1 Diversity and morphological variation in Formicidae in three strata of the Monteverde Cloud Forest. Peter A. Larson College of Biological Sciences, University of Minnesota Twin Cites ABSTRACT This study examines the arboreal presence of ants. Compari sons of diversity at three forest strata, canopy, mid level, and forest floor were analyzed. Measurements were also made of leg length to body length and compared across the strata. The hypothesis is that arboreal ants will have a higher leg to body rati o. Finally a representation of each ant species, except three underrepresented species was dropped from the canopy to test for the ability to glide . Conclusive results showed that ant diversity is greatest at the mid stratum H = 1.82. The remainin g two strata, canopy and floor, showed similar diversity values H = 1.04 and .86 respectively. Significance was found between mid level and forest floor, and mid level and canopy p < 0.05. Canopy level and mid level ants showed no difference in body length, leg length, or th e ratio between the two p > 0.05. Both, however, showed larger values than the forest floor ants in all three cases p < 0.05. None of the ants glided or showed any signs of being able to. Further study is required to solid ify these results but it seems evident that both greater diversity and a necessity for longer legs are consistent in arboreal ants . RESUMEN Este estudio examina la presencia arbÃ³rea de hormigas. Las comparaciones de la diversidad en tres pisos d e los estratos, del dosel, de nivel medio, y del piso del bosque fueron analizados. Se tomaron medidas de longitud de pierna a la longitud de cuerpo y comparadas a travÃ©s de los estratos. La hipÃ³tesis es que las hormigas arbÃ³reas tendrÃ¡n una pierna mÃ¡s gra nde en proporciÃ³n al cuerpo. Finalmente una representaciÃ³n de cada especie de hormiga, menos tres especies con baja representaciÃ³n fueron tiradas desde el dosel para para probar para que la capacidad planear . Los resultados concluyentes demostraron qu e la diversidad de la hormiga es mÃ¡s grande en el estrato medio H = 1.82. Los dos estratos restantes, dosel y piso, demostraron los valores similares de la diversidad H = 1.04 y .86 respectivamente. La significancia fue encontrada entre el piso y nivel medio, y entre el nivel medio y dosel < de p; 0.05. El nivel del dosel y el nivel medio las hormigas no demostraron ninguna diferencia en longitud de cuerpo, longitud de pierna, o un cociente entre los dos > de p; 0.05. Ambos, sin embargo, demostraron valores mÃ¡s grandes que las hormigas del piso del bosque en los tres casos < de p; 0.05. Ningunas de las hormigas planearon o demostraron cualquier muestra de poder hacerlo. Estudios adicionales se requieren para solidificar estos resultados pero parece evidente que la mayor diversidad y una necesidad para piernas mÃ¡s largas son constantes en hormigas arbÃ³reas. INTRODUCTION Worldwide ants boast 16 subfamilies, 300 genera, and 8,800 species Longino and Hanson 1995. In Costa Rica there are seven subfa milies, 81 genera, and 620 species Longino and Hanson 1995. Such a prolific family is sure to inhabit many of the vertical strata of a tropical rainforest. Tobin 1991 conducted a study in which he fogged two trees in Peru. He counted 28,279 arthropo ds, 69.7 percent 19,702 consisting of five
2 subfamilies and 52 species in 28 genera Tobin 1991. In fact, one tree crown may possess up to 50 species of ants Longino and Hanson 1995. Tobin s study provides the necessary base information to warrant th is study. The environment in the crown of the tree is such that animals inhabiting it probably possess morphological structures that are different than their ground dwelling relatives. This study is three fold: it examines the diversity of ants at three h eights, compares their leg and body lengths across the three levels, and addresses whether or not controlled glide behavior is common in arboreal ants. The study performed by Tobin shows that arboreal ants are numerous and diverse, this study compares the ir diversity to that of two other stratum. The purpose for testing whether or not arboreal ants are capable of gliding is based on a paper by Yanoviak et al. 2006, who found a species of gliding ant present in the canopy. Leg length and body length may play a significant role in determining where the ant spends the majority of her time foraging and caring for the brood. I hypothesize that there will be a significant difference in both diversity and in leg length when comparing ants across stratum. I al so hypothesize that none of the ants will exhibit the ability to glide, regardless of what level they inhabit. Ants in the canopy or mid level will be more diverse because they have more resources at their disposal and those resources are at a close proxi mity to their nest site. Because the crown of a tree displays considerable environmental and biological differences than the forest