|USFLDC Home | Tropical Ecology Collection [Monteverde Institute]||| RSS|
This item is only available as the following downloads:
Altitudinal variation in Ithomiine (Nymphalidae) color patterns Nicholas Sullender Department of Engineering, Columbia University and Daniel Winokur Department of Biology, Colorado College ABSTRACT Ithomiine butterflies contain several color comp lexes that fly at different heights in the forest, presumably in response to light conditions (Papageorgis 1975; Burd 1994). If so, these complexes should also respond to altitude, as light conditions in the forest change altitudinally. A previous study by Haber (1978) showed that color complexes do respond to altitude, but in a way inconsistent with light response. Either the previous study failed to incorporate forest conditions, like openness, which also alter light levels, or ithomiine color complexe s are responding to altitude for different reasons, including the possibility that each color complex had a different center of origin corresponding with different elevations. We caught butterflies in closed forest conditions along both slopes of the conti nental divide in Monteverde, Costa Rica at seven different sites along an altitudinal gradient, from 800 m to 1600 m. Two of the three color complexes responded to altitude, but in ways inconsistent with our expected light response. Clearwings, which fav or dark conditions, were more abundant at high altitudes while tiger stripe species, favoring strong, direct and sun fleck light conditions, were only found at lower altitudes on both slopes. Pacific slope forests generally had more open canopies but the patterns were nearly symmetrical on both slopes. Therefore, we conclude that the change was more likely a result of color complexes evolving at different altitudes and secondarily migrating out of these altitudinal bands. Additionally, light and altitude seem to cause a more complicated relationship than expected, probably because of increased cloud cover and epiphytic growth with increased altitude. RESUMEN Las mariposas de ithomiine contienen varios patrones de color que vuelan en diferentes alturas probablemente en respuesta a la luz (Papageorgis 1975; Burd 1994). Si es as estos complejos deben tambin responder a la altitud, como en las condiciones de luz atravez del cambio del bosque a nivel altitudinal. Un estudio anterior de Haber (1978) demon str que los patrones de color responden a la altitud, pero en una manera contraria con respuesta a la luz. Sin embargo, el estudio anterior no pudo incorporar las condiciones del bosque, como transparencia que tambin alteran los niveles de luz, o los pat rones del color del ithomiine estn respondiendo a la altitud por diferentes razones, incluyendo la posibilidad de que cada patrn del color tuvo un centro de origen diferente correspondiente con diversas elevaciones. Capturamos mariposas en condiciones ce rradas del bosque a lo largo de ambas pendientes de la division continental en Monteverde, Costa Rica en siete diversos sitios a lo largo de un gradiente altitudinal, a partir el 800 m a el 1600 m. Dos de los tres patrones del color respondieron a la altit ud, pero de maneras contraria con nuestra
hipotesis de luz prevista. Mariposas de alas de vidrio, se favorecen por las condiciones oscuras, fueron ms abundante en las altas altitudes mientras que las especies de la raya del tigre, las condiciones de luz f uertes, directas y de las moteadas del sol, fueron encontradas solamente en altitudes ms bajas en ambas cuestas. Los bosques del Pacifico de la cuesta tuvieron generalmente ms en lo zonas abierto pero los patrones eran casi simtricos en ambos cuestas. P or lo tanto, concluimos que el cambio era ms probablemente un resultado de los patrones del color que se desarrollaban en diferentes altitudes y despus que emigraban de estas vendas altitudinal. Adems, la luz y la altitud parecieron exhibir una relacin menos pronunciada que la esperada, probablamente debido a que la cobertura nubosa y crecimiento epfito se incrementa al incrementos de la altitud. INTRODUCTION Ithomiinae (Nymphalidae) is a common Neotropical butterfly subfamily (DeVries 1987). All ith omiine genera are chemically protected by pyrrolizidine alkaloids (Brown 1981; Masters 1990, 1992), which they sequester from several plant families, including Asteraceae, Boraginaceae, Caesalpinaceae (Fabaceae), and Orchidaceae (Haber 1978). These alkalo ids make ithomiines unpalatable to predators (Masters 1990), a fact that is reflected in their aposematic patterning (Bates 1862; Mller 1879). However, ithomiines' first line of predation defense is crypsis, an idea proposed by Papageorgis (1975). In Costa Rica, ithomiines converge to five major color patterns: clearwing, tiger stripe, black and yellow, black and rust, and golden translucent (Papageorgis 1975; Burd 1994; Levitt 1999, Haber 1978). Several genera converge on each of these distinct patte rns, and previous studies have shown that each complex flies in one of three different strata in the forest (Papageorgis 1975; Burd 1994; Levitt 1999). Clearwings fly in the understory, black and yellow and golden translucents in the middle stratum, and t iger stripes and black and rust in the upper stratum (Papageorgis 1975). According to Papageorgis, this