USF Libraries

Functional differences among color morphs of Impatiens walleriana (Balsaminaceae)

MISSING IMAGE

Material Information

Title:
Functional differences among color morphs of Impatiens walleriana (Balsaminaceae)
Translated Title:
Las diferencias funcionales entre los morfos de color de Impatiens walleriana (Balsaminaceae) ( )
Physical Description:
Book
Language:
English
Creator:
Morris, Anne, B
Publication Date:

Subjects

Subjects / Keywords:
Flowers--Color   ( lcsh )
Pollination by animals   ( lcsh )
Costa Rica--Puntarenas--Monteverde Zone   ( lcsh )
Flores--Colores
Polinizado por animales
Costa Rica--Puntarenas--Zona de Monteverde
Tropical Ecology Spring 2005
Ecología Tropical Primavera 2005
Genre:
Reports   ( lcsh )
Reports

Notes

Abstract:
Functional differences between color morphs of Impatiens walleriana were studied in Monteverde, Costa Rica to see if they influenced the fitness of the color morphs. Nectar removal rates, nectar quantity, nectar quality, length of pollen availability, flower longevity, and seed set were compared between magenta, salmon, pink, and white color morphs. Red, purple, magenta, pink, salmon, and white cultivated Impatiens and magenta, salmon, pink, and peach I. walleriana were observed in the Monteverde Butterfly Garden to see if color preferences occur among butterflies. In addition, magenta color morphs were studied along an altitudinal gradient to see if altitude caused functional differences within a single color morph and to see if instances of nectar robbery varied on an altitudinal scale. Among I. walleriana color morphs nectar removal rates were found to be affected by nectar robbery, all morphs suffered equally from nectar robbery. Color morphs also produced similar quantities of nectar. Magenta and salmon morphs produced nectar with higher sucrose concentrations than pink and white morphs. Color morphs did not exhibit differences in the length of pollen availability, but pink and white flowers were found to live longer than magenta and salmon flowers. In terms of seed set, all flowers were found to have high occurrences of seed set despite differences in nectar removal, robbery, and quality and in flower longevity among morphs. When cultivated Impatiens were observed for removal rates, pink and white had the most nectar removed. Heliconiinae butterflies showed species specific color preferences to the cultivated Impatiens. Eueides isabella preferred pink, white and salmon colors. Heliconius hecale preferred pink and magenta flowers. Dryas iulia preferred white and salmon colors and Heliconius charitonius preferred the colors pink and salmon. When I. walleriana color morphs were observed in the butterfly garden, salmon was the morph with the least nectar removed. Two Heliconiinae butterflies showed specific color preferences to I. walleriana. Eueides isabella disliked the color salmon. Heliconius hecale preferred salmon and peach. Although, Dryas iulia and Heliconius charitonius also visited I. walleriana they did not demonstrate specific color preferences. In addition, magenta color morphs were observed in San Luis (1105 m), on la Trocha (1305 m), and in Montverde (1535 m). A significant difference was found between nectar removal in San Luis and la Trocha; this may have been due to nectar robbery, which only occurred in San Luis. Differences in pollen availability were also found between the three sites. Along the gradient, almost all flowers set seed. Differences in color (nectar removal), nectar concentration, flower longevity, length of pollen availability, and nectar robbery cannot explain the high occurrence of seed set in all flowers; therefore, I. walleriana color morphs have similar fitness. It is likely that most flowers are being pollinated, and thus set seed, due to the presence of many different pollinators whose preferences encompass all the color morphs.
Abstract:
Las diferencias funcionales entre las formas de color de Impatiens walleriana fueron estudiadas en Monteverde, Costa Rica, para determinar si ellos influían en las adaptaciones de las formas de color. Las tasas de eliminación, la cantidad y calidad del néctar, el tiempo de disponibilidad del polen, la longevidad de la flor y la producción de las semillas se compararon entre las formas de color magenta, salmón, rosado, y blanco. Las formas cultivadas de Impatiens de colores rojo, púrpura, magenta, rosado, salmón y blanco, y las formas silvestres de I. walleriana de colores magenta, salmón, rosado y melocotón se observaron en el Jardín de Mariposas de Monteverde para ver si las mariposas preferían alguna variación de color. Además, la forma de color magenta fue estudiada en un gradiente altitudinal para determinar si la altitud causó las diferencias funcionales dentro de un solo color y para ver si los casos de robo de néctar variaban con la elevación. Se encontró que entre las formas de color magenta, salmón, rosado y blanco, las tasas de la eliminación del néctar fueron afectadas por el robo de néctar; pero no se encontró ninguna diferencia significativa en el robo de néctar entre las formas de color. La cantidad de néctar tampoco varió significativamente entre las formas de color. Las formas de color magenta y salmón produjeron néctar con concentraciones de sacarosa mayores que las formas rosadas y blancas. Las formas de color no exhibieron ninguna diferencia en el tiempo de disponibilidad de polen; pero las formas rosadas y blancas presentaron una longevidad mayor que las formas magenta y salmón. Todas las flores presentaron una producción de semillas grande, a pesar de diferencias en la eliminación, robo y calidad del néctar, y en la longevidad de la flor entre las diferentes formas. Las formas blancas y rosadas cultivadas mostraron las tasas más altas de eliminación de néctar. Las mariposas de la subfamilia Heliconiinae mostraron preferencias específicas de color en las Impatiens cultivadas. Eueides isabella prefirió las formas rosadas, blancas y salmón. Heliconius hecale prefirió las flores rosadas y magenta. Dryas iulia prefirió los colores blancos y salmon, y Heliconius charitonius prefirió las rosadas y salmón. El color salmón presentó la tasa más baja de eliminación de néctar en el jardín. Dos mariposas de la subfamilia Heliconiinae mostraron preferencias específicas de color por I. walleriana. Eueides isabella presentó aversión al color salmón. Heliconius hecale prefirió el salmón y el melocotón. Aunque Dryas iulia y Heliconius charitonius también visitaron I. walleriana, no mostraron preferencias específicas de color. La forma de color magenta fue observada en San Luis (1105 m), en la Trocha (1305 m) y en Monteverde (1535 m); se encontraron diferencias significativas entre las flores de San Luis y las de la Trocha. Esta diferencia puede haber sido debida a que el robo de néctar sólo ocurrió en San Luis; también se encontraron diferencias en la disponibilidad de polen entre los tres sitios. Casi todas las flores produjeron semillas en las tres localidades. Ya que las diferencias en color (y, por ende, la eliminación del néctar), la concentración de néctar, la longevidad de la flor, el tiempo de disponibilidad de polen, y el robo de néctar no pueden explicar la producción alta de semillas en todas las variaciones de color, se puede concluir que las formas de color magenta, salmón, rosado y blanco de I. walleriana son funcionalmente semejantes en términos de adaptación. Aunque se demostró cierta preferencia por color, deben existir preferencias diferentes en los polinizadores para asegurar que todas las formas de color sean polinizadas, ya que la mayoría de las flores fue polinizada y produjo semillas.
Language:
Text in English.
General Note:
Born Digital

