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 M39-00041
Sirivanchai, Sara M.L.
Compromisos de defensa Anti-herbvoro: una comparacin de Passiflora capsularis, Passiflora helleri, y Passiflora sexflora (Passifloraceae)
Anti-herbivore defense trade-offs: a comparison of Passiflora capsularis, Passiflora helleri, and Passiflora sexflora (Passifloraceae)
Plants in the Tropics must evolve strategies for combating high levels of herbivory with few available nutrients. This results in possible tradeoffs between chemical and structural defenses. Three species of tropical passion vines, Passiflora capsularis, Passiflora helleri, and Passiflora sexflora (Passifloraceae) were examined for possible tradeoffs in cyanide, leaf toughness, glands and pubescence as anti-herbivore defenses. One young and one old leaf were collected from ten individuals of each species and analyzed for cyanide concentration and leaf toughness. In addition, observations were made regarding habitat, pubescence and presence or absence of blade glands. From the results it is clear that each species analyzed has adapted different anti-herbivore defense strategies. Passiflora helleri (mean = 119.88 g/g leaf, Std Err. 39.54) demonstrated markedly higher concentrations of cyanide production as well as blade gland than P. capsularis (mean = 24.04 g/g leaf, Std Err. 18.15) and P. sexflora (mean = 0.05 g/g leaf, Std Err. 0.006), which did not have blade glands (N = 59, t = 2.01, P = 0.05). Passiflora sexflora showed observably more pubescence than either of the other two species. And P. capsularis exhibited an observable tradeoff in the levels of cyanide produced in young versus old leaves (N = 8, F = 5.14, P = 0.04, DF = 1). All species were found in the understory along well cleared trails. These data suggest that the anti-herbivore defense strategies employed by Passiflora spp. are diverse and represent evolutionary tradeoffs between different defenses under similar selective pressures.
Las plantas en los trpicos deben evolucionar estrategias para combatir los altos niveles de herbivora con los pocos recursos disponibles. Esto puede resultar en compensacin diferencial entre las defensas qumicas y estructurales. Se examinaron tres especies de bejucos, Passiflora capsularis, Passiflora helleri, y Passiflora sexflora, para evaluar la posibilidad de compensacin entre el cianuro, la dureza de las hojas, las glndulas y pubescencia como defensas anti-herbivora.
Text in English.
Cloud forest ecology--Costa Rica
Ecologa del bosque nuboso--Costa Rica
Tropical Ecology 2008
Ecologa Tropical 2008
t Monteverde Institute : Tropical Ecology
1Anti-Herbivore Defense Trade offs: A Comparison of Passiflora capsularis , Passiflora helleri , and Passiflora sexflora (Passifloraceae). Sara M.L. Sirivanchai Department of Biology, University of Colorado Denve r ABSTRACT Plants in the Tropics must evolve strategies for co mbating high levels of herbivory with few available nutrients. This results in possible tradeoffs betw een chemical and structural defenses. Three species of tropical passion vines, Passiflora capsularis , Passiflora helleri , and Passiflora sexflora (Passifloraceae) were examined for possible tradeoffs in cyanide, le af toughness, glands and pubescence as anti-herbivo re defenses. One young and one old leaf were collecte d from ten individuals of each species and analyzed for cyanide concentration and leaf toughness. In addit ion, observations were made regarding habitat, pubescence and presence or absence of blade glands. From the results it is clear that each species ana lyzed has adapted different anti-herbivore defense strate gies. Passiflora helleri (mean = 119.88 Âµg/g leaf, Std Err. Â± 39.54) demonstrated markedly higher concentr ations of cyanide production as well as blade gland than P. capsularis (mean = 24.04 Âµg/g leaf, Std Err. Â± 18.15) and P. sexflora (mean = 0.05 Âµg/g leaf, Std Err. Â± 0.006), which did not have blade glands (N = 59, t = 2.01, P = 0.05). Passiflora sexflora showed observably more pubescence than either of the other two species. And P. capsularis exhibited an observable tradeoff in the levels of cyanide produc ed in young versus old leaves (N = 8, F = 5.14, P = 0.04, DF = 1). All species were found in the understory along well cleared trails. These data suggest that the anti-herbivore defense strategies employed by Passiflora spp. are diverse and represent evolutionary tradeoffs between different defenses under similar selective pressures. RESUMEN Las plantas en los trÃ³picos deben evolucionar estra