floor, ants that inhabit the crown will also have longer legs. No ants will display the ability to glide, as that is a lar ge evolutionary step that intuitively appears to be difficult to achieve. METHODS Location and Dates A single fig tree, Ficus insipida Moraceae was sampled at three heights: the canopy at 30 m from the floor, mid level at 18 22 m, and the forest flo or at zero m. The research site was located in the Monteverde Cloud Forest Reserve, a lower montane wet forest, and the study tree was located at 1550 meters above sea level. Each site was sampled for eight hours at various times ranging from 9:30 a.m. t o 3:00 p.m. Collection and observation were performed on the days of July 22 nd , 23 rd , 27 th , 29 th , and 30 th . The weather for all days was clear with no precipitation except for the occasional short lived mist shower. Each site was sampled on at least two separate days. Sampling and Collection Methods Two methods of sampling were utilized, baiting and direct collection Bestelmeyer et al. 2000. Bait sampling was used to attract out of reach ants and also because dietary generalists represent a signific ant portion of ant faunas worldwide and these generalists can be used to examine patterns in ant communities Bestelmeyer et al. 2000 Longino and Hanson 1995. Tuna and honey were used as baits as they attract both carnivorous and non carnivorous predat ors. During each sampling period two honey and two tuna baits were set. Because baiting the same area often attracts the same ants Bestelmeyer et al. 2000 all four baits were placed at different spots during each sampling period.
3 Direct sampling was performed with the aid of collecting forceps and an aspirator. Epiphytes within reach were searched as well as any vines or lianas. Hollow twigs and other hollow structures were broken apart and searched as well. All collected ants, including those from the bait traps, were placed in plastic vials and kept alive. Ants from each morpho species, with the exception of morpho species 2, 7, and 11 were dropped from the crown using the techniques of Yanoviak et al. 2006. Note was taken of whether or not th e animal performed a J shape arc back to the tree, or showed any signs of controlling their descent. Once in the laboratory, an alcohol soaked rag >75% EtOH was placed in the vials to kill the ants. Laboratory Methods All ants were separated based o n morphology and categorized into morpho species using an Olympus SZ51 dissecting microscope. Separation by morpho species was permissible because I was sampling a small geographic area one tree, which means there would be little morphological variation among any given species Agosti et al. 2000. As suggested by Agosti et al. 2000 I preliminarily separated ants based on color, general shape, and size of each individual. Coloration in the class Insecta is very plastic at the specific level so furthe r examination was needed. Colonies of ants are widely known to exhibit caste behavior in which each caste may represent many different size variations. However, the majority of worker castes are monomorphic HÃ¶lldobler and Wilson 1990. This legitimizes separation by size. Further observations were necessary to make sure all ants were categorized correctly as some species displayed polymorphism. Ants collected at each level were kept separate. Characters suggested by Agosti et al. 2000 were used for morpho species identification. Three major characters had to be consistent across the specimens if they were to be considered the same species. Major characters include: the head, general shape, eyes, mandible and mouth parts, and antennal length versus body length; the petiole, shape and presence or absence of a peduncle; the gaster, examined laterally to count number of tergites and their separation; and the alitrunk, presence or absence of spikes or other notable nodules or protrusions. Figures 2 1 a nd 2 3 and accompanying descriptions by HÃ¶lldobler and Wilson 1990 were used to aid in morphological clarification. Statistical Tests Shannon Wiener index was used to quantify diversity, modified student t tests were run to determine if t he values differed, and an ANOVA was run to test for differences among leg to body length with respect to collection site. Ants that were collected at more than one site only three individuals were placed into the mid level for calculations. RESULTS N o ants glided, exhibited controlled aerial descent , or otherwise showed any indication of free falling aerial control. Of all ants collected, only one Morpho 6 showed any
