Do seed harvesting ants threaten the viability of a critically endangered non-myrmecochorous perennial plant population? A complex interaction

This study analyzes the fate of seeds removed by ants, as well as the seed supply, seedling survival, and the ability to form persistent seed banks in the soil, in a critically endangered population of the non-myrmecochorous perennial halophyte Helianthemum polygonoides, in order to determine the extent to which seed removal by ants represents a real bottleneck for recruitment and thus a threat for long-term population viability. Apparently, the impact of seed-harvester ants was dramatic: the primary seed shadow was reduced by up to 93% by the activity of Messor bouvieri and Aphaenogaster dulcineae. However, M. bouvieri lost 7% of fruits collected by dropping them on trunk trails, mostly under plant cover, the most propitious microenvironment for recruitment. In addition, some seeds were mistakenly rejected in refuse piles, although here recruitment was extremely low. A significant fraction of seeds (≈40%) remained viable in the soil for over 2 years. In addition, H. polygonoides had the ability to form short-term persistent seed banks, in spite of the intense ant seed removal. In general, seedling recruitment was very low, denoting the scarcity of safe sites in the habitat. Overall, the data corroborate that seed removal by ants, although highly intense, does not compromise the viability of perennial-plant populations, because seed supply and seed reserve in the soil are enough to exploit current and future safe sites. A population viability analysis demonstrated that other threats affecting the survival of reproductive H. polygonoides plants actually constitute the real risk for the conservation of this critically endangered species.     

Utilization on extrafloral nectaries and fruit domatia of Canavalia lineata and C. cathartica (Leguminosae) by ants

Utilization of the extrafloral nectaries (EFNs) and fruits of Canavalia lineata and C. cathartica by ants was investigated at 30 sites in Japan. The fruits of C. lineata and C. cathartica were inhabited by five and eight ant species, respectively. Ant nesting periods and their utilization of EFNs differed between C. lineata and C. cathartica. Canavalia lineata flowers once a year, and periods of EFN-utilization and fruit-nesting by ants do not overlap. The fruit-nesting ants on C. lineata seem to invade the plant from the holes made by moth larvae or breaches made by decay. The ants nesting on the fruits of C. lineata may defend the plant against seed herbivores because they feed on moth larvae. Canavalia cathartica flowers several times over a year, and fruits are found throughout the year; therefore, periods of EFN-utilization and fruit-nesting by ants are overlapped. Canavalia cathartica offers year-round nesting sites and food for ants, and therefore may receive a higher defensive effect from ants than C. lineata. Handling editor: Graham Stone.     

Invasive interactions: can Argentine ants indirectly increase the reproductive output of a weed?

The direct and indirect interactions of invasive ants with plants, insect herbivores, and Hemiptera are complex. While ant and Hemiptera interactions with native plants have been well studied, the effects of invasive ant–scale insect mutualisms on the reproductive output of invasive weeds have not. The study system consisted of Argentine ants (Linepithema humile), boneseed (Chrysanthemoides monilifera monilifera), and sap-sucking scale insects (Hemiptera: Saissetia oleae and Parasaissetia nigra), all of which are invasive in New Zealand. We examined the direct and indirect effects of Argentine ants on scale insects and other invertebrates (especially herbivores) and on plant reproductive output. Argentine ants spent one-third of their time specifically associated with scale insects in tending behaviours. The invertebrate community was significantly different between uninfested and infested plants, with fewer predators and herbivores on ant-infested plants. Herbivore damage was significantly reduced on plants with Argentine ants, but sooty mould colonisation was greater where ants were present. Herbivore damage increased when ants were excluded from plants. Boneseed plants infested with Argentine ants produced significantly more fruits than plants without ants. The increase in reproductive output in the presence of ants may be due to increased pollination as the result of pollinators being forced to relocate frequently to avoid attack by ants, resulting in an increase in pollen transfer and higher fruit/seed set. The consequences of Argentine ant invasion can be varied; not only does their invasion have consequences for maintaining biodiversity, ant invasion may also affect weed and pest management strategies.     