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - M39-00343
usfldc handle - m39.343
System ID:
SFS0001512:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Functional Differences Among Color Morphs of Impatiens walleriana (Balsaminaceae) Anne B. Morris Department of Biology, Fairfield University ABSTRACT Functional differences between color morphs of Impatiens walleriana were studied in Monteverde, Costa Rica to see if they influenced the fitness of the color morphs. Nectar removal rates, nectar quantity, nectar quality, length of pollen availability, flower longevity, and seed set were compared between magenta, salmon, pink, and white color morphs. Red, purple, magenta, pink, salmon, and white cultivated Impatiens and magenta, salmon, pink, and peach I. walleriana were observed in the Monteverde Butterfly Garden to see if color preferences occur among butterflies. In addition, magenta color morphs were studied along an altitudinal gradient to see if altitude caused functional differences within a single color morph and to see if instances of nectar robbery varied on an altitudinal scale. Among I. walleriana color morphs nectar removal rates were found t o be affected by nectar robbery, all morphs suffered equally from nectar robbery. Color morphs also produced similar quantities of nectar. Magenta and salmon morphs produced nectar with higher sucrose concentrations than pink and white morphs. Color mor phs did not exhibit differences in the length of pollen availability, but pink and white flowers were found to live longer than magenta and salmon flowers. In terms of seed set, all flowers were found to have high occurrences of seed set despite differenc es in nectar removal, robbery, and quality and in flower longevity among morphs. When cultivated Impatiens were observed for removal rates, pink and white had most nectar removed. Heliconiinae butterflies showed species specific color preferences to the cultivated Impatiens. Eueides isabella preferred pink, white and salmon colors. Heliconius hecale preferred pink and magenta flowers. Dryas iulia preferred white and salmon colors and Heliconius charitonius preferred the colors pink and salmon. When I. walleriana color morphs were observed in the butterfly garden salmon was the morph with the least nectar removed. Two Heliconiinae butterflies showed specific color preferences to I. walleriana. Eueides isabella disliked the color salmon. Heliconius hec ale preferred salmon and peach. Although, Dryas iulia and Heliconius charitonius also visited I. walleriana they did not demonstrate specific color preferences. In addition, magenta color morphs were observed in San Luis (1105 m), on la Trocha (1305 m), and in Montverde (1535 m). A significant difference was found between nectar removal in San Luis and la Trocha; this may have been due to nectar robbery, which only occurred in San Luis. Differences in pollen availability were also found between the thre e sites. Along the gradient, almost all flowers set seed. Differences in color (nectar removal), nectar concentration, flower longevity, length of pollen availability, and nectar robbery cannot explain the high occurrence of seed set in all flowers; ther efore, I. walleriana color morphs have similar fitness. It is likely that most flowers are being pollinated, and thus set seed, due to the presence of many different pollinators whose preferences encompass all the color morphs. RESUMEN Las diferencias funcionales entre las formas de color de Impatiens walleriana fueron estudiadas en Monteverde, Costa Rica, para determinar si ellos influan en las adaptaciones de las formas de color. Las tasas de eliminacin y la cantidad y calidad del nctar, el tiempo de disponibilidad del polen, la longevidad de la flor y la producin de la semilla se compararon entre las formas de color magenta, salmn, rosado, y blanco. Las formas cultivadas de Impatiens de colores rojo, prpura, magenta, rosado, salmn y blanco, y l as formas silvestres de I. walleriana de colores magenta, salmn, rosado y melocotn se observaron en el Jardn de Mariposas de Monteverde para ver si las mariposas preferan alguna variacin de color. Adems, la forma de color magenta fue estudiada en un gradiente altitudinal para determinar si la altitud caus las diferencias funcionales dentro de un solo color y para ver si los casos de robo de nctar variaban con la elevacin Se encontr que entre las formas de color magenta, salmn, rosado y blanco, l as tasas de la eliminacin del nctar fueron afectadas por el robo de nectar; pero no se encontr ninguna

