Predator activities may lead to the accumulation of nutrients in specific areas of terrestrial habitats where they dispose of prey carcasses. In their feeding sites, predators may increase nutrient availability in the soil and favor plant nutrition and growth. However, the translocation of nutrients from one habitat to another may depend on predator identity and diet, as well as on the amount of prey intake. Here we used isotopic (15N) and physiological methods in greenhouse experiments to evaluate the effects of the identity of predatory ants (i.e., the consumption of prey and nest sites) on the nutrition and growth of the bromeliad Quesnelia arvensis. We showed that predatory ants with protein-based nutrition (i.e., Odontomachus hastatus, Gnamptogenys moelleri) improved the performance of their host bromeliads (i.e., increased foliar N, production of soluble proteins and growth). On the other hand, the contribution of Camponotus crassus for the nutritional status of bromeliads did not differ from bromeliads without ants, possibly because this ant does not have arthropod prey as a preferred food source. Our results show, for the first time, that predatory ants can translocate nutrients from one habitat to another within forests, accumulating nutrients in their feeding sites that become available to bromeliads. Additionally, we highlight that ant contribution to plant nutrition may depend on predator identity and its dietary requirements. Nest debris may be especially important for epiphytic and terrestrial bromeliads in nutrient-poor environments.
Citation: Gonçalves AZ, Oliveira RS, Oliveira PS, Romero GQ (2016) Species-Specific Effects of Ant Inhabitants on Bromeliad Nutrition. PLoS ONE 11(3): e0152113. https://doi.org/10.1371/journal.pone.0152113
Editor: Peter Schausberger, University of Vienna, AUSTRIA
Received: November 25, 2015; Accepted: March 9, 2016; Published: March 22, 2016
Copyright: © 2016 Gonçalves et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data is available in the paper.
Funding: This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo scholarship to A.Z. Gonçalves (2011/10137-8), and by Fundação de Amparo à Pesquisa do Estado de São Paulo and Conselho Nacional de Desenvolvimento Científico e Tecnológico research grants to P.S. Oliveira, G.Q. Romero and R.S. Oliveira (2010/17204-0). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Ecological research has increasingly focused on the links between species interactions and ecosystem functioning [1, 2]. Predators have a significant influence on ecosystems by affecting prey populations, and controlling the dynamics of nutrients, either by cascading effects through herbivore-plant-decomposers or directly through their excreta and activity of nutrient translocation . As predators capture prey and transport them to their feeding sites, they may redistribute nutrients between habitats [3, 4]. Nutrient translocation by predators has been observed in riparian zones by bears that release marine and freshwater nutrients through deposition of salmon carcasses, raising ammonium and nitrate concentrations in the soil . Similarly, seabirds and sea turtles feeding on marine organisms may excrete nutrients or leave egg remains in their breeding nests and nourish nearby soils [5, 6]. In addition, crows that feed in urban areas transfer nutrients to their roosts in adjacent forests . These translocated nutrients may become available for plants and impact their performance and community diversity and structure .
The identity of predators and their hunting and feeding modes play a key functional role within communities and can determine the amount of nutrients that are translocated to soils [2, 8]. However, little is known about how predator identity and hunting mode can affect the nutrition of plants. In the absence of predators, some plants are able to capture or shelter a diversity of species that are not used by them as a direct source of nutrients. For example, the carnivorous plant Roridula captures prey but lack digestive enzymes, and depends on the activity of predatory hemipterans for prey digestion and release of nutrients . Similarly, Romero et al. [10, 11] showed that while predatory spiders and frogs provide nutrients to their host bromeliads, entire prey carcasses contribute less to plant nutrition. Predatory ants may also contribute to plant nutrition, and many plants have modified hollow structures (i.e., domatia) that provide shelter to ant colonies, establishing symbiotic relationships [12–16]. Because ants have hunting sites often distinct from their feeding sites (i.e., colonies), ants can translocate nutrients from one habitat to another, similarly to bears, seabirds, sea turtles and the crows.
