1Diversity and morphological variation in Formicidae in three strata of the Monteverde Cloud Forest. Peter A. Larson College of Biological Sciences, University of Minne sota Twin Cites ABSTRACT This study examines the arboreal presence of ants. Comparisons of diversity at three forest strata, c anopy, mid-level, and forest floor were analyzed. Measure ments were also made of leg length to body length a nd 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, excep t three underrepresented species was dropped from t he canopy to test for the ability to Â“glideÂ”. Conclus ive results showed that ant diversity is greatest a t the midstratum (HÂ’ = 1.82). The remaining two strata, ca nopy and floor, showed similar diversity values (HÂ’ = 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, o r thge ratio between the two (p > 0.05). Both, howev er, 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 evi dent that both greater diversity and a necessity fo r longer legs are consistent in arboreal ants. RESUMEN Este estudio examina la presencia arbrea de hormig as. Las comparaciones de la diversidad en tres piso s de los estratos, del dosel, de nivel medio, y del piso del bosque fueron analizados. Se tomaron medidas d e longitud de pierna a la longitud de cuerpo y compar adas a travs de los estratos. La hiptesis es que las hormigas arbreas tendrn una pierna ms grandes en proporcin al cuerpo. Finalmente una representaci n de cada especie de hormiga, menos tres especies con baja representacin fueron tiradas desde el dosel para para probar para que la capacidad Â“planearÂ”. Los r esultados concluyentes demostraron que la diversida d de la hormiga es ms 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 do sel y el nivel medio las hormigas no demostraron ninguna dif erencia en longitud de cuerpo, longitud de pierna, o un cociente entre los dos (> de p; 0.05). Ambos, si n embargo, demostraron valores ms grandes que las hormigas del piso del bosque en los tres casos (< d e p; 0.05). Ningunas de las hormigas planearon o demostraron cualquier muestra de poder hacerlo. E studios adicional se requieren para solidificar est os resultados pero parece evidente que la mayor divers idad y una necesidad para piernas ms largas son constantes en hormigas arbreas. INTRODUCTION Worldwide ants boast 16 subfamilies, 300 genera, an d 8,800 species (Longino and Hanson 1995). In Costa Rica there are seven subfam ilies, 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) cond ucted a study in which he fogged two trees in Peru. He counted 28,279 arthropods, 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 Hanso n 1995). TobinÂ’s study provides the necessary base information to warrant this study. The environment in the crown of the tree is such that animals inhabiting it probably po ssess morphological structures that are different than their ground dwelling relatives. This study is three fold: it examines the diversity of ants at three heights, compares their leg and body lengths across the thre e levels, and addresses whether or not controlled glide behavior is common in arboreal ant s. The study performed by Tobin shows that arboreal ants are numerous and diverse, this study compares their diversity to that of two other stratum. The purpose for testing whether or not arboreal ants are capable of gliding is based of 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 diff erence in both diversity and in leg length when comparing ants across stratum. I also hypothesize that none of the ants will exhibit the ability to glide, regardless of what le vel they inhabit. Ants in the canopy or mid-level will be more diverse because they have mo re resources at their disposal and those resources are at a close proximity to their n est 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 large evolutionary step that intuitive ly 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 floor at zero m. The research site was located in the Monteverde Cloud Forest Res erve, 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. to 3:00 p.m. Collection and observation were performed on the days of July 22nd, 23rd, 27th, 29th, and 30th. The weather for all days was clear with no precipitatio n 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-re ach ants and also because dietary generalists represent a significant 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 bait s as they attract both carnivorous and non-carnivorous predators. During each samplin g period two honey and two tuna baits were set. Because baiting the same area ofte n attracts the same ants (Bestelmeyer et al. 2000) all four baits were placed at different s pots during each sampling period.
