HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Knudsen, J. T. Search for Related Content PubMed Articles by Knudsen, J. T. Agricola Articles by Knudsen, J. T.

Variation in floral scent composition within and between populations of Geonoma macrostachys (Arecaceae) in the western Amazon¹

Botanical Institute, Göteborg University, Box 461, SE 405 30 Göteborg, Sweden

Received for publication February 21, 2002. Accepted for publication May 16, 2002.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
In geonomoid palms floral scent is both an important pollinator attractant and an important factor in reproductive isolation. However, little is known about intraspecific variation in floral scent composition in these as well as in other plants. In this study the level of variation in floral scent composition found within and among five populations of Geonoma macrostachys var. macrostachys in the western Amazon is documented. Floral scent samples were collected using head-space adsorption and were analyzed by gas chromatography-mass spectrometry. Most of the 108 compounds recorded were of isoprenoid origin, but only 28 of the compounds were found in all 62 samples analyzed. No differentiation was found between the studied populations, confirming that G. macrostachys var. macrostachys is outbreeding and indicating that the individual populations are part of a metapopulation linked by sufficient gene flow to avoid local differentiation. However, a negative correlation between distance and similarity of floral scent chemistry indicates a case of clinal variation within the distribution area of G. macrostachys. Male euglossine bees are infrequent visitors to G. macrostachys, while other groups of insects are abundant. However, the level of variation and the chemical composition lend support to a suggested importance of male euglossine bees in long-distance pollen flow in G. macrostachys. Other insect groups are probably important in securing pollination of most flowers with pollen from nearby sources.

Key Words: Arecaceae • clinal variation • floral scent • Geonoma macrostachys var. macrostachys • metapopulation • pollination • speciation • variation

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Several studies indicate that floral scent may be of importance for reproductive isolation among closely related species (e.g., Groth, Bergström, and Pellmyr, 1987 ; Bergström et al., 1992 ; Borg-Karlson et al., 1993 ; Knudsen, 1994 , 1999a , b ; Tollsten, Knudsen, and Bergström, 1994 ; Azuma et al., 1997 ; Barkman, Beaman, and Gaga, 1997 ; Levin, Raguso, and McDade, 2001 ), but so far studies of intraspecific variation in floral scent composition involving more than a few individuals have been conducted only in species of Cycnoches, Epidendrum, Ophrys, Platanthera, Stanhopea (Orchidaceae: Gregg, 1983 ; Whitten and Williams, 1992 ; Moya and Ackerman, 1993 ; Tollsten and Bergström, 1993 ; Schiestl et al., 1997 ), Parkia (Mimosaceae: Pettersson and Knudsen, 2001 ), Conopodium (Apiaceae: Tollsten and Øvstedal, 1994 ), Arum (Araceae: Kite, 1995 ), and Magnolia (Magnoliaceae: Azuma, Toyota, and Asakawa, 2001 ).

The neotropical genus Geonoma consists of small understory palms. They are monoecious, and because staminate and pistillate flowering usually is separated in time, most species need a vector to transport pollen to pistillate flowers for successful reproduction to occur. Most species of Geonoma produce a distinct floral scent, which is regarded to function as a reproductive isolation mechanism among species of geonomoid palms (Knudsen, 1999a , b ). In Geonoma macrostachys var. macrostachys, a field experiment showed that inflorescence dummies "loaded" with floral scent attracted a large number of insect visitors (Knudsen, Andersson, and Bergman, 1999 ). The insects belonged to the same orders and families as those reported as natural flower visitors in earlier pollination studies of G. macrostachys in Ecuador and Peru (Olesen and Balslev, 1990 ; Listabarth, 1993 ), i.e., bees (Hymenoptera: Apidae, Trigonidae, Halictidae), beetles (Coloptera: Curculionidae, Chrysomelidae, Nitidulidae), and flies (Diptera: Drosophilidae, Syrphidae). It was concluded that floral scent was more important than color in attracting potential pollinators to the inflorescences of G. macrostachys and that the nonrewarding pistillate flowers probably were visited because of their similarity in floral scent composition to that of the rewarding staminate flowers. Only scent-collecting male euglossine bees are also rewarded at the pistillate flowers.

Studies of features important for reproductive isolation and their variability may supply us with important information on the level of variation to be expected for reproductive isolation to occur. Though it will never be possible to investigate all populations in a reproductive community, information relevant for reproductive isolation obtained from a number of populations may still give us insight into the evolutionary processes at the population level and the level of variation to be expected.