pattern indicates that these distinct strata are optimal environments for crypsis of the corresponding ithomiine complexes. Clearwings are least visib le in the low light environment of the understory, as their translucent wings have low contrast to the vegetation. Black and yellow and golden translucents, typically brown, gold, yellow and red, have somewhat translucent wings and are most cryptic in the brown, wood dominated conditions and diffuse light of the middle stratum. The tiger stripes and black and rust patterns are most cryptic in the bright, sunfleck prone upper stratum of the forest. Light conditions in closed canopy tropical forest also cha nge with altitude (Whitmore 1989). Canopy tree height decreases, the density of large trees decreases, and leaves get smaller (Bruijnzeel and Veneklaas 1998), a trend consistent on both slopes of the continental divide in Costa Rica ( Nadkarni et al. 2000). Changes in the composition of the forest equates to changes in the amount and kind of light that gets through the forest canopy (Lee 1987). Integrating the vertical stratification of ithomiine complexes with this trend in light availability over an alti tudinal gradient, we predict that changes in altitude should correlate to changes in the relative abundance of ithomiine color complexes there. Higher elevation forests with more direct light coming through the canopy should result in an increased relativ e abundance of the tiger stripes and black and rust patterns, and a decrease in the clearwing complex. We expect golden translucents and black and yellows to remain relatively constant because diffuse light is common to all forest habitats (Endler 1993).
Indeed, ithomiines in Costa Rica appear to respond to altitude (Haber 1978), though not in a way consistent with predicted patterns. Instead, clearwings are more abundant at higher altitudes on both the Pacific and Atlantic slopes and decrease with elevat ion, while tiger stripes and black and rust patterns are more abundant at lower altitudes and decrease with increasing altitude (Haber 1978). It is possible that this trend reflects the study sites selected, as changes in forest canopy cover could obscure altitudinal impacts on light. Alternatively, the pattern of Haber (1978) may be best explained by the center of radiation model (Mora 2003), in which the area where species originate contains the highest species richness, and richness declines proportiona lly to distance from this locus, reflecting the radiation of species from this point (Karlson et al. 2004; Jablonski et al. 2006). Neotropical butterflies may have evolved and eventually speciated on mountain tops during the Pleistocene (Brown 1987). T he forest refugia theory (Haffer 1979) postulates that isolated wet "refuges" were created by glacial climate change that dried the lowland Tropics. Thus, separate mountaintops became refuges that reproductively separated many populations, due to allopatr ic speciation (Haffer 1979). Drawing upon this theory, we predict that ithomiine clearwings may have speciated in such refuges at high elevations, thus accounting for their continued greater abundance at higher altitudes (Haber 1978). Clearwing species s ubsequently radiated downwards on both slopes, accounting for their decreasing richness with decreasing altitude. Other color complexes may have speciated at lower refuges, and therefore still prefer these altitudes. The intentions of our study are thus to reevaluate the trends found in previous work using only closed forest sites on both slopes. If ithomiine color complexes respond to light in closed forests, clearwings should become less common at higher altitudes in lighter environments, compared to ot her, more light adapted color patterns. If ithomiine complexes are responding to light, ratios of color complexes for a given altitude should differ for a given altitude in a predictable way. If color complexes evolved, speciated and secondarily moved out of their preferred altitudinal band, ratios of color complexes should decline with increased distance from the center of radiation, but without regard to slope. MATERIALS AND METHODS Our study took place over an altitudinal gradient near Monteverde, Cost a Rica during the last three weeks of the dry season. We selected collection sites in closed canopy forest ranging from 800 m elevation on the Atlantic slope to 800 m on the Pacific slope, crossing over the continental divide at 1700 m in the Cordillera Ti lern. Of importance to our project was that our sites should be lek sites, for three important reasons. When ithomiines lek, the vertical stratification of flight heights is disrupted and all colors complexes occur closer to the forest floor (Haber 1978). This makes the capture of high flying species possible with a hand net. Additionally, ithomiine leks show great species richness, as individuals of many different species and genera are attracted to these aggregations (DeVries 1987). This increases our ab ility to accurately describe the species present in closed forest at a given altitude. Lastly, lek sites are not related to host plant availability (Haber 1978). This last characteristic diminishes the risk of a microhabitat bias that could occur if a coll ection was taken in close proximity to a species' host plant.
