Differences in Atta cephalotes Foraging Rate and Amount of Substrate Harvested Following the Introduction of an Antifungal Agent Benjamin Hedin Department of Environmental Science, Allegheny College ABSTRACT Optimal foraging theory dictates animals w ill behave in the most energetic favorable fashion, maximizing energy gained while minimizing energy lost. However, exceptions do exist, such as mating behaviors and predator avoidance. I show that oat flakes contaminated with antifungal powder, simulatin g secondary compounds, are selected less and at a lower rate by a colony of Atta cephalotes . This demonstrates the ability of the colony to recognize and discriminate against the contaminant to protect their symbiotic fungus. It also suggests that leaf cu tters, to an extent, can detect the amount or toxicity of the secondary compound. The willingness of the ants to take the antifungal flakes suggests cleaning is involved, possibly with minims, providing support for the hitchhiking cleaning process. Final ly, the harvesting of sub optimal resources might represent an attempt to sustain yields over an extended period of time. These conclusions suggest that we need to reconsider our current models of Atta cephaloies foraging behaviors for they might be more complex than we currently assume . RESUMEN La teorÃa de forrajeo Ã³ptimo dice que los animales deben tener un comportamiento energÃ©tico de la manera mÃ¡s favorable, maximizando la energÃa ganada y minimizando la energÃa perdida. Sin embargo, excepciones ex isten, como el comportamiento reproductivo y evitar a los depredadores. Muestro que hojuelas de avena contaminadas con un talco antifÃºngico, simulando compuestos secundarios, son seleccionados en menor cantidad y en una menor taza por una colonia de Atta cephalotes . Esto demuestra la habilidad de la colonia para reconocer y discriminar en contra de contaminantes para proteger a su hongo simbiÃ³tico. TambiÃ©n sugiere que las zompopas, pueden detectar la cantidad o toxicidad de compuestos secundarios. La disposiciÃ³n de las hormigas para tomar las hojuelas con la sustancia anti hongos sugiere una limpieza, posiblemente por las mÃnimas, apoyando la teorÃa que dice esto. Finalmente, la colecta de recursos por debajo del Ã³ptimo puede representar un intento p ara aumentar el rendimiento sobre perÃodos de tiempo extensos. Estas conclusiones sugieren que tenemos que reconsiderar los modelos actuales de forrajeo de Atta cephalotes , y que estos pueden ser mÃ¡s complejos de lo que pensamos. INTRODU C TION Every orga nism strives to maximize its energy gain to energy loss ratio while foraging or hunting (MacArthur & Pianka 1966). Optimal Foraging Theory suggests that a behavior will continue as long as the energy gained is greater than the energy spent (MacArthur & Pi anka 1966). The activities that maximize energy gain will be then be allocated more time. In theory, natural selection will favor those organisms that participate in more energetically favorable activities, maximizing gains while minimizing losses, and u ltimately realizing increased reproductive success. Mechanisms to increase this energy ratio are therefore modified and refined from generation to generation (MacArthur & Pianka 1966).