PAGE 2

2 diferencia significativa en el robo de nctar entre formas de color. La cantidad de nctar tampoco vari significativamente entre las formas de color. Las formas de color magenta y salmn produjeron nctar con concentraciones de sucrosa mayores que las formas rosadas y blancas. Las formas de color no exhibieron ninguna diferencia en el tiempo de disponibilidad de pollen; pero las formas rosadas y blanc as presentaron una longevidad mayor que las formas magenta y salmn. Todas las flores presentaron una produccin de semillas grande, a pesar de diferencias en la eliminacin, robo y calidad del nctar, y en la longevidad de la flor entre las diferentes for mas. Las formas blancas y rosadas cultivadas mostraron las tasas ms altas de eliminacin de nctar. Las mariposas de la subfamilia Heliconiinae mostraron preferencias especficas de color en las Impatiens cultivadas. Eueides isabella prefiri las formas r osadas, blancas y salmn. Heliconius hecale prefiri las flores rosadas y magenta. Dryas iulia prefiri los colores blancos y salmon, y Heliconius charitonius prefiri las rosadas y salmn. El color salmn present la tasa ms baja de eliminacin de nectar en el jardn. Dos mariposas de la subfamilia Heliconiinae mostraron preferencias especficas de color por I. walleriana. Eueides isabella present aversin al color salmon. Heliconius hecale prefiri el salmn y el melocotn. Aunque Dryas iulia y Helicon ius charitonius tambin visitaron I. walleriana, no mostraron preferencias especficas de color. La forma de color magenta fue observada en San Luis (1105 m), en la Trocha (1305 m) y en Montverde (1535 m); se encontraron diferencias significativas entre la s flores de San Luis y las de la Trocha. Esta diferencia puede haber sido debida a que el robo de nctar slo ocurri en San Luis; tambin se encontraron diferencias en la disponibilidad de polen entre los tres sitios. Casi todas las flores produjeron semi llas en las tres localidades. Ya que las diferencias en color (y, por ende, la eliminacin del nctar), la concentracin de nctar, la longevidad de la flor, el tiempo de disponibilidad de polen, y el robo de nctar no pueden explicar la produccin alta de semillas en todas las variaciones de color, se puede concluir que las formas de color magenta, salmn, rosado y blanco de I. walleriana son funcionalmente semejantes en trminos de adaptacin. Aunque se demonstr cierta preferencia por color, deben existi r preferencias diferentes en los polinizadores para asegurar que todas las formas de color sean polinizadas, ya que la mayora de las flores fue polinizada y produjo semillas. INTRODUCTION Balsaminaceae contains four genera and 500 600 species (Heywood e t al. 1996) including the common ornamental genera Impatiens Impatiens walleriana has escaped from East Africa (Haber 2000). This roadside herb is if often found along disturbed forest edges and streambeds in Monteverde, Cost a Rica (Haber 2000). Local observations suggest that, while I. walleriana flowers are visited by hummingbirds, bees, and butterflies; they are pollinated by butterflies (Haber 2000). Color polymorphism is common within the genera, including within I. walle riana The most abundant morph found in the Monteverde region has magenta colored flowers, but various shades of red, salmon, purple, pink, peach, and white have also been observed. Variable flower morphology can directly affect components of the mating s ystem such as pollinator visitation (Murcia et al. 1989). This may be because plants that depend on animal pollinators for pollen exchange among individuals are likely to be influenced by variation in floral signals such as corolla shape, petal color, and odor (Stanton 1987). Variation in floral signals would be expected to increase floral attractiveness to pollinators which would then enhance plant fitness, and ultimately, reproductive success (Stanton 1987 & Waser 1983). Color, therefore, may affect r eproductive success among different color morphs because pollinators can see different colors and patterns on the petals. This may further impact differences among morphs if pollinators respond to different color cues by showing preference to one color mo rph over another. A study comparing different color morphs of Platystemon californicus found that different morphs present different visual cues to pollinators and color patterns displayed in visible light apparently correspond to patterns visible to poll inators (Hannon 1981). In terms of butterflies, a known pollinator of I. walleriana, they often fly to

PAGE 3

3 bright colors like pink (Howe and Westley 1988). Butterflies are also known to see red (Barth 1991). Color might also affect species fitness if pollina tors are not only responding to color, but also demonstrate certain color preferences. Pollinator response to color variation can be due to certain morphs being visited more often than others due to innate or learned pollinator selectivity (Stanton 1987). In a study of Linaria canadensis (Scrophulariaceae), Sellers and Wolfe (1997) suggested that the patchwork of different color morphs found might be a result of selection driven by pollinator preference to color. Stanton (1987), however, suggests that in many cases, color preferences are short term and inconsistent. This suggests that color preference may not be the only factor influencing differences between color morphs. Characteristics like nectar production are associated with flower color and might obscure the significance of color as a method of pollinator attraction (Hannon 1981). Nectar is one of the rewards that flowers can produce in order to ensure that they attract and re attract pollinators (Masters 2005). A study conducted between differe nt color and scent morphs of Polemonium viscosum found that there were no differences in sugar concentration or rate of nectar production among the morphs (Galen and Kevan 1980). This is not to say, however, that, within an Impatiens system, nectar quanti ty and quality are not affecting nectar removal by pollinators and, thus, affecting plant fitness. Beyond color, flower longevity and length of pollen availability could also be influencing fitness among different color morphs. The longer each flower ha s pollen and the longer each flower lives, provides more time for butterflies, and other visitors, to find and pollinate each flower. A study between Impatiens capensis and Impatiens pallida suggested that when both species were present, the lower number of I. pallida flowers found per meter cubed was inconsequential because it has a longer flowering period and, thus, has more time to attract pollinators and successfully produce seeds (Rust 1977). Therefore, in patches containing multiple color morphs the morph that lives the longest has the greatest chance of being pollinated. Monteverde encompasses a wide range of elevations, 700 1600 meters, and two Holdridege life zones, premontane wet forest and lower montane wet forest (Haber 2000). I. walleriana occur throughout this climactic range. It is therefore, important to be sure that altitude is not affecting the individual fitness of the plants regardless of their color. Regardless of what factors are interacting to make individual color morphs differen t from each other, the ultimate test of fitness is the species ability to reproduce by setting seed. The evolutionary impact of relative attractiveness of color morphs to pollinators depends on the degree to which the plant species relies on insect pollin ation for seed set (Hannon 1981). A study conducted between deep purple and light blue/lavender color morphs of Linaria canadensis found that there were no significant differences between the number of fruit per plant (T test = .004) or in seed germinatio n (T test = .122 and Sellers and Wolfe 1997). This suggests that there is no overall difference in fitness between color morphs of L. canadensis Similar results were found even when differential visitation (color preference) was occurring between color morphs. Pollinator preference was observed between petal color morphs of Raphanus sativus, but no significant difference in fruit set occurred between the morphs (Stanton 1987). These studies suggest that although color polymorphism may occur and color m orphs may differ functionally, overall the polymorphism is not affecting the fitness of the plant. This study looks at how fitness of different color morphs of I. walleriana is affected by functional differences between morphs. First, pollinators often respond to different colors and

