| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
Department of Ecology and Evolutionary Biology, University of California, Irvine, California 92697-2525 USA
Received for publication October 20, 2004. Accepted for publication June 10, 2005.
ABSTRACT
Both changes in sex allocation and pollination mode may promote the separation of sexes in plant populations. Simultaneous evolution of wind pollination and dimorphism has occurred in Schiedea, where species with different female frequencies provide an opportunity to observe the effect of wind pollination on sex allocation and floral morphology. Differences among species in the ratio of anther to ovary volume were not the result of sex allocation trade-offs, but instead resulted from production of vestigial stamens in females; there were no changes in ovary volume in males and hermaphrodites (MH) of dimorphic species. Relative to hermaphroditic species, dimorphic species had more condensed inflorescences, a pattern often associated with wind pollination. Within dimorphic species, MH had longer filament lengths than females, and females had longer stigmas than MHs. These traits are characteristic of wind pollination, but there was no relationship between the degree of sexual dimorphism and female frequency. Ovary volume and ovule number and size had positive phenotypic correlations between females and MH of dimorphic species, making sex specialization more difficult. In dimorphic Schiedea species, selection for wind pollination may have a greater effect on floral traits than trade-offs in allocation between male and female function.
Key Words: Caryophyllaceae • dioecy • gynodioecy • Hawaiian Islands • Schiedea • sex allocation • sexual dimorphism • wind pollination
Flowering plants exhibit great diversity in floral morphology, mating systems, and mode of pollination. Several factors, including changes in sex allocation and changes in mode of pollination, may promote this diversification (reviewed in Geber et al., 1999
). Most flowering plant species are hermaphroditic, but dimorphic breeding systems, including dioecious breeding systems with separate male (staminate) and female (pistillate) individuals in a population, have evolved independently on a number of occasions (Renner and Ricklefs, 1995
). One common pathway to dioecy from hermaphroditism is through gynodioecy, the occurrence of females and hermaphrodites in a population (Lloyd, 1975
, 1979
; Charlesworth and Charlesworth, 1978
; Thomson and Brunet, 1990
; Geber et al., 1999
). With nuclear inheritance of male sterility, gynodioecy may be a transient intermediate stage in the evolution of dioecy, with strong selection on hermaphrodites to become more male-like, once females are present in a population (Charlesworth, 1999
). The evolution of dioecy from hermaphroditism and gynodioecy necessarily involves changes in allocation to primary sex traits (involving androecia and gynoecia) and differentiation of males and females, but it is not clear whether these changes will necessarily involve trade-offs between male and female investment, and few studies have documented the trade-offs assumed in sex allocation models (reviewed in Campbell, 2000
). In addition, changes in breeding system may affect not only primary sex traits but also secondary sex traits (e.g., sepals, petals, nectaries, inflorescence structure), particularly if a shift in the pollination system is involved (Lloyd and Webb, 1977
; Bell, 1985
; Delph et al., 1996
; Geber et al., 1999
). Several studies have reviewed patterns in sexual dimorphism and sex allocation in dioecious and gynodioecious species (reviewed in Geber et al., 1999
; Shykoff et al., 2003
). Some studies have used a phylogenetic approach (e.g., Miller and Venable, 2003
), although comparative approaches among dimorphic populations of the same species differing in sex ratio (e.g., Ashman, 1999
; Delph and Carroll, 2001
; Vaughton and Ramsey, 2002
; Ramula and Mutikainen, 2003
; Case and Barrett, 2004
) or among closely related dimorphic taxa (e.g., Costich and Meagher, 2001
) are more common. In most of these studies, the evolution of dimorphism is in the context of biotic pollination, and several of the theories developed for sex-specific patterns of selection (e.g., those related to attraction of pollinators) are most relevant for biotically pollinated plants (Delph et al., 1996
; Costich and Meagher, 2001
).