Symbiotic associations between plants and ants have been described for many species. Some plants are myrmecophytes (i.e., specialized ant-plants) that have modified hollow cavities (i.e., domatia) that house ant colonies in an obligate association, while other plants are myrmecophilous in a non-specialized association with ants [13–19]. In both cases, nest debris with food remains and ant feces are stored in contact with plant tissue, and nutrients can be obtained by plants [13–16]; ant-assisted plant nutrition is called myrmecotrophy. In the Neotropics, bromeliads are among the most used plants by ants [17, 18]. Some myrmecophytic bromeliads have bulb-shaped leaves that house ant colonies (e.g., Brocchinia acuminata, Tillandsia bulbosa, T. butzii), while others have ant nests constructed among roots or in inter-foliar cavities that do not accumulate water [18–21]. Tank bromeliads shelter many aquatic and terrestrial organisms, from microorganisms to vertebrates and retain leaves from the forest canopy [22, 23]. Therefore, tank bromeliads can be considered a complete ecosystem with debris, decomposers, detritivores, omnivores and predators . Because they possess foliar trichomes capable of absorbing water and nutrients, bromeliads can obtain nutrients from the trophic network they shelter [10, 11, 22, 24]. Many ants that occupy bromeliads are predators that forage in the vicinity of their host plants and bring arthropod prey into their colonies. Thus, nest debris can be a source of nutrients for bromeliads, as Leroy et al. [15, 16] suggested by comparing leaf nitrogen isotopes between bromeliads with and without ant colonies.
In the nutrient-poor soils of coastal Atlantic Forests , the bromeliad Quesnelia arvensis hosts colonies of Odontomachus hastatus, Gnamptogenys moelleri, Camponotus crassus and two other ant species . Among the ants that interact with this plant, nests of O. hastatus are located mainly among bromeliad roots [21, 27, 28]. On the other hand, G. moelleri and C. crassus have their nests among bromeliad leaves ([29, 30, 31, 32], AZG personal observations). While O. hastatus and G. moelleri have a predatory habit and feed mainly on arthropods, C. crassus feeds mostly on extrafloral nectaries, homopteran exudates, and fruits, but may also hunt for arthropods to supplement its diet [30–32]. As O. hastatus, G. moelleri and C. crassus capture and bring prey into their nests among bromeliad roots (O. hastatus) or bromeliad leaves (G. moelleri and C. crassus), they may translocate nutrients from one environment to another and improve plant nutrition, as shown for other predators [3–7]. Here we report the results of an experiment designed to evaluate the effects of the identity of ants and their diet (i.e., the consumption of prey and nest sites) on the nutrition and development of their host bromeliads. Assuming that these three ant species have nests with similar number of ants (see Methods), we hypothesized that O. hastatus and G. moelleri will contribute more to the nutrition and growth of Q. arvensis, since they have a predatory habit and their diet is based mainly on arthropods. On the other hand, we expect that C. crassus will contribute less to the nutrition and growth of host bromeliads because its diet is based on fewer prey items.
This study was conducted according to relevant national and international guidelines. Permit number 12.429/2011 for Ana Z. Gonçalves, issued by the Secretaria do Meio Ambiente and Instituto Florestal, in accordance with the Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis (IBAMA) and Ministério do Meio Ambiente (ICMBio MMA).
Organisms and field sampling
Colonies of Odontomachus hastatus, Gnamptogenys moelleri and Camponotus crassus were collected in the restinga forest of the Parque Estadual da Ilha do Cardoso (Cardoso Island), São Paulo State, Brazil (25°04’ S, 47°55’ W). The understory of the restinga is covered mostly by Quesnelia arvensis, whereas Vriesea procera is the most common epiphytic species [21, 23]. Nests of O. hastatus are frequently constructed in V. procera, but Q. arvensis can also host colonies . Quesnelia arvensis is a tank bromeliad that obtains nutrient not only by its roots, but also by foliar trichomes that are capable of absorbing nitrogen compounds (e.g., amino acids) . Among the ants that interact with Q. arvensis, O. hastatus has the greatest size (≈ 1.3 cm) and biomass (mean ± SE, 7.89 ± 1.57 mg) and their nests are abundant (ca. 33 colonies. ha-1) among bromeliad roots [21, 27, 28]. This nocturnal ant species is arboreal and occurs from Central to South America ; its nests are rarely found on the ground [27, 34]. Rodrigues and Oliveira  showed that O. hastatus workers can move more than 8 m away from their nests, but nearly half of the foraging activity occurs within 3 m around the nest. Gnamptogenys moelleri has ≈ 0.5 cm in length, 1.15 ± 0.32 mg in biomass, occurs in Neotropical plains, its nests can be found on the ground and in terrestrial bromeliads among 1–3 leaves, and workers forage almost exclusively on their host plants [29, 30]. Camponotus crassus (≈ 0.5 cm in length, 1.04 ± 0.09 mg in biomass) nests among bromeliad leaves and feeds mostly on extrafloral nectaries, homopteran exudates and fruits, and can suppress herbivores from plants that have extrafloral nectaries ([31, 32, 35], AZG personal observations). On Cardoso Island, nests of this ant are found among 2–3 leaves of Q. arvensis, where there is no water accumulation (AZG personal observations).
