Ant Plant Coevolution: A Systematic Evaluation of the Myrmelachista haberi (Formicinae) and Ocotea sp. (Lauraceae) Symbiosis Murtaza Paghdiwala Department of Biology, Northeastern University ABSTRACT I have evaluated the relationship between Myrmelac hista haberi and Ocotea sp. (Los Llanos) to provide definitive evidence supporting the nature of the symbiosis and possible explanations in terms of coevolutionary history. T his interaction has been described in terms of benefit to the ant Ã domatia and nu trition via phloem feeder farming Ã but benefits from the plants perspective have gone largely unstudied. McNett (2009) attempted to quantify the nature of the relationship by monitoring the uptake and movement of stable isotopes through each of the three participating species. This study found that nutrients do not transfer between the organisms , but left the possibilities for alternative vectors of mutualism open. In our study I have systematically evaluated this interaction in terms of all conventi on al a nt plant services: (1) the tendency of the ants to clean the leaves of foreign (invasive) matter, (2) the foraging behavior of the ants, and (3) the effect the presence of ants has on herbivory and response to disturbance. Myrmelachista haberi do not remov e foreign matter from the leaves, do not prefer to feed on the leaves, nor do they respond to herbivory damage on the leaves. My results provide definitive evidence supporting that the nature of the relationship is not a mutualism. RESUMEN He evaluado l a relaciÂ—n entre Myrmelachista haberi y Ocotea sp. (Los Llanos) para proporcionar apoyo a pruebas definitivas de la naturaleza de la simbiosis y las posibles explicaciones en tÂŽrminos de coevolutionary historia. Esta interacciÂ—n ha sido descrita en tÂŽrmino s de beneficio para la hormiga domatia Floema y nutriciÂ—n a travÂŽs de la agricultura alimentador pero los beneficios de la perspectiva de las plantas han pasado en gran medida sin estudiarse. McNett (2009) intentÂ— cuantificar la naturaleza de la relaci Â—n por el control de la absorciÂ—n y circulaciÂ—n de los isÂ—topos estables a travÂŽs de cada una de las tres especies. Este estudio encontrÂ— que los nutrientes no se transfieren entre los organismos, pero dejÂ— las posibilidades de vectores alternativos de mut ualismo abierto. En nuestro estudio he evaluado sistemÂ‡ticamente esta interacciÂ—n en tÂŽrminos convencionales de todos los servicios de hormiga planta: (1) la tendencia de las hormigas para limpiar las hojas de los extranjeros (invasoras), la materia, (2) e l comportamiento de forrajeo de las hormigas, y ( 3) el efecto de la presencia de hormigas sobre la herbivorÂ’a y la respuesta a las perturbaciones. Myrmelachista haberi no eliminar materias extraÂ–as de las hojas, no prefieren que se alimentan de las hojas, ni responder a los daÂ–os herbivorÂ’a en las hojas. Nuestros resultados proporcionan pruebas definitivas de que el apoyo a la naturaleza de la relaciÂ—n no es un mutualismo. INTRODUCTION Cospeciation of plants and insects has contributed strongly to the di sproportionately high diversity compared to other taxa of both groups (Thompson 1989). While most insect plant interactions are closely tied to pollination success, there are also many interactions that have evolved based on a plant's need for herbivor e defense and an insects need for domatia. Many ant plant interactions fall into this category; phylogenetic studies of ant plant associations exhibit host shifts and de novo colonizations (Quek, et al. 2004). Ant plant interactions are often divided into two broad categories: (1) direct ant plant interactions, in which the plant provides the ant with resources such as food and/or shelter, and (2) indirect ant plant interactions, which depend largely on the presence and farming of phloem feeders (H omoptera ) (Rico Gray & Oliveira 2007 ). Most direct interactions involve the provision of food resources and domatia by the plant in return for defense against herbivores and encroaching shrubs and
epiphytes. The now classic example of such an interaction is the Pse udomyrmex Acacia association. The plant provides the ants with domatia and carbohydrates and in return the ants aggressively defend the plant against herbivory and also clear away plant competitors (Janzen 1966). In indirect interac tions, the ants farm Hom optera n scale bugs within the hollowed out stems and trunks and utilize their honeydew for nutrition. Not all interactions, however, fit neatly into one category or the other ; rather there is a broad spectrum of associations that suggest an intricate coev olutionary past . In the case of the Azteca Cecropia interaction there is a complex association between three organisms . T he plant ( Cecropia) provides the ants ( Azteca ) with fats , the ants farm scale bugs (Homopterans) within the trunk for sugars, and lastl y the ants are left to forage on the leaves for herbivores as a protein source ( Janzen 1969 ). The range of ant plant interactions extends from opportunistic generalists , such as Crematogaster spp. , that will nest in almost any tree even dead ones solel y for domatia all the way to obligate mutualism where death of tree and death of the colony go hand in hand (Rico Gray & Oliveira 2007) . One understudied group of ant plant interactions is the Myrmelachista (Formicinae) association s . Myrmelachista ants a re best known for their creation of Ã” Devil's Gardens '. In these gardens, the ants use formic acid as an herbicide excreted from the gaster to kill all plants besides their host , thereby creating large monoculture stands within the Amazon (Fredrickson et al . 2005) . Myrmelachista also has symbiotic relationships with Ocotea (Lauraceae) species . Stout (1979) fir st described this relationship in M. flavocotea and what was probably a mix sampling of O. atirrrensis and , suggesting a tight interaction between plan t ant and scale bug dendrodaphne (Hammel 1986; Longino 2006) . Since then little work has been done on this particular association. Longino (2006) completed a taxonomic review of the Myrmelachista genus, allowing us to identify individuals to species leve l. Most recently, McNett (2009) attempted to quantify the nature of the interaction by using stable isotopes to monitor movement of nutrients through the system. McNett (2009) could not definitely provide evidence of mutualism or parasitism base d on the st able isotope uptake alone. In this study , I have performed a systematic evaluation of the interaction between Myrmelachista haberi and its host Ocotea sp. (Los Llanos) in order to provide definitive evidence describing the symbiotic nature of the associat ion. It is well known that the ants use these particular plants for domatia. The plants do not provide any other resources (e.g. food) to the ants. And it is known that the ants farm phloem feeders inside the hollow stems of the plants (Stout 1979) . The an t clearly benefits from this relationship. I have , for this reason, evaluated the association in terms of benefit to the plant. I studied: (1) the tendency of the ants to clean the leaves of foreign (invasive) matter, (2) the foraging behavior of the ants , and (3) the effect the presence of ants has on herbivory and response to disturbance . Each of these parameters demonstrate the ants tendencies towards their hosts in order to provide evidence as to where the symbiosis lies on a scale of parasitism , in whi ch one or more partners benefit at the net cost of the rest, to mutualism , where each partner provides a service as well as a reward such that the net fitness for all partners is increased (Bronstein 1994; 2001) . This study provides a missing piece to the co evolution of non pollination related plant insect interactions. I predict that M. haberi do es not clean or protect Ocotea sp. (Los Llanos) and that the tree is left to defend itself by other means besides the ants . I predict to find evidence pointing to wards commensalism or possibly parasitism. MATERIALS AND METHOD NATURAL HISTORY Ã Myrmelachista haberi is an obligate inhabitant of understory Lauraceae. This species is found in two general areas in Costa Rica: (1) Cordillera de TilarÂ‡n, 1000 Ã 1500m elev ation, and (2) Cordillera VolcÂ‡nica Central, 1100m elevation. Colonies reside within hollow stems of their host trees and are seen with low frequency on the leaves and stems of the tree. Colonies are monogynous, with the queen found alone in a separate cha mber near the shoot tip. The biology of this species is similar to the more widely studied M. flavocotea (Longino 2006). Ocotea sp. (Los Llanos) has potential to be a canopy tree (15 30 m) but is more commonly found as an understory treelet. This tree is e xtremely abundant in the Los Llanos, Monteverde area, 1200 1450m
elevation (see STUDY SITE ). Twigs are mottled brown, lenticilate and round. The leaves are thick and leathery (5 x 15 cm) with midvein raised above and below. The inflorescence is clustered n ear twig tips, but the tree flowers very rarely in the understory (Haber 2001 ). The species closely resembles O. tonduzii with a narrower leaf blade. STUDY SITE T his study was done in a patch of primary forest in Los Llanos located west of Santa Ele na de Monteverde, Costa Rica. This pre montane Atlantic Slope cloud forest ranges from 1300m 1500m in elevation and receives an annual 2.75m precipitation. This patch has approximately nine different species of Ocotea but I have focused solely on Ocotea sp . (Los Llanos ). The study was conducted from 20 April 2009 to 10 May 2009. Experiments were run in the