PAGE 4

4 have been demonstrated to show different color preferences. In terms of butterflies, a known pollinator of I. walleriana, they are often known to fly to bright colors like pink (Howe and Westley 1988). Butterflies are also known to see red (Barth 1991). Further, Heliconius (Heliconiinae) butterflies have been observed to visit and feed on I. walleriana in garden number two at the Monteverde Butterfly Garden. Heliconius charitonius is one species found in the Butterfly Garde n, and it has been found that this species can exhibit color preferences. Out of seven different colors, gray, and white paper models, the H. charitonius attempted to feed on blue/blue green and orange/red paper models most frequently (Swihart and Swihart 1970). Perhaps they will exhibit color preferences to I. walleriana If nectar is a reward for pollinator visitation, then color preference will be reflected in differing amounts of nectar removal from each color morph. Second, nectar quantity and quali ty are characteristics associated with color that may also affect fitness; and, thus, reduce the importance of color as an attracting force. Butterflies are known to like flowers that contain copious amounts of sucrose rich nectar and may be drawn to the c olor morphs that best exhibit this quality (Howe and Westley 1988). Aside from color entirely, flower longevity and the length of pollen availability increase fitness within a color morph by increasing the time allotted for pollination to occur. By compa ring flower fitness between one color morph at different altitudes, the effect of altitude on flower fitness will also be considered. Ultimately, the importance of nectar removal, quantity, and quality, the length of pollen availability and flower longevi ty, and the effects of altitude to different color morphs of I. walleriana can be demonstrated by the proportion of seed set between color morphs. Color morphs that set the most seed have the opportunity for the greatest reproductive success. MATERIALS AND METHODS Color and Nectar Removal in the Field Wild populations of magenta, salmon, pink, and white color morphs of I. walleriana were found throughout different locations, elevations, and Holdridge life zones in Monteverde, Costa Rica. San Luis has an altitude of 1105 meters and is located in the premontane wet forest life zone. Here, flowers were sampled from three patches that contained at least two color morphs. For all tests, magenta and salmon colored flowers were sampled in a patch at the en trance of the cabins belonging to the San Luis EcoLodge and Research Station. Pink and white flowers were sample on the paths leading down to the cabins and in a patch on the left side of the road where the Ro Aldonso crosses the main road. Wild populat ions of salmon, peach, and pink color morphs were collected on the roadside near the Friends School and near the Colina Lodge in Monteverde which are located at an altitude of 1400 meters in the premontane wet forest life zone. These flowers were transport ed to Garden 2 of the Monteverde Butterfly Garden. The Monteverde Butterfly Garden is located at 1340 meter and falls into the premontane wet forest life zone. Garden 2 is described as simulating a mid elevation climate. Wild magenta color morphs were al so sampled twenty meters above the lookout on La Trocha, at and altitude of 1305 meters in the premontane wet forest life zone, and in a patch above the lower lab at the Monteverde Biological Station which is located at 1535 meters and occurs in the lower montane wet forest life zone. I. walleriana flower throughout the year in this region. Color and Butterfly Foraging Behavior

PAGE 5

5 Salmon, peach, and pink color morphs were transplanted into black plastic pots from the roadside near the Friends School and the Colina Lodge in Monteverde. They were watered and allowed to recover for four days. Each pot contained approximately the same number of flowers. Potted salmon, red, magenta, purple, pink, and white color morphs of a cultivated Impatiens were purchased from a nursery. Each morph had approximately the same number of flowers. All of these potted color morphs were dispersed in the enclosed, environmentally controlled Garden 2 of the Monteverde Butterfly Garden. The cultivated Impatiens did not produce ne ctar. Using a syringe, 19.5% sucrose solution was added to the spur of each cultivated flower at the beginning and end of each observation period to ensure that the butterflies would be rewarded for attempting to remove nectar from the flowers. Since Gar den 2 contained many magenta color morph plants, a section with approximately the same number of flowers as the wild I. walleriana pots contained was observed. The other flowers that were actually observed for nectar removal attempts in the garden were th e other potted cultivated and wild Impatiens Six two hour periods of observation occurred over five days. Color preferences were determined on the basis of the number of nectar removal attempts to each color morph. A butterfly had to land of a flower f or at least five seconds for it to be considered a nectar removal attempt. Nectar Removal, Nectar Quantity, Pollen Availability, Flower Longevity, & Seed Set Within each flower patch, twenty buds of each color morph were bagged using small mesh bags to e nsure that no nectar was removed before the flowers were fully open. The bags were approximately 30.5 X 30.5 centimeters and had only one opening. Each bud was labeled with flagging tape that included the color, a unique number, and the date it was bagge d. The flagging tape was also used to secure the bags. Nectar is contained in a spur that is actually a modified calyx lobe (Elias and Gelb and 1997). Two days after opening, the flowers were uncovered and the height of the spur and of the nectar presen t were measured to the nearest centimeter using a caliper. This method was chosen because of apparent similarities in spur size and shape among color morphs. It is possible to see the height of nectar by backlighting by shining a light backwards through the spur to illuminate the nectar level (Inouye and Kearns 1993). The height at this initial measurement was considered the maximum nectar produced (nectar quantity) by the flower. The nectar level was measured everyday, for up to ten days, until the day of flower senescence. A reduction in the height of nectar within the spur was considered an occurrence of nectar removal. A flower was considered dead when all the petals had fallen off. This data was recorded along with a record of the length of polle n availability and whether seed set occurred. The length of pollen availability was considered the number of days that pollen was present on the flower. When pollen fell off the flower, an immature seedpod often remained. If the seedpod was present at t he time of flower senescence, seed set was considered to have occurred. Nectar Concentration A second set of twenty flowers was bagged for each wild color morph. Two days after opening, the spur of each flower was removed from the flower. The nectar was squeezed onto a refractometer and covered with a cover slip. The sucrose concentration was read from the refractometer.

PAGE 6

6 Altitudinal Gradient Twenty magenta flowers were bagged in patches containing at least two color morphs at altitudes of 1105 m, 1305 m, and 1535 m. These sites refer to the patches in San Luis, La Trocha, and Monteverde, respectively. The magenta morph was the focus organism in this test because it has been observed to be the most abundant color morph in the Monteverde region. The same data collected for the magenta color morphs by the cabins in San Luis was used in this test. Data for nectar removal, spur height, nectar height, length of pollen availability, day of senescence, and whether seed set occurred was collected. This experiment was conducted using the same RESULTS Color and Butterfly Foraging Behavior In Garden 2 of the Monteverde Butterfly Garden, both cultivated Impatiens and I. walleriana (wild) color morphs were observed for color preference and butterfly foraging behavior. Among the cultivated flowers, red, magenta, purple, pink, white, and salmon color morphs were observed for nectar removal. In total nectar removal occurred 428 times to the six color morphs by four species of Heliconiinae butterflies. Cultivated Impatiens were visited by Eueides isabella Dryas iulia Heliconius hecale Heliconius charit onius between each color morph (FIGURE 1). Pink and white were the most frequently visited cultivated morphs with 139 and 127 total visits, respectively. In this experiment, as well as in all the other experiments, it was found that color remained unchanged throughout the life of the flower. Species specific color preferences to the cultivated Impatiens in Garden 2 was also tested. All four of the butterfly species demonstr ated significant color preferences between the cultivated color morphs; also, different butterfly species preferred different color morphs. Isabellas visited 325 cultivated flowers and preferred the colors pink, white, and salmon (FIGURE 2). Spotted long wings visited 31 cultivated flowers and preferred the colors pink and magenta (FIGURE 2). Julias visited 59 cultivated Impatiens and preferred white and salmon colored flowers (FIGURE 2). Finally, Zebras visited 18 cultivated flowers. Pink and salmon we re their color preferences (FIGURE 2). I. walleriana were also observed for nectar removal in Garden 2 of the Monteverde Butterfly Garden. Magenta, salmon, pink, and peach color morphs were used in this experiment. In total nectar removal occurred 220 t imes to the four color morphs by four species of Heliconiinae butterflies. Wild I. walleriana were visited by Eueides isabella Dryas iulia Heliconius hecale Heliconius charitonius nt difference was found in the number of visits between each color morph (FIGURE 1). Salmon was the least favored color with only 30 visits.