In several cases, the evolution of dioecy has been associated with shifts to wind pollination, and selection presumably has favored traits typically associated with wind pollination, including greater pollen production with smaller pollen grains, fewer ovules, smaller but more numerous flowers, more exposed but highly condensed inflorescences, flowers with more exposed anthers on longer filaments or longer stigmas, and a decrease in structures associated with biotic pollination, such as nectaries and associated nectar production (Lloyd and Webb, 1977
; Delph et al., 1996
; Geber et al., 1999
; reviewed in Culley et al., 2002
). Relatively few studies have examined these changes between biotic and abiotic wind pollination in a comparative context (Steven and Waller, 2004
; reviewed in Culley et al., 2002
).
The endemic Hawaiian genus Schiedea (Caryophyllaceae) provides an opportunity to study how floral traits change with the simultaneous evolution of sexually dimorphic breeding systems and wind pollination. This monophyletic genus has 34 species, 10 of which are sexually dimorphic, with gynodioecious species with different frequencies of females, subdioecious species (female, male, and a few hermaphroditic individuals in a population), and dioecious species. Control of male sterility is nuclear (Weller and Sakai, 1991
). Species occurring in wet or mesic forest are hermaphroditic vines, herbs, or shrubs with selfing or biotic pollination, while species occupying dry habitats are typically woody shrubs, dimorphic, and often wind pollinated (Wagner et al., 2005
). Phylogenetic analyses suggest that dimorphic species of Schiedea evolved from a mesic forest, biotically pollinated hermaphroditic ancestor (Wagner et al., 1995
; Weller et al., 1995
; Soltis et al., 1996
; Sakai et al., 1997b
; Nepokroeff et al., 2005
). The clade that includes all but one of the dimorphic species has reasonable bootstrap support (77%), but relationships within the clade of 12 species remain unresolved (<50% bootstrap support in all but one node; Nepokroeff et al., 2005
), allowing a comparative approach but making a phylogenetic analysis of these traits premature.
Several factors may have affected the evolution of floral morphology in this genus. The shift from mesic to dry habitat may have resulted in a loss of pollinators and pollinator limitation (Weller and Sakai, 1990
; Weller et al., 1998
) and may explain the increased selfing rates and consequent expression of inbreeding depression in species occurring in dry habitats (Sakai et al., 1989
, 1997a
, b
; Weller and Sakai, 1990
, 2005
; Rankin et al., 2002
). In these species, females may be favored because their progeny are outcrossed and have higher fitness than inbred progeny of hermaphroditic individuals (Sakai et al., 1989
, 1997a
; Rankin et al., 2002
; Weller and Sakai, 2005
). The presence of females in a population is likely to favor the evolution of increased male function in hermaphroditic individuals (Charlesworth, 1999
). At the same time, selection for wind as a pollination vector may have allowed greater outcrossing and better pollination of females and thus led to an increase in the frequency of females in the population (Sakai et al., 1997a
; Weller et al., 1998
). Higher frequencies of females may be associated not only with greater sexual dimorphism, but also with more pronounced adaptations to wind pollination. In this study, we examined how floral morphology has changed with the evolution of dimorphic breeding systems and wind pollination. We addressed the following questions:
(1) How do dimorphic species differ from hermaphroditic species in traits related to sex allocation and wind pollination, and within dimorphic species, do these traits show sexual dimorphism consistent with selection for sex allocation and/or wind pollination? If phenotypic trade-offs in sex allocation are important, relative to hermaphroditic species, males and hermaphrodites of dimorphic species should have larger allocation to male function and reduced female function, and females should have larger ovaries and reduced stamens. Within dimorphic species, males and hermaphrodites of dimorphic species (referred to subsequently as MH) should have larger allocation to male function and reduced female function, relative to females. If wind pollination is a strong selective force, then relative to hermaphroditic species, dimorphic species should have reduced nectaries and nectar production in both sexes of dimorphic species, and more highly condensed inflorescences with smaller but more numerous flowers. Predictions for the effects of wind pollination on inflorescence structure and allocation patterns between the sexes of sexually dimorphic species are more difficult, but might involve the effects of flower size on inflorescence condensation. Wind pollination should also favor females with longer stigmas, and MH should have more exposed anthers on longer filaments, with more and smaller pollen grains relative to females or to plants of hermaphroditic species.