tegias para combatir los altos niveles de herbivorÃ a con pocos recursos disponibles. Esto puede resultar en compensaciÃ³n diferencial entre defensas quÃmicas y estructurales. Se examinaron tres especies de bejuc os, Passiflora capsularis , Passiflora helleri , y Passiflora sexflora , para evaluar la posibilidad de compensaciÃ³n entre cianuro, dureza de hojas, glÃ¡ndulas y pubescencia como defensas anti-herbivorÃa. De los r esultados es claro que cada especie analizada se ha adaptado de forma diferente con respecto a las defe nsas anti-hebÃvoras. P. helleri demostrÃ³ altas concentraciones de cianuro, al igual que glÃ¡ndulas ausentes en las otras 2 especies. P. sexflora demostrÃ³ mÃ¡s pubescencia que las otras 2 especies. En P. cap sularis ademÃ¡s se determinÃ³ que las hojas mÃ¡s jÃ³ven es contienen mayores concentraciones de cianuro. Los d atos sugieren que las defensas anti-herbivoras utilizadas por Passiflora son divergentes y represe ntan diferentes caminos evolutivos bajo las mismas presiones. INTRODUCTION In Tropical life zones herbivory rates are higher t han in Temperate life zones (Coley and Aide 1991), species richness is higher resulting in more specialized herbivores (Marquis and Braker 1994), and the soils of tropical forests are generally nutrient deficient (Baillie 1996, Harms et al. 2004, Powers 2004). Thus, plants in Tropical Rain and Moi st Forests must be more efficient at allocating the limited av ailable resources to growth,
2 reproduction, and defense. Because herbivory is hi gher, defense becomes an important element in the plantÂ’s energy budget. Young leaves are softer and have a higher nutrition al value than older leaves; therefore, herbivory is greater on young leaves than on old le aves (Coley and Aide 1991, Kursar and Coley 2003). As a result, plants have developed many mechanisms for minimizing herbivory on young leaves such as rapid expansion, delayed greening, extrafloral nectaries, tougher tissue, pubescence, and chemical defenses in the form of secondary metabolites (Coley 1983, Coley and Barone 1996, Kur sar and Coley 2003). Leaf toughness has been shown to be the most efficient d efense against herbivory. Once a leaf has toughened, by accumulating lignins and cellulos e, the leaves are no longer attacked. Therefore, plants utilize one or a combination of m any of these defense mechanisms to minimize herbivory of young leaves, before they are tough enough to eliminate herbivore risk (Kursar and Coley 2003). Members of the family Passifloraceae in the genus Passiflora have a long-standing plantherbivore relationship with nymphalid butterflies i n the genus Heliconius (Heliconiinae) (DeVries 1987, Feuilet 2004, Gilbert 1991). There are over 500 species of Passiflora , which can occur as small trees or shrubs but are ty pically vines or lianas, which are commonly referred to as passion vines (Feuilet 2004 ). This mostly tropical genus can be found in the forest understory or among secondary v egetation (Feuilet 2004, Mabberley 1993). Heliconius can be found in much of the Americas from the sout hern United States through South America and the West Indies (DeVries 1987). Although Heliconius adults primarily feed on the pollen of flowers in the gene ra Psiguria (DeVries 1987, Gilbert 1991) and Gurania (Gilbert 1991) in the family Cucurbitaceae, they l ay their eggs on the leaves and tendrils of vines in the family Passiflo raceae, mostly in the genus Passiflora (DeVries 1987, Gilbert 1991). When the larvae hatc h they use their host plant as a food source until they have become large enough to pupat e. This intense herbivory can remove considerable quantities of biomass from the host Passiflora plant leading to a reduced fitness for that individual (Gilbert 1991). This intense pressure is believed to be responsible for the evolution of a host of possible anti-herbivore defenses seen in Passiflora (Gilbert 1991). Passiflora and closely-related genera have cyanogenic glycosi des, a secondary metabolite (Feuilet 2004, Gilbert 1991). The cyanogenic glyco sides are effective at deterring general herbivores (Kursar and Coley 2003), however , they are not useful in preventing herbivory by heliconiine butterfly larvae. In fact, heliconiine butterflies and their larvae posses the same cyanogenic glycoside system as Passiflora and some species may sequester cyanide from the leaves, as well as benef iting nutritionally (Brown et al. 1991, Engler et al. 2000) The ability of heliconiines to overcome cyan ide defenses has resulted in a coevolutionary arms race between the Passiflor aceae and Heliconiinae (Feuilet 2004, Gilbert 1991, Gilbert 1971). In addition to cyanog enic glycosides, Passiflora spp. have evolved variable numbers of other defenses such as extrafloral nectaries, variable leaf shapes, egg mimics (Feuilet 2004, Gilbert 1991), an d pubescence on the leaves (Gilbert 1971). Extrafloral nectaries have been shown to re duce herbivory in P. incarnata by attracting territorial ants that harvest the nectar ies and remove insect eggs and larvae