PAGE 7

7 Species specific color preferences to different I. walleriana color morphs was also tested for in Garden 2. T wo of the four butterfly species demonstrated significant color preferences between the I. walleriana color morphs, these species preferred different color morphs. Isabellas visited 186 flowers and disliked the color same salmon (FIGURE 3). Spotted longw ings visited 15 cultivated flowers and preferred the colors salmon and peach (FIGURE 3). Julias visited 8 I. walleriana flowers and Zebra longwings visited 2 flowers, these species, however, did not show significant color preferences. FIGURE 1. The r elationship between color morphs of cultivated Impatiens and I walleriana and total number of nectar removal events to each color morph. Butterflies were observed for twelve hours in Garden 2 of the Monteverde Butterfly Garden. A nectar removal attempt o ccurred when a butterfly stayed on a flower for at least five seconds. Flowers were visited by Eueides isabella, Dryas iulia, Heliconius hecale, and Heliconius charitonius. Butterflies preferred pink and white cultivated flowers (X 2 = 202.581, df = 5). A mong I. walleriana butterflies preferred magenta, pink, and peach colors (X 2 = 18.583 df = 3).

PAGE 8

8 FIGURE 2. Cultivated Impatiens color morphs were observed for 12 hours in Garden 2 of the Monteverde Butterfly Garden. Observations were made of Eueides is abella, Dryas iulia, Heliconius hecale, and Heliconius charitonius butterflies to see if individual species showed color preference to certain morphs. A nectar removal attempt occurred when butterflies stayed on a flower for at least five seconds. Eueide s isabella preferred the colors pink, white, and salmon (X 2 = 158.377, df = 5). Dryas iulia preferred white and salmon colored flowers (X 2 = 133.31 df = 5). Heliconius hecale preferred the colors pink and magenta (X 2 = 17.193, df = 5). Pink and salmon were the color preferences of Heliconius charitonius (X 2 = 20.665 df = 5).

PAGE 9

9 FIGURE 3. Impatiens walleriana color morphs were observed for 12 hours in Garden 2 of the Monteverde Butterfly Garden. Observations were made of Eueides isabella, Dryas iulia, He liconius hecale, and Heliconius charitonius butterflies to see if individual species showed color preference to certain morphs. A nectar removal attempt occurred when butterflies landed on individual flowers for at least five seconds. Eueides isabella dis liked the color salmon (X 2 = 19.162, df = 3). Heliconius hecale preferred the colors salmon and peach (X 2 = 10.334, df = 3). Dryas iulia and Heliconius charitonius butterflies did not demonstrate significant color preferences. Nectar Removal, Nectar Qu antity, Pollen Availability, Flower Longevity, & Seed Set In the field, magenta, salmon, pink, and white color morphs of I. walleriana were studied. When nectar removal was observed in the field it was observed that nectar was robbed from many flowers. Holes in the side of the spur were evidence that nectar robbery had occurred. There was no significant difference in the occurrence of robbery between the color magenta, salmon, pink, and white color morphs (X 2 = 3.47 df = 3). Nectar quantity measurement s were collected from bagged flowers. The average amount of nectar in magenta, salmon, pink, and white flowers was 2.363, 2.238, 2.038, and 2.186 centimeters, respectively. A One Way ANOVA test of found no significant difference in nectar quantity betwee n color morphs (FIGURE 4). In terms of the length of time that pollen was available on the flower, magenta morphs had pollen for 2.47 days, salmon morphs averaged 2.25 days, pink morphs averaged 2.62 days, and white morphs averaged 3.47 days. However, a One Way ANOVA test found no significant difference between the length of pollen availability between color morphs (FIGURE 5). Salmon flowers, interestingly, held pollen for a significantly less amount of time than white flowers. Salmon had pollen for an average of only 2.250 days while white flowers had pollen for an average of 2.950 days (FIGURE 5). Flower longevity was also measured for each color morph. A One Way ANOVA found that there were significant differences in flower longevity between color mo rphs (FIGURE 6). Magenta and salmon colored flowers tend to live 1 3 days less than pink and white

PAGE 10

10 less than white flowers (FIGURE 6). Similarly, magenta and s almon flowers lived significantly shorter lives than pink flowers (FIGURE 6). The likelihood of flowers to set seed was also measured between the four color morphs. No significant difference was found between the number of flowers that set seed between c olor morphs (FIGURE 7). Magenta and salmon morphs set seed 90% of the time, pink set seed 86% of the time, and white set seed 70% of the time. FIGURE 4. The relationship between the amounts of nectar produced by magenta, salmo n, pink, and white color morphs of I. walleriana I. walleriana in San Luis, Costa Rica were bagged to prevent nectar removal before the flowers were fully opened. Two days after opening, the nectar height was measured using a caliper by backlighting thr ough the spur to see the nectar level. A One way ANOVA showed that there was no significant difference in nectar quantity between the color morphs (F = 1.327; df = 3; P = .2679).