Optimal Foraging Theory can be applied to many aspects of behavior. S ituations suggested to be governed by Optimal Foraging Theory fall into four broad categories; (1) diet; (2) optimal feeding location; (3) time allocation in different feeding locations; and (4 ) patterns and speed of movements (Pyke et al. 1977). This the ory can be observed in tropical ecosystems while looking at l eaf cutting ants . These ants often travel long distances to harvest leaf fragments for their subterranean fungus gardens and will forage more selectively when further from the nest ( H Ã¶ lldo b ler & Wilson 1990) . Food with higher nutritional and caloric value is necessary to offset the energy used in the increased foraging time. However, if we remove ourselves from a solely energy based thought process, we observe that these ants are incredibly sel ective in their foraging habits for other possible reasons. Although leaf cutting ants are the dominant herbivores in tropical forests ( HÃ¶lldobler & Wilson 1990 ), a colony of Atta cephalotes ants have been observed to take just 17 of 332 plant species in a given patch of forest ( HÃ¶lldobler & Wilson 1990 ). It has been observed that leaf cutter ants choose plants selectively while foraging. Some qualities of a plant that may influence whether the colony forages on a particular plants are: secondary compoun ds in the leaves, or physical characteristics such as toughness, water content and abundance of sap (Rockwood and Hubbell 1987). Although all of these factors play a role in selection, this process is thought to be most influenced by secondary compounds ( Rockwood & Hubbell 1987, Howard 1987, HÃ¶lldobler & Wilson 1990 ). Secondary compounds consist of chemicals produced by plants to deter herbivory ( HÃ¶lldobler & Wilson 1990). Most of these detrimental secondary compounds are terpenoids and are detrimental t o either for forager or fungus ( HÃ¶lldobler & Wilson 1990 ). While secondary compounds can be specific, such as antifungal agents (Rockwood & Hubbell 1987), leaves can also carry microbial contaminants that deter foraging (Griffiths & Hughes 2010). The leav es leaf cutter ants harvest are used to promote the growth of a symbiotic fungus (Basidiomycota) that is fed to the growing ant larvae. An antifungal secondary compound introduced into these fungus gardens would kill the fungus and eventually destroy the colony. It seems as if foragers select substrates that are conducive for fungus survival and discriminate against those which do not ( HÃ¶lldobler & Wilson 1990 , Ridley, et al . 1996). Foragers, though, are not the only sub caste involved with the plant mate rial outside the nest. A caste of smaller ants, minims, can be observed inside and outside the nest. A frequent behavior outside the nests involves these small ants hitchhiking on leaf fragments carried by foragers. This behavior occurs for a number of proposed reasons, one of them being sanitation, reducing the contaminant load of leaf fragments (Griffiths & Hughes2010). In this study, A . cephalotes ants were tested for their response to a collectable, but contaminated food source. A question still rem ains as to how A . cephalotes foraging columns respond to and deal with contaminants. The rejection of food sources is not as complete as might be suggested, with A . cephalotes initially accepting food sources with secondary compounds or contaminants and th en rejecting them after feedback loops have time to operate. Howard et al . 1988, Howse & Jackson 1996). In this experiment I observed the response of an A. cephalotes colony when presented with a food source contaminated with a small amount of antifungal talc powder to act as a secondary compound. The response will give us clues on how secondary compounds are detected, the degree to which they are discriminated against, and if cleaning is a possible coping strategy.
METHODS Study Site A colony of A. c ephalotes in San Luis, Monteverde, Costa Rica was used in this experiment during the end of the wet season. Data was collected from November __ to November __ 2010. The study area was located in the premontane wet Holridge life zone at 1100 meters above sea level. The colony had multiple active foraging trails and the visible nest was about four meters in diameter. The trail used was through a section of secondary growth forest in a cow pasture with a low and relatively open canopy consisting of mostly guava (Myrtaceae: Psidium guajava) and lemon trees (Rutaceae: Citrus limon ). The ant column remained un obscured for 15 meters through the patch of forest with no major blockages to change trail dynamics for the three weeks of data collection. Introduct ion of an Antifungal Agent To test my question, I used oat flakes to simulate leaves because ants will readily accept them as a food source ( Hoelldohler and Wilson 1990). I used BioLand B ran R ich W hole O atmeal in this study. A control (oat flakes soaked in water) and two treatments , talc powder ( Johnson a nd Johnson baby powder) and antifungal pow d er (Neutrodor powder with an antifungal agent) were used . One tablespoon of each powder type was mixed with half a cup of water and an equal amount of either gr een or blue food dye was added for identification purposes . Talc treatments were typically green, antifungal blue, and half of the control sample size was dyed green and the other half blue. The dye was the easiest way to identify the different treatment s in the field. Oat flakes were then dipped in these solutions and dried either overnight or in the sun on the day of the trials. After the flakes were completely dry, trials were carried out on the same ant column 12 to 13 meter s from the nest entrance . Only two oat flake treatments were used in a trial at one time. Oat flakes were placed on opposite sides of the trail in two parallel columns. Flakes were placed so that ants traveling on the trail would encounter them but far enough from the center t hat no interference occurred. Equal numbers of each treatment were used in each row. For example, 20 antifungal and 20 control flakes were used, 10 in each column. The flakes were places three centimeters apart and the treatments were mixed randomly wit hin columns (Fig. 1). Observations were taken every two to three minutes noting the number of oat flakes remaining from each treatment group. Each trial was run until either all the oat flakes were taken or until one hour had passed since the start of th e trial. The order of the trials was randomized to avoid recruiting effects, and therefore changed trail dynamics. If there was residue powder on the trail, a different part of the trail was used, still within the one meter test area. Weather conditions and time were noted at the beginning of every trial. antifungal leaves. Ten of each treatment were used and set up the same way as mentioned before, but with 5 in each c olumn and not randomly placed. Every time a flake was taken, it was replaced with one of the same treatment consistently offering the ants with consistently the same flakes to choose from. Additionally, every 5 minutes the front oat flake from each colum n was moved to the back to avoid edge effects. The test was run until I was out of oat flakes of one treatment.