The lek sites we found were all in closed canopy forest, within close proximity to a stream or shallow river. We observed that ithomiines only displayed lekking behavior when the weather was sun ny and warm therefore, we only collected between 10:00 am and 2:00 pm, the warmest hours of the day, and collected on days with full sun whenever possible. We attempted to collect at least 40 butterflies per site, and used handheld nets (1 and 2 m handle lengths) to do so. Our estimated maximum reach was 5 m in height, however, we observed that the majority of lekking ithomiines flew below 3 m high, well within our range. We identified the butterflies to species using DeVries (1987). Study Sites We col lected at three different sites on the Atlantic slope. The first site was in the trail network behind Refugio Eladio (Peas Blancas valley, Bosque Eterno de los Nios), between 800 and 820 m elevation. We found ithomiines lekking in the woods between two s treams that connect with the Rio Peas Blancas. The second Altlantic site was along Sendero Tabacn, a trail below the Estacin San Gerardo (Bosque Eterno de los Nios), between 1150 and 1190 m elevation. There were several lek sites along this 2km loop, a ll near the streams that cross the trail heading down the mountain. The third Altantic site was in the trail network above the Selvatura skywalks, between 1610 and 1640 m elevation. Although we did not observe lekking behavior at this site, we believe that this was due to the poor weather when we visited. During good weather, we would expect to see butterflies lekking in the flat, densely vegetated areas alongside the stream that runs parallel to the main trail at this elevation. On the Pacific slope, we co llected at four different sites. The first site was in the Bajos de San Luis, along the Rio Guacimal just below the confluence of the Quebrada Socorro, between 810 and 820 m elevation on the Mata Family farm. We collected in the forest between the river an d the pastures on the hillside. The second site was along the trail to the San Luis waterfall, on the Leitons' property, at between 1160 and 1190 m elevation. The majority of lekking butterflies flew around the forest edge right along the Rio San Luis. The third site was near Monteverde, along the Rio Guacimal below the Bajo del Tigre reserve on the Trostle property, between 1350 and 1370 m elevation. We observed lekking behavior in a flat, depressed area adjacent to a bend in the river. The fourth site was along the Quebrada Maquina, above the Estacin Biolgica Monteverde, between 1540 and 1590 m elevation. We observed occasional lekking behavior along Sendero Jilguero, which parallels the stream for a stretch. All seven of our sites were within closed can opy forests, within 100 m of moving water. Closed canopy was an important requirement, as we wanted to avoid possible bias from the butterflies often seen along disturbed forest edges and roadsides, as well as migratory species that can be found in greater abundances in clearings and disturbances (DeVries 1987). RESULTS Over the three week span of our data collection, we captured and identified 342 ithomiines (see Appendix for complete listing) representing 37 of the 61 known Costa
Rican species accordi ng to DeVries (1987). All species we collected can be found on both slopes, except for six that are unique to the Atlantic slope. We found species from all five color complexes described by Haber (1978), however due to pattern similarities and rarity of tw o of the complexes, we merged the black and yellow and golden translucent color patterns into a single complex called tawny, and the tiger stripe and black and rust color patterns into the tiger stripe complex. All three mimicry patterns were present in f our of the seven sites, two of the patterns in two of the sites, and only one pattern in our high elevation Atlantic slope site. Our data showed that the relative abundance and richness of clearwing ithomiines generally increased with elevation, while tho se of the tiger stripes generally decreased. The relative abundance and richness of tawny ithomiines varied but showed no clear pattern other than being omnipresent over the study sites (Figures 1 and 2). We defined abundance as the number of individuals o f all species within a mimicry complex, and richness as the number of species within a color complex. Abundances were significantly different when considered at each site individually and over all sites. Richnesses, however, were only significantly differe nt for two of the three color complexes and two of the seven sites (see Tables 1a and 1b). The trends we found in relative abundances agree with the results of Haber (1978). Although not statistically supported, the trends in relative richnesses correspond to species elevation data. Using DeVries (1987), we calculated the midpoints of the elevation ranges for all ithomiines in Costa Rica. We then grouped this data by mimicry complex and calculated the mean range midpoint for each complex (Figure 3). These m eans approximated the elevations of greatest species richness for the corresponding color pattern. In concordance with the trends from our relative species richness data, these means demonstrated that clearwings were most rich on average at higher elevatio ns and tiger stripes at lower elevations.