4 morphological traits, enlarged or flanged tarsi and high gagster mobility Yanovi ak et al 2006, which would suggest the ability to glide. The canopy had the smallest population N = 16 Fig. 1. In contrast, the middle level had the greatest number of specimens collected N = 48 Fig. 1. Similarly the forest floor was represent ed by a high number of individuals N = 36 Fig.1. The middle level was also defined by the highest species diversity compared to the other two levels H = 1.82. Both the canopy and the forest floor expressed similar Shannon Weiner values H = 1.04 and H = .86 respectively. When comparing diversity across levels the mid level shows more diversity than the canopy and more diversity than the forest floor p < 0.001 for both analyses. There was no significant difference in diversity between the can opy and the forest floor p = 0.5. Body length, leg length, or the ratio of the two was not significantly different between the canopy and the mid elevation Fig. 2, 3, 4. Both the canopy and mid level species are larger than the floor species p = 0.03, df effect = 2, F = 6.56, df error = 94 Fig. 3. Leg length and leg to body length ratio produced identical results with even greater p value correlations, p = .002, df effect = 2, F = 3.52, df error = 94 and p < .0001, df effect = 2, F = 40.99, df error = 94 respectively Fig. 2 and Fig.4 respectively. DISCUSSION Formicidae Diversity There is evidence for a greater diversity of ants at the mid level strata compared to the forest floor. The results comparing diversity between the canopy an d the floor did not prove to be significantly different, however they appear to be inconclusive and more testing is necessary. The canopy does show a higher degree of diversity than the floor as shown in the results but only three species were collected at each of those two levels Table 1. I feel that not enough species were collected at either level to accurately depict the true diversity. The reasons for a greater diversity of ants in the upper strata are many but only two will be discussed at any length. The strongest evidence of ant plant mutualism is domatia, plant constructed shelters to house ant colonies HÃ¶lldobler and Wilson 1990. Similarly many epiphytic plants are known to harbor ant colonies including but not limited to the bromeliads Bromeliacea and the orchids Orchidaceae HÃ¶lldobler and Wilson 1990. I observed many such domatia including epiphytes and counted no fewer than what appeared to be 15 different species growing on this single tree. Each plant is a potential host to a different species of ant. Ant gardens were also a common sight. Ants that live in carton nests place seeds of epiphytic plants inside their nest. The roots become integrated with the nest and the ants feed off of the roots; bromeliads are a common choi ce HÃ¶lldobler and Wilson 19990. A study performed by Nieder et al. 2000 noted that the majority of epiphytes examined 51 percent were ant garden epiphytes. A high number of epiphytes support a high number of ant gardens of diverse ant species. One possible reason the high ant community diversity in the canopy is the concentration of food. Tobin 1991 claims that ants probably attain more nutrients from plant matter than previously thought. It is possible that ant species specialize on different e piphytes for sustenance. Those ants that are carnivorous or generalists have a wealth of insect larvae, insects and other sources of protein at their disposal pers. obs..
5 Further studies assessing biodiversity should be concerned primarily with only t wo strata, the canopy and the forest floor. In my opinion the middle level and the canopy level are not far enough apart and both should be grouped collectively as canopy. The crown as a whole should be assessed and then compared to the forest floor. Or a greater separation between the crown and the mid level could be made. Examination of Leg and Body length and the ratio between the two The results of this study suggest that there is indeed a difference in absolute leg length Fig. 2. Morpho specie s 14 exhibits significantly longer legs than the other two terrestrial species, species 14 was identified as a leaf cutter ant, Atta spp , in which having long legs to climb trees would be beneficial. Explanations for the evolution of long legs in arboreal ants have not been studied. Though no ants displayed the ability to glide, long legs may serve a similar purpose. Arboreal ants can easily become displaced from their host tree by harsh weather conditions, predators or a misstep Yanoviak et al. 2006. Increased leg length may offer an opportunity for a greater chance at regaining contact with the host tree after the animal has become dislodged. Longer legs could reach out to leaves, branches or any other substrate within the ants reach. Terrestrial an ts have no such use for longer legs. Previously, a common assumption was that as animal body size gets smaller the ratio of leg length to body length diminishes. The size grain hypothesis presented by Kaspari and Weisner 1999 suggest that not only are t he legs of ants unnaturally long for their body length, but that as the ant grows their legs not only keep up with their growth but increase at a rate that further raises the ratio of leg length