|USFLDC Home | Tropical Ecology Collection [Monteverde Institute]||| RSS|
This item is only available as the following downloads:
Spatial d istribution of n ectar in Lantana camara and v isitation by Heliconius b utterflies Kaitland Harvey Department of Forestry, University of Kentucky ABSTRACT The quality, quantity, timing and placement of nectar can impact pollinator behavior. Here, I investigate the impact of the dis tribution of nectar in a patch: random, regular and clumped treatment A 5 x 5 spatial grid was setup using wooden stands and Lantana camara inflorescences in the Monteverde butterfly garden. Nectar was added to 6 flowers in 8 of 25 stands, resulting in 8 stands with and 17 without nectar. Butterflies visited nectar and non nectar stands equally in all three patterns. However, when comparing no n nectar stands across all treatments, t he regular treatment decreased total butterfly visits to non nectar stands ( 151 visits) and the random treatment increased total butterfly visits to non nectar stands (215 visits) (chi square=11.24, df=5 p=0.0468 ). Also, n on nectar stands had a higher average number of flowers visited per stand ( 5.95 0.306 ) than those with nectar ( 5.08 0.361 visits ) Moreover, the rank of visit length (1 5) was longer on necta r stands (1.92 0.091) than non nectar stands (1.62 0.083). Distance moved within the grid (regular 1.5 0.092, clumped 1.6 0.086, random 1.5 0.078) and off the grid was not impacted by pattern type. In brief, Heliconius butterflies do not demonstr ate strong preferences between nectar and non nectar flowers in a closely spaced, high density pattern. The distribution of nectar in small patches has little impact on pollinator behavior, perhaps because inflorescences are so closely spaces that there is no benefit to differentiate between them. RESUMEN La calida, cantidad, tiempo y ubicacin puede influir el comportamiento de los polinizadores. Aqu, investigu el impacto de diferentes parches de dispersin (azaroso, regular y agregado) de nctar. Una cuadrcula espacial de 5 x 5 se ubico usando estacas de madera e inflorescencias de Lantana camara en el Jardn de Mariposas de Monteverde. Se aadi nctar a 6 flores en 8 de 25 tratamientos, resultando en estacas 8 con y 17 sin nctar. Las maripos as visitaron flores con y sin nctar igualmente en los tres patrones. Sin embargo, la dispersin regular disminuye las visitas en las estacas sin nctar (151 visitas) y las estacas sin nctar fueron ms visitadas de lo esperado (215 visitas) en el patrn de dispersin azarosa (chi cuadrado = 11.24, gl = 5, p = 0.0468). Los tratamientos sin nctar tienen un mayor nmero de visitas por flor que los que poseen nctar. Adems, las flores con nctar reciben en promedio 5.08 0.361 visitas por flor y las flore s sin nctar reciben en promedio 5.95 0.306 visitas. Adems, el rango de visitas (1 5) fue mayor en tratamientos con nctar (1.92 0.091) que sin nctar (1.62 0.083). La distancia recorrida dentro de la cuadrcula (regular 1.5 0.092, agregado 1.6 0.086, azaroso 1.5 0.078) y fuera de la misma no tuvo impacto por el patrn. En resumen,las mariposas del gnero Heliconius no muestran preferencias fuertes entre flores con o sin nctar en un espacio cerrado y muy d enso. La distribucin de nctar en parches pequeos tiene poco impacto en el comportamiento de los polinizadores, sin embargo debido a que las inflorescencias estn tan poco espaciadas no hay beneficio en diferenciarlas entre ellas. INTRODUCTIO N Nectar i s a common reward for pollination of flowering plants. P lant s may manipulate pollinator behavior by altering the quality, quantity, timing and placement of nectar in flowers ( Real & Rathcke 1988; Klinkhamer et al. 2001) In this push and pull mutualism, a plant does not want a pollinator to become to o full too fast before enough flowers have been visited to ensure
pollination ( Kevan & Baker 1983 ). And o n the other hand, the plant does not want the pollinator to lose interest in a patch of flowers with inadequate rewards (Kevan & Baker 1983). There are many ways plants try to maintain this balance P lants with greater nectar production have been shown to have higher approach rates by bumble bees where more flowers visited in sequence on the same inflore scence, and longer time spent at a flower ( Klinkhamer et al 200 1). Quantity of decision to visit an additional flower depends on amount of reward from the previous f lower (Pleasants and Zimmerman 1979). Furthermore, t he spatial distribution of nectar can also be an important plant trait that alters pollinator visitation rates, pollinator movements, and as a result the overall plant fitness (Real & Rathcke 1988). V ar iation of nectar can be viewed on different