of natural complexity. Young autumnal moth larvae were placed on the trees and their survival was monitored during the larval period.
3. The larvae survived longer in more complex trees if predation by ants was prevented with a glue ring, whereas in control trees smaller canopy size improved survival times in one experiment. The density of ants observed in the trees was not affected by canopy size but spider density was higher on smaller trees. The effect of canopy structure on larval parasitism was weak; larger canopy size decreased parasitism only in one year. Until the fourth instar the larvae travelled shorter distances in trees with reduced branchiness than in trees with reduced foliage or control treatments. Canopy structure manipulation by pruning did not alter the quality of leaves as food for larvae.
4. The effect of canopy structure on herbivore survival may depend on natural enemy abundance and foraging strategy. In complex canopies herbivores are probably better able to escape predation by ambushing spiders but not by actively searching ants.     

Effects of habitat fragmentation on the three-way interaction among ants,aphids and larvae of the giant purple emperor, <Emphasis Type="Italic">Sasakia charonda</Emphasis> (Hewitson), a near-threatened butterfly

Survival rates of both early and middle instar larvae of the nymphalid butterfly, Sasakia charonda, were estimated to be lowest on test trees planted in a meadow (site A), intermediate in a small, narrow secondary deciduous broadleaf forest (small patch, site B) and highest in a large secondary deciduous broadleaf forest (large forest, site C). The larval mortality rates due to predation by tree-climbing predators from the ground (tree climbing predator) such as ants and the larvae of carabids were estimated to be greater at sites A and B than those at site C. The number of predatory ants climbing test trees was significantly greater at sites A and B than at site C, and the ants harvested honeydew from aphids living on tree leaves at those two sites. Aphid densities were significantly higher on trees at sites A and B than at site C, and aphid densities and numbers of predatory ants were significantly and positively correlated at sites A and B. In an experiment controlling aphid density per branch on test trees, the numbers of ants and the mortality rates of S. charonda larvae were greater on branches with high aphid densities than on those with low aphid densities at both sites A and B. These results suggest that the aphid density per host tree was higher in the meadow and the small patch than in the large forest; at both sites these higher aphid densities attracted higher numbers of predatory ants to test trees, and as a result, mortality rates of S. charonda larvae were increased.     

Interaction between the non-myrmecochorous herb Galium aparine and the ant Formica polyctena

Non-myrmecochorous plant species are able to colonize Formica polyctena mounds. Galium aparine has been chosen as a model to study this kind of ant-plant association. In order to understand benefits to ants and plants, we have studied removal rates of fruits by ant workers, and compared soil seed material, seedling pattern and development of plants of the nest versus forest.Ant workers remove fruits at very low rates; removal time for half of the sample of 10 seeds was on average 5 h. Removed fruits were usually lost within 20–30 cm distance of the ant route. The soil samples from the nests contained a higher number of fruits, than those from randomly chosen sites in the forest. Seeds start to germinate in late August; on the F. polyctena mounds significantly earlier than at surrounding places. The seedling density was higher within experimental quadrats located on mound sites especially in late September–early October. The highest density of seedlings was on the slope of the mound. Plants from mounds are significantly longer, have a higher number of nodes during all stages of development and a higher mortality rate than that at surrounding places.The interactions between G. aparine and F. polyctena ants are asymmetrical. In spite of higher mortality rate of seedlings on mounds, ants provide some benefits to the plant. Firstly, workers are additional dispersal agents for G. aparine fruits. Furthermore, the soil peculiarities of ant mounds are beneficial for seed germination and seedling growth. G. aparine does not provide ants with energetically valuable food like plants with elaiosome-bearing seeds. Moreover, the high density of seedlings and plants on the mounds shades the nest.     

