| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
2Centre d'Ecologie Fonctionnelle et Evolutive, CNRS, 1919 route de Mende, 34293 Montpellier CEDEX 5, France; and 3Laboratoire de Biologie et Physiologie Végétales, Université de La Réunion, 15 avenue René Cassin, BP 7151, 97715 Saint Denis Messag, CEDEX 9, Ile de La Réunion, France
Received for publication June 22, 1998. Accepted for publication January 29, 1999.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: altitude • Dombeya • leaky dioecy • oceanic islands • phenotypic gender • rare species • Sterculiaceae
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Thomson and Barrett, 1981
), it also appears to have evolved as a mechanism to increase the efficiency of resource allocation (Willson, 1979
; Bawa, 1980
; Givnish, 1980
; Thomson and Brunet, 1990
; Freeman et al., 1997
).
A particular situation in which the presence and evolution of dioecy have fascinated evolutionary biologists concerns the abundance of dioecious species on oceanic islands (Carlquist, 1966
; Baker, 1967
; Bawa, 1982
; Baker and Cox, 1984
; Sakai et al., 1995a
, b), e.g., on Hawaii where 14.7% of species are dioecious (Sakai et al., 1995a
). From initial suggestions that dioecy may have evolved autochtonously from self-compatible colonists (Baker, 1955
, 1967) it has become clear that although some groups have evolved dioecy on islands (e.g., see Mayer and Charlesworth, 1992
; Weller, Wagner, and Sakai, 1995
) most dioecious species on Hawaii were already dioecious prior to colonization (Sakai et al., 1995b
). To explain this ability of dioecious plants to colonize oceanic islands several ecological correlations have been invoked. These include fleshly fruits, which due to bird dispersal allow for colonization of remote islands, multiseeded dry fruits that facilitate population establishment in isolated situations, generalist pollinators, and longevity (woody habit) (see Thomson and Brunet, 1990
). Another potentially important component of the colonization strategy of dioecious plants on islands is that many taxa considered truly dioecious may have occasional hermaphrodite or opposite sex flowers, i.e., show leaky dioecy (Baker and Cox, 1984
). These authors provide several clear illustrations of this for dioecious plants on Hawaii. Data from other tropical volcanic islands are not, however, available to examine the general significance of this possibility.
Baker and Cox (1984)
also reported that the percentage of dioecious species on oceanic islands was closely (82% of the variance) related to the maximum elevation of the islands and their distance from the equator, the percentage of dioecy increasing with island height, and proximity to the equator. Baker and Cox (1984)
interpreted this correlation as being due to differences in the percentage of dioecy in the sources of island floras. It is, however, possible that the greater abundance of dioecy on tropical high islands is due to an ecological segregation, with dioecy being more abundant at high altitude for ecological causes. For example, Delph (1990)
showed an increase in unisexuality with altitude in the genus Hebe on New Zealand. Again the full significance of this result remains open due to the absence of work that compares the distribution of dioecy among the different vegetation communities on tropical high islands. Other studies on variation in the presence of dioecy in different community types have been made in temperate environments (e.g., Freeman, McArthur, and Harper, 1984
; Freeman, Sanderson, and Tiedemann, 1993
; Renner and Ricklefs, 1995).
We have thus begun a study of the abundance, ecology, and evolution of dioecious groups of species on the oceanic island of La Réunion, which with Mauritius and Rodrigues forms part of the Mascarene archipelago, situated 800 km east of Madagascar in the southern Indian Ocean. La Réunion (55°39' E, 21°00' S) is a geologically young volcanic island (2.5–3 x 106 yr ago) (Doumenge and Renard, 1989
), which despite its small surface area (2500 km2), has natural vegetation over a wide altitudinal gradient, up to 3069 m. La Réunion was thus used as one of the "high" islands in the analyses by Baker and Cox (1984)
, who unfortunately, by estimating dioecy from Cordemoya (1895)
at 4%, grossly underestimated the true figure, which lies between 15 and 20% (T. Pailler, unpublished data). The altitudinal gradient on La Réunion, combined with great variation in topography, has created an enormous diversity of ecological conditions. Seven main plant communities have been described (Cadet, 1977
; Dupouey and Cadet, 1986
): littoral vegetation, lowland wetlands, lowland tropical moist forest, lowland tropical dry forest, mid-altitude tropical moist forest, cloud forest, and ericoid vegetation. A large component of the indigenous flora (80%) shows taxonomic affinities to that of Madagascar, although many species appear to have originated in East Africa, Asia, and Australasia (Cadet, 1977
).
