BIOTROPICA 46(1): 78–82 2014
10.1111/btp.12073
Bromeliad-associated Reductions in Host Herbivory: Do Epiphytic Bromeliads Act as Commensalists or Mutualists? Edd Hammill1,2,3, Paloma Corvalan1, and Diane S. Srivastava1 1
University of British Columbia, 6270 University Boulevard, Vancouver, British Columbia, V6T 1Z4, Canada
2
University of Queensland, Goddard Building (8), St Lucia, 4072, Australia
ABSTRACT Many members of the family Bromeliacae are able to adopt epiphytic lifestyles and colonize trees throughout the Neotropics. Bromeliacae do not extract nutrients from their hosts and confer relatively minor costs on their host plants. We suggest that bromeliads, however, may benefit their hosts by providing habitat for predators of host plant herbivores. We report a correlation between bromeliad presence and a reduction in herbivore damage in orange trees, an effect that is increased when bromeliads are colonized by ants. Our results may have important implications for agricultural systems in the Neotropics, where bromeliads are often removed in the belief they are parasitic. We instead demonstrate that bromeliads may impart a benefit to their hosts, and speculate that under particular circumstances they may be part of a three-species mutualism. Abstract in Spanish is available in the online version of this article. Key words: agro-ecology; ants; bromeliad; community ecology; herbivory; indirect effects; mutualism; predator facilitation.
AGRICULTURE
AND BIODIVERSITY CONSERVATION ARE OFTEN PER-
(Tilman et al. 2001, Kleijn & Sutherland 2003). Conversion of land to agriculture typically involves clearing of locally native vegetation, degrading the landscape and reducing biodiversity (Barlow et al. 2007). Historically, natural landscapes close to agricultural lands were perceived as threats, and not treated favorably (Maybery et al. 2005). In recent years, the attitudes of agricultural managers toward natural areas have changed following the discovery that native biodiversity can provide important agricultural services such as increased pollination rates (Blanche et al. 2006), and increased top–down control of herbivores by parasitoids and predators (Landis et al. 2000). These services provided by natural ecosystems, however, tend to be spatially dependent, declining with distance from natural areas (Carvalheiro et al. 2010). In tropical forest ecosystems, predatory ants may represent the most important consumers of herbivorous arthropods (Floren et al. 2002). The ability of ants to substantially reduce herbivore populations has lead to many flowering plants actively producing incentives to attract ants, either in the form of shelter or nutrients (Heil & McKey 2002). The anti-herbivory effect of ants is highly beneficial to plants and has led to several tropical trees evolving close mutualisms with predatory ant species (Janzen 1966, Heil & McKey 2002, Heil 2008). Previous empirical studies investigating the relationships between ants and plants have shown that ants are able to reduce herbivory to such an extent that plant growth is increased (Frederickson et al. 2012), and alter the structure of epiphyte communities (Catling 1997). Reductions in herbivory
CEIVED AS CONFLICTING DEMANDS ON A LANDSCAPE
Received 5 April 2013; revision accepted 3 August 2013. 3
Corresponding author; e-mail:
[email protected]
78
associated with the presence of ants have also been shown to benefit economically significant species such as almonds (Rickson & Rickson 1998), and coffee (Philpott 2005). By reducing herbivores in agricultural systems, ants may act as important agents of biological control, meaning measures encouraging their presence may yield rewards in financial and ecological terms (Philpott 2005). The family Bromeliacae are an important component of the epiphyte community and are themselves important habitats for many animal species (Frank & Lounibos 2009). Bromeliads extract no nutrients from host plants, instead using their hosts merely as a platform from which to grow (Cascante-Marin et al. 2006). This lack of nutrient extraction suggests that the bromeliad/host relationship is commensal (+/0), rather than parasitic (+/ ). Despite not parasitizing their hosts, bromeliads at high densities may still reduce the fitness of their hosts through shading of leaves, breaking of limbs due to the weight of the plant (Benzing 1990), and reducing growth of new branch shoots (Monta~ na et al. 1997). Although bromeliads do not remove nutrients from their hosts, they are regularly removed from agricultural plantations due to the mistaken belief that they are parasitic (Toledo-Aceves et al. 2012), a labour intensive practice that may incur a financial cost to farmers. Bromeliads provide habitat and shelter for a variety of taxa and have been shown to affect the community composition of terrestrial arthropods within arboreal ecosystems (Yanoviak et al. 2011). Predatory ants in the genera Odontomachus and Platythyrea (Hymenoptera: Forminacae), as well as fungal-growing species in the genera Cyphomyrmex and Pheidole can be found nesting within bromeliads (Bluthgen et al. 2000). The strength of ant-bromeliad relationships differs between bromeliad genera, and certain ª 2013 The Association for Tropical Biology and Conservation