0 0.2 0.4 0.6 0.8 1 1.2 Refugio Eladio (n=72) Estacin San Gerardo (n=79) Selvatura (n=13) Estacin Biolgica Monteverde (n=42) Monteverde Community (n=58) San Luis Waterfall (n=43) Bajos de San Luis (n=35) Collection site Proportion of mimicry pattern, by number of individuals (n) 0 200 400 600 800 1000 1200 1400 1600 1800 Elevation (m) Clearwing Tawny Tiger-stripe Elevation Figure 1. Relative abundances of the three mimicry complexes at the seven collection sites, based on the number of individuals collected. The first three sites are on the Atlantic sl ope, the other four on the Pacific slope. Elevations of these sites are plotted on the secondary y axis. The stippled bar corresponding to the Selvatura site represents the fact that the collection size was less than our goal (see discussion).
0 0.2 0.4 0.6 0.8 1 1.2 Refugio Eladio (s=19) Estacin San Gerardo (s=23) Selvatura (s=5) Estacin Biolgica Monteverde (s=9) Monteverde Community (s=5) San Luis Waterfall (s=8) Bajos de San Luis (s=9) Collection site Proportion of mimicry pattern, by number of species (s) 0 200 400 600 800 1000 1200 1400 1600 1800 Elevation (m) Clearwing Tawny Tiger-stripe Elevation Figure 2. Relative abundances of the three mimicry complexes at the seven collection sites, based on species richness. Table 1a. 2 tests of mimicry complexes over all seven sites. A p value of less than 0.05 indicates a signific ant difference in the number of individuals (first column) or species (second column) in the given complex. Entries in italics indicate the expected values for that test fell below 5, violating one of the rules of thumb of the 2 test. Complex By number o f individuals By number of species Clearwing 81.78 d.f. = 6 p < 0.05 13.95 d.f. = 6 p < 0.05 Tawny 61.52 d.f. = 6 p < 0.05 8.08 d.f. = 6 p > 0.05 Tiger stripe 58.67 d.f. = 6 p < 0.05 20.00 d.f. = 6 p < 0.05 Table 1b. 2 tests of each site for all mimicry complexes. A p value of less than 0.05 indicates a significant difference in the number of individuals (first column) or species (second column) in the three complexes at each site. Site By number of individuals By number of species Refugio Eladi o 21.00 d.f. = 2 p < 0.05 0.11 d.f. = 2 p > 0.05 Estacin San Gerardo 32.68 d.f. = 2 p < 0.05 5.83 d.f. = 2 p > 0.05 Selvatura 26.00 d.f. = 2 p < 0.05 10.00 d.f. = 2 p < 0.05 Estacin Biolgica Monteverde 21.14 d.f. = 2 p < 0.05 6.00 d.f. = 2 p < 0.05 Monteverde Community 98.93 d.f. = 2 p < 0.05 2.80 d.f. = 2 p > 0.05 San Luis Waterfall 34.37 d.f. = 2 p < 0.05 1.75 d.f. = 2 p > 0.05 Bajos de San Luis 23.03 d.f. = 2 p < 0.05 0.67 d.f. = 2 p > 0.05
300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 Clearwing (s=18) Tawny (s=20) Tiger (s=18) Mimicry complex Elevation (m) Figure 3. Means of ra nge midpoints for all Costa Rican ithomiine species, grouped by mimicry complex. Error bars are standard deviations. In terms of relative abundance, two sites deviated from these trends. At the Selvatura site (Atlantic slope), we only collected 13 ithomi ines, all of which were part of the clearwing mimicry complex. This was perhaps due to the cloudy conditions that day. Although we fell shy of our collection goal at this site, we hesitate to dismiss the data. We have observed butterflies of the tawny comp lex flying in this environment on non cloudy days, and we believe that a larger sample size taken on a sunny day would demonstrate proportions similar to that of our high elevation Pacific slope site, showing the joint presence of the clearwing and tawny c omplexes with a much higher relative abundance of clearwing individuals. Another site that deviated significantly from the relative abundance trends we found was the Monteverde community site (Pacific slope). We collected many fewer clearwing individuals t han we expected at this site. This site was a large, flat area in closed canopy forest, but with noticeably less understory plant density. We have found the majority of clearwing species in dense understory, where they are better able to take advantage of their cryptic patterning. We suspect that this habitat difference was responsible for the anomaly, especially because the relative richness did not deviate from the trend. Another trend shown by our data was that the Atlantic slope sites (except Selvatura) generally had a higher species richness and diversity of all three color complexes. The trend was apparent in both Margalef (Figure 4) and Shannon Wiener indices (Figure 5). (We measured species richness using Margalef indices to ensure that the trend was not an artifact of larger sample sizes on the Atlantic slope.) We suspect that the higher species