morning during peak foraging activity. The a nts were active rain or shine, as was I. REMOVAL OF FOREIGN MATTER Ã Ants involved in mutualisms with trees often demonstrate maintenance behavior s towards their plant . I quantified the presence of such a behavior in our system of interest. I selected 15 Ocotea sp. (Los Llanos) individuals and put a small (>0.5 cm 2 ) smudge of clay on five leaves per plant. I mo nitored response and removal of the clay over a 1 hour period and then again after 24 hours. I then repeated the same experiment on the apical meristem of each of the 15 trees and monitored cleaning over the same time periods. I counted a cleaning as the p resence of clay moved to a location on or off the tree different from the initial placement. FORAGING Ã I then studied the foraging abilities and preferences of the ants. S tudies have shown that tuna in oil is a standard ant bait (McNett 2009, Sagers, et al. 2000) . I placed canned tuna (< 1g) on one leaf per plant and monitored the ability o f the ants to find and remove the tuna baits, aga in checking at one and 24 hours after placement on the leaf . The same amount of tuna was also placed on 15 apical meris tems of each plant and monitored in the same manner as on the leaves . Apical meristem and leaf trials were run independently of each other. ANT EXCLUSION AND HERBIVORY Ã I then sought to study the effect of ant presence or absence on the rate of leaf her bivory by other insects. I selected 26 of the newest leaves (still red from delayed greening) found on 13 plants (one control and one exclusion per plant) within our study site. Of the 26 leaves, ants were excluded from 13 of them by applying sticky tack r emoved from paper mouse traps around the stem or petiole. The remaining 13 were left unmanipulated. Each leaf was photographed with a Nikon D50 and a 60mm macro lens before and after treatment. The experiment was allowed to run for one week and change in l eaf area was calculated using a standardized 1cm 2 grid and Photoshop CS2. RESPONSE TO DISTURBANCE Ã Finally , I measured increased ant activity upon disturbance at both the apical meristem and at the leaves. To simulate predation at the apical meristem I shook 20 plants by the trunk for 10 seconds each and then watched for increased ant presence within a five minute period. To simulate leaf herbivory I cut 30 leaves across 15 plants and watched for increased ant presence within a five minute and 1 hour per iod. I counted an increase in ant presence as an increase of three individuals in the disturbed area within the two time periods. DATA ANALYSIS Ã Removal of foreign matter, foraging preference, and response to disturbance data were analyzed by ! 2 comparisons. Changes in herbivory rates for exclusion and control were compared using a paired t test. RESULTS CLEANING Ã Figure 1 A shows removal of foreign matter from both the leaves and the apical meristems. In th e case of the leaves, only one leaf out of 75 showed any evidence of clay removal ( ! 2 = 71.05, P < .0001 , df = 1). On the apical meristem , clay was re moved away from 11 of the 15 stems , however the
number of plants that clay was removed from was not more t han predicted by chance ( ! 2 = 3.27, P < .071 , df = 1; Fig. 1B) . FORAGING Ã Baits were placed on the leaves and on the apical meristems to survey foraging behavior. Two of seven baits placed on leaves were found by th e ants in a 24 hour period but ants did not removed baits more often than chance from leaves ( ! 2 = 1.29, P < .26, df = 1 Fig. 2A ). A ll seven baits placed on the stem had ants foraging on them within an hour ( ! 2 = 7, P < .008, df = 1; Fig. 2 B ). FIGURE 1. (A) Clay was placed on five leaves and the apical meristem per plant. Evidence of moved clay (balls made by workers or clay found away from initial position) was counted as a removal . (B) Ants moved clay away the apical m eristems (11 of 15) in some cases but almost no cleaning of leaves (1 of 75). Number of Individual Leaves Number of Individual Stems A B
FIGURE 2. Foraging behavior based on the ants' ability to find baits on (A) leaves and (B) stems. Ants were able to locate baits on apical meristems in under an hour (7 of 7) bu t only 2 of 7 baits were found on the leaves after 24 hours. Number of Individual Leaves Number of Individual Stems A B
HERBIVORY AND PREDATION DISTURBANCE Ã After one week, leaves where ants were excluded and were not excluded did not differ in amount of herbivory (t = .035, df = 12, P = .381; Fig. 3). In response to simulated herbivory and disturbance, ants d i d not show increased attendance to either leaves or apical meristem (Leaves: ! 2 = 19.2, P < .0001, df = 1; Meristems: ! 2 = 5, P < .025, df = 1; Figs. 4A and B). FIGURE 3. Herbivory on leaves compared between leaves with ants excluded and leaves with ants present. Rates of change in herbivory do not vary significantly between the two treatments (n = 26) . Ants d o not play a role in leaf defens e. Change in Leaf Area (cm 2 )