3 from leaves. Variable leaf shape is believed to th wart heliconiine femalesÂ’ ability to identify Passiflora as suitable for oviposition (Feuilet 2004, Gilbert 1991). Likewise, because Heliconius spp. that lay their eggs on passion vines typically hav e cannibalistic young they will only oviposit one egg on a leaf or tendril at a time (DeVries 1987). It has been suggested that some species of Passiflora have adapted small yellow structures, leaf glands, that mimic Heliconius eggs in order to deter oviposition and the subsequ ent herbivory by larval heliconiine (Feuilet 2004, Gilb ert 1991, Williams and Gilbert 1981). Finally, the presence of pubescence on leaves makes it difficult for the larvae to move across the leaf (Coley 1983) and when the pubescenc e are hook shaped they can even result in the death of the larva as seen in P. adenopoda (Gilbert 1971). Tropical species of Passiflora must deal with increased herbivory in nutrient poo r environments. Therefore, it may not be energetical ly possible to develop an antiherbivory strategy that employs all of the defense mechanisms that have been described for this genus. Instead, plants may have trade-off s where investment in one defense precludes others (Coley 1983, Coley and Barone 1996 , Gilbert 1991, Kursar and Coley 2003 Marquis 1994). Additionally, given that young leaves experience higher levels of herbivory, it would seem evolutionarily advantageou s to invest more energy in the protection of young leaves than old leaves. The pu rpose of this study is to examine possible tradeoffs made in resources allocated to a ntiherbivore defense in three tropical passion vine species, Passiflora capsularis , Passiflora helleri , and Passiflora sexflora . METHODS Study Site Leaf samples of P. capsularis , P. helleri , and P. sexflora were collected from a Montane Tropical Moist Forest in Monteverde, Costa Rica (el evation between 1450-1750 m) around the EstaciÃ³n BiolÃ³gica Monteverde. All leaf samples were collected from plants that were found along well maintained walking trail s. Leaf Collection and Identification Samples were collected for three days on April 30, May 1, and May 3, 2008. One young and one old leaf were collected from ten different vines of each species sampled. Leaves were only collected from vines when the apex of the vine could be located. When possible, vines were followed to their end point in order to ensure that the same individual was not sampled more than once. The age of the leaf was determined relative to the tip of the vine. The third leaf from the ap ex was collected as the young leaf sample. To determine the old leaf sample, leaves w ere counted down from the position of the young leaf and the farthest down or the tent h leaf was collected, which ever came first (Fig. 1). The nearest old leaf collected was the seventh leaf from the young leaf. The majority of the old leaves collected were in the te nth position from the young leaf sample with the exception of P. sexflora , in which the old leaf sample was more consistentl y the eighth leaf. Leaf samples from the same species we re all collected and tested the same
4 day. William Haber and Willow Zuchowski, local bot anists, were consulted for the identification of the three species used in this st udy. Toughness The toughness of the leaves was determined using a penetrometer, which measures the toughness of the leaf. Leaf samples were placed bet ween two metal plates that were 10 x 7.5 cm. Each plate had a hole drilled through the center. The hole in the bottom plate had a diameter of 3 mm and the hole in the top plat e had a diameter of 2 mm. The bottom plate had two metal posts, one in each oppos ite corner that matched up with two holes in the top plate, this worked to keep the pla tes and sample from moving during the test. A plastic plate, diameter of 7.4 cm, with a metal rod, length 1.5 cm and diameter 1.1 mm, in the center was set on top of the sample thro ugh the hole in the top plate (Fig. 2). In order to ensure that a proper measurement was ta ken, the leaf sample