The number of ants per colony did not differ among treatments of the experiment described below (One-way ANOVA, P = 0.261; mean ± SE, 129.3 ± 7.6 ants O. hastatus colonies; 141.7 ± 4.7 ants in G. moelleri colonies; and 149.0 ± 11.4 ants in C. crassus colonies). However, the nest biomass differed among species (One-way ANOVA, P < 0.001; mean ± SE, 1020.1 ± 60.1 mg of O. hastatus; 162.9 ± 5.5 mg of G. moelleri; and 154.9 ± 11.8 mg of C. crassus).
Previously to the experiment, Tenebrio larvae (simulating prey of ants) were enriched with 15N stable isotope to quantify the flux of nitrogen from Tenebrio to bromeliads. Tenebrio were grown in a substrate in the proportion of 100 g of rat chow Labina-Purina®, 100 g cassava flour and 20 mL of a solution of 5 g enriched ammonium sulfate [(15NH4)2SO4, 10% excess atoms, Cambridge Isotope Laboratories, MA] per liter of distilled water. This substrate was dried for 24 h at 60°C before being offered to Tenebrio.
In order to test whether ants alter the availability of nutrients to their host bromeliads due to prey intake and nest sites, a two-month (10-Jan-2012 to 18-Mar-2012) greenhouse experiment was carried out at the Department of Plant Biology at Universidade Estadual de Campinas, Brazil. Since O. hastatus can be found in Q. arvensis , this species was chosen for the experiment and was obtained at Veiga Arquitetura e Paisagismo®, CEASA, Campinas, Brazil. All bromeliads were young and had similar biomass and size (e.g., foliar length varying from 25 to 30 cm). Bromeliads were planted in pots (14.5 cm in diameter, 14.5 cm high) with Pinus sp. bark (simulating the poor soil of restinga forest; Martins et al. 2015) and were watered with limited amounts of water to avoid desiccation through an automatic irrigation system with a capacity of 8L.h-1, which worked for 10 min every 2 h. Each pot was kept individually in a white plastic tray (40.7 x 60.8 x 9.8 cm) with Tanglefoot® resin around its border to prevent ants leaving their host bromeliads. All pots had four holes at their bases (approximately 1 x 1 cm) to allow ants to access the bromeliad roots. Colonies of O. hastatus collected in the field were placed in the plastic trays and quickly sought refuge and entered the holes of the pots in which bromeliads were planted. In about 30 min, all ants entered the pots and carried eggs and pupae to establish the colony within the roots. Colonies of G. moelleri and C. crassus were manually placed in the rosette of potted bromeliads, protecting eggs and pupae with Pinus sp. barks (the same used inside pots among bromeliad roots).
Experimental bromeliads had the following treatments (n = 10 each treatment): (1) control; (2) no ants but with Tenebrio larvae placed among roots (to test the amount of nitrogen that bromeliads acquire through roots); (3) no ants but with Tenebrio placed on leaves (to test the amount of nitrogen derived from insects that fall into the rosettes); (4) with C. crassus colony and Tenebrio placed on leaves (to test if ants accelerate the acquisition of nitrogen through leaves); (5) with G. moelleri colony and Tenebrio placed on leaves (to test if ants accelerate the acquisition of nitrogen through leaves); and (6) with O. hastatus colony among roots and Tenebrio placed on leaves (to test if ants accelerate the acquisition of nitrogen through roots). Two Tenebrio larvae were cut into six pieces and were applied every other day on each bromeliad (except on control plants), in the center of the rosette above the tank in treatments 3, 4, 5 and 6. Tenebrio was the only nutrient source for the colonies, and ants were very fast at collecting Tenebrio on leaves. Tenebrio was placed among roots (treatment 2) through a PVC pipe (2 cm in diameter) inserted in the root mass. At the end of the experiment, two leaves of the second inner node of each bromeliad rosette were collected for isotope analyses, and three leaves of the fifth inner node of the rosettes were collected for soluble protein analyses.