diversity in our Atlantic slope lek sites reflected the timing of our study, which took place at the end of the dry season. Ithomiines migrate from the Pacif ic slope to the less seasonal Atlantic slope to pass the dry season (Haber 1978), and therefore we would expect to. As evidence of ithomiine butterflies' migratory nature, Sorensen's quantitative indices showed comparable species composition overlap betwee n sites on both the same and opposite sides of the continental divide (Table 2). This indicated that species are not completely isolated to either side of the divide, a result consistent with the fact that migratory butterflies are able to cross the divide The Sorensen's quantitative indices also showed a significant linear relation (R 2 = 0.19; p < 0.05; n = 21) between composition overlap and elevation difference between sites (Figure 6), indicating that two sites of similar elevation, regardless of slope had more species in common than two sites of dissimilar elevation. Although the species were mobile across the divide, they showed an altitudinal preference on both slopes. This was consistent with our findings of altitudinal trends in color complex abun dances. 0 0.5 1 1.5 2 2.5 3 3.5 Refugio Eladio (n=72, s=19) Estacin San Gerardo (n=79, s=23) Selvatura (n=13, s=5) Estacin Biolgica Monteverde (n=42, s=9) Monteverde Community (n=58, s=5) San Luis Waterfall (n=43, s=8) Bajos de San Luis (n=35, s=9) Collection site Margalef's index 0 200 400 600 800 1000 1200 1400 1600 1800 Elevation (m) Clearwing Tawny Tiger-stripe Elevation Figure 4. Species richness in terms of Margalef indices.
0 0.5 1 1.5 2 2.5 Refugio Eladio (n=72, s=19) Estacin San Gerardo (n=79, s=23) Selvatura (n=13, s=5) Estacin Biolgica Monteverde (n=42, s=9) Monteverde Community (n=58, s=5) San Luis Waterfall (n=43, s=8) Bajos de San Luis (n=35, s=9) Collection site Shannon-Wiener index (H') 0 200 400 600 800 1000 1200 1400 1600 1800 Elevation (m) Clearwing Tawny Tiger-stripe Elevation Figure 5. Diversity in terms of Shannon Wiener (H') indices. Table 2. Species composition overlap between communities, in terms of Sorensen's quantitative indices. Refugio Eladio Estacin San Gerardo Selvatura Estacin Biolgica Monteverde Monteverde Community San Luis Waterfall Bajos de San Luis Refugio Eladio 0.15 0.00 0.04 0.03 0.12 0.11 Estacin San Gerardo 0. 28 0.46 0.07 0.52 0.21 Selvatura 0.40 0.03 0.07 0.00 Estacin Biolgica Monteverde 0.08 0.35 0.18 Monteverde Community 0.06 0.13 San Luis Waterfall 0.23 Bajos de San Luis
y = -0.0003x + 0.277 R 2 = 0.1913 0 0.1 0.2 0.3 0.4 0.5 0.6 0 100 200 300 400 500 600 700 800 900 Magnitude of elevation difference Sorensen's quantitative index Figure 6. Species composition overlap (Sorensen's quantitative indices) versus elevation difference between communities. Stippled line represents linear regression. DISCUSSION Studies have shown that separate color complexes occupy distinct vertica l strata (Burd 1994) in order to utilize light conditions unique to each vertical zone to become more cryptic (Papageorgis 1975). Additionally, in mountainous areas in the Tropics, increasing altitude leads to a general decrease in canopy height and leaf size (Bruijnzeel and Veneklaas 1998), suggesting an increase in the amount of direct sunlight. If there is more available light at higher elevations, and the optimal strata for tiger stripes is the strata of the forest that receives the most light, then t iger stripes should be the most prevalent complex at higher elevations, and clearwings the least. Our data exhibit different trends for relative abundances of ithomiine color complexes across an altitudinal gradient, similar to those from Haber (1978). Spe cifically, clearwings are most abundant at higher altitudes, and decrease in abundance as altitude decreases. The inverse is true of tiger stripes: they are most abundant at low altitudes and increase in abundance as altitude increases (Figure 1). The sam e trends are also visible in altitudinal gradients of relative species richness (Figure 2), absolute species richness (Margalef indices, Figure 4), and diversity (Shannon Wiener indices, Figure 5). The trends refute our light hypothesis, suggesting that ot her factors influence available light under the canopy and may have obscured or changed the predicted pattern. Namely, at higher elevations there is increased cloud cover, which in fact decreases