RESULTS Ants accepted most of the oat flakes of every treatment as a food source when they were dry; 325 of 330 antifungal oat flakes, 331 of 335 talc oat flakes and 309 of 313 control oat flakes were taken. When comparing the slope of trend lines, control and talc powder flakes did not differ significantly in rate left at the study site ( Tukey Test at p =0.869, Fig. 2a) at 0.36 flakes/minu te and 0.34 flakes/minute, respectively. Control and antifungal flakes did not differ significantly in the rate left at 0.30 flakes/minute and 0.37 flake/minute respectively (p=0.121, Fig. 2b). Finally, talc and antifungal flakes differed significantly a t the rate at which they were left at 0.31 flakes/minute and 0.40 flakes/minute, respectively (p=0.037, Fig. 2c). When comparing the height of the trend lines, control and talc flakes did not differ significantly with elevation values of 13.29 oat flakes and 12.46 oat flakes, respectively (p=0.529, Fig. 2a). Control and antifungal treatments display a significant difference with elevation values of 9.71 and 15.40, respectively (Fig. 2b, p<0.001). Finally, talc and antifungal treatments differed significa ntly with elevation values of 9.87 and 16.63, respectively (Fig. 2c, p<0.001). The Smorgasboard Leaf Disk Assay resulted in 18 antifungal oat flakes and 34 talc oat flakes being taken. A A A A A T T T T T FIGURE 1. Antifungal oat flakes (blue and labeled A) and talc flakes (green and labeled T) are mixed randomly and presented in two parallel columns.
a. b. c. FIGURE 2. A) Cont rol (R 2 =0.408) and talc (R 2 =0.406) oak flakes showed no significant differences in either slope (t=0.166. DF=106, p=0.869) or elevation (t=0.632, DF=107, p=0.529). Equations of trend lines are given on graph. B) Control (R 2 =0.430) and antifungal (R 2 =0.70 2) flakes showed no significant difference in slope (t=1.558, DF=166, p=0.121) but a significant difference in elevat ion (t=6.879, DF=167, p<0.001). C) Talc (R 2 =0.757) and antifungal (R 2 =0.450) flakes showed significant differences in both slope (t=2.096, DF=194, p=0.037) and elevation (t=8.970, DF=195, p<0.001).