Analyses of plant protein
Total soluble protein concentration was determined using the colorimetric Bradford assay . Fresh leaves of Q. arvensis were cut into small pieces (1 cm2), and 1 g of these leaves was frozen in liquid nitrogen and homogenized with 3 mL of ultra-pure water. The homogenate was centrifuged at 12.000 rpm (g) for 10 min, and the supernatant (15 μL) was mixed with Comassie Brilliant Blue G-250 dye solution (185 μL), obtained from 100 mg of the dye dissolved in 95% ethanol and 100 mL of 85% phosphoric acid. The absorbance was measured using a spectrophotometer (Ultrospec 3000; Cambridge, England) at 595 nm, and the concentration of protein was determined by plotting the absorbance of the sample vs. a standard curve obtained with bovine serum albumin.
To determine the relative contribution of nitrogen derived from Tenebrio only or from ant nests on bromeliad growth, two leaves (i.e., fifth inner node of the rosettes) from each bromeliad were randomly chosen and their lengths were measured at the beginning and the end of the experiment. The bromeliad leaf length was directly related to the leaf biomass (Linear regressions: Q. arvensis: r2 = 0.73, P < 0.001); leaves showed continuous growth during the experiment and their relative growth rate (RGR, Ln(cm)/day) was calculated using the following equation: RGR = [ln(Lfinal)–ln(Linitial)]/(t2 − t1). The ln(Lfinal) and ln(Linitial) are, respectively, the natural logarithm of the foliar final length and the natural logarithm of the foliar initial length, with (t2 –t1) being the number of days between the initial and final measurements.
Isotopic and statistical analyses
The total N concentration (μg mg-1 dry leaf tissue) of bromeliad leaves and the δ15N of Tenebrio, ants and bromeliads were determined with an isotope ratio mass spectrometer (20–20 mass spectrometer; PDZ Europa, Sandbach, UK) after sample combustion to N2 at 1000°C by an on-line elemental analyzer (PDZ Europa ANCA-GSL) in the Stable Isotope Facility at the University of California, Davis. The nitrogen fraction in bromeliads with and without ants that received Tenebrio (fA) were calculated using mixing model equations with two sources of nitrogen (i.e., soil and Tenebrio) and one single isotopic signature (e.g., δ15N; see ). Since control bromeliads only had Pinus sp. bark substratum as a source of nutrients, leaves of these bromeliads were considered the soil end-member in the equation. The fractionation of 15N during its assimilation and metabolic processing in plants  was considered in the following equation: fA = (δM − δB − Δδ15N)/(δA − δB). The fA is the proportionate contribution of labelled Tenebrio absorbed by bromeliads (%), δM is the isotope ratio of bromeliads that received Tenebrio, δA is the isotope ratio of Tenebrio while δB is the isotope ratio of control bromeliads, and Δδ15N is the trophic shift for nitrogen between Tenebrio or control bromeliads and consumer (e.g., bromeliads). The values of Δδ15N used were + 1.4 ± 0.21‰ (mean ± SE) . All the response variables were compared using ANOVA and Tukey HSD post-hoc test were used for pair-wise comparisons.
The δ15N values of enriched Tenebrio larvae, of the ants Odontomachus hastatus, Gnamptogenys moelleri and Camponotus crassus that fed on the enriched Tenebrio indicate that these materials were enriched during the experiment (Table 1). As expected, Odontomachus hastatus contributed more to the nutrition of its host bromeliad, accounting for 19.3 ± 1.8% (mean ± SE) of the total nitrogen of Quesnelia arvensis (Tables 1 and 2; Fig 1A). Gnamptogenys moelleri contributed 16.1 ± 1.9% of the nitrogen of its host bromeliad, whereas C. crassus contributed with 10.7 ± 2.4% (Tables 1 and 2; Fig 1A). In the absence of ants, Tenebrio on leaves contributed with 5.3 ± 2.6% of the nitrogen of bromeliads, and Tenebrio on roots contributed with 3.9 ± 0.5% (Tables 1 and 2; Fig 1A). Despite differences among treatments in the contribution to bromeliad nutrition, only O. hastatus contributed to the increase of total nitrogen and soluble protein concentrations in bromeliad leaves (Table 2; Figs 1B and 2). Bromeliads with G. moelleri and C. crassus colonies had total nitrogen and soluble protein concentrations similar to bromeliads with prey carcasses or control bromeliads (Table 2; Figs 1B and 2). Additionally, O. hastatus and G. moelleri improved foliar growth of their host plants compared to control bromeliads while C. crassus had no effect on bromeliad growth (Table 2; Fig 3).