scales : with in an individual plant, a patch, or a cross a community. Likewise, t he way a pollinator forages might be affected by any of these scales For example, p atterns of nectar variation within small patch es are not differentiated by bees until each patch is a minimum of six meters apart (K linkhamer et al 2001) L arger patches of flowers have also been known to have higher nectar variation between flowers than small er more isolated patches because the cost of additional visits is low in high densities (Thakar et al. 2003). The patchiness or variation in nectar distribution can alter pollinator visitation. W ith a higher presence of nectarless flowers within a patch the overall total number of visits to the patch c an increase because the pollinator must visit more flowers to receive the same amount of nectar, ultimately promoting out crossing and plant fitness (Thakar et al. 2003). Alternatively, too many nectarless flow ers may cause the pollinator to l eave the patch too quickly. The spatial arrangement of nectar and pollinator response has been closely studied in bumblebees (Real & Rathcke 1988; Klinkhamer et al 2001), but remains understudied in other pollinator groups, like butterflies. Butterfly pollinated plants are also known to have high variation in nectar distribution (Anand et al 2007. For example, L. camara is a common butterfly pollinated plant in the tropics and is known to have a high variation of nectar within each patch. In one study, a sample of L. camara patches found that 44.14% of the flowers were nectarless (Anand et al 2007). This high percentage of nectarless flowers in one community can have significant effects on plant fitness, such as, higher visitation, enhanced cross pollination, and better contact to the flower (Thakar et al. 2003). Therefore, the impact of nectar distribution on pollinator res ponse needs further investigation, especially with plants with known nectar variation and their associated pollinators. To e xplore spatial arrangement of nectar further in an understudied group of pollinators, butterflies, I examine d the effects of diffe rent patterns of nectar distribution between inflorescences of a single patch and record ed butterfly visitation Evolutionarily speaking, nectar is a costly resource to produce (Pyke 1991), and if a plant cannot produce nectar in every flower could significantly increase pollinator visitation and plant fitness.
M ETHODS Study Organisms & Site Lantana camara is a naturalized, weedy butterfly pollinated plant found commonly along roadsides, pastures, and gardens in Costa Rica It is generally 1 2 m tall and found in an herbaceous or shrubby habit. It flowers all year round. Flowers are arranged in a compound inflorescence. They are yellow when they first open and gradually turn t o orange, and then red The inflorescence is referred to as a inflorescence with younger, yellow in the middle and older orangey attractiveness of the inflorescence and i ncreases the landing pad for the butterfly Each inflorescence last 2 3 days (Schemske 1983). Heliconius butterflies were selected as the focal pollinator for this study because they are very active visitors and known to be attracted to L. camara flowers ( Andersson and Dosbson 2003 ) T hey also have the ability to learn the locations of flowers ( Turner 1981 ) I n nature, Heliconius search for a few species of cucumber vines that they can knock pollen into the nectar and come back the next day to drin k this pollen rich nectar ( DeVries 1987 ) This unique diet requires them to be more sophisticated at spatial mapping because they must remember where these pollen rich nectar sources are in order to revisit them T his experiment was conducted in the M onteverde Butterfly Garden in Costa Rica All of the work for this project was completed in garden two, the Heliconius butterfly garden. This garden contains four species of Heliconius butterflies: erato charitonius sara, and Dryas julia However, the overwhelming majority of butterflies in garden were erato and charitonius only two Heliconius saras and one Dryas julia was observed during the study Experimental Design. A spatial grid was assembled using 25 wooden stands in a five by fi ve grid. The grid was 1m 30cm x 1m 50cm and the distance between each stand was 30 cm x 26cm (Figure 1). Small vases were attached to each stand with two inflorescences of L. camara Figure 1. Monteverde Buttefly Garden interior highlighting the dimensions of the spatial grid (1m 30cm x 1m 50cm). E ach stand was 26cm x 30cm apart with a vase of L camara inflorescences (n=2). Each inflorescence has a minimum of 3 yellow flowers with a baseline minimum of 6 yellow flowers per stand.