Pollen viability reduction as a potential cost of ant association for Acacia constricta (Fabaceae)

Field studies investigating the impact of ants on the reproduction of plants bearing extrafloral nectaries have traditionally focused on seed production, a component of female fitness. The purpose of this study was to test whether ants can affect the pollen viability, a component of male fitness, when they visit flowers of the shrub Acacia constricta. Acacia constricta inflorescences hand-pollinated with flowers over which Formica perpilosa ants had crawled set significantly fewer seed pods than inflorescences hand-pollinated by control flowers that had no contact with ants. Many ant species secrete antibiotic substances onto the integument that render pollen inviable, and these secretions are probably the mechanism for reduced pollen viability in this study. The ratio of seed pods produced by self-pollinated inflorescences to those produced by cross-pollinated inflorescences was 0.16, indicating that A. constricta is largely self-incompatible. Because F. perpilosa workers forage primarily on the acacia tree under which they nest, they are unlikely to serve as efficient vectors of outcrossing. Previous work showed that A. constricta shrubs with F. perpilosa ants produce approximately twice as many seeds as similarly sized plants not so associated. The results indicate that association with F. perpilosa could cause a reproductive trade-off for A. constricta: benefits to female function may be accompanied by costs to male function. Selection to discourage ant visitation to flowers may have affected the pollination biology of this and other ant-associated plant species.     

Plant quality and the presence of beneficiaries govern the larval distribution of the critically endangered Sinai Baton Blue butterfly (Pseudophilotes sinaicus)

The larval distribution of herbivorous insects play an important role in their development and hence future fitness. Here we study larval distribution of the critically endangered Sinai Baton Blue butterfly, Pseudophilotes sinaicus, which feeds exclusively on the buds and flowers of a single host plant, also endangered, the near-endemic Sinai thyme, Thymus decussatus. We studied the larval distribution over 131 plants, recording the size, quality and phenological stage of the plants along with the presence of beneficiaries. Larvae were found on plants with a high number of flowers, a relatively advanced flowering phenology and tending ants. This highlights the importance of the vitality and quality of the host plant to the spatial distribution of the Sinai Baton Blue. Future conservation plans might concentrate on improving the quality and quantity of the host plant in order to increase resources for this narrowly endemic species.     

Vertebrate Fruit Removal and Ant Seed Dispersal in the Neotropical Ginger Renealmia alpinia (Zingiberaceae)*

Plants frequently display fruit characteristics that support multiple seed‐dispersal syndromes. These ambiguous characteristics may reflect the fact that seed dispersal is usually a complex process involving multiple dispersers. This is the case for the Neotropical ginger Renealmia alpinia (Zingiberaceae). It was originally suggested that the aromatic fruits of R. alpinia located at the base of the plant are adapted for terrestrial mammal seed dispersal. However, the dark‐purple coloration of the fruits and bright orange aril surrounding the seeds suggest that birds may play a role in R. alpinia seed dispersal. At La Selva Biological Station, Costa Rica, we used camera traps to record vertebrate visits to infructescences of R. alpinia. Most visitors were toucans and aracaris (Ramphastidae). However fruits were also removed by terrestrial mammals (coatis and armadillos). In addition to vertebrate fruit removal, some of the fruits dehisce and the seeds that fall on the ground are dispersed by ants. Fruitfall traps showed that 77 percent of fruits are removed by vertebrates. However, 15 percent of fruits fall to the base of parent plants to be potentially dispersed by ants. Experiments using a laboratory ant colony showed that ants are effective seed dispersers of R. alpinia. Ant seed manipulation increased germination success and reduced time to germination. In conclusion, primary seed dispersal in the Neotropical ginger R. alpinia is mostly performed by birds, additionally ants are effective dispersers at short distances. Seed dispersal in R. alpinia is a complex process involving a diverse array of dispersal agents.     

A mutualism's indirect costs: the most aggressive plant bodyguards also deter pollinators

Plant defenses against herbivores may be costly if they exclude mutualists. Here, I test the hypothesis that aggressive ant bodyguards of plants deter pollinators, and explore mechanisms by which Ferocactus wislizeni , an extrafloral nectary bearing cactus, limits conflicts between its pollinators and bodyguards. Flower visitation by ants and pollinating bees differed among plants tended by four different ant species. The ant species most rarely found in flowers showed the strongest aversion to F. wislizeni flower petals in laboratory assays, suggesting that those structures may include an ant-deterrent. Species-specific estimates of mean ant abundance within flowers and aggressiveness towards other arthropods were used to distinguish the relative threat of ant attack in flowers on plants tended by each ant species. Pollinator surveys in 2003 and 2004 demonstrated that bee visitation rates and the duration of flower occupation differed among plants with different ant associates, decreasing as the threat of ant attack increased. Flowers on plants tended by Solenopsis xyloni , the best ant bodyguard, were more dangerous than those on plants tended by three milder species, due to that ants' greater aggressiveness and abundance within flowers. These flowers were visited by pollinators least frequently and for less time per visit, and produced fruits with significantly lower total seed mass, fewer seeds, and lighter individual seeds, relative to fruits from similarly-sized plants tended by three other ant species. As a result, the best bodyguard may indirectly constrain plant reproduction in some settings. Conflicts between mutualistic guilds may be particularly common in generalized systems, where there is variation in partner quality and in the relative importance of the protection and pollination mutualisms.     