was placed on the bottom plate so that no venation was centered a bove the hole. A container of known mass was placed on the plastic plate and water was added to the container until the metal rod punctured the leaf. The amount of water was me asured in milliliters and converted to grams. The mass of the water was added to the mass of the container and this was used as the measure of toughness in grams. When the mas s of the plastic plate was sufficient for puncturing the sample a toughness value of zero was recorded. This test was conducted on three different locations of each leaf sample (Fig. 3) and the average was calculated and used as the final toughness value. Habitat Preference and Anti-herbivore Defense Inven tory Habitat type was noted during collection, whole lea f weight was obtained, and pubescence and gland observations were recorded pri or to the toughness test. Pubescence was determined to be high, medium, or no ne in relation to the pubescence on the other species sampled as demonstrated in Figure 4. Cyanide Analysis The sodium picrate test was used to determine the c oncentration of cyanide in each leaf sample (Seigler, 1991). Sodium picrate paper was p repared by dipping 9 x 50 mm chromatography paper into a prepared aqueous soluti on of 0.5% (w/v) picric acid and 5% (w/v) sodium bicarbonate and allowed to dry fully. Each leaf sample was weighed out to the approximate weight of the lightest sample from that species to the nearest thousandth of a gram and placed in a 21 mm glass vial. The le af sample was macerated in the vial and three drops of toluene was added as a solvent. The sodium picrate paper was then suspended over the sample by attaching it to the ru bber stopper and the vial was placed in the window seal for ten minutes at room temperature . The presence of cyanide is indicated by a change in color from yellow to red-o range. The sodium picrate paper was then placed in 3 ml of de-ionized water for 30 sec and then removed. The solution was then placed into a cuvet and read in an MRC UV-200RS Ultraviolet and Visible spectrophotometer at 540 nm, the percent transmitta nce was recorded. The blank was prepared by following the same procedure without a leaf sample. A standard curve was
5 developed by plotting the percent transmittance aga inst known values of cyanide concentration in a serial dilution. A stock soluti on of aqueous potassium cyanide (KCN) was made by dissolving 1 g of KCN in 1 ml of de-ion ized water. Each subsequent solution was made by diluting 0.1 ml of the solutio n with 0.9 ml of de-ionized water so that there were five solutions that decreased in cy anide (CN) concentration by a factor of ten from 1 g CN ml-1 to 0.0001 g CN ml-1. A sodium picrate test was performed on 0.1 ml of each CN solution as described above. The con centrations of CN were divided by 0.1 in order to account for the volume tested and t hen the unit of measure was converted to micrograms. The percent transmittance was conve rted to absorbance for both the standard curve and sample. The sample absorbance w as then converted to concentration of CN using the standard curve and divided by the s ample weight to obtain the amount of CN per leaf reported in units Âµg / g. RESULTS As predicted, the younger leaves of all three speci es were not as tough as the older leaves sampled (Fig. 5, N = 59, F = 21.60, P < 0.0001, DF = 1). P. capsularis old leaves, with a mean +/SE of 64.3g +/2.76, were more than twice as tough as its young leaves , with a mean of 23.73g +/6.26 (N =10 for old and N = 9 f or young). Likewise, old leaves for P. helleri had a mean of 78.5g +/13.07, while young leaves were less than half as tough, with a mean of 31.48 +/12.88 (N = 10 for both). P. sexflora , on the other hand, had young and old leaves that were much closer to one a nother, in terms of toughness: old leaves were near the toughness of the other species , at 73.87g +/13.07, but young leaves were far tougher than the other Passiflora spp., ha ving a mean toughness of 53.83g +/6.22, which is nearly the toughness of old P. capsu laris leaves (see figure 5). In general, there was no difference observed in the mean toughn ess between each species with young and old leaves combined (Fig. 5, N = 59, F = 2.25, P = 0.12, DF = 2). The LSMeans differences studentÂ’s t test suggested that the mea n toughness of all P. helleri leaves combined was statistically equivalent to the toughn ess of both P. capsularis and P. sexflora leaves,. By contrast, P. capsularis and P. sexflora had statistically different mean