Bromeliads Reduce Herbivory in Orange Trees
members of the genus Tillandsia are considered truly myrmecophillic (‘ant-loving’), providing physiological structures specifically for colonization by ants, which then provide a reciprocal service by reducing herbivory (Dejean et al. 1995). However, many bromeliad species—especially members of the genera Guzmania and Vriesea—do not produce specialized structures or other incentives to attract ants. As such, colonization by ants is thought to be opportunistic rather than the result of a specialized association (Bluthgen et al. 2000). Citrus fruits represent the highest value fruit crop in the world. According to the Food and Agriculture Organisation of the United States, in 2007, the world produced over a 100 million tonnes of citrus fruit, with >50 percent consisting of varieties of the sweet orange Citrus sinensis (L. Osbeck). Orange trees in agricultural plantations are attacked by a suite of herbivorous insects, leading to the implementation of biological and chemical control measures to reduce herbivore damage (Timmer & Duncan 1999). Orange trees are also a host to a suite of epiphytes, including bromeliads (Catling 1997). Here we examine whether associations between bromeliads and ants have ‘spill over’ benefits for the agricultural trees in which the bromeliads occur. Such spill over effects could occur if the ants forage beyond the bromeliad plant where their nest is located. Functionally, this would turn the ant-bromeliad mutualism into a three-way mutualism whereby the tree benefits the bromeliad and ants by providing a structure for the bromeliad to grow on, the bromeliad benefits the ants by providing shelter, and the ants benefit the orange tree by reducing herbivory rates.
METHODS To isolate the effects of ant-bromeliad associations on orange leaf herbivory, we designed a survey that separated the effects of bromeliad presence from ant-bromeliad associations, and compared leaves near and distant from bromeliads. Our study was conducted in ‘Finca Yafa,’ a Del Oro-operated orange plantation in Guanacaste, Costa Rica. We selected 30 orange trees from unsprayed plantations as follows: 10 trees without bromeliads, 10 trees with bromeliads, and 10 trees with bromeliads inhabited by ants. The presence of ants was ascertained by removal and dissection of the bromeliad. We were specifically interested in predatory ants of the genus Odontomachus. These ants are believed to be generalist predators able to capture and consume a wide diversity of insect prey (Ehmer & Holldobler 1995). In addition to the bromeliad itself, we also inspected the host tree for the presence of colonies of large carnivorous ants, and rejected trees where ants were found. For our bromeliad treatments, we found trees containing a single large bromeliad (plant diameter > 30 cm, maximum water volume > 250 ml, genus Vriesea), between 180 cm and 250 cm from the ground. The bromeliad was one sampling point (‘position 1′) and a second sampling point (‘position 2’) was established on the opposite side of the tree at the same height. For trees without bromeliads, we randomly selected a point on the tree between 180 cm and 250 cm from the ground, and the
79
second point was immediately opposite the first. Our design therefore had two different levels of control, a comparison of trees with and without bromeliads, as well as near the bromeliad/ far from the bromeliad within tree comparison. In all cases, at each sampling position, a total of eight leaves were randomly selected by hand from within 50 cm of the sampling point. In an attempt to reduce bias, the researcher who selected the leaves then passed them to a second researcher who did not know which treatment the leaves represented for processing. Sampled leaves were photographed in the field. Leaf area was calculated by digitally tracing each leaf using ImageJ image analysis software (Rasband 