(A) Percentage of nitrogen derived from Tenebrio larvae and (B) total nitrogen concentration of Quesnelia arvensis leaves of different treatments (control; Tenebrio among roots; Tenebrio on leaves; Camponotus crassus ants and Tenebrio; Gnamptogenys moelleri and Tenebrio; and Odontomachus hastatus and Tenebrio). Bars indicate the standard error and letters indicate Tukey post-hoc test (α < 0.05).
Soluble protein concentration for Quesnelia arvensis leaves of different treatments (control; Tenebrio larvae among roots; Tenebrio on leaves; Camponotus crassus ants and Tenebrio; Gnamptogenys moelleri and Tenebrio; and Odontomachus hastatus and Tenebrio). Bars indicate the standard error and letters indicate Tukey post-hoc test (α < 0.05).
Relative growth rate of the Quesnelia arvensis leaves from different treatments (control; Tenebrio larvae among roots; Tenebrio on leaves; Camponotus crassus ants and Tenebrio; Gnamptogenys moelleri and Tenebrio; and Odontomachus hastatus and Tenebrio). Bars indicate the standard error and letters indicate Tukey post-hoc test (α < 0.05).
Average δ15N values of natural abundance and enriched Tenebrio larvae, Camponotus crassus, Gnamptogenys moelleri and Odontomachus hastatus ants, and Quesnelia arvensis leaves receiving the following treatments: control; Tenebrio among roots; Tenebrio on leaves; Camponotus crassus ants and Tenebrio; Gnamptogenys moelleri and Tenebrio; and Odontomachus hastatus and Tenebrio.
Analyses of variance (ANOVA) summarizing the effects of different treatments (control; Tenebrio larvae among roots; Tenebrio on leaves; Camponotus crassus ants and Tenebrio; Gnamptogenys moelleri and Tenebrio; and Odontomachus hastatus and Tenebrio) on the % of N derived from Tenebrio, the total N and soluble protein concentrations, and relative growth rate of Quesnelia arvensis leaves. Significance of P < 0.05 is highlighted in bold.
In this study, we assessed how the identity and diet of ant inhabitants can affect the nutrition of host plants through the translocation of prey carcasses and feces by predators. We found that the ant species with protein as their main food source (i.e., Odontomachus hastatus, Gnamptogenys moelleri) had a higher contribution to the performance of their host bromeliads (i.e., plant nutrition, production of soluble proteins, and growth). Regardless of the ant species, nitrogen physically stored in the body of prey becomes available to bromeliads.
We demonstrate experimentally that predatory ants play a role on bromeliad nutrition since they concentrate prey carcasses in their feeding sites (i.e., colonies in bromeliads) and release nutrients from prey bodies through nest debris and feces. We also demonstrate that predators that have arthropod prey as their main source of protein contribute more to the bromeliad performance (i.e., O. hastatus and G. moelleri). Although G. moelleri and O. hastatus have similar predatory habit [27, 30], the former species showed an intermediate contribution to its host plant nutrition, possibly due to its small biomass (similar to C. crassus). These results reinforce that the dietary requirements of each species must be taken into account to evaluate the potential nutrient-enrichment by predators in a community .
Tenebrio larvae (simulating prey that fall into the tank of bromeliads) contributed less to plant nutrition than the presence of predatory ants on bromeliads. This result emphasizes the role of ants in processing the organic matter of prey bodies, releasing compounds through the nest debris and their feces, which can be obtained by bromeliad roots (e.g., when plant interact with O. hastatus) or by leaf trichomes (e.g., when interacting with G. moelleri or C. crassus). Other studies also showed that predator-plant interactions could release organic matter from prey bodies retained by plants [10, 40, 41, 42]. In the absence of predators, Tenebrio need to be mineralized by microorganisms present in the bromeliad rosettes [24, 43], a process that should be slower than predator activity, which could explain the lower contribution of Tenebrio to bromeliad nutrition during the experiment.