An experiment on ant predation in soil using a new bait trap method

To examine which species of ant are underground predators, we invented a simple bait trap using live mealworms as bait. A total of 75 bait traps were buried in soil at various depths (10, 20, 30 and 50 cm) to investigate the vertical distribution of predatory ants. The ants captured in approximately 50% of the traps were mostlyTetramorium caespitum andSolenopsis japonica. Solenopsis japonica was captured at greater depths (>30 m) thanT. caespitum. Eight kinds of soil invertebrate (the earthworm, wood louse, pill bug, millipede, adults of the dung beetle and carabid beetle, larvae of a cetoniine beetle and the earwig), placed in traps at a depth of 10 cm, varied greatly in attractiveness and vulnerability to ant predation. The earthworm and the wood louse were both attractive and vulnerable to ant predation, whereas the pill bug, the dung beetle and the millipede were attractive but not vulnerable. The larvae of a cetoniine beetle, the carabid beetle and the earwig were mostly unattractive.     

The size–distance relationship in the wood ant Formica rufa

1. The size–distance relationship among honeydew‐collecting foragers of the red wood ant Formica rufa was investigated. Within the colony territory, the size (as measured by head width) and fresh weight of samples of foragers were determined for ants ascending and descending trees near, and farther from, the central nest mound. 2. The mean size of the ants was significantly higher at far trees than at near trees in six out of the seven colonies investigated, confirming the general presence of the size–distance relationship. 3. In three colonies, a load–distance relationship was also found. For a given head width, honeydew‐carrying ants descending far trees were significantly heavier than those descending near trees (i.e. they were carrying heavier loads from trees farther away from the central nest mound). 4. This is the first time that both load–distance and size–distance relationships have been reported in foraging workers from the same ant colony. 5. The combined effects of these characteristics suggest that colony foraging efficiency is enhanced by far trees being visited by the larger workers that then return with heavier loads of honeydew.     

Positive interactions between desert granivores: localized facilitation of harvester ants by kangaroo rats

Facilitation, when one species enhances the environment or performance of another species, can be highly localized in space. While facilitation in plant communities has been intensely studied, the role of facilitation in shaping animal communities is less well understood. In the Chihuahuan Desert, both kangaroo rats and harvester ants depend on the abundant seeds of annual plants. Kangaroo rats, however, are hypothesized to facilitate harvester ants through soil disturbance and selective seed predation rather than competing with them. I used a spatially explicit approach to examine whether a positive or negative interaction exists between banner-tailed kangaroo rat (Dipodomys spectabilis) mounds and rough harvester ant (Pogonomyrmex rugosus) colonies. The presence of a scale-dependent interaction between mounds and colonies was tested by comparing fitted spatial point process models with and without interspecific effects. Also, the effect of proximity to a mound on colony mortality and spatial patterns of surviving colonies was examined. The spatial pattern of kangaroo rat mounds and harvester ant colonies was consistent with a positive interspecific interaction at small scales (<10 m). Mortality risk of vulnerable, recently founded harvester ant colonies was lower when located close to a kangaroo rat mound and proximity to a mound partly predicted the spatial pattern of surviving colonies. My findings support localized facilitation of harvester ants by kangaroo rats, likely mediated through ecosystem engineering and foraging effects on plant cover and composition. The scale-dependent effect of kangaroo rats on abiotic and biotic factors appears to result in greater founding and survivorship of young colonies near mounds. These results suggest that soil disturbance and foraging by rodents can have subtle impacts on the distribution and demography of other species.     

Impact of predatory ants on two lepidopteran insect pests in Indian cauliflower agroecosystems