So far floral scent composition has only been determined in one or a few individuals of each species of Geonoma and hence the intraspecific variability in this trait is not known. Here I present a study of a wide sample of individuals and populations of Geonoma macrostachys Mart. var. macrostachys. Because of the occurrence of several local variants, Geonoma macrostachys var. macrostachys is a variable taxon. However, this study only includes floral scent from individuals of the typical variant of G. macrostachys var. macrostachys, which morphologically is rather uniform and geographically widespread on terra firme forest in the western Amazon. The aims were to obtain data on the variation in floral scent composition within and among populations and thereby obtain knowledge of the level of variation in a character of major importance for reproductive isolation in species of Geonoma.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Floral scent samples were collected from 42 staminate and 20 pistillate inflorescences in 54 individuals of Geonoma macrostachys var. macrostachys belonging to five different populations (Table 1). All the studied populations were located in Amazonian Ecuador: (1) Napo province, Estación Cientifica Yasuní on Río Tiputini 76°23' W, 00°40' S, altitude approximately 200 m ("Yasuní"); (2) Napo province, Tiputini Biodiversity Station on Río Tiputini South of Pañacocha 76°04' W, 00°38' S, altitude approximately 200 m ("Tiputini"); (3) Napo province, Jatún Sacha Biological Station, 8 km below Misahuallí, South of Rio Napo 77°30' W, 01°08' S, altitude approximately 500 m ("Jatún Sacha"); (4) Sucumbios province, 76°18' W, 00°01' N, altitude approximately 200 m ("Cuyabeno"); and (5) Morona-Santiago province, vicinity of Makuma village, 77°43' W, 02°10' S, altitude approximately 600 m ("Makuma").

Data analysis
Twenty-five chemical compounds were present in all floral scent samples collected (Table 1), and these were used to ordinate the samples in a principal component analysis (PCA) using SIMCA-P 8.0 (Umetrics: Eriksson et al., 1999 ). It is not known whether the individual compounds used in the ordination are attractive to insects. The original data were percentage values and before analysis all the values were arcsine square-root transformed to better approach normal distribution patterns. Prior to analysis all variables were scaled to unit variance, which makes all the variable axes have the same length and, thus, the same importance in the analysis. Coefficient of variation of each of the 25 most common compounds (standard deviation x 100/mean) was calculated on arcsine square-root transformed data for each of the three larger populations (Yasuni, Tiputini, and Jaun Sacha) and for all samples taken together.

Sørensen's index of similarity (SI: Mueller-Dombois and Ellenberg, 1974 ) was used to calculate the quantitative and qualitative similarities between all populations using all compounds detected in the floral scent. The significance of the correlation between distance and quantitative and qualitative similarities was tested using a Mantel test for association between two distance matrices. Prior to analysis, SI values were transformed to distance values (100-SI) and the tests were made after 9999 random permutations (PC-ORD, multivariate analysis of ecological data, MJM Software Design, Gleneden Beach, Oregon, USA).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
A total of 108 different compounds were detected in the 62 floral scent samples analyzed. Thirty-eight of the compounds occurred sporadically in ten or fewer individuals. Eighty compounds occurred in amounts of <1% in any individual. Only six compounds occurred in >1% relative amount in all samples, while the mean relative amount of 13 compounds per population was >1%. Twenty-eight compounds were found in all individuals, but three of these co-eluted with other compounds and could not be quantified adequately. The remaining 25 compounds and their relative amounts are given in Table 1.

Most of the compounds were of isoprenoid origin, but a few benzenoid and fatty acid-derived compounds were present in small amounts. (Z)- and (E)-ß-Farnesene could not be separated sufficiently to be quantified individually. They are reported together because none possessed any unique ions allowing single ion monitoring and quantification. For a detailed list of the chemical composition of the floral scent of G. macrostachys var. macrostachys, see Knudsen (1999a) .

The number of flowers from which floral scent was collected varied from 17 to 900 per inflorescence with a mean of 162 flowers. The mean amount of volatiles collected per inflorescence per hour varied from 8 to 1745 µg with a mean of 602 µg.