The terrestrial bromeliad Quesnelia arvensis is very abundant in the sandy nutrient-poor soils of Cardoso Island [23, 25]. Through the interaction with predators, this plant can obtain nutrients from the soil via its roots or leaf trichomes that are in contact with debris and feces from ant nests. In fact, we showed that the interaction between Q. arvensis and ants favored plant nutrition and enabled a higher production of soluble protein in its leaves. High soluble protein concentration is associated with the N availability in plants indicating their favorable nutritional status , which may have allowed a higher growth of Q. arvensis when interacting with ants. In addition, the increase in N concentration in leaves of bromeliads can favor the production of amino acids usually associated with N storage , and can be especially important for epiphytes whose roots are not in contact with soil and have intermittent access to water and nutrients. At Cardoso Island, nearly 70% of the O. hastatus nests were recorded in the epiphytic Vriesea procera . Therefore, it is expected that V. procera can benefit even more from interactions with ants than the terrestrial Q. arvensis.
The contribution of O. hastatus and G. moelleri to bromeliads were similar to the N values that the carnivorous plants Cephalotus, Drosera, Philcoxia and Sarracenia derived from insect digestion [46–48]. Despite the existence of carnivorous bromeliads (i.e., Catopsis, Brocchinia), Q. arvensis is considered a saprophytic species because it lacks an active way to attract, immobilize and digest prey . The interactions described here represent a by-product digestive mutualism between ants and bromeliads where ants can find a suitable habitat in bromeliads and, in turn, contribute nutritionally to their hosts [10, 11, 40]. Finally, some researchers have argued that ants preferably associate with bromeliads with longer leaves and larger rosettes and root mass [21, 30]. We suggest that the larger size of ant-inhabited bromeliads could be a result of this by-product mutualism, and not necessarily the cause mediating plant colonization by ants.
In conclusion, our results emphasize the effects of predator identity and diet in the translocation and processing of prey nutrients that contribute to plant nutrition. Predatory ants with that feed on prey as their main source of protein contributed more to the nutrition, protein production and growth of their host bromeliads. These results emphasize that ants can play a role in redistributing nutrients between habitats, from different areas of the forest to their feeding sites (i.e., colonies). We highlight the importance of predator activities in concentrating wastes near their feeding sites and their potential to provide nutrients to plants, and we also reinforce that the by-product digestive mutualism can be relevant to plants that occur in oligotrophic environments.
We thank Dr. Paula M. de Omena, Renan F. Awata, Wendell Correia, Rodolfo F. Silva and Dr. Gustavo C. Piccoli for help with sampling of ants in the field.
Conceived and designed the experiments: AZG PSO GQR. Performed the experiments: AZG. Analyzed the data: AZG RSO PSO GQR. Contributed reagents/materials/analysis tools: RSO PSO GQR. Wrote the paper: AZG RSO PSO GQR.
- 1. Ngai JT, Srivastava DS. Predators accelerate nutrient cycling in a bromeliad ecosystem. Science. 2006; 314: 963. pmid:17095695
- 2. Schmitz OJ, Hawlena D, Trussell GC. Predator control of ecosystem nutrient dynamics. Ecology Letters. 2010; 13: 1199–1209. pmid:20602626
- 3. Holtgrieve GW, Schindler DE, Jewett PK. Large predators and biogeochemical hotspots: brown bear (Ursus arctos) predation on salmon alters nitrogen cycling in riparian soils. Ecological Research. 2009; 24: 1125–1135.
- 4. Hocking MD, Reynolds JD. Impacts of salmon on riparian plant diversity. Science. 2011; 331: 1609–1612. pmid:21442794
- 5. Bouchard SS, Bjorndal KA. Sea turtles as biological transporters of nutrients and energy from marine to terrestrial ecosystems. Ecology. 2000; 81: 2305–2313.
- 6. Kolb GS, Palmborg C, Taylor AR, Baath E, Hambäck PA. Effects of nesting cormorants (Phalacrocorax carbo) on soil chemistry, microbial communities and soil fauna. Ecosystems. 2015; 18: 643–657.
- 7. Hilderbrand GV, Hanley TA, Robbins CT, Schwartz CC. Role of brown bears (Ursus arctos) in the flow of marine nitrogen into a terrestrial ecosystem. Oecologia. 1999; 121: 546–550.
- 8. Rudolf VHW, Rasmussen NL. Resolving the roles of body size and species identity in driving functional diversity. Proceedings of Royal Society B. 2014; 281: 1–8.
- 9. Anderson B, Midgley JJ. Digestive mutualism, an alternate pathway in plant carnivory. Oikos. 2003; 102: 221–224.