leaf toughnesses, in that P. helleri generally had tougher leaves (Fig. 5, N = 59, t = 2.01, P = 0.05). Interestingly, I did not observe an overall differe nce in the mean cyanide concentration between the young and old leaves of the three speci es (Fig. 6, N = 59, F = 0.77, P = 0.39, DF = 1). Passiflora Capsularis is the exception to this trend, the young leaves, with a mean +/SE of 47.75 Âµg / g leaf +/35.59, has 40 times the cyanide concentration as the old leaves, with a mean +/SE of 0.33 Âµg / g leaf +/0.09 (N = 10 for both). Two outliers were removed from the analysis of P. capsularis as they resulted in a very large standard error (Fig 7 and 8, [CN] = 351.97 with Tou ghness = 0 and [CN] = 113.37 with Toughness = 6). When these data were removed from the analysis the difference in mean cyanide concentration between young and old leaves of P. capsularis is statistically significant (N = 8, F = 5.14, P = 0.04, DF = 1, dat a not shown). Passiflora helleri young leaves, with a mean +/SE of 134.64 Âµg / g leaf +/ 53.01 Âµg / g leaf, had slightly higher cyanide concentrations than the old leaves, with a mean +/SE of 106.61 Âµg / g leaf +/63.71 (N = 9 for young leaves and N = 10 for old le aves). Passiflora sexflora produced
6 very little cyanide and the young leaves, with a me an +/SE of 0.05 Âµg / g leaf +/0.007, had equal concentrations of cyanide as the old leav es, with a mean +/SE of 0.05 Âµg / g leaf +/0.009 (N = 10 for both). Due to the equal concentrations of cyanide in the young and old leaves of P. sexflora , the data point for the old leaf is being covered by the data point for the young leaf in figure 6. At the speci es level, P. helleri produced greater concentrations of cyanide in both young and old lea ves than did P. capsularis and P. sexflora , which showed statistically equivalent concentrati ons of cyanide (Fig. 6, N = 59, t = 2.01, P = 0.05). A trend of decreasing cyanide concentrations with i ncreasing leaf toughness was observed for all three species of Passiflora (Fig. 9). Due to the large scale used on the yaxis in the linear regression, P. sexflora does not appear to follow this trend, however, when the scale is lowered the cyanide concentration s do show this trend despite the fact that nearly all of the data points for this species are near to zero. The linear regression also clearly shows that P. helleri produces larger quantities of cyanide. As with th e relationship between age and cyanide concentrations , P. capsularis was the only species to show statistical significance in this relationsh ip (N = 20, P = 0.02, R2 = 0.26, y = 95.02 Â– 1.61*x). No statistically significant relationsh ip was observed for P. helleri (N = 19, P = 0.13, R2 = 0.13, y = 200.28 Â– 1.43*x) and P. sexflora (N = 20, P = 0.35, R2 = 0.05, y = 0.06 Â– 0.00017*x), though all trends were in this d irection. Additional observations All species of Passiflora that where sampled were located along well cleared trails in understory habitat and often times occurred with in a few feet of another species of passion vine sampled (Table 1). Therefore, habita t type did not differ for the three species examined here. In terms of pubescence, Passiflora sexflora appeared to have substantially more pubescence than either of the tw o other species, whereas P. capsularis had relatively more pubescence than P. helleri , which appeared to have little to no pubescence (Fig. 4, Table 1). Finally, of the thre e species, P. helleri was the only one to possess observable glands on the leaf blades (Table 1). DISCUSSION Tradeoffs in anti-herbivore defenses assume a limit ed availability of resources that restricts the number and types of defenses any one plant species can employ to protect themselves (Coley and Aide 1991). Tropical species should show clear tradeoffs, as nutrients are generally limiting and herbivore pres sure is high. Given the large number of described anti-herbivore defense mechanisms utilize d by Passiflora spp . against their primary heliconiine herbivores, it is unlikely to b e energetically feasible for one species to employ all available mechanisms. Therefore, man y species in this genus will combine different mechanisms in order to evolve a subset of anti-herbivore defenses. Because the most effective protection a plant can have against herbivores is tougher, less nutrient rich tissue, such as that found in older leaves (Coley a nd Aide 1991, Kursar and Coley 2003), all plants should evolve to toughen old leaves, but may have to employ other strategies until leaves can acquire toughness.