2012). Lost or damaged areas of leaves were also traced, and their area added to the undamaged area to obtain an estimate of ‘total’ leaf size (the area the leaf would have been if it had not been damaged). The lost/damaged area was then divided by the ‘total’ area to obtain percent damaged/lost. Statistical analysis was conducted using R (R Development Core Team 2011). Data were analyzed using generalized linear mixed effects models with binomial error structures. To demonstrate differences between the three different treatments, two orthogonal contrasts were used. The first contrast compared the 10 trees without bromeliads to the 20 trees that contained bromeliads. The second contrast compared differences between trees containing ant-free bromeliads and bromeliads inhabited by ants. In each of the contrasts, ‘leaf ’ was the unit of replication, ‘treatment’ and ‘position’ (e.g., leaves from a 50 cm radius sphere near the bromeliad, or on the opposite side of the tree from the bromeliad) were included as fixed effects, and ‘tree identity’ was included as a random factor. We included tree identity as a random factor to account for the fact that multiple leaves were collected from each tree, and two leaves collected from the same tree are likely to be more similar than two leaves collected from different trees. We also included the interaction between ‘treatment’ and ‘position’. A significant interaction demonstrates that the effect of ‘position’ is different in the different treatments. Post-hoc analyses of each contrast were undertaken using the ‘glht’ function in the package ‘multcomp’ to perform a Tukey’s test to ascertain which treatments within a contrast were different from each other. Previous studies have shown that areas of rainforest provide ecosystem services to agriculture (Blanche et al. 2006), and the effects of these services decline as distance from natural areas increases (Carvalheiro et al. 2010). Finca Yafa borders the Area de Conservaci on Guanacaste, a national park and designated UNESCO World Heritage site. As we believed proximity to the park might affect orange leaf health, we included the distance from the park boundary as a covariate within our models.
RESULTS Bromeliads were associated with a reduction in orange leaf herbivory, but only on the side of the tree nearest the bromeliad. Specifically, there was a significant position*treatment interaction (z = 20.43, P < 0.001, Fig. 1A). The main effect of position was also found to be significant (z = 5.115, P < 0.001, Fig. 1A),
80
Hammill, Corvalan, and Srivastava
A
B
FIGURE 1. Proportion of leaf missing or damaged from trees in different treatments, accounting for proximity to national park boundary. For ‘bromeliad’ trees, ‘Position 1’ refers to leaves collected within a 50 cm radius of the bromeliad. Error bars represent 95% confidence intervals. A. First contrast comparing presence/absence of bromeliads B. Second orthogonal contrast comparing the effect of the presence of ants within the bromeliad. Different letters above the bars denote treatments that were significantly different from one another within a single contrast.
while the main effect of bromeliad was not (z = 1.064, P = 0.29). Post-hoc Tukey’s analysis revealed that orange leaves from trees with a bromeliad, but on the opposite side of the tree from the bromeliad, showed similar levels of damage to leaves from trees without bromeliads. However, orange leaves collected near bromeliads showed reduced damage levels (Fig. 1A). Leaf damage increased with distance from the national park boundary (z = 10.32, P < 0.001), and this was accounted for in the figure. We then compared the effects of ants within the bromeliad, a contrast that is orthogonal to the above comparison of bromeliad presence versus absence. Trees containing bromeliads were split depending on the presence/absence of ants within the bromeliad. We found a significant treatment*position interaction (z = 12.93, P < 0.001, Fig. 1B) showing that the presence of ants increased the difference between positions with and without bromeliads, suggesting the presence of ants is associated with further reductions in leaf damage. The main effects of position (z = 15.52, P < 0.001) and ants (z = 8.53, P < 0.001) were also found to be significant. Post-hoc Tukey’s analysis demonstrated that leaves closer to the bromeliad experienced lower herbivory rates, and herbivory rates were further reduced by the presence of ants (Fig. 1B). Leaf damage increases with distance from the park boundary (z = 8.95, P = 0.001), which was accounted for in the figure.