FIGURE 4. (A) Leaves were cut to simulate herbivory damage and (B) the stems were shaken to mimic predation, increased ant presence was monitored at each location post disturbance over 1 hour and 5 minute periods, respectively. There was no increase in ant abundance at either location. Number of Individual Stems Number of Individual Leaves B A
OBSERVATIONS Myrmelachista haberi lives in the hollowed out stems of Ocotea sp. (Los Llanos). I hypothesize that the ants hollow out the stems themselves. This is based on observations that saplings without ant colonies are solid (n = 2), very young saplings with ants are only partially hollowed ( n = 2) and the fact that none of the branches are hollow while the ants do not inhabit the branches. The ants can most often be seen at the apical meristem, where the main entrance hole is also located. Ants can also occasionally be seen on leaves most c ommonly on the youngest ones and lower on the trunk as well. Ants were also observed on the forest floor within close proximity of their host tree. The ants showed conspecific aggression upon the transfer of an individual from a different colony, but lit tle to no interspecific aggression was observed and different ant species were commonly seen on the leaves and trunk. No ants were seen in dead Ocotea sp. in Los Llanos (n=3). Ocotea sp. (Los Llanos) has a characteristic growth pattern where branches are completely parallel to the forest floor and leaves present only at the very top of the growing end, resulting in a long bare trunk and a large crown at the top. The leaves have a very strong characteristic odor and were plagued by epiphylls . M any individua ls were hosting vines and lianas as well. Growth was very slow in the understory and most plants held from zero to only two or three new leaves during the study period. The plants do not provide any resources to the ants besides domatia. DISCUSSION Ants protecting the ir host tree s from foreign matter with the potential to decrease photosynthetic ability, such as competitive epiphylls and epiphytes, is one of the basic services a nts can provide (Rico Gray & Oliveira 2007) . In the foreign matter removal ex periments I was able to confidently show that M. haberi does not clean the leaves of its host (Fig. 1 A ) . The result that M. haberi removed the clay from the apical meristem but not the leaves provides evidence that the ants noticed it as a foreign object ( Fig. 1B) . However, if M. haberi were in a mutualism with Ocotea. s p . they would have attempted to remove it from the leaves as well since foreign objects may have a direct fitness cost to the plant. In addition to the clay removal data, I observed heavy vi ne and epiphyll cover on the majority of sampled trees , this strengths the argument that the ants do not maintain the tree . The ants demonstrated no behavioral tendencies to remove foreign matter or plant competitors to release their hosts from competitive stress. I f the ants do not directly tend to the plant by cleaning it of debris and invaders, perhaps the plants benefit indirectly as a result of the ants feeding on herbivores from the leaves. If the ants eat the majority of their food from the leaves of the plants the plants should benefit from herbivore removal (Janzen 1966) . Additionally, some ant plant associations are characterized by plant uptake of ant derived nutrients; ants live within the plants and the plants are able to utilize solutes in the ant deposited fecal matter (Gay 1993 ; Sagers et al. 2000 ). A nt plant co evolutionary theory suggests that leaf defense varies with age: old leaves being mostly chemically defended but young leaves having mostly biotic defense (Folgarait and Davidson 1994 ; 1995 ). I found no preference in the ants foraging ability and behavior to forage on the leaves (Fig. 3) . McNett (2009) performed a stable isotope uptake analysis to evaluate the feeding of Ocotea by Myrmelachista but found no evidence to suggest that it oc curs. The ants do not prefer to forage on the leaves, nor are the plant s able to utilize nutrition from fecal deposition that results from ant foraging. Lastly, I looked at the ants response to herbivory and predation disturbance. Another characteristic a nt plant service is the aggressive defense against insect and mammalian herbivory. In the Pseudomyrmex Acacia interaction the ants respond quickly to disturbance at both the stems and the leaves by filing out of their domatia to defend the plant again s t at tack (Janzen 1966). Myrmelachista haberi did not respond to herbivory by increasing activity to investigate the disturbance or remove the potential predator. The youngest leaves were used for the same reason as above. After a week long exclusion, there was no statistical difference in herbivore damage between the treatment and control (Fig. 3). The exclusion took place in the start of the rainy season, a time where insects are abundant and