(2) Within dimorphic species, is there a correlation between the frequency of females in dimorphic species and the amount of sexual dimorphism? In insect-pollinated plants, the sex ratio of populations may affect the degree of pollen limitation of females and strength of selection on floral traits (Ashman and Diefenderfer, 2001
). In the initial stages of the evolution of wind pollination, pollen limitation for females may be a significant factor if insect pollinators are lost with a shift in habitat. Under these conditions, there may be strong selection for traits associated with wind pollination in dimorphic species. Depending on how rapidly these traits respond to selection, the frequency of females may reflect the relative stage of the population in the evolution of dimorphism and wind pollination, and species with higher frequencies of females may have relatively greater sexual dimorphism in allocation and/or greater adaptations to wind pollination.
(3) What are the phenotypic correlations between traits that might reflect underlying genetic constraints on the evolution of floral traits? The response to selection on floral traits is likely to reflect both direct selection as well as indirect selection on correlated traits. Negative correlations between male and female traits (e.g., pollen number and ovule number) may enhance differences in sex allocation, but positive correlations of homologous traits between the sexes (e.g., ovule number in females and in hermaphrodites) may constrain differentiation. If phenotypic correlations reflect underlying genetic correlations, then these phenotypic correlations may give a general indication of how easily changes in breeding system could evolve (Lande, 1980
; Meagher, 1992
, 1994
).
MATERIALS AND METHODS
All plants used in analyses were genetically distinct individuals grown in the greenhouse from field-collected seeds or cuttings. This study included 21 species, 9 of 10 sexually dimorphic species and 12 of 24 hermaphroditic species (Table 1). The frequency of females in natural populations was determined from field population surveys taken from 1985 to 1996 (Weller and Sakai, 1990
; unpublished data). The frequency of females in S. haleakalensis could not be determined because so few plants could be sexed in the field.
|
) and taking the mean of 10 hemacytometer counts under a compound microscope. Total ovary volume was calculated as a sphere with a radius that was the mean diameter of the length and width of the ovary. Anther volume was calculated as a cylinder. Total anther volume was calculated as the product of anther volume and number of anthers per flower. Total reproductive volume was calculated as the sum of total anther volume and total ovary volume, and total anther volume/ovary volume was used as the anther : ovary ratio.
|
Inflorescence lengths and number of flowers per inflorescence were based on inflorescences collected in the field. The condensation index was calculated as the ratio of number of flowers/length of the main axis of the inflorescence (Weller et al., 1998
). Data on nectar production and field data for inflorescence length, number of flowers, and inflorescence condensation at the species level were previously reported in Weller et al. (1998)
but not analyzed with respect to sex in dimorphic species.
Data analyses
Hermaphroditic species were compared separately to females and to males and/or hermaphrodites of dimorphic species (MH) using t tests, with the number of species in each breeding system as the sample size. The MH of dimorphic species included hermaphrodites of species with gynodioecious systems, males and hermaphrodites pooled into one sex for species with subdioecious breeding systems (S. globosa and S. kealiae), and males of species with dioecious breeding systems. Males and hermaphrodites of the two subdioecious species were pooled because it was difficult to distinguish between these two sexes in subdioecious species based on morphology without tests of sexual function, and in at least one subdioecious species (S. globosa), some plants switch between male and hermaphroditic function through time. Individual plant means were used to calculate the mean for each sex of each species, and the species means for each sex were used as the samples in these analyses. Each species was represented by only one population, with the exception of S. hookeri and S. salicaria, which had two populations each. In these species, the average of the means of two populations for each species was used as the species mean.