to body length. They discuss that small animals must walk th rough their environment, not around it or on top of it; this reinforces the previously held thought that as animals get smaller larger legs are too costly Kaspari and Weisner 1999. In the case of ants, having larger legs prevented them from accessing th eir interstitial world, however, having a larger body size and consequently possessing longer legs meant an increased resistance to desiccation, access to a larger food niche, and increased speed at a comparatively lower cost Kaspari and Weisner 1999. A rboreal ants generally forage at the tips of branches Yanoviak and Dudley 2006. Their nests are usually concentrated towards the center of the tree at the interface of epiphytic roots and thicker branches Longino 1991. Traveling from the nest to fora ging territory and back often forces the ant to expose herself to predators. Longer legs would allow the workers to cover ground more quickly as branches offer few places to hide from predators. Again referring to the Kaspari and Wiesner 1999 study, co ntact with intense tropical sunlight is greater in the canopy than on the forest floor. Thus, it is necessary to have a larger body size to aid in the avoidance of desiccation . Weaver ants inhabit only the Old World tropics and are large, long legged and solely arboreal HÃ¶lldobler and Wilson 1990. The weaver ants use these long legs in their nest construction. They link with one another; forming large bridges to pull leaves together, then utilize the silk produced by their larvae to glue the nest tog ether HÃ¶lldobler and Wilson 1990. Exceptionally long legged ants in the Neotropics may be an evolutionary step towards this kind of nest weaving behavior. A more likely explanation though is that the arboreal ants use their long legs to link together i n a similar
6 fashion creating bridges to access food or new foraging sites instead of for nest construction. Personal observations of solitary ants in the canopy reaching for a branch or leaf led me to the belief that it was entirely possible for ants to link with one another to access otherwise unreachable areas. More experiments and tests are needed. A larger data set is most certainly necessary and more trees and different species of trees need to be sampled. This is a start however and the prelimi nary results are promising in explaining not only those arboreal ants possess longer legs, but also why those long legs are necessary and much more pronounced in arboreal ants. ACKNOWLEDGEMENTS I would first of like to thank the people who funded me, Gr eg and Julene Larson, their money, but more importantly their unrelenting support, praise, and encouragement has propelled me to new levels. My advisor Tania ChavarrÃa who helped me significantly with running my statistical analyses and was always ready to help me when I needed it. Pablo Allen Monge who brought me to my study site and answered any and every random question I had. Karen Masters who has provided me with amazing experiences and opportunities and has opened my eyes to the spectacular world of tropical biology and ecology. JosÃ© Carlos CalderÃ³n for his constant positive attitude and assistance with my statistical testing. LITERATURE CITED BESTELMEYER, B. T., D. AGOSTI, L.E. ALONSO, C. R. F. BRANDAO, W. L. BROWN Jr., J. H. C. DELABIE, and R. SILVESTRE, 2000 Field Techniques for the Study of Ground Dwelling Ants: An Overview, Description, and Evaluation. In: ANTS: Standard Methods For Measuring And Monitoring Biodiversity , D. Agosti, J. D. Majer, L. E. Alonso, and T. R. Schultz eds. The S mithsonian Institution, Washington, DC, pp. 125 127. HÃ–LLDOBLER, B., and E. O. WILSON, 1990. the ANTS . The Belknap Press of Harvard University Press, Cambridge, MA, pp. 4 5, 312, 382, 534 536, 546. KASPARI, M., and M. D. WEISER, 1999. The size grain hyp othesis and interspecific scaling in ants. Functional Ecology 13 : 550 558. LATTKE, J. E., 2000 Specimen Processing: Building and Curating an Ant Collection. In: ANTS: Standard Methods For Measuring And Monitoring Biodiversity , D. Agosti, J. D. Majer, L. E . Alonso, and T. R. Schultz eds. The Smithsonian Institution, Washington, DC, pp. 157, 163 165. LONGINO, J. T., and P. E HANSON, 1995. The ants Formicidae. In: The Hymenopter of Costa Rica , P. E. Hanson, and I. D. Gauld eds. Oxford University Press , New York, NY, pp. 586, 592. LONGINO, J. T. 1991 The ants of Monteverde. In: Monteverde Ecology and Conservation of a Tropical Cloud Forest , N. M. Madkarni, and N. T. Wheelwright, eds. Oxford University Press, New York, NY, pp 135. NIEDER, J., E. ENGWAL D, M. KLAWUN, and W. BARTHLOTT, 2000. Spatial Distribution of Vascular Epiphytes including Hemiepipytes in a Lowland Amazonian Rain Forest Surumoni Crane Plot of Southern Venezuela. Biotropica 323 : 385 396. TOBIN, J. E. 1995. A Neotropical rainfores t canopy, ant community: some ecological considerations.