3 Direct sampling was performed with the aid of colle cting 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 searc hed 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 s pecies 2, 7, and 11 were dropped from the crown using the techniques of Yanoviak et al.( 2006). Note was taken of whether or not the animal performed a J-shape arc b ack to the tree, or showed any signs of controlling their descent. Once in the laborato ry, an alcohol soaked rag (>75% EtOH) was placed in the vials to kill the ants. Laboratory Methods All ants were separated based on morphology and cat egorized 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 spec ies (Agosti et al. 2000). As suggested by Agosti et al. (2000) I preliminarily s eparated ants based on color, general shape, and size of each individual. Coloration in the class Insecta is very plastic at the specific level so further examination was needed. Colonies of ants are widely known to exhibit caste behavior in which each caste may repr esent many different size variations. However, the majority of worker castes are monomorp hic (Hlldobler and Wilson 1990). This legitimizes separation by size. Further obser vations were necessary to make sure all ants were categorized correctly as some species dis played polymorphism. Ants collected at each level were kept separate. Characters suggested by Agosti et al. (2000) were u sed for morpho species identification. Three major characters had to be c onsistent across the specimens if they were to be considered the same species. Major char acters 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 peduncl e; 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. Fi gures 2-1 and 2-3 and accompanying descriptions by Hlldobler and Wilson (1990) were u sed to aid in morphological clarification. Statistical Tests Shannon-Wiener index was used to quantify diversity modified student t-tests were run to determine if the values differed, and an ANOVA w as 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 No ants glided, exhibited Â“controlled aerial descen tÂ”, or otherwise showed any indication of free falling aerial control. Of all ants collec ted, only one (Morpho 6) showed any
4 morphological traits, enlarged or flanged tarsi and high gaster mobility (Yanoviak et al 2006), which would suggest the ability to glide. The canopy had the smallest population (N = 16) (F ig. 1). In contrast, the middle level had the greatest number of specimens collecte d (N = 48) (Fig. 1). Similarly the forest floor was represented by a high number of in dividuals (N = 36) (Fig.1). The middle level was also defined by the highest specie s diversity compared to the other two levels (HÂ’ = 1.82). Both the canopy and the forest floor expressed similar ShannonWeiner values (HÂ’ = 1.04 and HÂ’ = .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 canopy and the forest floor (p = 0.5). Body length, leg length, or the ratio of the two wa s 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 resp ectively) (Fig. 2 and Fig.4 respectively). DISCUSSION Formicidae Diversity There is evidence for a greater diversity of ants a t the mid level strata compared to the forest floor. The results comparing diversity betw een the canopy and the floor did not prove to be significantly different, however they a ppear to be inconclusive and more testing is necessary. The canopy does show a highe r degree of diversity than the floor (as shown in the results) but only three species were c ollected at each of those two levels (Table 1). I feel that not enough species were col lected at either level to accurately depict the true diversity. The reasons for a greater dive rsity 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 do matia, plant constructed shelters to house ant colonies (Hlldobler and Wils on 1990). Similarly many epiphytic plants are known to harbor ant colonies including b ut not limited to the bromeliads (Bromeliacea) and the orchids (Orchidaceae) (Hlldo bler and Wilson 1990). I observed many such domatia including epiphytes and counted n o 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; bromel iads are a common choice (Hlldobler and Wilson 19990). A study performed by Nieder et al. (2000) noted that the majority of epiphytes examined (51 percent) were ant garden epi phytes. A high number of epiphytes support a high number of ant gardens of diverse ant species. One possible reason the high ant community diversit y in the canopy is the concentration of food. Tobin (1991) claims that an ts probably attain more nutrients from plant matter than previously thought. It is possib le that ant species specialize on different epiphytes for sustenance. Those ants that are carn ivorous 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 co ncerned primarily with only two strata, the canopy and the forest floor. In my opinion the middle level and the canopy level are not far enough apart and both shou ld be grouped collectively as canopy. The crown as a whole should be assessed and then co mpared to the forest floor. Or a greater separation between the crown and the mid-le vel could be made. Examination of Leg and Body length and the ratio be tween the two