7 In terms of toughness and cyanide, the three Passif lora spp examined are very different and seem to exhibit very different tradeoffs. Passiflora capsularis produces more cyanide when the leaves are untoughened and young, but have no cyanide in older, tougher leaves. Thus, as the leaf toughens, P. caps ularis stops its investment in cyanide defenses. Passiflora helleri did not show a trade off in the production of cyan ide with leaf age or toughness. In part, this is because ev en old, tough leaves retain high cyanide concentrations in P. helleri ; cyanide concentrations of P. helleri old leaves are equal to its young leaves and surpass even the young leaves of the other species. In P. sexflora , leaves are never very tough and yet their leaves co ntain little or no cyanide. Passiflora helleri may maintain high cyanide production in old leaves because they lack pubescence. Pubescence has been shown to be an eff ective anti-herbivore defense mechanism (Coley 1983, Gilbert 1971). However, sma ll chemical compounds, such as cyanide, are less energetically costly than a struc tural defense, such as pubescence (Coley and Aide 1991). Therefore, it maybe more efficient for P. helleri to invest in the consistent production of an inexpensive defense rat her than invest more energy at one time to a structural defense. Another possible exp lanation may be that the heliconiine butterflies that specialize on P. helleri lay their eggs on mature leaves as well as young leaves and therefore the older leaves require added protection. This is one possible scenario in the selection of the anti-herbivore def ense mechanisms employed by P. helleri. Passiflora sexflora, having neither cyanide nor tough leaves, appears to utilize yet another type of anti-herbivory defense strategy. Th is species seems to invest more in pubescence. Passiflora sexflora had considerably more pubescence than either of th e other two species sampled for this study. Gilbert, 1971, suggests that hooked pubescence is the ultimate deterrent against herbivory by Passiflora specialists heliconiine. Hooked pubescence have been shown to rip the soft flesh of new larvae causing them to lose hemolymph and eventually die (Gilbert 1971). The u se of pubescence may be so effective at deterring herbivory that P. sexflora has no reason to toughen its leaves or invest in cyanide defenses. Perhaps the abundance of pubescence observed on P. sexflora is indicative of an advantage over P. capsularis and P. helleri . It could be that P. sexflora is moving toward one of the most effective defense strategies agains t heliconiine lavae, hooked pubescence, and is farther along in the Passiflora Heliconius coevolutionary arms race than are P. capsularis and P. helleri . If this is the case, then perhaps P. capsularis is farther along in the coevolutionary arms race than P. helleri and the cyanide/toughness tradeoff observed in the former is not due to resource allocation but rather, is due to the eventual phasing out of a defense mechanism that is no longer releva nt. ACKNOWLEDGMENTS
8 I am thankful to Dr. Alan Masters for all of his gu idance and support, Dr. Karen Masters, Taegan McMahon, Phillip Burkholder, William Haber a nd Willow Zuchowski for their time and expertise, and Dr. Gregory Cronin and Dr. Sherri Jones for all of the predeparture brainstorming. I would like to thank La EstaciÃ³n BiolÃ³gica Monteverde. Finally, A special thanks goes to Pablo Allen for h elping me with my statistics and graphs more times than I care to admit, Thank you. LITERATURE CITED Baillie, I.C.. 1996. Chapter 10. Soils of the humid tropics In Richards, P.W.. 1996. The tropical rainforest second edition. pp. 256-286. Ca mbridge University Press, New York, New York. Brown, K.S. Jr.,J.R. Trigo, R.B. Francini, A.B. Bar ros de Morais, and P.C. Motta. 