Abstract:  The impact of predatory, agroecosystem-nesting ant species on the larval stages of the two major lepidopteran pests, the diamondback moth (DBM), Plutella xylostella and the black cutworm (BC), Agrotis ipsilon , in cauliflower agroecosystem in Uttar Pradesh, India, was investigated in farmers' fields. Field observations revealed that six ant species, the most important being Pheidole sp., preyed on the crop-damaging stages of DBM and BC. Bait experiments with larvae of DBM and BC, placed randomly on the ground, demonstrated significant active retrieval by Pheidole sp. Investigations on larval density of both the insect pests on cauliflower crops, located within 1–5 m distances from nests of Pheidole sp., revealed a significant increase in the level of infestation of the crop plants with increasing distance from the ant nests. While a significant negative correlation was found between infested plants with low larval densities, a positive correlation was found between plants with high larval densities with regard to the distance of crop plant location from the ant nest. Ant-inclusion and -exclusion experiments under field conditions further confirmed the role of predatory ants, with Pheidole sp. having the major impact in the direct retrieval of the lepidopteran larvae from the crop plants. A significant difference was found in the leaf area damaged by BC larvae between the ant-included and -excluded potted plants during the 6-h observation period. Thus, this study provides direct evidence of the role of ants in actively searching and directly preying on the larval stages of DBM and BC, on cauliflower plants. Moreover, we demonstrate that in ephemeral, annual cropping agroecosystems, predatory ants are important natural enemies of lepidopteran pests, playing a significant role in plant protection.     

Impact of Formica exsecta Nyl. on seed bank and vegetation patterns in a subalpine grassland ecosystem

The mound building ant Formica exsecta Nyl. is widely distributed in grassland ecosystems of the Central European Alps. We studied the impact of these ants on seed bank and vegetation patterns in a 11 ha subalpine grassland, where we counted over 700 active ant mounds. The mounds showed a distinct spatial distribution with most of them being located in tall‐grass, which was rarely visited by ungulates (red deer; Cervus elaphus L.). Heavily grazed short‐grass, in contrast, seemed to be completely avoided by ants as only few mounds were found in this vegetation type. The species composition of the ant mound and grassland seed banks was quite similar, i.e. from 15 common plant species 12 were found in both seed bank types. We found the same proportions of myrmecochorous seeds in ant mound and grassland soil samples. In contrast, the number of seeds was 15 times higher in mound compared with the grassland soil samples. Also, the vegetation growing on ant mounds significantly differed from the vegetation outside the mounds: graminoids dominated on ant mounds, herbaceous and myrmecochorous species in the grassland vegetation. We found significant continuous changes in vegetation composition on gradients from the ant mound centre to 1 m away from the mound edge. Overall, F. exsecta was found to have a considerable impact on seed bank and vegetation patterns in the grassland ecosystem studied. These insects not only altered grassland characteristics in the close surrounding of their mounds, but also seem to affect the entire ecosystem including, for example, the spatial use of the grassland by red deer.     

Effects of group size and pine defence chemicals on Diprionid sawfly survival against ant predation

The defence chemicals and behavioural adaptations (gregariousness and active defensive behaviour) of pine sawfly larvae may be effective against ant predation. However, previous studies have tested their defences against very few species of ants, and few experiments have explored ant predation in nature. We studied how larval group size (groups of 5 and 20 in Neodiprion sertifer and 10, 20 and 40 in Diprion pini) and variation in levels of defence chemicals in the host tree (Scots pine, Pinus sylvestris) affect the survival of sawfly larvae. Food preference experiments showed that ants do eat sawfly larvae, although they are not their most preferred food item. According to our results, ant predation significantly increases the mortality rate of sawfly larvae. Larval mortality was minor on pine tree branches where ant traffic was excluded. We also found that a high resin acid concentration in the host tree significantly decreased the mortality of D. pini larvae when ants were present. However, there was no such relationship between the chemical concentrations of the host tree and larval mortality for N. sertifer. Surprisingly, grouping did not help sawfly larvae against ant predation. Mortality risk was the same for all group sizes. The results of the study seemingly contradict previous understanding of the effectiveness of defence mechanisms of pine sawfly against ant predation, and suggest that ants (Formica exsecta in particular) are effective predators of sawfly larvae.An erratum to this article can be found at     

Susceptibility and behavioral response of red imported fire ant (Hymenoptera: Formicidae) to selected entomogenous nematodes (Rhabditida: Steinernematidae & Heterorhabditidae).