Are flower-visiting ants mutualists or antagonists? A study in a gynodioecious wild strawberry

Ants are common flower visitors, but their effects on plant reproductive fitness have not often been assessed. Flower-visiting ants were studied to determine whether they are antagonists or mutualists and whether they could influence floral or breeding system evolution in gynodioecious wild strawberry (Fragaria virginiana). Ant and flying pollinator (bees/flies) access to plants was manipulated, and visitation, fruit, and seed set were assessed. Ants visited flowers of hermaphrodites more often than those of females when bees and flies were excluded, but visited the sex morphs equally when they were present. Insect class did not influence fruit or seed set of hermaphrodites. In contrast, ants had both positive and negative effects on seed set in females. Females visited only by ants had 90% of the seed set of those visited only by bees/flies, and their seed set increased with ant visitation. The spatial pattern of seed set, however, suggests that ants may also damage pistils. Lastly, in contrast to bees and flies, ants failed to increase visitation with floral display size, suggesting that ant presence at flowers could reduce selection on this attractive trait. Findings suggest that when in high abundance, flower-visiting ants could affect breeding system and floral evolution in this gynodioecious plant.     

Ants in your plants: effects of nectar-thieves on pollen fertility and seed-siring capacity in the alpine wildflower, Polemonium viscosum

Flowers of the alpine skypilot, Polemonium viscosum , are attacked by nectar thieving ants of Formica neorufibarbus gelida . Ants exert selection on flower scent, size and shape in skypilots by damaging the pistils. Here, I report on the frequency and nature of contact between ants and pollen-bearing anthers and determine the consequences of such contact for pollen performance and pollen donor paternity. In laboratory trials, ants entered flowers with full intact anthers and emasculated (female) flowers equivalently. Similarly, flower visitation rates of ants foraging naturally were not affected by the frequency of male phase flowers per plant. Ants actively interacted with the pollen-bearing anthers during 21% of flower visits, on average. The rate at which such interactions occurred was predicted by the proportion of flowers in the male phase, under a random foraging model. The effect of ants on pollen fertility was tested experimentally by enclosing ants in male-phase flowers on intact inflorescences. Adjacent control flowers were left un-occupied. Pollen from flowers with a history of ant occupancy had significantly lower germination on virgin recipient stigmas than pollen from unoccupied control flowers. With hand-pollination, sufficient pollen was transferred from ant-occupied flowers to saturate seed set. However, a model based on the relationship between seed set and compatible pollen delivery by natural pollinators indicated that ant damage to pollen should reduce paternity accruing per flower visit by 20–26% on average, in nature. Results support the hypothesis that in P. viscosum , selection on floral traits by nectar-thieving ants operates through male as well as female function.     

极小种群浙江安息香的物候观测

浙江安息香(Styrax zhejiangensis)分布于浙江建德林场泷江林区桃花坞,是浙江省特有的极小种群野生植物。为了更好地解析该物种的濒危原因,对浙江安息香进行物候观测,了解其开花、结实和种子生物学特征等。结果表明:浙江安息香花叶同期。花器官和花粉特征与同属植物较为类似,结实率为21%~32%。种子存在形态变异,饱满度不足等现象,以及外种皮致密坚硬导致其透水性差有可能影响其萌发。温差及连续降雨虽然会加重其落花落果。因此,保护该物种需要更多了解其花果实的特性,并加以人工辅助。     

Bionomics ofPhyllonorycter (Lepidoptera,Gracillariidae) onQuercus. II. Effects of ants

The effects of ants on a population ofPhyllonorycter leaf miners were observed in an oak chaparral on the Ishikari Coast. The density of ant nests was higher at the chaparral's edge than in the interior, while the density of mines was lower at the edge than in the interior. The effects of ant predation were examined in a comparison between treated ant-free trees and untreated ant-rich trees. Percentages of the torn mines observed were higher on the ant-rich trees than on the ant-free trees, whereas the mortality due to parasitoids was higher on the ant-free trees than on the ant-rich trees. This shows that ants greatly contributed to the mortality of leaf miners.     