The genus Dombeya (Sterculiaceae) in the Mascarene islands provides a particularly interesting group for quantifying variation in the dioecious breeding system on oceanic islands. This genus, with more than 300 described tree species (Friedmann, 1987
; Seyani, 1991
), is confined to Africa, Saudi Arabia, the Yemen, the Comores, Madagascar, and the Mascarene islands. The genus is reported to be hermaphrodite in Africa (Seyani, 1991
). On Madagascar, although many species bear hermaphrodite flowers, unisexual flowers have also been reported (Arènes, 1959
). In the Mascarene archipelago, 13 dioecious species and one hermaphrodite species have been described (Friedmann, 1987
). The two species endemic to Mauritius and Rodrigues are extremely rare (only a handful of individuals now exist) while on La Réunion, of the 11 known endemic species, seven have local abundances that exceed more than ten trees of each sex. Previous studies of two of these endemic species (D. ciliata and D. delislei) have found that both species show cryptic dioecy, i.e., the retention of opposite sex structures within the flowers of each sex, and leaky dioecy (or inconstant males, see Lloyd, 1980
), i.e., the presence of fruits on a high percentage of male trees (Humeau, Pailler, and Thompson, 1999
; L. Humeau, unpublished data). These studies suggest that leaky dioecy may be particularly abundant in dioecious species of this genus in the Mascarene islands. On La Réunion, Dombeya are present over an extremely wide gradient of ecological conditions, occurring in both dry tropical forest over a wide range of altitudes (mean annual rainfull <500 mm) and in humid tropical forest (mean annual rainfull >6000 mm). This allows for a comparison of the occurrence of dioecy across a range of ecological conditions in a closely related group of species. What is more, in one of our studies the lowland populations showed leaky dioecy, while an upland population showed strict dioecy.
The purpose of this study was to extend our previous work on two species to all endemic Dombeya on the island of La Réunion, in order to generalize on the patterns of morphological and fertility variation. We examined three questions: (1) Is the expression of cryptic and leaky dioecy a consistent feature of dioecious Dombeya on La Réunion? (2) Does variation in the expression of dioecy between species in different habitats and communities parallel between-population variation observed in the previously studied D. ciliata? (3) What are the implications of the variation in the dioecious breeding system for the conservation of Dombeya species on an oceanic island whose natural vegetation is subject to intense human pressures.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
and in the personal field notes of this author. These data were used to produce distribution maps for each species. Four classes of abundance were then established depending on the number of individuals in populations and the number of populations on the island: (1) abundant and in large populations, (2) frequent but in small populations, (3) uncommon and in small populations, and (4) extremely rare.
For each species and population studied, we determined the relative abundance of the two sexes. All observed trees were in flower; only small and probably juvenile trees were not flowering. In each population, all accessible adult individuals were sampled and classified as male or female based on the a priori sex classification mentioned below. Sex ratios were analyzed using the G test for inequality of frequencies (Sokal and Rohlf, 1981
).
During the course of this study we became aware of a new taxon not previously described by Friedmann (1987)
, which is distinghished from all other species due to a combination of characters normally restricted to two different species (D. ficulnea and D. punctata). This taxon is sympatric with D. ficulnea but flowers at a different time of the year (L. Humeau, unpublished data). This new taxon was included in the study and referred to as Dombeya sp. (Table 1).
|
Five randomly selected flowering stems per individual were cut. Two leaves on the fourth and fifth node under the apex and axillary inflorescence were sampled on each of the five stems. Leaves were pressed and dried at ambient temperature in the herbarium, and length and width were measured. Leaf surface area was calculated as length x width. The number of inflorescences per stem was counted on five stems. On fresh stems, the number of flowers per inflorescence, inflorescence diameter, and peduncule length for two inflorescences were measured for each of the five stems per individual. Three flowers from different inflorescences per individual were collected and preserved in 70% ethanol prior to floral measurments. Corolla diameter and staminode, stamen, gynoecium, and stigma length were measured to 0.01 mm using digital calipers. The number of loculi, stigmas, and ovules were also counted in each flower.