DISCUSSION Leaf herbivory can have important fitness consequences for individual plants (Marquis 1984). Recent evidence shows that herbivore pressure increases toward the equator (Salazar & Marquis 2012), suggesting that the selection pressure for anti-herbivory strategies will be stronger in the tropics. In response to herbivore pressure, plants have evolved a suite of different defensive strategies (Walling 2000). Defensive strategies designed to directly repel
herbivores, however, may be costly, requiring the production of metabolically expensive chemicals that can affect plant fitness (van Hulten et al. 2006). To avoid paying these chemical costs, several plant species recruit ants to consume their herbivores resulting in a tri-trophic, indirect defence (Heil 2008, Stanton & Palmer 2011). The recruitment of predatory arthropods, however, might incur other costs in terms of chemical synthesis (Hoballah et al. 2004) or the production of novel structures to serve as habitat for arthropod predators (Stanton & Palmer 2011). The importance of ants as consumers of herbivores makes them ideal candidates to act as biological control agents in tropical agriculture (Rickson & Rickson 1998, Philpott 2005). Natural ecosystems have been shown to provide services to tropical agricultural systems in terms of pollination (Blanche et al. 2006) and pest control (Schauff et al. 1998). However, services provided by natural areas are often spatially dependent and decline as the distance from natural ecosystem increases (Carvalheiro et al. 2010). The spatial dependency of ecosystem services means they will be unavailable to crop plants in the centre of large plantations, as densities of pollinators and consumers of herbivorous pests decrease as distance from natural areas increases. Our results suggest, however, that the presence of natural epiphytes within agricultural plantations can provide ecosystem services to their host plants. We therefore propose that epiphytes may bring part of the forest to trees within a plantation, providing a habitat for ants and other predators of herbivores and conferring a benefit to their host tree. The bromeliads in our study appeared to benefit their host plants. Our results suggest, however, that the anti-herbivory benefits of bromeliads are limited to areas in close proximity to the bromeliad, with herbivory rates on the side opposite the bromeliad (150–250 cm away from the bromeliad) similar to those from the trees without epiphytes. It therefore appears that achieving anti-herbivory rates similar to those seen in trees with
Bromeliads Reduce Herbivory in Orange Trees
arthropod mutualists would require a high density of bromeliads spaced uniformly throughout the tree. However, this high density of bromeliads may also result in such costs such as increases in weight leading to limb damage (Benzing 1990). We therefore suggest that while the presence of bromeliads within a host tree could provide some ancillary benefit to trees, it is unlikely that there is a selection for promoting their establishment and growth. It is unclear from our study how bromeliads without ants are related to the reductions in herbivory rates we observed. In addition to providing habitat for ants, bromeliads may provide shelter for many other predatory arthropods that could consume herbivores; especially prevalent are spiders of the genus Bourguyia (Osses et al. 2007) and scorpions of the genus Centruroides (E. Hammill—pers. obs.). In addition, our ant-free bromeliads may have contained ants in the past, meaning they may have historically conferred anti-herbivory benefits. As the Odontomachus ants we focused on in this study were large and carnivorous, we may expect them to forage throughout the tree, reducing herbivory on leaves far from the bromeliad as well as nearby. However, we observed significantly higher rates of herbivory away from the bromeliad compared to close to the bromeliad, and can only speculate on what may be generating through this observation. The relatively high rates of herbivory away from bromeliads may result from herbivores moving away from the ant colony in an attempt to avoid being consumed. Alternatively, although ants may forage throughout the tree the amount of time they spend close to the bromeliad will be relatively high. The mechanism generating the increase in herbivory rates, however, remains unclear and represents an interesting avenue for future study. We propose that bromeliads may act as ‘indirect mutualists’ by conferring a benefit to their host trees through association with ant guards that may reduce herbivory. These benefits may also have important applied consequences—although bromeliads are often removed from trees within plantations due to the mistaken belief that they are parasitic (Toledo-Aceves et al. 2012), we suggest future studies to evaluate the possibility that leaving them may increase fruit yields.