Production of food bodies on the reproductive organs of myrmecophytic Macaranga species (Euphorbiaceae): effects on interactions with herbivores and pollinators

In protective ant–plant mutualisms, plants offer ants food (such as extrafloral nectar and/or food bodies) and ants protect plants from herbivores. However, ants often negatively affect plant reproduction by deterring pollinators. The aggressive protection that mutualistic ants provide to some myrmecophytes may enhance this negative effect in comparison to plant species that are facultatively protected by ants. Because little is known about the processes by which myrmecophytes are pollinated in the presence of ant guards, we examined ant interactions with herbivores and pollinators on plant reproductive organs. We examined eight myrmecophytic and three nonmyrmecophytic Macaranga species in Borneo. Most of the species studied are pollinated by thrips breeding in the inflorescences. Seven of eight myrmecophytic species produced food bodies on young inflorescences and/or immature fruits. Food body production was associated with increased ant abundance on inflorescences of the three species observed. The exclusion of ants from inflorescences of one species without food rewards resulted in increased herbivory damage. In contrast, ant exclusion had no effect on the number of pollinator thrips. The absence of thrips pollinator deterrence by ants may be due to the presence of protective bracteoles that limit ants, but not pollinators, from accessing flowers. This unique mechanism may account for simultaneous thrips pollination and ant defense of inflorescences.     

Do ant visitors to extrafloral nectaries of plants repel pollinators and cause an indirect cost of mutualism?

Plants and ants have widespread relationships that are commonly mediated by the offer of extrafloral nectar (EFN) to ants that protect plants against herbivores. However, these ant–plant interactions are highly facultative and vary in time and space, mainly depending on the characteristics of the ant species, such as density and aggressiveness. In general, the outcomes of these relationships are positive, but in some cases, the presence of ants is neutral or negative to plants. Some studies suggest that aggressive attacks or merely the presence of ants might reduce the visitation rate of insect pollinators, such as bees, to flowers. We used experimental manipulation in natural conditions to test the hypothesis that ants on flowers of EFN-bearing plants might be recognized as a danger by pollinators (bees) and reduce the plant fitness (fruit-set). Our results show that the avoidance that ant bodyguard species feeding on EFNs of the Malpighiaceae Heteropterys pteropetala cause in pollinators, is not enough to decrease plant fruit-set. However, ants were indeed identified as a danger to pollinators as hypothesized and as suggested for other plant–pollinator relationships: flowers with plastic ants placed on the petals produced significantly fewer fruits than other treatments (using instead a plastic circle) or the control (natural condition). Indirect costs of facultative mutualisms are the focus of few studies and have been performed only rarely in the Neotropics; our results show that mutualism must be considered in multitrophic interactions studies for a better understanding of the functioning of the system.     

Assessing ant seed predation in threatened plants: a case study

Erodium paularense is a threatened plant species that is subject to seed predation by the granivorous ant Messor capitatus. In this paper we assessed the intensity and pattern of ant seed predation and looked for possible adaptive strategies at the seed and plant levels to cope with this predation. Seed predation was estimated in 1997 and 1998 at the population level by comparing total seed production and ant consumption, assessed by counting seed hulls in refuse piles. According to this method, ant seed predation ranged between 18% and 28%. A more detailed and direct assessment conducted in 1997 raised this estimate to 43%. In this assessment spatial and temporal patterns of seed predation by ants were studied by mapping all nest entrances in the studied area and marking the mature fruits of 109 reproductive plants with a specific colour code throughout the seed dispersal period. Intact fruit coats were later recovered from the refuse piles, and their mother plants and time of dispersal were identified. Seeds dispersed at the end of the dispersal period had a greater probability of escaping from ant seed predation. Similarly, in plants with late dispersal a greater percentage of seeds escaped from ant predation. Optimum dispersal time coincided with the maximum activity of granivorous ants because, at this time, ants focused their harvest on other plant species of the community. It was also observed that within-individual seed dispersal asynchrony minimised seed predation. From a conservation perspective, results show that the granivorous ant–plant interaction cannot be assessed in isolation and that the intensity of its effects basically depends on the seed dispersal pattern of the other members of the plant community. Furthermore, this threat must be assessed by considering the overall situation of the target population. Thus, in E. paularense, the strong limitation of safe-sites for seedling establishment reduces the importance of seed predation.     