- 10. Romero GQ, Mazzafera P, Vasconcellos-Neto J, Trivelin PCO. Bromeliad-living spiders improve host plant nutrition and growth. Ecology. 2006; 87: 803–808. pmid:16676522
- 11. Romero GQ, Nomura F, Gonçalves AZ, Dias NYN, Mercier H, Conforto EC, et al. Nitrogen fluxes from treefrogs to tank epiphytic bromeliads: an isotopic and physiological approach. Oecologia. 2010; 162: 941–949. pmid:20024585
- 12. Rico–Gray V, Barber JT, Thien LB, Ellgaard EG, Toney JJ. An unusual animal–plant interaction: feeding of Schomburgkia tibicinis (Orchidaceae) by ants. American Journal of Botany. 1989; 76: 603–608.
- 13. Treseder KK, Davidson DW, Ehleringer JR. Absorption of ant-provided carbon and nitrogen by a tropical epiphyte. Nature. 1995; 357: 137–139.
- 14. Fischer RC, Wanek W, Richter A, Mayer V. Do ants feed plants? A 15N labelling study of nitrogen fluxes from ants to plants in the mutualism of Pheidole and Piper. Journal of Ecology. 2003; 91: 126–134.
- 15. Leroy C, Corbara B, Pélozuelo L, Carrias JF, Dejean A, Céréghino R. Ant species identity mediates reproductive traits and allocation in an ant-garden bromeliad. Annals of Botany. 2012; 109: 145–152. pmid:21984729
- 16. Leroy C, Carrias JF, Corbara B, Pélozuelo L, Dézerald O, Brouard O, et al. Mutualistic ant contributes to tank–bromeliad nutrition. Annals of Botany. 2013; 112: 919–926. pmid:23864002
- 17. Dejean A, Olmsted I, Snelling RR. Tree-epiphyte-ant relationships in the low inundated forest of Sian Kaan Biosphere Reserve, Quintana Roo, Mexico. Biotropica. 1995; 27: 57–70.
- 18. Blüthgen N, Verhaagh M, Goitiá W, Blüthgen N. Ant nests in tank bromeliads—an example of non-specific interaction. Insectes Sociaux. 2000; 47: 313–316.
- 19. Huxley C. Symbiosis between ants and epiphytes. Biological Reviews. 1980; 55: 321–340.
- 20. Benzing DH. Vascular epiphytes, general biology and related biota. Cambridge: Cambridge University Press; 1990.
- 21. Oliveira PS, Camargo RX, Fourcassié V. Nesting patterns, ecological correlates of polygyny and social organization in the neotropical arboreal ant Odontomachus hastatus (Formicidae, Ponerinae). Insectes Sociaux. 2011; 58: 207–217.
- 22. Benzing DH. Bromeliaceae: profile of an adaptive radiation. Cambridge: Cambridge University Press; 2000.
- 23. Romero GQ, Srivastava DS. Food-web composition affects cross-ecosystem interactions and subsidies. Journal of Animal Ecology. 2010; 79: 1122–1131. pmid:20584097
- 24. Gonçalves AZ, Hoffmann FL, Mercier H, Mazzafera P, Romero GQ. Phyllosphere bacteria improve animal contribution to plant nutrition. Biotropica. 2014; 46: 170–174.
- 25. Martins SC, Neto ES, Piccolo MC, Almeida DQA, Camargo PB, Carmo JB, et al. Soil texture and chemical characteristics along an elevation range in the coastal Atlantic Forest of Southeast Brazil. Geoderma Regional. 2015; 5: 106–116.
- 26. Camargo RX. Ecologia e comportamento da formiga arborícola Odontomachus hastatus (Hymenoptera: Formicidae: Ponerinae). M. Sc. Thesis, Universidade Estadual de Campinas. 2002. Available: www.bibliotecadigital.unicamp.br/document/?code=vtls000242817&opt=1
- 27. Camargo RX, Oliveira PS. Natural history of the Neotropical arboreal ant, Odontomachus hastatus: nest sites, foraging schedule, and diet. Journal of Insect Science. 2012; 12: 1–9.
- 28. Rodrigues PAP, Oliveira PS. Visual navigation in the Neotropical ant Odontomachus hastatus (Formicidae, Ponerinae), a predominantly nocturnal, canopy-dwelling predator of the Atlantic rainforest. Behavioral Processes. 2014; 109: 48–57.