ACKNOWLEDGMENTS We would like to especially thank Victor Hugo Chaves, Fulvio Arias, Freddy Vargas, and Del Oro Costa Rica for their cooperation and help during this project. Without their assistance and facilities, this work would not be possible.
LITERATURE CITED BARLOW, J., T. A. GARDNER, I. S. ARAUJO, T. C. AVILA-PIRES, A. B. BONALDO, J. E. COSTA, M. C. ESPOSITO, L. V. FERREIRA, J. HAWES, M. M. HERNANDEZ, M. S. HOOGMOED, R. N. LEITE, N. F. LO-MAN-HUNG, J. R. MALCOLM, M. B. MARTINS, L. A. M. MESTRE, R. MIRANDA-SANTOS, A. L. NUNES-GUTJAHR, W. L. OVERAL, L. PARRY, S. L. PETERS, M. A. RIBEIRO-JUNIOR, M. N. F. da SILVA, C. D. S. MOTTA, AND C. A. PERES. 2007. Quantifying the biodiversity value of tropical primary, secondary,
81
and plantation forests. Proc. Natl. Acad. Sci. U. S. A. 104: 18555– 18560. BENZING, D. H. 1990. Vascular epiphytes. Cambridge University Press, Cambridge. BLANCHE, K. R., J. A. LUDWIG, AND S. A. CUNNINGHAM. 2006. Proximity to rainforest enhances pollination and fruit set in orchards. J. Appl. Ecol. 43: 1182–1187. BLUTHGEN, N., M. VERHAAGH, W. GOITIA, AND N. BLUTHGEN. 2000. Ant nests in tank bromeliads - an example of non-specific interaction. Insect. Soc. 47: 313–316. CARVALHEIRO, L. G., C. L. SEYMOUR, R. VELDTMAN, AND S. W. NICOLSON. 2010. Pollination services decline with distance from natural habitat even in biodiversity-rich areas. J. App. Ecol. 47: 810–820. CASCANTE-MARIN, A., J. H. D. WOLF, J. G. B. OOSTERMEIJER, J. C. M. den NIJS, O. SANAHUJA, AND A. DURAN-APUY. 2006. Epiphytic bromeliad communities in secondary and mature forest in a tropical premontane area. Basic Appl. Ecol. 7: 520–532. CATLING, P. M. 1997. Influence of aerial azteca nests on the epiphyte community of Some Belizean orange orchards. Biotropica 29: 237–242. DEJEAN, A., I. OLMSTED, AND R. R. SNELLING. 1995. tree-epiphyte-ant relationships in the low inundated forest of Sian-Kaan Biosphere Reserve, Quintana-Roo, Mexico. Biotropica 27: 57–70. EHMER, B., AND B. HOLLDOBLER. 1995. Foraging behavior of Odontomachus bauri on Barro Colorado Island, Panama. Psyche 102: 215–224. FLOREN, A., A. BIUN, AND K. E. LINSENMAIR. 2002. Arboreal ants as key predators in tropical lowland forest trees. Oecologia 131: 137–144. FRANK, J. H., AND L. P. LOUNIBOS. 2009. Insects and allied associated with bromeliads: a review. Terres. Arthro. Rev. 1: 125–153. FREDERICKSON, M. E., A. RAVENSCRAFT, G. A. MILLER, L. M. A. HERNANDEZ, G. BOOTH, AND N. E. PIERCE. 2012. The direct and ecological costs of an ant-plant symbiosis. Am. Nat. 179: 768–778. HEIL, M. 2008. Indirect defence via tritrophic interactions. New Phytol. 178: 41–61. HEIL, M., AND D. MCKEY. 2002. Protective ant-plant interactions as model systems in ecological and evolutionary research. Ann. Rev. Ecol. Evol. Syst. 34: 425–453. HOBALLAH, M. E., T. G. KOLLNER, J. DEGENHARDT, AND T. C. J. TURLINGS. 