Flower diameter and petal length were also measured for a small number of individuals of two rare dioecious species on La Réunion: D. populnea (N = 2 males and 3 females), D. ferruginea (N = 4 males and 3 females), and one hermaphrodite species, D. acutangula (N = 10 hermaphrodite individuals). All three species have different inflorescence architecture from other species and are in a separate evolutionary clade. Two species, D. blatiolens and D. umbellata, could not be sampled due to their rarity and the inaccessibility of known trees.
Pollen grain diameter was quantified for ten plants of each sex in the eight species for which several individuals of each sex could be sampled. Mature undehisced anthers were collected from buds, pollen from two anthers from a single flower were mixed, and 20 pollen grains were measured under a light microscope at 1000x magnification. The pollen viability of ten individuals of each sex of each species was quantified by counting 100 pollen grains per individual stained with 2% lacto-phenol blue aniline. A pollen grain was considered as viable when it was entirely stained, i.e., had a normally formed cytoplasm (Stone, Thomson, and Dent-Acosta, 1995
).
Floral traits were analyzed with a two-way mixed-model analysis of variance (Sokal and Rohlf, 1981
) using PROC GLM in SAS (1990)
.
Fecundity
Two to three months after the flowering period, when fruits were mature, the number of infructescences per stem was determined on five randomly sampled stems per plant of all individuals of each species on which floral traits were measured. The number of fruits per inflorescence was counted on two inflorescences on each of the five stems per individual. The number of seeds per fruit was determined for five fruits on each of the ten infructescences for the females and from one (when only one fruit was found) to 50 fruits for males. These fecundity data were collected for three consecutive years for D. ciliata (L. Humeau et al., unpublished data), two years for D. delislei (L. Humeau et al., unpublished data), and in a single year for the remaining species (none of which show any variation within each sex, see Results). Fecundity data were analyzed with a two-way mixed-model analysis of variance (Sokal and Rohlf, 1981
) using PROC GLM in SAS (1990)
with sex and species as fixed effects.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) indicate that D. blatiolens (which we were unable to sample) may be very closely related to D. ciliata. The only discernible difference between the two species is their flowering periods. Two species flower in June to August (D. blatiolens and D. ficulnea), two species in December or January (D. reclinata and D. pilosa), and seven species (D. delislei, D. elegans, D. punctata, D. umbellata, D. populnea, D. ferruginea and D. acutangula) in March to June. Populations of D. ciliata at low altitude flower in September, while those at high altitude flower in January–February. There was no significant difference in the abundance of males and females for any species (Table 1).
|
). Male flowers have significantly larger stamens, staminodes, and pollen grains than females in all studied species (Table 2). In females, all observed pollen grains were found to be nonstainable, and although females bear anthers, they never dehisce. In all species, the presence of morphologically well-developed male structures in females, and to a lesser extent gynoecia in male flowers, indicate that all species are cryptically dioecious.
|
In all species, males and females showed significant differences in inflorescence diameter, flower diameter, and petal length but not in peduncule length and number of flowers per inflorescence (Table 2). Differences in inflorescence diameter between sexes are essentially due to differences in flower size since there were no differences in total flower number per inflorescence between sexes (Table 2). Inflorescence diameter and petal length showed a significant species by sex interaction, which illustrates that flower size dimorphism between sexes is greater in some species than in others (e.g., see Fig. 2). No significant correlation was found in any species or across species between the number of flowers per inflorescence and petal length or flower diameter. There was no significant sexual dimorphism in leaf area (Table 2).