Consequences of ants and extrafloral nectar for a pollinating seed‐consuming mutualism: ant satiation,floral distraction or plant defense?

Non‐pollinating consumers of floral resources, especially ants, can disrupt pollination and plant reproductive processes. As an alternative food resource to flowers, extrafloral nectar (EFN) may distract and satiate ants from flowers, thereby reducing their antagonistic effects on plants. Yet, EFN may actually attract and increase ant density on plants, thus increasing the disruption of pollination and/or their defense of plants. In this study, we tested the effects of ants and EFN on pollinating seed‐consuming interactions between senita cacti and senita moths in the Sonoran Desert. Prior study of senita showed that EFN can distract ants from flowers, but consequences for plant–pollinator interactions remain unstudied. In our current study, ant exclusion had no effect on pollination or oviposition when moths were abundant (>85% flower visitation). Yet, in an ant by EFN factorial experiment under lower moth abundance (<40% visitation), there was a significant effect of ant exclusion (but not EFN or an ant × EFN) on pollination and oviposition. In contrast with our predictions, ant presence increased rather than decreased pollination (and oviposition) by moths, indicating a beneficial effect of ants on plant reproduction. While ant density on plants showed a saturating response to continuous experimental variation in EFN, in support of ant satiation and distraction, the probability of pollination and oviposition increased and saturated with ant density, again showing a beneficial effect of ants on plant reproduction. Ants showed no significant effect on fruit set, fruit survival, or fruit production of oviposited flowers in the ant exclusion experiment. Ants did not affect the survival of moth larvae, but there was a marginally significant effect of ants in reducing wasp parasitism of moths. We suggest that EFN may not only distract ants from disrupting plant–pollinator interactions, but they may also enhance plant–pollinator interactions by increasing pollination and reducing wasp parasitism. Though often considered dichotomous hypotheses, ant distraction and plant defense may be synergistic, though the mechanism(s) for such positive ant effects on plant–pollinator interactions needs further study.     

Interaction between a threatened endemic plant (<Emphasis Type="Italic">Anchusa crispa</Emphasis>) and the invasive Argentine ant (<Emphasis Type="Italic">Linepithema humile</Emphasis>)

Seed dispersal mutualisms are essential to ensure the survival of diverse plant species and communities worldwide. Here, we investigated whether the invasive Argentine ant can replace native ants by fulfilling their functional role in the seed dispersal of the rare and threatened endemic myrmecochorous plant, Anchusa crispa, in Corsica (France). Our study addressed the potential of Linepithema humile to disperse elaiosome-bearing seeds of A. crispa, examining L. humile’s effects on (1) the composition of communities of ants removing seeds, (2) the number of seed removals, (3) seed preference, (4) the distance of seed dispersion, and (5) seed germination. We caught seven native species at the control site, but only the Argentine ant at invaded sites. L humile removed A. crispa seeds in greater numbers than did native ants, respectively 66 and 23%, probably due to their higher worker density. The invader was similar to native ants with respect to distance of seed transport. Finally, rates of seed germination were not significantly different between seeds previously in contact with either Argentine ants or not. Taken all together, these results suggest that the Argentine ant is unlikely to pose a threat to A. crispa population. These results have important implications for the management of this rare and threatened endemic plant and provide an example of non-negative interactions between invasive and native species.     

Nectar- and pollen-gathering Cephalotes ants provide no protection against herbivory: a new manipulative experiment to test ant protective capabilities

The question if ant behaviour and biological limitations should be considered before generalisations about the ant’s defensive capabilities in ant–plant relationships was explored through a new experimental manipulation. In the Brazilian tropical savanna, we tested the protective action of Cephalotes pusillus Klug on the extrafloral nectar-bearing plant Ouratea spectabilis Engl. (Ochnaceae). Three treatments were performed: control (free ant access), Cephalotes-treatment (access permitted only to C. pusillus), and ant free treatment (no ants). No difference was found in the levels of leaf herbivory among experimental stems. Visitation by different ant species to control stems translated into significantly greater fruit and seed production by this stem category than by ant-free and Cephalotes-treated stems. Thus, results showed that an investigation of system’s natural history, ant’s morphological traits, defensive capabilities and behaviour are needed before a protective role is inferred to each associated ant species.     