2004. Costs of induced volatile production in maize. Oikos 105: 168– 180. van HULTEN, M., M. PELSER, L. C. van LOON, C. M. J. PIETERSE, AND J. TON. 2006. Costs and benefits of priming for defense in Arabidopsis. Proc. Roy. Soc B 103: 5602–5607. JANZEN, D. H.. 1966. Coevolution of mutualism between ants and acacias in Central America. Evolution 20: 249–275. KLEIJN, D., AND W. J. SUTHERLAND. 2003. How effective are European agrienvironment schemes in conserving and promoting biodiversity? J. Appl. Ecol. 40: 947–969. LANDIS, D. A., S. D. WRATTEN, AND G. M. GURR. 2000. Habitat management to conserve natural enemies of arthropod pests in agriculture. Annu. Rev. Entomol. 45: 175–201. MARQUIS, R. J. 1984. Leaf herbivores decrease fitness of a tropical plant. Science 226: 537–539. MAYBERY, D., L. CRASE, AND C. GULLIFER. 2005. Categorising farming values as economic, conservation and lifestyle. J. Econ. Psych. 26: 59–72. ~ , C., R. DIRZO, AND A. FLORES. 1997. Structural Parasitism of an MONTANA Epiphytic Bromeliad upon Cercidium praecox in an Intertropical Semiarid Ecosystem. Biotropica 29: 517–521. OSSES, F., E. G. MARTINS, AND G. MACHADO. 2007. Oviposition site selection by the bromeliad-dweller harvestman Bourguyia hamata (Arachnida: Opiliones). J. Ethol. 26: 233–241. PHILPOTT, S. M. 2005. Changes in arboreal ant populations following pruning of coffee shade-trees in Chiapas Mexico. Agroforest. Syst. 64: 219– 224. R DEVELOPMENT CORE TEAM. 2011. R: A language and environment for statistical computing. R. F. f. S. Computing (Ed.), Vienna.
82
Hammill, Corvalan, and Srivastava
RASBAND, W. S. 2012. ImageJ 1.46. U.S. National Institute of Heath (Ed.). Bethesda, Maryland, USA. RICKSON, F. R., AND M. M. RICKSON. 1998. The cashew nut, Anacardium occidentale (Anacardiaceae), and its perennial association with ants: Extrafloral nectary location and the potential for ant defense. Am. J. Bot. 85: 835–849. SALAZAR, D., AND R. J. MARQUIS. 2012. Herbivore pressure increases toward the equator. Proc. Natl. Acad. Sci. U. S. A 109: 12616–12620. SCHAUFF, M. E., J. LASALLE, AND G. A. WIJESEKARA. 1998. The genera of chalcid parasitoids (Hymenoptera: Chalcidoidea) of citrus leafminer Phyllocnistis citrella Stainton (Lepidoptera: Gracillariidae). J. Nat. Hist. 32: 1001–1056. STANTON, M. L., AND T. M. PALMER. 2011. The high cost of mutualism: effects of four species of East African ant symbionts on their myrmecophyte host tree. Ecology 92: 1073–1082.
TILMAN, D., J. FARGIONE, B. WOLFF, C. D’ANTONIO, A. DOBSON, R. HOWARTH, D. SCHINDLER, W. H. SCHLESINGER, D. SIMBERLOFF, AND D. SWACKHAMER. 2001. Forecasting agriculturally driven global environmental change. Science 292: 281–284. TIMMER, L. W., AND L. W. DUNCAN. 1999. Citrus health management. The American Phytopathological Society, St Pauls, MN. TOLEDO-ACEVES, T., J. G. GARCIA-FRANCO, A. HERNANDEZ-ROJAS, AND K. MACMILLAN. 2012. Recolonization of vascular epiphytes in a shaded coffee agroecosystem. App. Veg. Sci. 15: 99–107. WALLING, L. L. 2000. The myriad plant responses to herbivores. J. Plant Growth Regul. 19: 195–216. YANOVIAK, S. P., S. M. BERGHOFF, K. E. LINSENMAIR, AND G. ZOTZ. 2011. Effects of an epiphytic orchid on arboreal ant community structure in panama. Biotropica 43: 731–737.