PAGE 11

11 FIGURE 5. The amount of time that polle n was produced/presented by magenta, salmon, pink, and white color morphs of I. walleriana was tested in San Luis, Costa Rica. Twenty buds of each flower morph were bagged to prevent nectar removal before the flowers were fully opened. Two days after ope ning, the bags were removed. Flowers were checked daily to see if pollen was available. A One Way ANOVA showed that overall there was no significant difference in the number of days that flowers had pollen available between the color morphs (F = 1.887; d f = 3; P = .1388). There was however, a significant difference in the length of time that salmon and FIGURE 6. Day of flower senescence was observed in magenta, salmon, pink, and white color morphs of I. walleriana in San Luis, Costa Rica. Twenty buds of each morph were bagged to prevent nectar removal before the flowers were fully open. Two days after the flowers opened, the bags were removed. Flowers were checked dail y to see if the flowers were still alive, flowers were considered dead when all petals had fallen off. A One Way ANOVA showed that overall there was a significant difference in flower longevity between color morphs (F = 15.121; df = 3; P < .0001). Furthe r, pink flowers lived significantly longer than magenta and salmon flowers

PAGE 12

12 FIGURE 7. The proportion of plants that set seed among different color morphs of Impatiens walleriana in San Luis, Costa Rica. Twenty buds of each morph were bagged to prevent nectar removal before the flowers were fully open. Two d ays after the flowers opened, the bags were removed. Flowers were checked daily to see if the flowers were still alive, flowers were considered dead when all petals had fallen off. Flowers where an immature seedpod remained after the petals had fallen of f were considered to have set seed. No significant difference was found between the number of flowers that set seed between color morphs (X 2 = .6461; df = 3). Magenta and salmon morphs set seed 90% of the time, pink set seed 86% of the time, and white se t seed 70% of the time. Nectar Concentration The average nectar concentration in magenta, salmon, pink, and white flower morphs differed significantly. (FIGURE 8, One Way ANOVA, F = 3.405; df = 3; P = .0224). Magenta and salmon color morphs had higher sugar concentrations than pink and white flowers. Pink flowers had significantly lower sugar concentrations than magenta and salmon flowers (FIGURE 8). White flowers also showed significantly lower sugar concentrations than magenta and salmon flowers (F IGURE 8).

PAGE 13

13 FIGURE 8. Sucrose concentration of nectar was collected from magenta, salmon, pink, and white color morphs of I. walleriana in San Luis, Costa Rica. Twenty buds were bagged until flowers had been open for two da ys. Spurs were then collected and sucrose concentration was measured from nectar samples. Magenta and salmon flowers had higher sucrose concentrations. = .0155 Altitudinal Gradients Magenta color morphs were observed in San Luis, on La Trocha, and at the Monteverde Biological Station. Significant differences were found, using a Kruskal Wallis test, in the amount of nectar removal between the three sites (FIGURE 9, P < .0001). The majority of the flowers in San Luis were affected by nectary robbery while no instance s of robbery were found on either La Trocha or in Monteverde. In the absence of nectar robbers, flowers on La Trocha found no significantly differences between th e amount of nectar removal between Monteverde and San Luis or between Monteverde and La Trocha. Nectar quantity was also compared between the three altitudinal sites. Overall, the three sites did not exhibit significant differences in nectar quantities ( One Way ANOVA; F = 2.553; df = 2; P = .0863), but flowers on La Trocha had an average nectar height of 2.112 cm and produced significantly less nectar than flowers in of length of pollen availability, overall, there was a significant difference in the time that each site lost their pollen, but they all lost their pollen within two and a half to three days. (One Way ANOVA; F = 3.196; df = 2; P = .0480). Pollen was lost in San Luis around the 2.47 th day and on La Trocha around the 3.05 th .0147). The average day of senescence was also recorded. All flowers lived between four and a half and five days (One Way ANOVA; F = 1.028; df = 2; P = .3640). In San Luis, they lived 4.65 days, on La Trocha they lived 5.05 days, and in Monteverde, flowers lived an average of 4.55 days. Likewise, almost all flowers, in all three locations, set seed (FIGURE 10). While, there was no significant difference in the seed set (X 2 = .533, df = 2), it is noticeable that almost all of the plants set seed. In San Luis 90%, on La Trocha 95%, and in Monteverde, 100% of the flowers set seed (FIGURE 10).

PAGE 14

14 FIGURE 9. Nectar removal was compared along an altitudinal gradient for magenta color morphs of I. walleriana in the Monteverde region. Twenty buds at each site were bagged until the flowers had been open for two days, then nectar height was measured da ily using a caliper and backlighting. A reduction in the nectar level meant that pollen had been removed by a potential pollinator. A Kruskal Wallis test showed that there was a significant difference in the amount of nectar being removed between the thr ee sites (P < .0001). Flowers on La Trocha were .0147). FIGURE 10. The proportion of magenta flowers that set seed was compared between three altitudinal sites in the Monteverde region. Twenty flowers at each site were bagged and allowed to flower for two days before the bags were removed to prevent nectar removal before the flowers were fully open. Flowers were observed, and the day of senescence was recorded. Flowers that died, dropped all of their petals, without dropping their immature seedpod were considered to have set seed. There was no significant difference in the number of flowers that set seed between San Luis, La Tro cha, and Monteverde (X 2 = .533 df = 2). Flowers set seed 90%, 95%, and 100% of the time in San Luis, La Trocha, and Monteverde.