For each treatme nt nectar was artificiall y added to 8 of the 25 stands using a 20% sucrose solution and a small syringe. For each nectar stand (n=8), a drop of nectar w as added to th r e e yello w flowers on each inflorescence ( 2 inflorescences per sta nd ) For each treatment I distributed th e nectar stands in a clumped, dispersed, and random pattern (n=3 treatments Figure 2 ). A regular distribution was create d by placing the eight nectar stands evenly throughout the grid. A clumped distribution consisted of all eight nectar stands clustered in a contiguous grouping A r andom distribution was conducted by assigning all stands a number and by drawing numbers out of a bag to determine which the eight stands would have nectar For each treatment, fou r parameters were measured: number of butterfly visits per stand, number of flowers visited per stand, length of visit per stand, and distance moved from each stand after visitation Length of visit was recorded using a qualitative rank scale from one to f ive, where one was the shortest visit observed and five was longest visit observed. Each parameter was observed for at least ten minute s or until at least ten visits to nectar stands and ten visits to non nectar stands were recorded Data were collected from 10 November to 17 November 2010 during sunny mornings. To avoid any bias based on daily activity levels or feeding habits measurements were recorded at three different times of the day: early morning 800 to 930, mid morning 930 to 1100, and late morning 1100 to 1230. Each nectar distribution (n=5) w as run at least one time during all three different morning observation times. But each distribution also has to be missing one of these periods. RESULTS D istribution of stands with nectar and patterns of butterfl y visit ation Butterflies visited stands with and without nectar equally in each treatment. When comparing nectar stands across all treatments there was no significant difference in total butterfly visits across treatments (chi square=0.35, df=5, p= 0.9966 ). However, when comparing non nectar st ands across all treatments there was a significant difference in number of butterfly visits, in Figure 2. Patterns of nectar dispersion for Lantana camara stands in the Monteverde Butterfly Garden containing 2 inflorescences with a minimum of six yellow flowers. Each circle is a stand a nd solid circles are those which a 20% sucrose solution has been added to 6 of the flowers. Open circles have flowers with no appreciable nectar. Distances between circles was approximately 26cm x 30 cm.
which the regular pattern was visited less than expected (151visits) and the random pattern was visited more than expected (215 visits) and the clumped pattern was visited the same as the expected value (191 visits) (chi square=11.24, df=5, p=0.0468; figure 1). Nectar distribution changed the mean number of butterfly visits per station for one of the three treatment s The mean number of butterfly visit s to non nectar stands in a random pattern was significantly different when co mpared to all other treatments (F = 4.57, df =5, p = 0.04). In a random nectar distribution, non nectar stands received a higher number of mean butterfly visits ( 3.32 0.165 ) th an nectar stands within the same treatment ( 2.51 0.247) a s well as both other treatments ( F=4.57 d f=5 p=0.0005 ; Figure 4). How does the distribution of nectar change how Figure 3 Placement of nectar and non nectar stations of L. camara in the Monteverde Butterfly Garden did not impact total butterfly visits within a treatment per stand ( dispersed chi square=0.18, clumped chi square=0 .58, and random chi square=2.92) E ach stand is made up of one vase with two inflorescences and are in a 5 x5 spatial grid. Total butterfly visits on nectar stands (n=8) did not reveal a significant difference when compared across all treatments: regular (80 visits), clumped (86 visits), and random (87 visits) (chi square=0.35, df=5, p= 0.9966 ) But there was a significant difference for non nectar stands (n=17), when they were compared across all treatments, in which a regular pattern received less visits than expected (115 visits) and a random pattern received more vis its than expected (215 visi ts) (chi square=11.24, df=5, p=0.0468). These significant differences are denoted by a asterisk. Figure 4. The mean number of butterfly visits to non nectar stands (n=17) in a random distribution ( 3.32 0.165) was significantly different than nectar stands (n=8) within the distribution and both other nectar treatments, regular and clumped ( F=4.57, df=5, p=0.0005 ).