DISCUSSION Foraging co lumns of Atta cephalotes are able to partially detect and discriminate against food The ability of ants to discriminate between food sources is well documented ( Howard et al. 1988, Rid ley et al. 1996) , but it is common for a negative feedback loop to act on the ants as a way to control foraging habits. Foraging cues can be from the symbiotic fungus or inferred when foragers die from toxic secondary compounds. These cues, however, take time (Ridley et al. 1996). In my study antifungal leaves were always accepted at a lower rate or lower amount right from the beginning. The ability of ants to sense contaminants on their first en counter with new foraging material counters evidence that a nts most likely rely on relatively simple stimuli or stimulus substrate harvested through feedback loops ( Howard et al . 1988, Ridley et al . 1996). The antifungal tre atment acted as a secondary compound ( Howard et al . 1988) detrimental for the fungus but not the forager. Secondary compounds are chemicals inside leaves that are harmful to either the ant or fungus. In experiments involving secondary compounds, complete rejection is follows an initial period of acceptance (Howard et al. 1988, Ridley et al. 1996). Because these secondary compounds strongly influence harvesting behavior, it would appear that A. cephalotes ants are perhaps more adept at detecting them tha n we currently assume because of their ability to discriminate against secondary compounds, as seen in this study. Although the flakes may have been rejected once inside the nest, plants with secondary compounds are subject to complete rejection after neg ative feedback loops have time to occur (Ridley et al. 1996). In the absence of a negative feedback loop, this decision would have been made based on the amount of contaminants and possibly even the type of contaminant. Results suggest that A. cephalote s ants can take a more active approach to deal with contaminants rather than just passively avoiding them. Eventual rejection and column avoidance of a food source is a viable option for ants to avoid secondary compounds (Ridley et al . 1996). In my study though, the secondary compounds were externally located, possibly subjecting them to cleaning by minims. The oat flakes without antifungal power were taken at a higher rate than those with antifungal powder suggesting avoidance. Although the rate of sele ction may have been lower, eventually all 330 antifungal oat flakes, except for five, were taken from the study area within one hour. Why would these flakes with antifungal compounds eventually be taken? Minims hitchhiking behavior is correlated to decrea sed spore contaminant levels on leaves, indicating cleaning (Griffiths & Hughes 2010). The role of minims in cleaning leaves with spores and microorganisms probably needs to be expanded to include secondary compounds. Although minims were not observed in this study, the antifungal components on oat flakes needed to be dealt with before entering the fungus garden. The ability of Atta ants to actively deal with low levels of secondary compounds possibly explains why any of the antifungal treatment oat flak es were even taken. The selection of contaminated oat flakes suggests that they will be dealt with in an active, rather then passive manner, possibly minim cleaning. The significant difference in the rate of selection between antifungal flakes and talc decision to harvest leaves with less contaminants would maximize their energy gain by minimizing the amount of energy needed for cleaning and preparing the lea f for the fungus garden. From this statement it would appear logical that only leaves with no contaminants would be harvested, this is clearly not the case. Most oat flakes were eventually harvested with the
antifungal flakes being taken less and at a lo preference is for leaves with lower contaminant loads, sub optimal leaves will be selected, but in a lower amount. Although this behavior might lower the energy yield in the short run, it might maintain energy gai ns in the long run. Ants prefer young leaves and forage more often in the canopies of trees (HÃ¶lldobler & Wilson 1990, Griffiths & Hughes 2010). In the absence of total defoliation, selection of leaves from other parts of the tree than just new leaves in the canopy would reduce losses to the host tree. Because lower leaves on trees generally have more external contaminants because of rain splash and contaminants being washed off canopy leaves and carried to lower branches in rain, my results suggest that foraging can occur in sub canopy environments. The decreased loss to the tree would result in increased growth the next year; more so than if just new canopy leaves had exclusively been selected. Ants can be observed skipping suitable foraging trees close to their nest, leaving them completely foliated, supposedly gain a longer sustained yield (HÃ¶lldobler & Wilson 1990). My data, along with this observation, suggest that ants act in such a way to sustain yields over a long period of time. The selection o f leaves needing more energetic output to clean maintains the availability of food for future generations of the ant colony. The simplicity we assign to leaf cutter ant systems needs to be re evaluated. The willingness of A. cephalotes ants to select sub par substrate suggests that active responses, such as the cleaning of leaf fragments, are employed to defend against secondary compounds. The amount of contaminated leaves taken was less in every trial in this study yet a vast majority was eventually take n, although at a slower rate. The eventual removal of all antifungal oak flakes suggests A. cephalotes is able to discriminate against and deal with contaminated food sources. My study suggests that Atta cephalotes colonies display complex behaviors in t he selection of appropriate fungus substrate and have evolved strategies to deal with contaminants in order to ensure the survival of the colony for an extended period of time. ACKNOWLEDGMENTS I would like to thank Hugo and Leila in San Luis for providi ng me with housing and a ridiculous amount of amazing food for the duration of my project. I would also like to thank Anjali Kumar for guidance with the project and helping me with statistics. I thank Pablo Allen, Alan Masters, Moncho CalderÃ³n and Raquel Martinez for additional help. Finally, a thank you to Mark Bliss for helping me with my project and keeping me company during data collection. LITERATURE CITED Griffiths, H. M. and W. O. Hughes. 2010. Hitchhiking and the Removal of Microbial Contami nants by the Leaf cutting Ant Atta colombica . Ecological Entomology. 35: 529 537. Hoelldobler, B. and E. O. Wilson. 1990. The Ants. The Belknap Press of Harvard University Press. Cambridge, Massachusetts. 732. Howard, J. 1987. Leafcutting Ant Di et Selection: The Role of Nutrients, Water, and Secondary Chemistry. Ecology. 68: 503 515.