available light in the understory due to a dramatic decrease in the amount of direct light reaching the canopy (Endler 1993). Thus, higher altitude habitats, such as cloud forest, might actually be darker in spite of a shorter canopy. Furthermore, in cloud forest and rain shadow environments, the higher mist and pr ecipitation lead to increased epiphyte growth, possibly compounding the effects of the cloud cover. With more epiphytic growth, not only is there a higher amount of green light transmittance into the understory (Endler 1993) but also there is also more veg etation blocking direct light from passing through the canopy strata. These factors would all lead to a lower light environment at higher elevations, and one in which spectral quality is diminished, equating to a more conducive light environment for the c learwings (Papageorgis 1975; Lee 1987). Interestingly, the majority of butterflies we observed in all environments on cloudy days were clearwings, an observation that helps to support this argument. We believe that our initial light hypothesis was incomple te in its evaluation of forest light conditions the effects of decreased canopy height and leaf size with increasing elevation are likely offset by increased cloud cover and epiphytic growth. The magnitude of this offset needs to be addressed by light av ailability studies. One interesting result from our data is that the proportion of the tawny complex did not show any consistent pattern with elevation. We suggest that because the tawnies' optimal light environment is diffuse light, they will be ubiquitou s through the forest habitats, as we observed diffuse light in all of the forest habitats we visited. This one trend supports our original light hypothesis. Apropos of the light discussion above, cloud cover and epiphytic growth should not significantly al ter the presence of diffuse light under the canopy. Center of radiation models also appear to be pertinent in explaining our data. The clearwing complex is most prevalent at higher elevations and decreases in species richness across a descending altitudina l gradient. Such a gradient typically denotes a center of radiation, as highest richness is found at the geographical center and decreasing richness at further distances from this point (Karlson et al. 2004). We hypothesize that ithomiine clearwings' cent er of radiation was at higher elevations, perhaps reinforced by light conditions optimal to their cryptic patterning. From this elevation, they radiated down both slopes of the divide. The same model can be applied to the tiger stripes. We hypothesize tha t this complex had a center of radiation at a lower altitude, also perhaps due to optimal available light, and radiated up both slopes from the lowlands. Considering the center of radiation model, the tawny complex is more complicated due to the lack of a clear richness gradient. The tawnies perhaps originated from a center like the other two patterns, but radiated much further due to the fact that their optimal light conditions (diffuse light) are present in a wide range of forest habitats. This center of radiation could also have been much broader in altitudinal range than the other complexes' centers. Alternately, the tawnies might have diverged from the clearwings or the tigers, as the tawny patterning visually falls somewhere between the other two patte rns. In conclusion, we believe that the variance in color pattern proportions over the altitudinal gradient was due to a combination of two factors. First, we suspect different origins of radiation for the clearwings and tigers, in regard to the opposing directions of these two complexes' diversity gradients. We also suspect that these speciation events were influenced by the second factor: a gradient of available light conditions. Vertical