Variation in male fruit set
In all species, females show no male function; they achieve 100% of their reproductive effort through ovules. Among species, males varied. In some species (namelly Dombeya reclinata, D. punctata, D. ficulnea, D. elegans), all males examined produced no fruit (Table 3) and were therefore strict males. In other species, e.g., D. ciliata, D. delislei, and D. pilosa, some males are able to set a small amount of fruit (Table 3).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Anderson and Symon, 1989
; Willson, 1991
). Most Dombeya are visited by bees (Apis melifera) and although butterflies (Euploea goudotii, Henostesia narcissus, Parnana marchalii) and birds (Zosterops borbonica, Zosterops olivaceus) visit some species for nectar (T. Pailler and L. Humeau, unpublished observations), anthers could be an important component of pollination attraction—hence their retention in females. The variation among species in the retention of female structures by males reflects in fact two phenotypic manifestations of dioecy in Mascarene Island Dombeya. In one group, males do not produce fruit and species are strictly dioecious. This includes species in which males do not have either ovules (D. ficulnea, D. punctata, D. sp.) or styles (D. reclinata). Hence, although our observations were made in a single year, it is very unlikely that gender variation could occur in these species. Although there was no difference in the expression of dioecy in D. elegans in three studied populations, it is always possible that continued observations over several years may reveal limited fruit set on some male trees of this species due to the presence of ovules and styles in male flowers.
In the second group of species, males produce a small number of fruits with seeds and species show leaky dioecy, e.g., D. ciliata, D. pilosa, and D. delislei. Studies on fruit set were made during 2 or 3 yr for some species (D. ciliata, D. delislei, and D. punctata), none of which showed variation in the expression of phenotypic gender, i.e., males that set fruit in one year did so in consecutive years (L. Humeau et al., unpublished data). Males are, however, phenotypically different from females since the stigma lobes remain at least partially closed thoughout the life of the flower. Hence their fruit set is low compared to females. In two of the rare species (D. populnea and D. umbellata), although we could not carry out detailed observations, field observations (L. Humeau and T. Pailler) show that males of both species set fruit. Elsewhere in the Sterculiaceae, there are dioecious genera in the Mascarene archipelago (i.e., Astyria and Ruizia, which are endemic to Mauritius and La Réunion, respectively) (Friedmann, 1987
) and in Madagascar and Africa (Pterygota and Hildegardia; Arènes, 1959
). Hermaphroditism is also known both in Dombeya acutangula on La Réunion and in Africa (Friedmann, 1987
; Seyani, 1991
) and other Sterculiaceae such as Sterculia (Taroda and Gibbs, 1982
), Cola (Jacob, 1973, 1980
), and Theobroma (Cope, 1962
). It is thus unclear whether the leaky dioecy represents a breakdown of dioecy or a stage in the evolution of dioecy.
The distribution of dioecy and leaky dioecy in Dombeya on La Réunion
Our data on the distribution of the eight Dombeya species on La Réunion suggest an interesting trend in the presence of dioecy. Species with strict dioecy almost exclusively occur in large populations in mid- to high-altitude moist tropical cloud forest, while species in small fragmented and highly isolated relict populations at lower altitudes (either in dry tropical forest or lowland wet tropical forest) almost all show leaky dioecy. The species that show leaky dioecy thus occur in sites where human disturbance via habitat destruction and fragmentation is greatest. The trend is such that of the nine strictly dioecious populations, eight are in cloud forest and one is in a semidry fragmented tropical forest, but all nine are above 1000 m. Of the four leaky dioecious populations, three are in semidry fragmented tropical forest, while one is above 1000 m in cloud forest. In addition to these species, D. populnea occurs as very rare individuals in lowland sites, and observations suggest leaky dioecy as two males have been observed with fruits. Finally, the single hermaphrodite species on La Réunion, D. acutangula, also occurs in extremely small fragmented populations in lowland semidry tropical forest. The latter two species thus add weight to the trend for leaky dioecy to occur at low altitude and/or in small fragmented populations.