1991. In Price, P.W., T.M. Lewinsohn, G.W. Fernandes, W.W. Benson (Eds.). Plant Animal Interactions. pp. 403-427. John Wiley & Sons , Inc., New York, New York. Coley, P. 1983. Herbivory and defensive characteris tics of tree species in a lowland tropical forest. Ecological Monographs. 53:209-233. Coley P. D., and T.M. Aide. 1991. Comparison of her bivory and plant defenses in temperate and tropical broad-leaved forests. In Price, P.W., T.M. Lewinsohn, G.W. Fernandes, W.W. Benson (Eds.). Plant-Animal In teractions. pp. 25-49. John Wiley & Sons, Inc., New York, New York. Coley, P.D., and J. Barone. 1996. Herbivory and pla nt defenses in tropical forests. Annu. Rev. Ecol. Syst. 27: 305-335 Davis, K. 2007. Light-induced anti-herbivory defens es in Passiflora biflora (Passifloraceae).Tropical Ecology and Conservation Council on International Educational Exchange Monteverde, Costa Rica. pp. 11 4-120 DeVries, P.J. 1987. The butterflies of Costa Rica a nd their natural history Papilionidae, Pieridae, Nymphalidae. pp. 192-198. Princeton Unive rsity Press, New Jersey. Engler, H.S., K.C. Spencer, and L.E. Gilbert. 2000. Preventing cyanide release from leaves. Nature. 406:144 Feuilet, C.. 2004. Passifloraceae (passion flower f amily). In Smith, N., A. Henderson, D.W.M. Stevenson, S.V. Heald. Flowering plants of t he Neotropics. pp. 286-287. Princeton University Press, Princeton, New Jersey. Gilbert, L.E. 1971. Butterfly-plant coevolution: Ha s Passiflora adenopoda won the selectional race with heliconiine butterflies? Scie nce.172(3983): 585-586 Gilbert, L.E. 1991. Biodiversity of a central Ameri can Heliconius community: patterns, process, and problems. In Price, P.W., T.M. Lewinsohn, G.W. Fernandes, W.W. Benson (Eds.). Plant-Animal Interactions. pp. 403-4 27. John Wiley & Sons, Inc., New York, New York. Harms, K.E., J.S. Powers, and R.A. Montgomery. 2004 . Variation in small sapling density, understory cover, and resource availabilit y in four neotropical forests. Biotropica. 36(1): 40-51. Kursar, T.A. and P.D. Coley. 2003. Convergence in d efense syndromes of young leaves in tropical rainforests. Biochemical Systematics an d Ecology. 31: 929-949.
9 Mabberley, D.J. 1993. The Plant-Book: A Portable Di ctionary of the higher plants. pp. 434. the Press Syndicate of the University of C ambridge. Cambridge, New York, New York. Marquis, R.J. and H.E. Braker. 1994. Plant-herbivor e interactions: diversity, specificity, and impact. In McDade, L.A., K.S. Bawa, H.A. Hespenheide, and G.S . Hartshorn. 1994. La Selva ecology and natural history of a neo tropical rain forest. pp. 261281. The University of Chicago Press. Chicago, Illi nois. Powers, J.S. 2004. New Perspectives in comparative ecology of neotropical rain forest: reflections on the past, present, and future. Biotr opica. 36(1): 2-6 Seigler, D.S. 1991Cyanide and cyanogenic glycosides . In Rosenthal, G.A., and M.R. Berenbaum (Eds.). Herbivores: their interactions wi th secondary plant metabolites, 2nd Edition, Volume 1: The chemical participants, pp. 35-75. Academic Press, New York. Williams, K.S., and L.E. Gilbert. 1981. Insects as selective agents on plant vegetative morphology: egg mimicry reduces egg laying by butte rflies. Science. 212:467469.
10 Figure 1. Leaf collection points, P. helleri vine, arrows indicate the approximate positions of leaves collected from the sampled Passiflora , young (yellow) and old (orange).
11 Figure 2. Penetrometer technique used to determine the toughness of each Passiflora leaf sampled.
12 Figure 3. Locations of toughness test shown on P. helleri leaf, arrows indicate the three positions of the leaf toughness test, first Â– yello w, second Â– orange, and the third Â– purple. The average toughness was calculated and reported as th e sample toughness.