The first three principal components, PC1, PC2, and PC3 explained 65.9% of the variation in the data set, i.e., 32.5, 23.5, and 10.0%, respectively. One outlier, a female flower sample from Yasuni, was found. However reanalyzing the data matrix without this individual did not change the outcome, and the original data set was therefore used. Menthatrienes had high positive and ß-farnesenes high negative loadings along PC1; sesquiterpenes 2 and 3 had high positive and myrcene high negative loadings along PC2; and ipsdienol and sesquiterpene 4 had high positive and ß-farnesenes high negative loadings along PC3 (Table 2). The relative amounts of some compounds varied several orders of magnitude among the studied individuals (Table 1), but the variation seemed to occur irrespective of the geographical origin of the samples and no clear groups can be discerned in the score plots in Fig. 1. The variation found among the Tiputini and Yasuní individuals covered the entire range of variation found in the whole sample. Individuals from Cuyabeno formed a cluster in the score plot of PC1 and PC3, but they still overlapped with the individuals from Tiputini and Yasuní. The two population samples with the farthest geographical distance, i.e., the individuals from Cuyabeno and Makuma were well separated along PC1, but not along the remaining two PCs (Fig. 1).

View larger version (13K): [in this window] [in a new window]   Fig. 1. Score plots of the first three principal components, PC1, PC2, and PC3, from an ordination of 25 compounds present in all individuals and populations of Geonoma macrostachys var. macrostachys

The overall similarity between the means of the five populations using all 108 compounds varied quantitatively from 68% between Cuyabeno and Makuma to 89% between Yasuní and Tiputini and qualitatively from 35% between Yasuní and Makuma to 81% between Yasuní and Tiputini (Tables 3 and 4). There was a significant negative correlation (r = –0.88) between distance and quantitative similarity (transformed 100-SI, r = 0.88, Z = 79 088, P = 0.009), and the data, although not significant, suggest a negative correlation (r = –0.73) between distance and qualitative similarity (transformed 100-SI, r = 0.73, Z = 148 044, P = 0.059).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Intraspecific variation in Geonoma macrostachys and in other species
The present study of variation in floral scent composition is among the largest ever performed, and the only one involving a tropical lowland rain forest species studied in situ. Furthermore, G. macrostachys differs from other studied species in habit, life form, and partly also in ecology. In nectar-rewarding species, three patterns of variation have been found among populations. In Parkia biglobosa Jacq., a bat-pollinated tree of tropical West Africa, no apparent differentiation and relatively little variation in floral scent composition was found among three populations (Pettersson and Knudsen, 2001 ). In Conopodium majus (Gouan) Loret, a promiscuously pollinated perennial herb of Scandinavia, the studied populations could be differentiated on the basis of floral scent composition (Tollsten and Øvstedal, 1994 ). In two moth-pollinated Platanthera species a highly variable floral scent composition was found, and individuals were characterized by either linaloolic, lilac, geraniolic, or benzenoid compounds, irrespective of population and species (Tollsten and Bergström, 1993 ).

Two disparate explanations were offered for the observed patterns. In Parkia biglobosa the lack of differentiation was suggested to reflect the fact that some gene flow between populations is occurring or has occurred in the near past, while in Conopodium majus and Platanthera spp. the results were suggested to reflect local differentiation or isolation. In the case of C. majus the differentiation was suggested to reflect local pollination ecotype differentiation in an otherwise promiscuous species, while in Platanthera spp. variation in scent composition was suggested to be a result of adaptation to the local guild of available moth pollinators and thus reflect local variations in composition or preferences of these.

In scent-rewarding, male-euglossine-bee-pollinated species in the genera Catasetum and Cycnoches, the similarity between populations was >65%, while the similarity between male and female flowers of Cycnoches densiflorum was as high as 95% (Gregg, 1983 ; Whitten and Williams, 1992 ). A high similarity is regarded as a necessity for this odor-driven pollination system to function, and deviations may lead to speciation (Whitten and Williams, 1992 ).

A highly variable floral scent composition was found among non-rewarding flowers of the beetle-pollinated Magnolia kobus DC. (Azuma, Toyota, and Asakawa, 2001 ) and the moth-pollinated orchid Epidendrum ciliare L. (Moya and Ackerman, 1993 ). In female-stage flowers of M. kobus, no single floral scent compound was found in all individuals. Furthermore, individuals or whole populations were characterized either by oxygenated terpenoids and benzenoids, N-compounds, and benzenoids, or by terpenoid hydrocarbons combined with benzenoids and/or oxygenated terpenoids. In E. ciliare, the similarity in floral scent composition among four populations was likewise very low, ranging from 12 to 21%. Pollination in both species is suggested to rely partly or wholly on deception. However, in M. kobus the large variation is suggested to reflect low importance of scent compared to visual cues in attracting pollinators, while the deception in E. ciliare is suggested to rely on disruption of the learning processes in the exploratory behavior of food-searching, naïve moths.