Two reasons for this differential distribution of species that show strict and leaky dioecy can be suggested. The first and most obvious reason for this trend is that ecological differences favor strict dioecy in particular sites. An example of this phenomena on oceanic islands can be seen in work by Sakai and Weller (1991)
who showed that some male individuals of Schiedea globosa become hermaphrodite under better growing conditions (including higher temperatures) in the field. The propensity for labile sex expression was in this case under both environmental and genetic control. Cool temperatures in high-altitude cloud forest could be important in this context. Alternatively, the pollination environment may favor dioecy in the cloud forest if pollination is more by generalist pollinators that increase selection pressures for sex separation (Beach and Bawa, 1980
). Indeed pollinator faunas are well known to vary over altitudinal gradients (Arroyo, Primack, and Armesto, 1982
). An example of variation of gender dimorphism with altitude occurs in the genus Hebe in New Zealand, which varies from gynodioecy to dioecy with altitude (Delph, 1990
). Furthermore, our work on variation in floral traits in two Rubiaceae on La Réunion provides further illustrations of altitudinal differences that may relate to differences in the abiotic and/or pollination environment. In distylous Gaertnera vaginata, flower size decreases with altitude (Pailler and Thompson, 1997
). In dioecious Chassalia corallioides male flowers are smaller and wider in altitudinal populations, where bees may be the most important pollinators, than in low-altitude populations where hawk moths are more abundant (Pailler et al., 1998
).
A second possible reason concerns the population characteristics of the leaky dioecious populations. These are all small populations in disturbed areas where population extinction and recolonization may occur frequently or where small population size may constrain reproduction. If such populations have been historically small these characteristics may have favored males capable of setting fruit, as metapopulation models suggest (Pannell, 1997
). However, because we have no data on what population sizes were like prior to human-induced habitat disturbance and fragmentation, we cannot be sure that population sizes have been small long enough for selection to have favored leaky dioecy.
For the single species that occurs in both cloud forest and lowland rain forest (i.e., D. ciliata) the trend for strict dioecy in cloud forest and leaky dioecy in lowlands is repeated (L. Humeau et al., unpublished data). Two lowland fragmented rain forest populations show leaky dioecy and the large unfragmented cloud forest population is strictly dioecious, a result that was consistent over 3 yr of study. Current work on the flora as a whole indicates that there is a greater percentage of dioecy in cloud forest than in lowland rain forest or in tropical dry forest (T. Pailler, unpublished data). The trend for strict dioecy to be more abundant in cloud forest and leaky dioecy to be more common in lowland rain forest or semidry tropical forest thus occurs at three levels, within a single species (D. ciliata), between closely related Dombeya species, and at the community level.
Our data showing an increasing tendency for dioecy to occur at high altitude on La Réunion are particularly interesting in the context of a correlation presented by Baker and Cox (1984)
. These authors showed that maximum height of tropical oceanic islands and proximity to the equator account for 82% of the variance in percentage of dioecism on oceanic islands. They interpreted this trend as caused by the proximity of floras to their tropical mainland sources, i.e., forests for high islands vs. littoral vegetation for low islands. Interestingly, if the true percentage of dioecy on La Réunion had been included in an analysis such as that performed by Baker and Cox (1984)
(who put the percentage of dioecy on La Réunion at 4% while it is actually betwen 15 and 20%; T. Pailler, unpublished data), their correlation would have been even stronger (see their Fig. 1). Our data suggest that the high percentage of dioecy on high islands may result from dioecy being favored at high altitude on such islands. Data on the percentage of dioecy on islands have never taken into account the possibility that the percentage of dioecy may vary across environments on a single island. Our results suggest that perhaps future work should do so.
Implications for the conservation of Mascarene island Dombeya
Our data on the distribution and abundance of endemic Dombeya on La Réunion in relation to variation in the expression of dioecy have important ramifications for the conservation of a flora that is under ever-increasing threats from man. First, several Dombeya species are key components of tropical moist and cloud forests on La Réunion. They play an important role in the maintenance of the structure of such communities and in ecosystem function where many trees harbor an extremely diverse epiphytic flora and fauna. These species are strictly dioecious. Second, several species are important in relictual fragments of dry tropical forest and at low altitude. In small or fragmented populations, chance factors may lead to small groups of same-sex plants becoming isolated from plants of the opposite sex, with disastrous results. What is more, some Dombeya species are now very rare and occur as either isolated individuals or in botanical gardens. What is interesting here is that most of the species that are rare (i.e., D. populnea and D. acutangula studied here plus D. mauritiana and D. rodriguesiana, endemic to Mauritius and Rodriges, respectively) appear, based on morphological traits, to be very closely related and distinct from the more abundant strictly dioecious species. Although such rarity is clearly due in major part to the disappearance of the type of habitat these species occupy, the potential relatedness of these rare species points to a potential phylogenetic significance of rarity for species in particular environments.
|
|
|
FOOTNOTES |
|---|
4 Author for correspondence (CEFE-CNRS, 1919 route de mende, 34293 Montpellier Cedex 5, France, (thompson{at}cefe.cnrs-mop.fr
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Arènes, J. 1959 Sterculiacées, Flore de Madagascar et des Comores, 131eme famille. Firmin-Didot et Cie, Paris.