13 Figure 4. Relative pubescence of (a) Passiflora capsulari , (b) P. helleri , and (c) P. sexflora . A B C
14 0 10 20 30 40 50 60 70 80 90 100 P. capsularisP. helleriP. sexflora Species Old Young Figure 5. The difference between the means of toug hness as a relationship of species (N = 59, F = 2.2 5, P = 0.12, DF = 2), age (N = 59, F = 21.60, P < 0.0001 , DF = 1), and the interaction between species and age (N = 59, F = 1.11, P = 0.34, DF = 2) for Passiflora capsularis (young: N = 10, mean = 23.73 Â± 6.26 Std. Err., old: N = 10, mean = 64.3 Â± 2.76 Std. Err.), P. helleri (young: N = 9, mean = 31.48 Â± 12.88 Std. Err., old: N = 10, mean = 78.5 Â± 11.30 Std. Err.), and P. sexflora (young: N = 10, mean = 53.83 Â± 6.22 Std. Err., old: N = 10, mean = 73.87 Â± 13.07 Std. Err.). Lette rs represent significantly different means between species (N = 59, t = 2.01, P = 0.05). A AB B
15 0 20 40 60 80 100 120 140 160 180 200 P. capsularisP. helleriP. sexflora Species Old Young Figure 6. The difference between the means of cyani de concentration (Âµg / g leaf) for three species of Passiflora from Monteverde, Costa Rica (N = 59, F = 6.46, P = 0.003, DF = 2), age (N = 59, F = 0.77, P = 0.39, DF = 1), and the interaction between species and age (N = 59, F = 0.23, P = 0.79, DF = 2) for Passiflora capsularis (young: N = 10, mean = 47.75 Â± 35.59 Std. Err., ol d: N = 10, mean = 0.33 Â± 0.09 Std. Err.), P. helleri (young: N = 9, mean = 134.64 Â± 53.01 Std. Err., ol d: N = 10, mean = 106.61 Â± 63.71 Std. Err.), and P. sexflora (young: N = 10, mean = 0.05 Â± 0.007 Std. Err., old : N = 10, mean = 0.05 Â± 0.009 Std. Err.). Young leaf samples shown as a green square a nd old leaf samples shown as a gold diamond. Lette rs represent significantly different means between spe cies (N = 59, t = 2.01, P = 0.05). A B B
16 0102030405060708090 Toughness 0 50 100 150 200 250 300 350 Cyanide Â“g / g Leaf C y a n i d e Âµ g / g L e a f Figure 7. Cyanide concentrations ( m g / g leaf) of Passiflora capsularis against leaf toughness with outliers included ([CN] = 351.97 wit h Toughness = 0 and [CN] = 113.37 with Toughness = 6, N = 20, P = 0.02, R2 = 0.26, y = 95.02 Â– 1.61*x).
17 0102030405060708090 Toughness 0 1 2 3 4 5 Cyanide Â“g / g Leaf C y a n i d e Âµ g / g L e a f Figure 8. Decreasing cyanide concentrations ( m g / g leaf) with increasing toughness of Passiflora capsularis having removed two outlier points, [CN] = 351.97 w ith Toughness = 0 and [CN] = 113.37 with Toughness = 6 (N = 18, P = 0.0017, R2 = 0.47, y = 2.67 Â– 0.04*x).
18 020406080100120140160 Toughness -50 0 50 100 150 200 250 300 350 400 450 500 550 Cyanide Â“g/ g Leaf C y a n i d e Âµ g / g L e a f Figure 9. Plots of cyanide concentration (Âµg/g Leaf ) against leaf toughness of Passiflora capsularis (open green diamond and green regression line, N = 20, P = 0.02, R2 = 0.26, y = 95.02 Â– 1.61*x), P. helleri (open blue circle and blue regression line, N = 19 , P = 0.13, R2 = 0.13, y = 200.28 Â– 1.43*x), and P. sexflora (open red square and red regression line, N = 20, P = 0.35, R 2 = 0.05, y = 0.06 Â– 0.00017*x). P. helleri P. sexflora P. capsularis
19 Table 1. Observed habitat type at time of sample co llection, the presence of pubescense in relative amounts, and the presence of glands on the blades o f Passiflora capularis. P. heleri, and P. sexflora in Monteverde, Costa Rica Cloud Forest habitat be tween 1450 and 1600 meters . Species Habitat Pubescence Glands on the Blade P. capsularis Understory along trail Medium None P. helleri Understory along trial In trees along trail None Yes P. sexflora Understory along trail High None