stratification experiments have demonstrated that an available lig ht gradient is an important microhabitat determinant for ithomiines. However, the quantity and quality of available light over an altitudinal gradient appears to be more complex than the vertical stratification experiments, possibly influenced by factors s uch as cloud cover and epiphytic growth. Further studies need to examine the genetics of the mimicry complexes, lending more insight into the radiation models. Additionally, the formulation of a better light hypothesis requires studies of available light o ver an altitudinal gradient. ACKNOWLEDGEMENTS We would like to thank Alan and Karen Masters, for all of their help with the project, especially regarding significance of our trends; Bill Haber, for kindly lending us his thesis; the Leitons, for the use o f their trail to the San Luis Waterfall; Eladio, for his Refugio; and John T.M.T.' Thurstman and Matt Kleinart, for getting dirty and swinging nets with us. LITERATURE CITED Bates, H.W. 1862. Contributions to an insect fauna of the Amazon valley (Lep idoptera: Heliconidae) Transactions of the Linnaean Society of London 23 :495 566 Brown, K.S. 1981. The biology of Heliconius and related genera. Annual review of Entomology 26 :427 456. Brown, K.S. 1987 Biogeography and evolution of neotropical butterfli es. Biogeography and quaternary history in tropical America (eds. Whitmore, T.C. & Prance, G.T.), pp. 66 104, Oxford, UK: Oxford University Press Bruijnzeel, L.A. and E.J. Veneklaas. 1998. Climatic conditions and tropical montane forest productivity: the f og has not lifted yet. Ecology 79 :3 9. Burd, M. 1994. Butterfly wing colour patterns and flying heights in the seasonally wet forest of Barro Colorado Island, Panama. Journal of Tropical Ecology 10 :601 610. DeVries, P.J. 1987. The Butterflies of Costa Rica, and their Natural History. Princeton University Press, New Jersey. 214 216. Endler, J.A. 1993. The color of light in forests and its implications. Ecological Monographs 63 :1 27. Haber, W.A. 1978. Evolutionary ecology of tropical mimetic butterflies (Lepidoptera:Ithomiidae). The University of Minnesota, Ph.D. Haffer, J. 1979. General aspects of the refuge theory. In : Prance, G.T. (Ed.). Biological Diversification in the Tropics. Columbia University Press, New York, NY. 6 24. Jablonski, D., K. Roy, an d J.W. Valentine. 2006. Out of the tropics: evolutionary dynamics of the latitudinal diversity gradient. Science 314 :102 106. Karlson, R.H, H.V. Cornell, and T.P. Hughes. 2004. Coral communities are regionally enriched along an oceanic biodiversity grad ient. Nature 429 :867 870. Lee. D.W. 1987. The spectral distribution of radiation in two neotropical rainforests. Biotropica 19 :161 166. Levitt, E. 1999. Vertical stratification in color complexes of Ithomiine (Nymphalidae) butterflies. CIEE Independent Project. Masters, A.R. 1990. Transparent butterflies. In : Nadkarni, N.M. and N.T. Wheelwright. Monteverde: Ecology and Conservation of a Tropical Cloud Forest. Oxford University Press, New York. 120 121. Masters, A.R. 1992. Variable chemical defense and mimicry. In : Nadkarni, N.M. and N.T. Wheelwright. Monteverde: Ecology and Conservation of a Tropical Cloud Forest. Oxford University Press, New York. 121 122. Mora, C., P.M. Chittaro, P.F. Sale, J.P. Kritzer, and S.A. Ludsin. 2003. Patterns and proce sses in reef fish diversity. Nature 421 :933 936.
Mller, F. 1879. Ituna and Thyridia ; a remarkable case of mimicry in butterflies (translated by R. Meldola). Procedings of the Royal Entomological Society of London. 20 29. Nadkarni, N.M., R.O. Lawton, K.L. Clark, T.J. Matelson, and D. Schaefer. 2000. Ecosystem Ecology and Forest Dynamics. In : Nadkarni, N.M. and N.T. Wheelwright. Monteverde: Ecology and Conservation of a Tropical Cloud Forest. Oxford University Press, New York. 303 308. Papageorgis, C. 1 975. Mimicry in neotropical butterflies. American Scientist 63 :522 532. Whitmore, T.C.W. 1989. Tropical forest nutrients, where do we stand? A tour de horizon. In : Proctor, J. (Ed.). Mineral nutrients in tropical forest and savanna ecosystems. Blackwell Scientific Press, Oxford, UK. 1 13.