new leaves are most vulnerable. It is possible Ocotea sp. possesses chemical defenses Ã perhaps the strong odor of the leaves supports this as well, however chemical defense in Ocotea needs to be further examined. I then simulated both herbivory at the leaves and predatio n on the ant colony at the stem. I n both cases I fo und that the ants failed increase presence in response to the disturbance (Fig s . 4 A and B ) . The majority of ant species involved in ant plant mutualisms have morphological characteristics that equip them for defense of the plant they inhabit (Rico Gray & O liveira 2007). Myrmelechista haberi is an extremely small ant and ill equipped for defense Ã lacking both biting mandibles and stingers. It therefore is logical that the ants would not rush out to attempt to defend the tree as they would prove to be a wea k line of defense. Observationally, interspecific aggression was also non existent. Various other ant and insect species were often observed on the sample trees. The presence of M. haberi does not have an effect on herbivory nor do the ants rush to protect the tree against damage. Ant plant associations are characterized by specific vectors of interaction. For example, the plant provides a limited nutritional array and the ants complete their diet by foraging for herbivores (Janzen 1966, Downhower 1975). The plant may adapt a hollow growth form to facilitate the residence of ants. Some plants can utilize ant derived nutrients deposited in the interior of the plants. Some ants will reside in a tree dead or alive. I have systematically evaluated the most com mon vectors of ant plant interaction in the Myrmelachista haberi Ocotea sp. (Los Llanos) association in order to provide evidence to support the nature of this particular symbiosis. I have found definitive evidence across the board that the ants do not pro tect or aid the plant in any conventional way. It has been suggested that the scale bug interaction may play a larger role in this system than is known, the ants may even consume the coccoids they farm for a protein source (Stout 1979; McNett et al. 2009) . The scale bug interaction also explains, why , if the ants only use the tree for domatia, they were not present in dead individuals. The farming of phloem feeders would be impossible in a dead tree. The ants were observed on the forest floor and I hypothe size that the ants forage on the floor at least as much as the y forage on the tree Ã reducing the herbivo re removal effect even further. In spite of weak to no biotic defense the plant is able to grow abundantly and defend itself. Studies need to be done t o evaluate the loss of plant fitness as a result of phloem farming and stem hollowing, these factors could imply a parasitism. I have provided evidence that suggests that this interactio n appears to be outside of the mutualism range on the symbiotic spectr um . ACKNOWLEDGMENTS I would like to thank Eladio Cruz, without whom I would never have been able to locate my plants. I would like to thank Moncho for his willingn ess to help with and provide supplies. Without Yimen's statistical knowledge and general i nterest in my project it would not have gone as smoothly as it did Ã thank you Yimen. Thank you Anjali for advising me and helping me to masticate this once arduous task. LITERATURE CITED B RONSTEIN , J. L. 1994 . Our current understanding of mutualism. Q . Rev. Biol., 69: 31 Ã 51 . B RONSTEIN , J. L. 2001 . The exploitation of mutualisms. Ecol. Lett., 4 : 277 Ã 287 . F OLGARAIT , P. AND D. W. D AVIDSON . 1994. Biotic and chemical defenses of myrmecophytic Cecropia in different light regimes. Oikos 71: 305 320. F OLGARAIT , P. AND D. W. D AVIDSON . 1995. Myrmecophytic Cecropia : antiherbivore defenses in different nutrient regimes. Oecologia 104:189 206. GAY H. 1993. Animal fed plants: an investigation into the uptake of ant derived nutrients by the far eastern epiphytic fern Lecanopteris Reinw . (Polypodiaceae). Biol. J. Linn. Soc . 50: 221 233. HABER , W.A. 2001. Plants of Monteverde: Identification and other resources. < www.cs.umb.edu/efg> JANZEN , D. H. 1966 . Coevolution of mutualism between ants and acacias in Central Americ a. Evolution 20 : 249 Ã 275 . JANZEN D.H. 1969. Allelopathy by myrmecophytes: the ant Azteca as an allelopathic agent of Cecropia . Ecology. 50 : 147 153. LONGINO , J. T. (2006). A Taxonomic review of the genus Myrmelachista (Hymenoptera: Formicidae) in Costa Ric a . Zootaxa 1141 :1 54.