The results of this study suggest that there is ind eed a difference in absolute leg length (Fig. 2). Morpho species 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. Expl anations 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 co ntact 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 ants 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 ar e the 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 rati o of leg length to body length. They discuss that small animals must walk through 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 their interstitial world, howev er, having a larger body size and consequently possessing longer legs meant an increa sed resistance to desiccation, access to a larger food niche, and increased speed at a co mparatively lower cost (Kaspari and Weisner 1999). Arboreal ants generally forage at the tips of branc hes (Yanoviak and Dudley 2006). Their nests are usually concentrated toward s the center of the tree at the interface of epiphytic roots and thicker branches (Longino 19 91). Traveling from the nest to foraging territory and back often forces the ant to expose herself to predators. Longer legs would allow the workers to cover ground more q uickly as branches offer few places to hide from predators. Again referring to the Kas pari and Wiesner (1999) study, contact with intense tropical sunlight is greater in the ca nopy than on the forest floor. Thus, it is necessary to have a larger body size to aid in the avoidance of dessication. Weaver ants inhabit only the Old-World tropics and are large, long-legged and solely arboreal (Hlldobler and Wilson (1990). The weaver ants use these long legs in their nest construction. They link with one anothe r; forming large bridges to pull leaves together, then utilize the silk produced by their l arvae to Â“glueÂ” the nest together (Hlldobler and Wilson 1990). Exceptionally long l egged ants in the Neotropics may be an evolutionary step towards this kind of nest weav ing behavior. A more likely explanation though is that the arboreal ants use th eir long legs to link together in a similar
6 fashion creating Â“bridgesÂ” to access food or new fo raging sites instead of for nest construction. Personal observations of solitary an ts in the canopy Â“reachingÂ” for a branch or leaf led me to the belief that it was entirely p ossible for ants to link with one another to access otherwise unreachable areas. More experiments and tests are needed. A larger da ta set is most certainly necessary and more trees and different species of t rees need to be sampled. This is a start however and the preliminary results are promising i n explaining not only that arboreal ants possess longer legs, but also why those long l egs are necessary and much more pronounced in arboreal ants. ACKNOWLEDGEMENTS I would first of like to thank the people who funde d me, Greg and Julene Larson, their money, but more importantly their unrelenting support, praise, and encouragement has propelled me to new levels. My advisor Tania Chavarra who helped me significantly with running my statistical analyses and was alway s ready to help me when I needed it. Pablo Allen Mon ge 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 spectac ular world of tropical biology and ecology. Jos Carlos Caldern 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 f or 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. S chultz (eds.) The Smithsonian Institution, Washington, DC, pp. 125-12 7. HLLDOBLER, 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-grai n hypothesis and interspecific scaling in ants. Functional Ecology 13 : 550-558. LATTKE, J. E., 2000 Specimen Processing: Building a nd 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, Was hington, 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 Universi ty Press, New York, NY, pp. 586, 592. LONGINO, J. T. 1991 The ants of Monteverde. In: Mon teverde Ecology and Conservation of a Tropical Cloud Forest N. M. Madkarni, and N. T. Wheelwright, eds. Oxfor d University Press, New York, NY, pp 135. NIEDER, J., E. ENGWALD, M. KLAWUN, and W. BARTHLOTT 2000. Spatial Distribution of Vascular Epiphytes (including Hemiepipytes) in a Lowland Ama zonian Rain Forest (Surumoni Crane Plot) of Southern Venezuela. Biotropica 32(3) : 385-396. TOBIN, J. E. 1995. A neotropical rainforest canopy, ant community: some ecological considerations.
7 In: Ant-Plant Interactions C. R. Huxley and D. F. Cutler, eds. Oxford Univer sity Press, New York, NY, pp. 536-538 YANOVIAK, S. P., and R. DUDLEY, 2006. The role of v isual cues in directed aerial descent of Cephalotes atratus workers (Hymenoptera:Formicidae). Journal of Exper imental Biology 209 : 1777-1783.
8 Table 1. Abundance of ant morphospecies at each f orest stratum in Ficus insipida The corresponding 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 nr
9 Figure 2. The average leg length of ants at each le vel. (df effect = 2, F = 3.52, df error = 94) Figure 3. The average body length of ants at each level. (df effect = 2, F = 6.56, df error = 94) n nr n rn nr n
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) rn nr n
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