Variation in floral scent in relation to pollination system
The level of variation recorded in floral scent composition within and between populations of G. macrostachys is in the same range as that found in male-euglossine-bee-pollinated orchids and in bat-pollinated P. biglobosa. Bats and euglossine bees are both strong fliers. Plants pollinated by them are prone to form large metapopulations, in which gene exchange by means of pollen flow takes place every generation between individual populations. Despite the fact that male euglossine bees make up only a small fraction of the insect visitor fauna of G. macrostachys (Olesen and Balslev, 1990 ; Listabarth, 1993 ; Knudsen, Andersson, and Bergman, 1999 ) they may nevertheless be highly important for long-distance pollen dispersal in this species, reflected by high similarity in floral scent composition among populations located far apart.

Another avenue to obtain a large metapopulation is through long-distance seed or fruit dispersal either apart or in combination with pollination. The mature fruits of G. macrostachys are black and contrast against the orange infructescence axis, and the mesocarp becomes somewhat juicy at maturity. These features are suggestive of bird dispersal, but so far no reports of seed or fruit dispersal have been made in species of Geonoma.

At first glance, pollination of G. macrostachys appears to be promiscuous because of visits by insects from three different insect orders: beetles, flies, and bees. However, only male euglossine bees are rewarded both by staminate and pistillate flowers, while beetles, flies, and other bees, who visit the flowers to collect or forage on pollen, are rewarded only at staminate flowers. For the latter pollinator groups, Geonoma macrostachys may represent an intraspecific or Bakerian mimicry system in which the non-rewarding pistillate flowers are visited because of their similarity to and lower frequency than the rewarding staminate flowers (Olesen and Balslev, 1990 ; Listabarth, 1993 ; Knudsen, Andersson, and Bergman, 1999 ). In such systems a high similarity in floral scent composition between gender separated either in space or time increases the liability for pollination and has been found among species with non-rewarding female flowers, such as in phytelephantoid palms (Ervik, Tollsten, and Knudsen, 1999 ), in G. macrostachys (Knudsen, Andersson, and Bergman, 1999 ), and in Carica papaya (Knudsen and Tollsten, 1993 ). In G. macrostachys, the similarity in scent composition between staminate and pistillate flowers was found to be 88% (Knudsen, Andersson, and Bergman, 1999 ).

The compounds consistently found in G. macrostachys are mainly myrcene and related oxidized compounds, such as ipsdienone and ipsdienol, menthatrienes, (Z)- & (E)-ß-farnesene, and an epoxide of the latter. Of these ipsdienol has been shown to be very attractive to some species of male euglossine bees (Whitten, Hills, and Williams, 1988 ; Whitten and Williams, 1992 ). (E)-ß-Farnesene has been reported in the floral scent of a number of male-euglossine-bee-pollinated orchids (Gerlach and Schill, 1991 , 1993 ), likewise myrcene. In a field test myrcene did not, however, attract male euglossine bees (Whitten and Williams, 1992 ). The farnesene has not yet been tested as an attractant.

Geonoma macrostachys produces amounts of volatiles in the range of strongly scented moth-pollinated species, and during field work flowering specimens were often detected by olfaction before they were seen visually.

Taken together, the level of variation in floral scent composition, the constituting compounds and the strong scent all point to the importance of male euglossine bees as pollinators of G. macrostachys. Visitation by male euglossine bees may secure some long-distance pollen flow, which, in combination with visitation by other insect groups, secures that most or all flowers are safely pollinated, at least with pollen from nearby sources.

Methodological considerations
It is seldom obvious how to meaningfully handle a complex data set like the present one, and it is not possible to be totally objective in doing so. Only using compounds that are present in all individuals shows the least variable picture. However, the commonly occurring compounds represent the floral scent composition quite well because they quantitatively on average constituted 95% of the scent (Table 1). When considering all compounds the quantitative similarity in floral scent composition among populations was equally high ranging from 68% between the two populations farthest away from each other (Cuaybeno and Makuma) to 89% between those two closest to each other (Yasuní and Tiputini, Table 3). This is in the range of similarity in floral scent composition found between male and female flowers of Salix as well as in many other palm species (Tollsten and Knudsen, 1992 ; Knudsen, Tollsten, and Ervik, 2001 ). However, the qualitative similarity was much lower, ranging from 35% between the Yasuní and Makuma populations to 81% between Yasuní and Tiputini (Table 4). In the present material, 38 of the compounds were detected sporadically in ten or fewer individuals. These compounds will add to the dissimilarity found between the mean of each population, especially for the Jatún Sacha and Makuma populations, because very few individuals were available, and the chance to encounter rare compounds was thus low.