PAGE 15

15 DISCUSSION It was found that functional differences do in fact occur between different color morphs, however these differenc es do not separate the color morphs in terms of fitness. Color (through nectar removal in the Butterfly Garden), nectar concentration, and flower longevity all differed significantly between different color morphs of I. walleriana. Four species of butter flies were also found to have color preferences between color morphs of cultivated Impatiens and two species of butterflies showed color preferences among I. walleriana. In terms of an altitudinal gradient, nectar removal rates differed significantly betw een low, middle, and high elevations in the Monteverde region. Also, although there was no significant difference between the percent of seed set that occurred between the three altitudinal sites, almost all flowers set seed. Even the magenta flowers tha t were affected by nectar robbery set seed 90% of the time. Species specific color preference was demonstrated among cultivated Impatiens. First, this demonstrates that multiple color morphs may exist because there are enough different pollinators, who like different colors, to ensure that all the color morphs will be pollinated. Second, this test controlled for nectar quantity and concentration, and, thus, even though butterflies are known to like lots of sucrose rich nectar, they may choose to visit flowers that are a certain color regardless of their nutritive value. This may help explain why, when nectar removal was tested among I. walleriana in the Butterfly Garden, Eueides isabella disliked salmon color flowers, flowers that were shown to have hi gh nectar concentrations. Nectar quantity was not shown to play a major role in pollinator attraction when tested among the four I. walleriana color morphs, in the future the volume of the nectar within the spur should actually be measured in order to str ength the argument that nectar quantity does not indirectly affect fitness. In fact, data suggest that plants do not really need nectar in order to have reproductive success. Even when nectary robbery occurred, 90% of the magenta flowers in San Luis stil l set seed. It seems that flower longevity and the length of pollen availability may be more important than color in increasing fitness, but even this importance is somewhat negated by the fact that regardless of how long flowers were available for polli nation, seed set still tended to occur. In San Luis, there was a significant difference between flower longevity between colors, but there was no significant difference in the length of pollen availability. Along the altitudinal gradient, there was no si gnificant difference in flower longevity, but there was a significant difference in length of pollen availability. These results suggest that having one of these qualities does in fact increase fitness by increasing the length of time that pollination can occur. However, it does not explain how even though salmon and magenta I. walleriana flowers died significantly earlier than pink and white flowers they still set seed 90% of the time. Since, differences in color (nectar removal), nectar concentration flower longevity, length of pollen availability, and nectar robbery cannot explain the high occurrence of seed set in all flowers it can be concluded that magenta, salmon, pink, and white color morphs of I. walleriana are functionally similar in terms of fitness. Even though some color preference was demonstrated, enough pollinators with different preferences to ensure that all of the color morphs are still being pollinated must exist because the majority of flowers were still pollinated and able to set seed. Also, seed set occurred in flowers where nectar robbery occurred negating the significance that some flowers had stronger sucrose concentrations than others. Further, longevity of flowers and pollen seems to increase the chance that pollination woul d occur to flowers that lacked other advantageous functions, but failed to explain why seed set also

PAGE 16

16 occurred to flowers with short life spans. Therefore, functional differences between color morphs do not explain fitness differences between color morphs. Instead, it seems that in spite of functional differences, different color morphs possess similar fitness. This could have potentially negative consequences on the entire population of I. walleriana if pollinator limitation begins to occur. Parker (200 5) found that when pollinator limitation occurs seed set is negatively affected. In terms of this study, this means that all flowers would be equally affected by the pollinator limitation because they are all being pollinated approximately equally. Theref ore, the collective population of all the color morphs would all be in danger if pollinator limitation occurred. If the color morphs did not possess equal fitness, however, than the color morph that was best at attracting pollinators would be least affect ed by pollinator limitation and produce the most seedpods. Future studies should investigate the effects of pollinator limitation on color morph fitness. Perhaps, in the presence of pollinator limitation, functional differences would become important in determining reproductive success between color morphs. ACKNOWLEDGEMENTS Special thanks to Fabricio at the San Luis EcoLodge and Research Station and to Jim Wolf at the Monteverde Butterfly Garden for the gracious use of their facilities. Thanks to my a dvisor, Alan Masters and to Karen Masters for their expertise. Thank you to Javier Mndez for his translating expertise and thanks to Oliver Hymen and Matt Gasner for their inexhaustible patience. LITERATURE CITED Barth, Friedrich G. 1991. Insects and flowers the biology of a partnership. Princeton University Press, Princeton, New Jersey. Elias, Thomas S. and Hellan Gelband. 1977. Morphology, anatomy, and relationship of extrafloral nectaries and hydathodes in two species of Impatiens (B alsaminaceae). Botanical Gazette 138: 206 212. Galen, Candace and Peter G. Kevan. 1980. Scent and color floral polymorphism and pollination biology in Polemonium viscosum Nutt. American Midland Naturalist, The University of Notre Dame. Ha ber, William A. 2000. Plants and vegetation. In Nadkarni, Nalinin M. and Nathaniel T. Wheelwright (Eds.). Monteverde ecology and conservation of a tropical could forest, pp. 39 94. Oxford University Press, Inc., New York. Hannan, Gary L. 1981. Flower color polymorphism and pollination Biology of Playtstemon californicus Benth. (Papaveraceae). American Journal of Botany, Botanical Society of America. Heywood, V. ed. 1996. Flowering plants of the world. Mayflower Books, Inc., Ne w York, New York. Howe, Henry F. and Lynn C. Westley. 1988. Ecological relationships of plants and animals. Oxford University Press, Inc., New York. Inouye, David William and Carol Ann Kearns. 1993. Techniques for pollination biologists. Univ ersity Press of Colorado, Colorado. Masters, Alan. 2005. Tropical community ecology reader. CIEE, Spring. Murcia, Carolina, Claudia Sobrevila, and Lorne M. Wolfe. 1989. Flower polymorphism in the beach plant,

PAGE 17

17 Ipomoea imperti (Vahl.) Gris ebach (Convulvulaceae). Biotropica 21: 1, 84 88. Parker, Alison. 2005. Pollinator limitation and seed set in Impatiens walleriana CIEE, Spring. Rust, Richard W. 1977. Pollination in Impatiens capensis and Impatiens pallida (Balsaminaceae). Bullet in of the Torrey Botanical Club 104: 4, 361 367. Sellers, Susan E. and Lorne M. Wolfe. 1997 Polymorphic floral traits in Linaria Canadensis (Scrophulariaceae). American Midland Naturalist, The University of Notre Dame. Stanton, Maureen L. 1987. Reproductive biology of petal color variants in wild populations of Rhapanus sativus : I. pollinator response to color morphs. Journal of Botany, Botanical Society of America. Swihart, C.A. and S.L. Swihart. 1970. Colour selection and lear ned feeding preferences in the butterfly, Heliconius charitonius Linn. Animal Behavior 18: 60 64. Waser N. M., 1983. Competition for pollination and floral character differences among sympatric plant species; a review of evidence. In C.E. J ones and R.J. Little (Eds.). Handbook of experimental pollination biology. Van Nostrand Reinhold, New York. In John L. Randall and Khidir W. Hilu. 1990. Interference through improper pollen transfer in mixed stands of Impatiens capensis an d I. pallida (Balsaminaceae). American Journal of Botany 77: 7, 939 944.