Arroyo, M. T. K., R. Primack, and J. Armesto. 1982 Community studies in pollination ecology in the high temperate andes of central Chile. I. Pollination mechanisms and altitudinal variation. American Journal of Botany 69: 82–97.[CrossRef][ISI]
Baker, H. G. 1955 Self-compatibility and establishment after "long-distance" dispersal. Evolution 9: 347–348.[CrossRef][ISI]
———. 1967 Support for Baker's law—as a rule. Evolution 21: 853–856.[CrossRef][ISI]
———, and P. A. Cox. 1984 Further thoughts on dioecism and islands. Annals of the Missouri Botanical Garden 71: 244–253[CrossRef][ISI]
Bawa, K. S. 1980 Evolution of dioecy in flowering plants. Annual Review of Ecology and Systematics 11: 15–39.
———. 1982 Outcrossing and the incidence of dioecism in islands floras. American Naturalist 119: 866–871.[CrossRef][ISI]
Beach, J. H., and K. S. Bawa. 1980 Role of pollinators in the evolution of dioecy from distyly. Evolution 34: 1138–1142.[CrossRef][ISI]
Cadet, T. 1977 La végétation de l'île de La Réunion: étude phytosociologique et phytoécologique. Thèse de doctorat. Université d'Aix-Marseille, France.
Carlquist, S. 1966 The biota of long-distance dispersal. I. Principles of dispersal and evolution. Quarterly Review of Biology 41: 247–70.[CrossRef][Medline]
Charlesworth, B., and D. Charlesworth. 1978 A model for the evolution of dioecy and gynodioecy. American Naturalist 112: 975–997.[CrossRef][ISI]
Cope, F. W. 1962 The mechanism of pollen incompatibility in Theobroma cacao L. Heredity 17: 157–182.[ISI]
Cordemoya, E. J. de. 1895 Flore de l'île de La Réunion. P. Klincksiet, Paris.
Delph, L. F. 1990 The evolution of gender dimorphism in New Zealand Hebe (Scrophulariaceae) species. Evolutionary Trends in Plants 4: 85–97.
Doumenge, C., and Y. Renard. 1989 La conservation des écosystèmes forestiers de l'île de La Réunion. International Union for the Conservation of Nature and Natural Ressources, Gland, Switzerland.
Dupouey, J. L., and T. Cadet. 1986 Subdivision de la forêt de bois de couleur à l'île de La Réunion. Annales des Sciences Forestières 43: 105–115.
Freeman, D. C., J. L. Doust, A. El-Keblawy, K. J. Miglia, and E. D. McArthur. 1997 Sexual specialization and inbreeding avoidance in the evolution of dioecy. Botanical Review 63: 65–92.
———, E. D. McArthur, and K. T. Harper. 1984 The adaptive significance of sexual lability in plants using Atriplex canescens as a principal example. Annals of the Missouri Botanical Garden 71: 265–277.[CrossRef][ISI]
———, S. C. Sanderson, and A. R. Tiedemann. 1993 The influence of topography on male and female fitness components of Atriplex canescens. Oecologia 93: 538–547.
Friedmann, F. 1987 Flore des Mascareignes, La Réunion, Maurice, Rodrigues, Sterculiacées. Royal Botanical Gardens, Kew, 53: 1–50.
Givnish, T. J. 1980 Ecological constraints on the evolution of breeding systems in seed plants: dioecy and dispersal in gymnosperms. Evolution 34: 959–972.[CrossRef][ISI]
Humeau, L., T. Pailler, and J. D. Thompson. 1999 Variation in the breeding system of two sympatric Dombeya species on La Réunion island. Plant Systematics and Evolution, in press.
Jacob, V. J. 1973 Self-incompatibily mechanism in Cola nitida. Incompatibility Newsletter 3: 60–61.