The floral scent of Geonoma macrostachys contains many compounds (108) compared to most other plant species investigated (personal observation). In a recent review of floral scent in the palm family, the qualitative similarity was always much lower between the sexes in species containing many (34–44) floral scent compounds compared to those containing relatively few (6–20) (Knudsen, Tollsten, and Ervik, 2001 ). Thus, there may be an inherent problem in using qualitative similarity because very rare and unique compounds influence on the similarity may be out of proportion compared to their biological significance. However, this discrepancy may also reflect a methodological problem, because samples were collected from a highly variable number of flowers in the field and thus varied in the amounts of floral scent collected (Table 1). More compounds are usually detected in strong compared to weaker floral scent samples, because compounds found in minor and trace amounts in strong samples may, even if they are present, be below the analytical detection threshold in weaker samples.

Perspectives
The present study shows that floral scent composition in the typical variant of Geonoma macrostachys var. macrostachys is homogenous over large distances. In the light of earlier studies, which showed a differentiation in floral scent composition among adjacent, but phenotypically different populations of G. macrostachys var. macrostachys (Knudsen, 1999a , b ), the combined findings may indicate a local differentiation at several different localities, which eventually will lead to new species.

	FOOTNOTES

 
¹ The author thanks F. Borchsenius and B. Ståhl for constructive discussions and criticisms of the manuscript; F. Borchsenius for performing the Mantel test; P. Bergman, S. Andersson, E. Pettersson, N. Dideriksen, B. Bro, J. B. Petersen, P. Hecht, H. Balslev, C. Asmussen, F. Borchsenius, and B. Ståhl for pleasant company during various field trips in Ecuador; INEFAN for collection permits; and the staff at the Jatún Sacha, Yasuní, and Tiputini Biological Stations, as well as the Makuma villagers, for logistical support. Financial support was received from the Swedish Natural Research Council and Carlsbergsfonden.

	LITERATURE CITED

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Azuma H. M. Toyota Y. Asakawa 2001 Intraspecific variation of floral scent chemistry in Magnolia kobus DC. Journal of Plant Research 114: 411-422[CrossRef][ISI]

Azuma H. M. Toyota Y. Asakawa R. Yamaoka J. G. Garcia-Franco G. Dieringer L. B. Thien S. Kawano 1997 Chemical divergence in floral scents of Magnolia and allied genera (Magnoliaceae). Plant Species Biology 12: 69-83

Barkman T. J. J. H. Beaman D. A. Gaga 1997 Floral fragrance variation in Cypripedium: implications for evolutionary and ecological studies. Phytochemistry 44: 575-582[CrossRef][ISI][Medline]

Bergström G. G. Birgersson I. Groth A. Nilsson 1992 Floral fragrance disparity between three taxa of lady's slipper Cypripedium calceolus L. (Orchidaceae). Phytochemistry 31: 2315-2319[CrossRef][ISI]

Borg-Karlson A.-K. I. Groth L. Ågren B. Kullenberg 1993 Form-specific fragrances from Ophrys insectifera L. (Orchidaceae) attract species of different pollinator genera. Evidence of sympatric speciation. Chemoecology 4: 39-45

Eriksson L. E. Johansson N. Kettaneh-Wold S. Wold 1999 Introduction to multi- and megavariate data analysis using projection methods (PCA & PLS). Umetrics AB, Box 7960, SE 90719 Umeå, Sweden

Ervik F. L. Tollsten J. T. Knudsen 1999 Floral scent chemistry and pollination ecology in phytelephantoid palms (Arecaceae). Plant Systematics and Evolution 217: 279-297[CrossRef][ISI]

Gerlach G. R. Schill 1991 Composition of orchid scents attracting euglossine bees. Botanica Acta 104: 379-391[ISI]

Gerlach G. R. Schill 1993 Die Gattung Coryanthes Hook. (Orchidaceae). Eine monographische Bearbeitung unter besonderer Berucksichtigung der Blütenduftstoffe. Tropische und subtropische Pflanzenwelt 83

Gregg K. B. 1983 Variation in floral fragrances and morphology: incipient speciation in Cycnoches?. Botanical Gazette 144: 566-576[CrossRef]