Pathogenicity of infective juveniles of selected Steinernema spp. and Heterorhabditis spp. toward developing and reproductive stages of the red imported fire ant, Solenopsis invicta Buren, was tested under laboratory conditions. At 10(3)-10(5) infective juveniles per Petri dish, mortality of reproductive larvae, pupae, and alates ranged from 28 to 100% at higher doses after 96 h at 23-25 degrees C. Steinernema carpocapsae All was the most consistent species tested; this nematode caused mortality of fire ant larvae, pupae, and alates of 82-94, 64-96, and 38-99%, respectively. Although not susceptible to nematode infection, worker ants vigorously preened nematodes from brood, alates, and themselves. In a field study, S. carpocapsae (5 x 10(6) and 2 x 10(6) drench, 2 x 10(6) infective juvenile infection) was applied to active fire ant mounds in 3.8-liter suspensions. Hydramethylnon (75 ml), a water drench, a water injection, and untreated fire ant mounds were marked and treated. Overall activity in mounds treated with nematodes of hydramethylnon ranged from 40 to 48%. Satellite mound activity accounted for 32-44% of overall activity in mounds treated with nematodes 2 wk after treatment. However, 6 wk after treatment, activity in mounds treated with hydramethylnon was 44%; activity of mounds treated with nematodes ranged from 52 to 80%. Satellite mound activity accounted for 0-24% of overall activity. Whereas a soil drench of S. carpocapsae showed potential as a control method for the red imported fire ant, colony relocation after nematode treatment could limit overall efficacy unless application techniques are developed to overcome or take advantage of the movement.     

Interactions of the ant Crematogaster scutellaris with the fig/fig wasp mutualism

Abstract. 1. A classical example of specialised pollination mutualisms is the relationship between fig trees and their pollinating wasps, in which each partner depends completely on the other for its reproduction; however the fig/fig wasp association is also the target of a great diversity of other species, ranging from specialised parasites to opportunistic foragers, among them ants.
2. The ant community and the sources of ant attraction observed on the Mediterranean fig tree Ficus carica were characterised.
3. A guild of ants attracted by homopterans tended on the plant was distinguished from a second guild composed of two co-dominant ant species ( Crematogaster scutellaris and Pheidole pallidula ) that prey mostly on pollinating wasps, abundant during certain parts of the fig cycle.
4. Foraging workers of C. scutellaris search for prey on the fig inflorescence (syconium), capturing pollinating wasps mostly at the peak of wasp emergence and at a rate estimated to reach 600 prey per day for an entire tree.
5. Detailed study of the predatory sequences displayed under experimental conditions showed that ant workers captured 100% of the pollinating wasps offered, while they captured only 5.5% of the parasitoid wasp specific to the pollinator. The respective impacts of the interaction on ants and on the figs are discussed, as well as several behavioural traits of predation by the dominant ant on pollinators.     

Predation risk alters interactions among species: competition and facilitation between ants and nesting birds in a boreal forest

Although interactions between species are often assumed to be fixed, theory and empirical evidence suggest that they may be quite variable, changing in the presence of other species or environmental conditions. The interaction between ants and nesting birds exhibits such variability, ants sometimes being predators of bird nests and other times protectors of them. Hypothesizing that predation risk might be a critical factor in altering the interaction of ants with birds, I investigated the interaction of wood ants Formica aquilonia with nesting birds under different levels of predation risk. In a controlled field experiment, I allowed tits ( Parus major, P. caeruleus, P. ater ) and pied flycatchers ( Ficedula hypoleuca ) to select nest boxes in trees with ants (ant trees) or trees without ants. I found that birds usually nested in trees without ants, apparently to avoid the danger of injury from encounters with ants. Nesting in ant trees occurred mainly in the habitat where risk of predation was highest (along the forest edge), and with the bird taxa that lost nests most frequently in trees without ants (tits). Tits nesting on the forest edge achieved significantly greater nesting success, and fledged significantly more young, in ant trees compared with trees without ants. As the season progressed, ant traffic increased in trees without nesting birds, but decreased in trees with nesting birds, indicating that the outcome of interference competition between ants and nesting birds was reversed under increased predation risk. These results support the idea that predation risk can shift species interactions from predominately competitive processes to predominately facilitative processes.