MCNETT , K., et al . 2009. Stable isotope investigation of a cryptic ant plant association. In progress. QUEK , S.P., et al. 2004. Codiversification in an Ant Plant Mutualism. Evolution, 58 : 3 , 554 570. RICO GRAY V. and Oliveira P.S. 20 07. The Ecology and Evolution of Ant Plant Interactions. University of Chicago Press, Chicago, USA. SAGERS , C.L., et al. 2000. Carbon and nitrogen isotopes trace nutrient exchange in an ant plant mutualism. Oecologia . 124:4. 582 586. STOUT , J. 1979. An as sociation of an ant, a mealy bug, and an understory tree from a Costa Rican rain forest. Biotropica . 11:4. 309 311. THOMPSON , J. N. 1989 . Concepts of coevolution. Trends Ecol. Evol, 4 : 179 Ã 183 . THRALL , P. H., et al. 2006. Coevolution of symbiotic mutualists and parasites in a community context. Ecology . 22:3. 120 126. VASCONCELOS H.L. and Casimiro A.B. 1997. Influence of Azteca alfari ants on the exploitation of Cecropia trees by a leaf cutting ant. Biotropica 29 :84 92.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00103
Co-evolucin entre la planta y la hormiga: una evaluacin sistemtica del Myrmelachista haberi (Formicinae) y el simbiosis de Ocotea sp. (Lauraceae)
Ant-plant coevolution: a systematic evaluation of the Myrmelachista haberi (Formicinae) and Ocotea sp. (Lauraceae) symbiosis
I have evaluated the relationship between Myrmelachista haberi and Ocotea sp. (Los Llanos) to provide definitive evidence
supporting the nature of the symbiosis and possible explanations in terms of co-evolutionary history. This interaction has been described in terms of benefit to the ant-domatia and nutrition via phloem-feeder farming but benefits from the plants perspective have gone largely unstudied. McNett (2009) attempted to quantify the nature of the relationship by monitoring the uptake and movement of stable isotopes through each of the three participating species. This study found that nutrients do not transfer between the organisms, but left the possibilities for alternative vectors of mutualism open. In our study I have systematically evaluated this interaction in terms of all conventional ant-plant services: (1) the tendency of the ants to clean the leaves of foreign (invasive) matter, (2) the foraging behavior of the ants, and (3) the effect the presence of ants has on herbivory and response to disturbance. Myrmelachista haberi do not remove foreign matter from the leaves, do not prefer to feed on the leaves, nor do they respond to herbivory damage on the leaves. My results provide definitive evidence supporting that the nature of the relationship is not a mutualism.
He evaluado la relacin entre Myrmelachista haberi y Ocotea sp. (Los Llanos) para proporcionar apoyo a pruebas definitivas de la naturaleza de la simbiosis y las posibles explicaciones en trminos de la historia co-evolutiva.
Text in English.
Mutualism (Biology)--Case studies
Mutualismo(Biologa)--Estudios de casos
Tropical Ecology Spring 2009
Ecologa Tropical 2009
t Monteverde Institute : Tropical Ecology