many flowers are visited per inflorescence, distance moved between stands, and length of visit? When comparing differences of mean number of flowers visited by treatment, only the regular treatment revealed a significant difference between nectar stands ( 5.14 0.669) and non nectar stands (6.903 0.517). T he overall impact of nectar pattern on mean number of flowers visited across all treatments was non significant (F= 1.87 df=2, p=0.1553 ; Figure 5 ). Mean number of flowers visited per stand is better explained by the presence or a bsence of nectar on an individual stand than treatment type ( F=2.14 d f=5 p=0.0605 ; Figure 5). Butterflies had a higher number of average flower s visited for non nectar stands ( 5.95 0.306) t han nectar stands (5.08 0.361). Mean length of visit was longer on nectar stands ( 2.32 0 .091 ) than on non nectar stands ( 1.63 0.083) based on a qualitative rank scale where one is the shortest visit and five is the longest visit ( F=7.58 df=5 p= < 0.0001 ). Within treatments, both regular and clumped have a significantly higher mean length of visit for nectar stands (regular 2 .43 0.166, clumped 2.07 0.155) than non nectar stands (regular 1.65 0.145, clumped 1.71 0.156) Figure 5. T he overall effect of nectar pattern on mean number of flowers visited across all treatments was non significant (F= 1.87, df=2, p=0.1553 ). However within treatments, the regular treatment received a significantly higher number of flowers visited per stand on non nectar stands ( 6.903 0.517 ) than on nectar stands ( 5.14 0.669 ). However, overall m ean number of flowers visited per stand is better explained by the presence or absence of nectar on an individual stand (F=2.14, df=5, p=0.0605 ). Mean number of flowers visited for non nectar stands was higher ( 5.95 0.306) than nectar stands (5.08 0.361).
Distance between mean number of stations a butterfly moved within the grid was not statistically significant for any treatment (F=0.193, df=5, p=0.965) Each treatment had a similar mean distance moved ( regular 1.46 0.092 stations, clumped 1.55 0.08 6 stations, and random 1.53 0.078 stations ) Furthermore, there was no significant difference in distance moved when butterflies moved from a nectar or a non nectar flowers stand ( F=0.193, df=5, p=0.7455 nectar 1.50 0.082 non nectar 1.52 0.055 ). Spatial distribution of nectar did not affect the foraging distance of the butterfly when moving from flower to flower. In addition t he likelihood that a butterfly would leave th e grid was not influenced by treatment (chi square=0.004, df =2, p= 0.8395). T hus, the number of butterflies leaving the grid did not significantly differ between nectar stands and non nectar stands more than expected by chance for any of the three treatme nts. DISCUSSION Butterflies visited stands with and without nectar equally for each treatment, except for non nectar stands in a random pattern (which they visit more) and in a regular pattern (which they visit less). Dispersion of nectar between stands impacted visitation by increasing visits to stands without nectar and increasing visiting time to stands with nectar but did not impact movement between stands. Overall, Heliconius butterflies did not distinguish between nectar and non nectar stands in a clo sely spaced, high density patter n. This could be because w hen a butterfly visits a high density patch, there is little cost to visit additional flowers because energy and travel distance is minimal. In nature, plants with a high den sity flowers, like L. camara, have a greater likelihood of et al. 2003). Nectarless flowers benefit from growing in a close community with other nectar rich flowers because the p lant is able minimize nectar p roduction and maximize visitation, because pollinator s will not discriminate if foraging cost is low. Therefore, from an evolutionarily perspective one could infer it might be better to Figure 6. Mean length of visit was longer on nectar stands (2.32 0.091) than on non nectar stands (1.63 0.083) based on a qualitat ive rank scale where 1 is the shortest visit and 5 is the longest visit.