——— 1980 Pollination, fruit setting and incompatibility in Cola nitida. Incompatibility Newsletter 12: 50–56.
Lloyd, D. G. 1980 The distribution of gender in four angiosperm species illustrating two evolutionary pathways to dioecy. Evolution 34: 123–134.[CrossRef][ISI]
Mayer, S. S., and D. Charlesworth. 1992 Genetic evidence for multiple origins of dioecy in the Hawaiian shrub Wikstroemia (Thymeleaceae). Evolution 46: 207–215.[CrossRef][ISI]
Pailler, T., L. Humeau, J. Figier, and J. D. Thompson. 1998 Reproductive trait variation in the functionally dioecious and morphologically heterostylous island endemic Chassalia corallioides (Rubiaceae). Biological Journal of the Linnean Society, 64: 297–313.
———, and J. D. Thompson. 1997 Distyly and incompatibility variation in Gaertnera vaginata (Rubiaceae). American Journal of Botany 84: 315–327.[Abstract]
Pannell, J. 1997 Variation in sex ratios and sex allocation in androdioecious Mercurialis annua. Journal of Ecology 85: 57–69.
Renner, S. S., and R. E. Ricklefs. 1995 Dioecy and its correlates in the flowering plants. American Journal of Botany 82: 596–606.[CrossRef][ISI]
Sakai, A. K., W. L. Wagner, D. M. Ferguson, and D. R. Herbst. 1995a Origins of dioecy in the Hawaiian flora. Ecology 76: 2517–2529.[CrossRef][ISI]
———, ———, ———, and ———. 1995b Biogeographical and ecological correlates of dioecy in the Hawaiian flora. Ecology 76: 2530–2543.[CrossRef][ISI]
———, and S. G. Weller. 1991 Ecological aspects of sex expression in subdioecious Schiedea globosa (Caryophyllaceae). American Journal of Botany 78: 1280–1288.[CrossRef][ISI]
SAS 1990 SAS/STAT guide for personal computers, version 6, 4th ed. SAS Institute, Cary, NC.
Seyani, J. H. 1991 The genus Dombeya (Sterculiaceae) in Continental Africa. Meise, National Botanical Garden of Belgium.
Sokal, R. R., and J. R. F. Rohlf. 1981 Biometry, 2nd ed. W. H. Freeman, New York, NY.
Stone, J. L., J. D. Thomson, and S. J. Dent-Acosta. 1995 Assesment of pollen viability in hand-pollination experiments: a review. American Journal of Botany 82: 1186–1197.[CrossRef][ISI]
Taroda, T., and P. E. Gibbs. 1982 Floral biology and breeding system of Sterculia chicha St. Hil. New Phytologist 90: 735–743.[CrossRef][ISI]
Thomson, J. D., and S. C. H. Barrett. 1981 Selection for outcrossing, sexual selection and the evolution of dioecy in plants. American Naturalist 118: 443–449.[CrossRef][ISI]
———, and J. Brunet. 1990 Hypotheses for the evolution of dioecy in seed plants. Trends in Ecology and Evolution 5: 11–16.
Weller, S. G., W. L. Wagner, and A. L. Sakai. 1995 A phylogenetic analysis of Schiedea and Alsinidendron (Caryophyllaceae: Alsinoideae): implication for the evolution of breeding system. Systematic Botany 20: 315–337.[CrossRef][ISI]
Willson, M. F. 1979 Sexual selection in plants. American Naturalist 113: 777–790.[CrossRef][ISI]
———. 1991 Sexual selection, sexual dimorphism and plant phylogeny. Evolutionary Ecology 5: 69–87.
This article has been cited by other articles:
|
|
 |
|
  G. J. Anderson, G. Bernardello, M. R. Opel, A. Santos-Guerra, and M. Anderson Reproductive biology of the dioecious Canary Islands endemic Withania aristata (Solanaceae) Am. J. Botany, September 1, 2006; 93(9): 1295 - 1305. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Davis Madagasikaria (Malpighiaceae): a new genus from Madagascar with implications for floral evolution in Malpighiaceae Am. J. Botany, April 1, 2002; 89(4): 699 - 706. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