Assessing ant seed predation in threatened plants: a case study

Erodium paularense is a threatened plant species that is subject to seed predation by the granivorous ant Messor capitatus. In this paper we assessed the intensity and pattern of ant seed predation and looked for possible adaptive strategies at the seed and plant levels to cope with this predation. Seed predation was estimated in 1997 and 1998 at the population level by comparing total seed production and ant consumption, assessed by counting seed hulls in refuse piles. According to this method, ant seed predation ranged between 18% and 28%. A more detailed and direct assessment conducted in 1997 raised this estimate to 43%. In this assessment spatial and temporal patterns of seed predation by ants were studied by mapping all nest entrances in the studied area and marking the mature fruits of 109 reproductive plants with a specific colour code throughout the seed dispersal period. Intact fruit coats were later recovered from the refuse piles, and their mother plants and time of dispersal were identified. Seeds dispersed at the end of the dispersal period had a greater probability of escaping from ant seed predation. Similarly, in plants with late dispersal a greater percentage of seeds escaped from ant predation. Optimum dispersal time coincided with the maximum activity of granivorous ants because, at this time, ants focused their harvest on other plant species of the community. It was also observed that within-individual seed dispersal asynchrony minimised seed predation. From a conservation perspective, results show that the granivorous ant–plant interaction cannot be assessed in isolation and that the intensity of its effects basically depends on the seed dispersal pattern of the other members of the plant community. Furthermore, this threat must be assessed by considering the overall situation of the target population. Thus, in E. paularense, the strong limitation of safe-sites for seedling establishment reduces the importance of seed predation.     

Pollen limitation in a narrow endemic plant: geographical variation and driving factors

Pollen limitation may have important consequences for the reproduction and abundance of plant species. It may be especially harmful to endangered and endemic plants with small populations. In this study, we quantify the effect of pollen limitation on seed production and seedling emergence in an endangered narrow endemic crucifer, Erysimum popovii. We conducted a pollen addition experiment across the entire geographic distribution of the species, and explored the effect of pollinator assemblage, plant population size and density, and other habitat variables on pollen limitation intensity in 13 populations. We supplemented flowers in 20 plants per population with allogamous pollen. To account for potential resource reallocation, we used two types of control untreated flowers: internal control flowers from the same individual as the supplemented flowers, and external control flowers from other individuals. Our results indicate that E. popovii is pollen-limited in most of the populations studied, but only through seed production, since pollen supplementation did not enhance seedling emergence. Beefly abundance was associated with among-population differences in pollen limitation intensity. Populations in which beeflies were more abundant were less pollen-limited. In contrast, the abundance of other flower visitors, such as large bees or butterflies, was not associated with pollen limitation. Annual rainfall and bare soil cover were associated with the intensity of pollen limitation across populations.     

Novel interactions of non-pollinating ants with pollinators and fruit consumers in a tropical forest

The tropical ants Ectatomma ruidum and E. tuberculatum (Formicidae) regularly patrol leaves, flowers, and fruits of the understory shrub, Psychotria limonensis (Rubiaceae), on Barro Colorado Island, Panama. Ant and pollinator exclusion experiments elucidated both positive and negative effects of ant attendance on plant reproductive success, including pollination, fruit set, fruit loss, and fruit removal. Ants did not pollinate flowers but did contribute to higher pollination success, probably by increasing the relocation frequency of winged pollinators and thus the rate of flower visitation. Ants also prevented fruit loss to herbivorous insects which were common during the early stages of fruit development. Thus, ant attendance strongly improved both pollination and fruit set whereby plants with ants set more fruit per flower and also lost fewer fruits during fruit maturation. In contrast, ants had a negative effect on the removal of ripe fruits by avian frugivores. Thus, ant attendance has a non-trivial influence on plant reproduction, this interaction being beneficial at some stages of the plant reproductive cycle and carrying costs at another stage. A tight ecological or co-evolved relationship between these Ectatomma spp. and P. limonensis is unlikely given that ant attendance of plants is detrimental to fruit removal. Received: 18 May 1998 / Accepted: 1 March 1999