grow in a gregarious patch with other plants to reduce total amount of nectar necessary to ensure pollination success. However, the success for this type of spatial arrangement depends a great deal on the composition of the rest of the community. For example, if there is a patch of flowers with high and low nectar production and a patch of flowers with only high production at a certain distance away pollinators might begin to discriminate between patches based on level of variation. Bumblebees have been found to discriminate between patches of flowers with varying necta r prod uction and placement starting at a distance of 6 meters (Klinkhamer et al .2001). throughout the community. These spatial patterns have been studied in bumblebees on a small scale, but it would be interesting to understand how nectar production and placement applies to other pollinators like butterflies, and on varying ecological scales. Acknowledgements I would like to thank Alan Masters first and foremost, for his excellent advising, life advice, and positivity! I also want to thank Alan and Moncho Caldern for helping me get this project in the ground, and many thanks to Anjali Kumar and Pablo Allen for supporting me and the butterflie s. And m ost importantly, to the people that made all of this possible at the THE BUTTERFLY GARDEN especially Seba s Allen, Quin, Julio, and Marta for the extra encouragement and support! Lastly, thanks to the weather gods for providing me with consecutive days of beautiful butterfly weather in the rainy season! LITERATURE CITED Anand, C., C. Umranikar, P. Shintre, A. Damle, J. Kale, J. Joshi, and M. Watve. 2007. Presence of two types of flowers with respect to nectar sugar in two gregariously floweri ng species. J. Biosci. 32: 769 774. Andersson, S, and H. E. M. Dobson. 2003. Behavioral foraging responses by the butterfly Heliconius melpomene to Lantana camara floral scent. J. Chem. Ecol. 29: 2303 2318. Barrows, E. M. 1976. Nectar Robbing and Pollination of Lantana camara (Verbenaceae). Biotropica. 8: 132 135. Devries P. J. 1985. The Butterflies of Costa Rica and their Natural History. Volume I: Papilionidae, Pieridae, Nymyphalidae Princeton University Press, Pri nceton, New Jersey. pp. 192 193. Fenster, C. B., W. S. Armbruster, P. Wilson, M. R. Dudash, and J. D. Thomson 2004. Pollination syndromes and floral specializations. Annu. Rev. Evol. Syst. 35: 375 403. Kevan, P. G., and H G. Baker 1983. Insects as f lower visitors and pollinators. Ann. Rev. Entomol. 28:407 453.
Klinkhamer, P. G., T. J. Jong, and L. A. Linnebank. 2001. Small scale spatial patterns determine ecological relationships: A n experimental examples using nectar production rates. Ecology Le tters. 4:559 576. Pleasants, J. M., and M. Zimmerman. 1978. Patchiness in the dispersion of nectar resources evidence for hot and cold spots. Oecologia. 41: 283 288. Pyke G. H. 1991. What does it cost a plant to produce floral nectar?. Nature. 350: 5 8 59. Real, L. A. & Rathcke, B. J 1988.Patterns of individual variability in floral resources. Ecology. 69:728 735. Thakar, J.D., K. Kunte, A. K. Chauhan, A. V. Watve, and M. G. Watve. 2003. Nectarless flowers: ecological correlates for evolutionary s tability. Oecologica. 136: 565 570. Turner, J. R. G. 1981. Adaptation and evolution in Heliconius: A Defense of NeoDarwinism. Annual Review of Ecology and Systematics. 12: 99 121. Schemske, D. W. 1983. Lantana Camara (Cinco Negritos, Lantana). In: Co sta Rican Natural History. D. H. Janzen, ed. The University of Chicago Press, Chicago, IL, pp. 266 268.