Howard, J. J., C. Cazin, and D. F. Wiemer. 1988. Toxicity of Terpenoid Deterrents to the Feafcutting Ant Atta cephalotes and its Mutualistic Fungus. Journal of Chemical Ecology. 14: 59 70. MacArthur, R. and E. Pianka. 1966. On Optimal Use of a Patchy Environment. The American Naturalist. 100: 603 611. Manlove, L. 2009. Atta cephalotes prefer pioneer over shade tolerant plant species in San Luis, Montever de, Costa Rica. Tropical Ecology and Conservation. Spring 2009: 74 79. Pyke, G. H., H. R. Pulliam, and E. L. Charnov. 1977. Optimal Foraging: A Selective Review of Theory and Tests. The Quarterly Review of Biology. 52: 137 139. Ridley, P., P. E. Ho wse, C. W. Jackson. 1996. Control of the Behavior of Leaf Fungus. Experientia. 52: 631 635 Rockwood, L. and S. Hubbell. 1987. Host plant Selection, Diet Diversity, and Optimal Foraging in a Tropical Leafcutting Ant. Oecologia. 74: 55 61.
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 M39-00274
Las diferencias en la tasa de forrajeo de Atta cephalotes y la cantidad de sustrato que se cosecho despus de la introduccin de un agente anti fngico
Differences in Atta cephalotes foraging rate and amount of substrate harvested following the introduction of an antifungal agent
Optimal foraging theory dictates animals will behave in the most energetic favorable fashion, maximizing energy gained while minimizing energy lost. However, exceptions do exist, such as mating behaviors and predator avoidance. I show that oat flakes contaminated with antifungal powder, simulating secondary compounds, are selected less and at a lower rate by a colony of Atta cephalotes. This demonstrates the ability of the colony to recognize and discriminate against the contaminant to protect their symbiotic fungus. It also suggests that leafcutters, to an extent, can detect the amount or toxicity of the secondary compound. The willingness of the ants to take the antifungal flakes suggests cleaning is involved, possibly with minims, providing support for the hitchhiking cleaning process. Finally, the harvesting of sub-optimal resources might represent an attempt to sustain yields over an extended period of time. These conclusions suggest that we need to reconsider our current models of Atta cephalotes foraging behaviors for they might be more complex than we currently assume.
La teora de forrajeo ptimo dice que los animales deben tener un comportamiento energtico de la manera ms favorable, maximizando la energa ganada y minimizando la energa perdida. Sin embargo, las excepciones existen, como el comportamiento reproductivo y evitar a los depredadores. Muestro que las hojuelas de avena contaminadas con un talco anti fngico, simulando compuestos secundarios, son seleccionados en menor cantidad y en una menor tasa por una colonia de Atta cephalotes. Esto demuestra la habilidad de la colonia para reconocer y discriminar en contra de los contaminantes para proteger a su hongo simbitico. Tambin sugiere que las zompopas, pueden detectar la cantidad o toxicidad de los compuestos secundarios. La disposicin de las hormigas para tomar las hojuelas con la sustancia anti-hongos sugiere una limpieza, posiblemente por las mnimas, apoyando la teora que dice esto. Finalmente, la colecta de recursos por debajo del ptimo puede representar un intento para aumentar el rendimiento sobre los perodos de tiempo extensos. Estas conclusiones sugieren que tenemos que reconsiderar los modelos actuales de forrajeo de Atta cephalotes, y que estos pueden ser ms complejos de lo que pensamos.
Text in English.
Costa Rica--Puntarenas--Monteverde Zone--San Luis
Alimentos de animales
Costa Rica--Puntarenas--Zona de Monteverde--San Luis
Tropical Ecology Fall 2010
Ecologa Tropical Otoo 2010
t Monteverde Institute : Tropical Ecology