7 In: Ant Plant Interactions , C. R. Huxley and D. F. Cutler, eds. Oxford University Press, New York, NY, pp. 536 538 YANOVIAK, S. P., and R. DUDLEY, 2006. The role of visual cues in directed aerial descent of Cephalotes atratus workers Hymenoptera:Formicidae. Journal of Experimental Biology 209 : 1777 1783.
8 Table 1. Abundance of ant morphospecies at each forest stratum in Ficus insipida . The correspond ing heights are canopy 30 m, middle 18 22 m, forest floor 0 m . Figure 1. Depicts values of ants collected at each collection site. Canopy Middle Forest Floor Morpho 1 13 Morpho 2 1 Morpho 3 2 Morpho 4 10 Morpho 5 4 Morpho 6 6 Morpho 7 1 Morpho 8 1 Morpho 9 9 Morpho10 12 Morpho 11 5 Morpho 12 26 Morpho 13 2 Morpho 14 6 0 10 20 30 40 50 60 Canopy Mid level Floor Number of Specimens Strata
9 Figure 2. The average leg length of ants at each level. df effect = 2, F = 3.52, df error = 94 Figure 3. The average bod y length of ants at each level. df effect = 2, F = 6.56, df error = 94 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 1 Average length mm Level Leg Length canopy mid ground 0 1 2 3 4 5 6 1 Average Body Length mm Level Body Length canopy mid ground
10 Figure 4. The ratio of leg length compared to body length of ants at each level. df effect = 2, F = 40.99, df error = 94 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Ratio of Leg length/Body length % Level Ratio of Leg to Body length canopy mid ground
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 M39-00002
Larson, Peter A.
La diversidad y la variacin morfolgica en Formicidae en tres estratos del Bosque Nuboso de Monteverde
Diversity and morphological variation in Formicidae in three strata of the Monteverde Cloud Forest
This study examines the arboreal presence of ants. Comparisons of diversity at three forest strata, canopy, mid-level, and forest floor were analyzed. Measurements were also made of leg length to body length and compared across the strata. The hypothesis is that arboreal ants will have a higher leg to body ratio. Finally a representation of each ant species, except three under-represented species was dropped from the
canopy to test for the ability to glide. Conclusive results showed that ant diversity is greatest at the mid-stratum (H = 1.82). The remaining two strata, canopy and floor, showed similar diversity values (H =
1.04 and .86 respectively). Significance was found between mid-level and forest floor, and mid-level and canopy (p < 0.05). Canopy level and mid level ants showed no difference in body length, leg length, or
the ratio between the two (p > 0.05). Both, however, showed larger values than the forest floor ants in all three cases (p < 0.05). None of the ants glided or showed any signs of being able to. Further study is
required to solidify these results but it seems evident that both greater diversity and a necessity for longer legs are consistent in arboreal ants.
Este estudio examina la presencia arbrea de hormigas. Se analizaron las comparaciones de la diversidad en tres estratos del bosque, el dosel, el nivel medio y el suelo del bosque. Se realizaron mediciones de la longitud de las piernas con la longitud del cuerpo y fueron comparados a travs de los estratos.
Text in English.
Monteverde Cloud Forest Reserve
Reserva Bosque Nuboso Monteverde
Tropical Ecology 2008
Ant diversity--Costa Rica--Monteverde
Ant morphology--Costa Rica--Monteverde
Ecologa Tropical 2008
Diversidad de hormigas--Costa Rica--Monteverde
Morfologa de hormigas--Costa Rica--Monteverde
t Monteverde Institute : Tropical Ecology