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 M39-00280
La distribucin espacial de nctar en Lantana camara y las visitas por las mariposas Heliconius
Spatial distribution of nectar in Lantana camara and visitation by Heliconius butterflies
The quality, quantity, timing, and placement of nectar can impact pollinator behavior. Here, I investigate the impact of the distribution of nectar in a patch: random, regular and clumped treatment. A 5 x 5 spatial grid was setup using wooden stands and Lantana camara inflorescences in the Monteverde butterfly garden. Nectar was added to 6 flowers in 8 of 25 stands, resulting in 8 stands with and 17 without nectar. Butterflies visited nectar and non-nectar stands equally in all three patterns. However, when comparing non-nectar stands across all treatments, the regular
treatment decreased total butterfly visits to non-nectar stands (151visits) and the random treatment increased total butterfly visits to non-nectar stands (215 visits) (chi-square=11.24, df=5, p=0.0468 ). Also, non-nectar stands had a higher average number of flowers visited per stand (5.95 0.306) than those with nectar (5.08 0.361 visits ). Moreover, the rank of visit length (1-5) was longer on nectar stands (1.92 0.091) than non-nectar stands (1.62
0.083). Distance moved within the grid (regular 1.5 0.092, clumped 1.6 0.086, random 1.5 0.078) and off the grid was not impacted by pattern type. In brief, Heliconius butterflies do not demonstrate strong preferences between nectar and non-nectar flowers in a closely spaced, high density pattern. The distribution of nectar in small patches has little impact on pollinator behavior, perhaps because inflorescences are so closely spaces that there is no benefit to differentiate between them.
La calidad, cantidad, tiempo y ubicacin pueden influir en el comportamiento de los polinizadores. Aqu, investigu el impacto de los diferentes parches de dispersin (azaroso, regular y agregado) de nctar. Una cuadrcula espacial de 5 x 5 se ubic usando estacas de madera e inflorescencias de Lantana camara en el Jardn de Mariposas de Monteverde. Se aadi nctar a 6 flores en 8 de 25 tratamientos, resultando en 8 estacas con nctar y 17 sin. Las mariposas visitaron las flores con y sin nctar en los tres patrones por igual. Sin embargo, la dispersin regular disminuye las visitas en las estacas sin nctar (151 visitas) y las estacas sin nctar fueron las ms visitadas de lo esperado (215 visitas) en el patrn de dispersin azarosa (chi cuadrado = 11.24, gl = 5, p = 0.0468). Los tratamientos sin nctar tienen un mayor nmero de visitas por flor que los que poseen nctar. Adems, las flores con nctar reciben en promedio 5.08 0.361 visitas por flor y las flores sin nctar reciben en promedio 5.95 0.306 visitas. Adems, el rango de visitas (1-5) fue mayor en los tratamientos con nctar (1.92 0.091) que sin nctar (1.62 0.083). La distancia recorrida dentro de la cuadrcula (regular 1.5 0.092, agregado 1.6 0.086, azaroso 1.5 0.078) y fuera de la misma no tuvo impacto por el patrn. En resumen, las mariposas del gnero Heliconius no muestran fuertes preferencias entre las flores con o sin nctar en un espacio cerrado y muy denso. La distribucin de nctar en los parches pequeos tiene poco impacto en el comportamiento de los polinizadores, tal vez porque las inflorescencias estn tan poco espaciadas que no hay ningn beneficio en diferenciarlas entre ellas.
Text in English.
Pollination by insects
Costa Rica--Puntarenas--Monteverde Zone
Polinizado por insectos
Costa Rica--Puntarenas--Zona de Monteverde
Tropical Ecology Fall 2010
Ecologa Tropical Otoo 2010
t Monteverde Institute : Tropical Ecology