- 29. Lattke JE. Revision of the ant genus Gnamptogenys in the New World (Hymenoptera: Formicidae). Journal of Hymenoptera Research. 1995; 4: 137–193.
- 30. Cogni R, Oliveira PS. Patterns in foraging and nesting ecology in the neotropical ant, Gnamptogenys moelleri (Formicidae, Ponerinae). Insectes Sociaux. 2004; 51: 123–130.
- 31. Oliveira PS, Freitas AVL. Ant-plant-herbivore interactions in the neotropical cerrado savanna. Naturwissenschaften. 2004; 91: 557–570. pmid:15551026
- 32. Sendoya SF, Freitas AVL, Oliveira PS. Egg-laying butterflies distinguish predaceous ants by sigh. American Naturalist. 2009; 174: 134–140. pmid:19456265
- 33. Kempf WW. Catálogo abreviado de formigas neotropicais (Hymenoptera, Formicidae). Stud Entomol. 1972; 15: 1–344.
- 34. Gibernau M, Orivel J, Delabie JHC, Barabe D, Dejean A. An asymmetrical relationship between an arboreal ponerine ant and a trash–basket epiphyte (Araceae). Biological Journal of Linnean Society. 2007; 91: 341–346.
- 35. Davidson DW, Cook SC, Snelling RR, Chua TH. Explaining the abundance of ants in lowland tropical rainforest canopies. Science. 2003; 300: 969–972. pmid:12738862
- 36. Bradford MM. Rapid and sensitive method for quantification of microgram quantities of protein utilizing principle of protein-dye binding. Analytical Biochemistry. 1976; 72: 248–254. pmid:942051
- 37. Phillips DL, Gregg JW. Uncertainty in source partitioning using stable isotopes. Oecologia. 2001; 127: 171–179. pmid:24577646
- 38. McCutchan JHJ, Lewis WMJ, Kendall C, McGrath C. Variation in trophic shift for stable isotope ratios of carbon, nitrogen and sulfur. Oikos. 2003; 102: 378–390.
- 39. Caut S, Angulo E, Courchamp F. Variation in discrimination factors (Δ15N and Δ13C): the effect of diet isotopic values and applications for diet reconstruction. Journal of Applied Ecology. 2009; 46: 443–453.
- 40. Gonçalves AZ, Mercier H, Mazzafera P, Romero GQ. Spider-fed bromeliads: seasonal and interspecific variation in plant performance. Annals of Botany. 2011; 107: 1047–1055. pmid:21385776
- 41. Ellis AG, Midgley JJ. A new plant-animal mutualism involving a plant with sticky leaves and a resident hemipteran insect. Oecologia. 1996; 106: 478–481.
- 42. Anderson B, Midgley JJ. It takes two to tango but three is a tangle: mutualists and cheaters on the carnivorous plant Roridula. Oecologia. 2002; 132: 369–373.
- 43. Inselsbacher E, Cambui CA, Richter A, Stange CF, Mercier H. Microbial activities and foliar uptake of nitrogen in the epiphytic bromeliad Vriesea gigantea. New Phytologist. 2007; 175: 311–320. pmid:17587379
- 44. Barneix AJ, Causin HF. The central role of amino acids on nitrogen utilization and plant growth. Journal of Plant Physiology. 1996; 149: 358–362.
- 45. Endres L, Mercier H. Amino acid uptake and profile in bromeliads with different habitats cultivated in vitro. Plant Physiology and Biochemistry. 2003; 41: 181–187.
- 46. Schulze W, Schulze ED, Pate JS, Gillison AN. The nitrogen supply from soils and insects during growth of the pitcher plants Nepenthes mirabilis, Cephalotus follicularis and Darlingtonia californica. Oecologia. 1997; 112: 464–471.
- 47. Ellison AM, Gotelli NJ. Evolutionary ecology of carnivorous plants. Trends in Ecology and Evolution. 2001; 16: 623–629.
- 48. Pereira CG, Almenara DP, Winter CE, Fritsch PW, Lambers H, Oliveira RS. Underground leaves of Philcoxia trap and digest nematodes. PNAS. 2012; 109: 1154–1158. pmid:22232687
- 49. Givnish TJ, Burkhardt EL, Happel RE, Weintraub JD. Carnivory in the bromeliad Brocchinia reducta, with a cost/benefit model for the general restriction of carnivorous plants to sunny, moist, nutrient-poor habitats. American Naturalist. 1984; 124: 479–497.