Appendix. Number of individuals of each species collected at each site, and the corresponding mimicry complex. Site Mimicry Complex Refugio Eladio (n=72) Estacin San Gerardo (n=79) Selvatura (n=13) Estacin Biolgica Monteverde (n=42) Monteverde Community (n=58) San Luis Waterfall (n=43) Bajos de San Luis (n=35) Callithomia hezia hezia 1 Tiger stripe Dircenna chiriquensis 1 Tawny Dircenna klugii 1 1 2 Tawny Dircenna relata 1 1 Tawny Episcada salvinia 2 1 1 Clearwing Godyris zavaleta sorites 16 Tawny Greta andromica lyra 4 1 Clearwing Greta anette 11 8 8 Clearwing Greta nero 6 1 1 Clearwing Greta oto 1 3 Clearwing Greta polissena umbrana 7 1 1 Cl earwing Hyalyris excelsa decumana 2 1 Tiger stripe Hypoleria cassotis 3 Clearwing Hyposcada virginiana evanides 6 2 Tiger stripe Ithomia bolivari 1 Clearwing Ithomia diasa hippocrenis 2 3 Clearwing Ithomia heraldica 1 24 1 2 3 30 7 Tawny Ithomia patilla 3 1 4 1 Clearwing Ithomia xenos 1 7 Tawny Mechanitis ethra lilis 1 Tiger stripe Mechanitis lysimnia doryssus 3 Tiger stripe Mechanitis menapis saturata 4 1 4 Tiger stripe Mechanitis polymnia isth mia 1 Tiger stripe Melinaea ethra lilis 1 Tiger stripe Napeogenes cranto paedaretus 14 Tawny Napeogenes tolosa amara 1 1 Tiger stripe Oleria rubescens 2 2 Clearwing Oleria vicina 3 2 10 2 Clearwing Oleria zelica pagasa 1 1 1 Tawny Olyras insignis insignis 9 Tawny Pseudoscada utilla pusio 2 1 Clearwing Pteronymia artena artena 1 3 1 1 Clearwing Pteronymia fulvescens 52 2 Tawny Pteronymia notilla 2 14 Tawny Pteronymia parva 1 Clearwi ng Pteronymia simplex simplex 1 1 Clearwing Tithoria tarricina pinthias 1 Tiger stripe
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 M39-00139
Variacin en la altitud en los patrones de color de Ithomiinae (Nymphalidae)
Altitudinal variation in Ithomiine (Nymphalidae) color patterns
Ithomiine butterflies contain several color complexes that fly at different heights in the forest, presumably in response to light conditions (Papageorgis 1975; Burd 1994). If so, these complexes should also respond
to altitude, as light conditions in the forest change altitudinally. A previous study by Haber (1978) showed that color complexes do respond to altitude, but in a way inconsistent with light response. Either the previous study failed to incorporate forest conditions, like openness, which also alter light levels, or ithomiine color complexes are responding to altitude for different reasons, including the possibility that each color complex had a different center-of-origin corresponding with different elevations. We caught butterflies in closed forest conditions along both slopes of the continental divide in Monteverde, Costa Rica at seven different sites along an altitudinal gradient, from 800 m to 1600 m. Two of the three color
complexes responded to altitude, but in ways inconsistent with our expected light response. Clearwings, which favor dark conditions, were more abundant at high altitudes while tiger stripe species, favoring strong, direct and sun-fleck light conditions, were only found at lower altitudes on both slopes. Pacific slope forests generally had more open canopies but the patterns were nearly symmetrical on both slopes. Therefore, we conclude that the change was more likely a result of color complexes evolving at different altitudes and secondarily migrating out of these altitudinal bands. Additionally, light and altitude seem to cause a more complicated relationship than expected, probably because of increased cloud cover and epiphytic growth with increased altitude.
Las mariposas Ithomiinae contienen varios complejos de colores que vuelan a differentes alturas en el bosque, presumiblemente en respuesta a condiciones de luz (Papageorgis 1975; Burd. 1994). Si es as, estos complejos tambin debe responder a la altitud, a como cambian altitudinalmente las condiciones de luz en el bosque. Un estudio previo realizado por Haber (1978) demostr que los complejos de color responden a la altitud, pero de una manera incompatible con la respuesta de la luz. O bien el estudio previo no incorporo las condiciones del bosque, como la apertura, que tambin alteran los niveles de luz, o los complejos de color Ithomiinae estn respondiendo a la altitud por diversas razones, incluyendo la posibilidad de que cada complejo de color tena otro centro de origen correspondiente con diferentes alturas. Cogimos a las mariposas en condiciones de bosque cerrado a lo largo de ambas vertientes de la divisin continental en Monteverde, Costa Rica en siete puntos diferentes a lo largo de un gradiente altitudinal, desde 800m hasta 1600m.
Text in English.
Environmental impact analysis
Anlisis de impacto ambiental
Tropical Ecology 2007
Ecologa Tropical 2007
Patrones de color
t Monteverde Institute : Tropical Ecology