Groth I. G. Bergström O. Pellmyr 1987 Floral fragrances in Cimicifuga: chemical polymorphism and incipient speciation in Cimicifuga simplex. Biochemical Systematics and Ecology 15: 441-444[CrossRef]

Kite G. C. 1995 The floral odour of Arum maculatum. Biochemical Systematics and Ecology 23: 343-354[CrossRef]

Knudsen J. T. 1994 Floral scent variation in the Pyrola rotundifolia complex in Scandinavia and Western Greenland. Nordic Journal of Botany 14: 277-282[ISI]

Knudsen J. T. 1999a Floral scent chemistry in geonomoid palms (Palmae: Geonomeae) and its importance in maintaining reproductive isolation. Memoirs of the New York Botanical Garden 83: 141-168

Knudsen J. T. 1999b Floral scent differentiation among co-flowering, sympatric species of Geonoma (Arecaceae). Plant Species Biology 14: 137-142

Knudsen J. T. S. Andersson P. Bergman 1999 Floral scent attraction in Geonoma macrostachys, an understorey palm of the Amazonian rain forest. Oikos 85: 409-418[CrossRef][ISI]

Knudsen J. T. L. Tollsten 1993 Trends in floral scent chemistry in pollination syndromes: floral scent composition in moth pollinated taxa. Botanical Journal of the Linnean Society 113: 263-284[CrossRef]

Knudsen J. T. L. Tollsten F. Ervik 2001 Flower scent and pollination in selected neotropical palms. Plant Biology 3: 642-653[CrossRef][ISI]

Levin R. A. R. A. Raguso L. A. McDade 2001 Fragrance chemistry and pollinator affinities in Nyctaginaceae. Phytochemistry 58: 429-440[CrossRef][ISI][Medline]

Listabarth C. 1993 Pollination in Geonoma macrostachys and three congeners, G. acaulis, G. gracilis, and G. interrupta. Botanica Acta 106: 496-506[ISI]

Moya S. J. D. Ackerman 1993 Variation in the floral fragrance of Epidendrum ciliare (Orchidaceae). Nordic Journal Botany 13: 41-47

Mueller-Dombois D. H. Ellenberg 1974 Aims and methods of vegetation ecology. Wiley, New York, New York, USA

Olesen J. M. H. Balslev 1990 Flower biology and pollinators of the Amazonian monoecious palm, Geonoma macrostachys: a case of Bakerian mimicry. Principes 34: 181-190

Pettersson S. J. T. Knudsen 2001 Diurnal and populational variation in floral scent composition in bat pollinated Parkia biglobosa (Mimosaceae). Botanical Journal of Linnean Society 135: 97-106[CrossRef]

Schiestl F. P. M. Ayasse H. F. Paulus D. Erdmann W. Francke 1997 Variation of floral scent emission and postpollination changes in individual flowers of Ophrys sphegodes subsp. sphegodes. Journal of Chemical Ecology 23: 2881-2895[CrossRef][ISI]

Tollsten L. L. G. Bergström 1993 Fragrance chemotypes of Platanthera (Orchidaceae)—the result of adaptation to pollinating moths?. Nordic Journal of Botany 13: 607-613[ISI]

Tollsten L. J. T. Knudsen 1992 Floral scent in dioecious Salix (Salicaceae)—a cue determining the pollination system. Plant Systematics and Evolution 182: 229-237[CrossRef][ISI]

Tollsten L. J. T. Knudsen L. G. Bergström 1994 Floral scent in generalistic Angelica (Apiaceae)—an adaptive character?. Biochemical Systematics and Ecology 22: 161-169

Tollsten L. D. O. Øvstedal 1994 Differentiation in floral scent chemistry among populations of Conopodium majus (Apiaceae). Nordic Journal of Botany 14: 361-367[ISI]

Whitten M. W. H. G. Hills N. H. Williams 1988 Occurrence of ipsdienol in floral fragrances. Phytochemistry 27: 2759-2760[CrossRef][ISI]

Whitten M. W. N. H. Williams 1992 Floral fragrances of Stanhopea (Orchidaceae). Lindleyana 7: 130-153

This article has been cited by other articles:

				G. P. Svensson, M. O. Hickman Jr, S. Bartram, W. Boland, O. Pellmyr, and R. A. Raguso Chemistry and geographic variation of floral scent in Yucca filamentosa (Agavaceae) Am. J. Botany, October 1, 2005; 92(10): 1624 - 1631. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Knudsen, J. T. Search for Related Content PubMed Articles by Knudsen, J. T. Agricola Articles by Knudsen, J. T.