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 M39-00343
040
FHM
0 041
eng
049
FHmm
1 100
Morris, Anne, B.
242
Las diferencias funcionales entre los morfos de color de Impatiens walleriana (Balsaminaceae)
245
Functional differences among color morphs of Impatiens walleriana (Balsaminaceae)
260
c 2005-05
500
Born Digital
3 520
Functional differences between color morphs of Impatiens walleriana were studied in Monteverde, Costa Rica to see if they influenced the fitness of the color morphs. Nectar removal rates, nectar quantity, nectar quality, length of pollen availability, flower longevity, and seed set were compared between magenta, salmon, pink, and white color morphs. Red, purple, magenta, pink, salmon, and white cultivated Impatiens and magenta, salmon, pink, and peach I. walleriana were observed in the Monteverde Butterfly Garden to see if color preferences occur among butterflies. In addition, magenta color morphs were studied along an altitudinal gradient to see if altitude caused functional differences within a single color morph and to see if instances of nectar robbery varied on an altitudinal scale. Among I. walleriana color morphs nectar removal rates were found to be affected by nectar robbery, all morphs suffered equally from nectar robbery. Color morphs also produced similar quantities of nectar. Magenta and salmon morphs produced nectar with higher sucrose concentrations than pink and white morphs. Color morphs did not exhibit differences in the length of pollen availability, but pink and white flowers were found to live longer than magenta and salmon flowers. In terms of seed set, all flowers were found to have high occurrences of seed set despite differences in nectar removal, robbery, and quality and in flower longevity among morphs. When cultivated Impatiens were observed for removal rates, pink and white had the most nectar removed. Heliconiinae butterflies showed species specific color preferences to the cultivated Impatiens. Eueides isabella preferred pink, white and
salmon colors. Heliconius hecale preferred pink and magenta flowers. Dryas iulia preferred white and salmon colors and Heliconius charitonius preferred the colors pink and salmon. When I. walleriana color morphs were observed in the butterfly garden, salmon was the morph with the least nectar removed. Two Heliconiinae butterflies showed specific color preferences to I. walleriana. Eueides isabella disliked the color salmon. Heliconius hecale preferred salmon and peach. Although, Dryas iulia and Heliconius charitonius also visited I. walleriana they did not demonstrate specific color preferences. In addition, magenta color morphs were observed in San Luis (1105 m), on la Trocha (1305 m), and in Montverde (1535 m). A significant difference was found between nectar removal in San Luis and la Trocha; this may have been due to nectar robbery, which only occurred in San Luis. Differences in pollen availability were also found between the three sites. Along the gradient, almost all flowers set seed.
Differences in color (nectar removal), nectar concentration, flower longevity, length of pollen availability, and nectar robbery cannot explain the high occurrence of seed set in all flowers; therefore, I. walleriana color morphs have similar fitness. It is likely that most flowers are being pollinated, and thus set seed, due to the presence of many different pollinators whose preferences encompass all the color morphs.
Las diferencias funcionales entre las formas de color de Impatiens walleriana fueron estudiadas en Monteverde, Costa Rica, para determinar si ellos influan en las adaptaciones de las formas de color. Las tasas de eliminacin, la cantidad y calidad del nctar, el tiempo de disponibilidad del polen, la longevidad de la flor y la produccin de las semillas se compararon entre las formas de color magenta, salmn, rosado, y blanco. Las formas cultivadas de Impatiens de colores rojo, prpura, magenta, rosado, salmn y blanco, y las formas silvestres de I. walleriana de colores magenta, salmn, rosado y melocotn se observaron en el Jardn de Mariposas de Monteverde para ver si las mariposas preferan alguna variacin de color. Adems, la forma de color magenta fue estudiada en un gradiente altitudinal para determinar si la altitud caus las diferencias funcionales dentro de un solo color y para ver si los casos de robo de nctar variaban con la elevacin. Se encontr que entre las formas de color magenta, salmn, rosado y blanco, las tasas de la eliminacin del nctar fueron afectadas por el robo de nctar; pero no se encontr ninguna diferencia significativa en el robo de nctar entre las formas de color. La cantidad de nctar tampoco vari significativamente entre las formas de color. Las formas de color magenta y salmn produjeron nctar con concentraciones de sacarosa mayores que las formas rosadas y blancas. Las formas de color no exhibieron ninguna diferencia en el tiempo de disponibilidad de polen; pero las formas rosadas y blancas presentaron una longevidad mayor que las formas magenta y salmn. Todas las flores presentaron una produccin de semillas grande, a pesar de diferencias en la eliminacin, robo y calidad del nctar, y en la longevidad de la flor entre las diferentes formas. Las formas blancas y rosadas cultivadas mostraron las tasas ms altas de eliminacin de nctar. Las mariposas de la subfamilia Heliconiinae mostraron preferencias especficas de color en las Impatiens cultivadas. Eueides isabella prefiri las formas rosadas, blancas y salmn. Heliconius hecale prefiri las flores rosadas y magenta. Dryas iulia prefiri los colores blancos y salmon, y Heliconius charitonius prefiri las rosadas y salmn. El color salmn present la tasa ms baja de eliminacin de nctar en el jardn. Dos mariposas de la subfamilia Heliconiinae mostraron preferencias especficas de color por I. walleriana. Eueides isabella present aversin al color salmn. Heliconius hecale prefiri el salmn y el melocotn. Aunque Dryas iulia y Heliconius charitonius tambin visitaron I. walleriana, no mostraron preferencias especficas de color. La forma de color magenta fue observada en San Luis (1105 m), en la Trocha (1305 m) y en Monteverde (1535 m); se encontraron diferencias significativas entre las flores de San Luis y las de la Trocha. Esta diferencia puede haber sido debida a que el robo de nctar slo ocurri en San Luis; tambin se encontraron diferencias en la disponibilidad de polen entre los tres sitios. Casi todas las flores produjeron semillas en las tres localidades. Ya que las diferencias en color (y, por ende, la eliminacin del nctar), la concentracin de nctar, la longevidad de la flor, el tiempo de disponibilidad de polen, y el robo de nctar no pueden explicar la produccin alta de semillas en todas las variaciones de color, se puede concluir que las formas de color magenta, salmn, rosado y blanco de I. walleriana son funcionalmente semejantes en trminos de adaptacin. Aunque se demostr cierta preferencia por color, deben existir preferencias diferentes en los polinizadores para asegurar que todas las formas de color sean polinizadas, ya que la mayora de las flores fue polinizada y produjo semillas.
546
Text in English.
650
Flowers--Color
Pollination by animals
Costa Rica--Puntarenas--Monteverde Zone
4
Flores--Colores
Polinizado por animales
Costa Rica--Puntarenas--Zona de Monteverde
653
Tropical Ecology Spring 2005
Ecologa Tropical Primavera 2005
655
Reports
720
CIEE
773
t Monteverde Institute : Tropical Ecology
856
u http://digital.lib.usf.edu/?m39.343