In a second analysis including only dimorphic species, the amount of dimorphism between the sexes was compared in a two-way ANOVA with post-hoc Tukey tests, with sex (female or MH) and species both considered as fixed effects and with plant means for floral traits used as cases (Proc GLM; SAS Institute, 2001
). For S. salicaria, plants from two populations were pooled together. Traits included sepal length, anther and filament length, stigma length, ovary size, ovule number and size, anther : ovary volume, nectary length, and the inflorescence length, number of flowers per inflorescence, and condensation index.
To address the question of how the sexes diverged from one another with increasing frequencies of females, Pearson product-moment correlation coefficients were used to measure the correlation between the frequency of females in populations on both absolute values as well as the standardized differences between the sexes in dimorphic species. The standardized difference was used because of changes in flower size among species and was calculated as the difference of the MH value minus the female value for that trait (e.g., sepal length) divided by the MH value of the trait for each species. For most morphological traits, the sample size was nine dimorphic species, although nectar analyses had fewer species. Phenotypic correlations between traits were also calculated using Pearson product-moment correlation coefficients for specific pairs of traits within and between species in hermaphroditic and dimorphic species.
RESULTS
Dimorphic vs. hermaphroditic species
Allocation in males and hermaphrodites (MH) of dimorphic species vs. hermaphroditic species
Although somewhat smaller, the total reproductive volume of females and MH of dimorphic species was similar to that of hermaphroditic species (Table 2). Males and hermaphrodites of dimorphic species had a far greater proportion of reproductive volume in anthers (anther : ovary ratio = 0.54) than hermaphroditic species (0.22), consistent with greater specialization in male function in the dimorphic species. Males and hermaphrodites of dimorphic species produced 1.46 times as much pollen as hermaphroditic species, although this difference was not statistically significant (Table 2). For female traits, allocation in MH of dimorphic species was similar to hermaphroditic species in ovary volume, ovule number, and width (Table 2) and did not show the predicted reduction in size for female traits.
|
Traits potentially related to wind pollination
Both sexes of dimorphic species had much smaller flowers (shorter sepal length) than hermaphroditic species, and these smaller flowers were borne on more highly condensed inflorescences (higher condensation indices) than in hermaphroditic species (Table 2). These higher condensation indices occurred because the inflorescence length of dimorphic species was only about one-quarter that of hermaphroditic species, while having the same number of flowers per inflorescence. Differences in floral traits were consistent with evolution of wind pollination in dimorphic species. The MH of dimorphic species and hermaphroditic species had similar filament lengths, but because MH have shorter sepals, the anthers are more exserted, a situation that should favor better wind dispersal of pollen. The vestigial filaments of females were only 15% the length for hermaphroditic species (Table 2). Stigmas of females were longer than those of MH, but hermaphroditic species had the largest flowers and the longest stigmas (Tables 2, 3). Total nectary length was greatly reduced in each sex of dimorphic species relative to hermaphroditic species, but hermaphroditic species and each sex of dimorphic species were similar in 24-h nectar volume, in part because of the variability in nectar production of hermaphroditic species (Table 2).
|
Allocation patterns and sexual dimorphism
Total reproductive volume varied by species, but did not differ between females and MH of dimorphic species. In contrast, there was a significant interaction between sex and species for anther volume (Table 3; Fig. 2B), while ovary volume showed significant differences among species but with no evidence of sexual dimorphism (Table 3; Fig. 3C). As a consequence, the anther : ovary ratio depended on both sex and species (significant interaction of species and sex; Fig. 2D), indicating that the amount of sexual dimorphism depended upon the species. For example, in S. adamantis, a gynodioecious species with 39% females, hermaphrodites and females differed greatly in anther : ovary ratio (0.31 and 0.11, respectively). In dioecious S. ligustrina, the differences were even more pronounced, with a ratio of 0.98 for males and 0.10 for females. The female traits of ovary volume, ovule number, and ovule size (width) showed no evidence of sexual dimorphism, although there were significant differences among species (Table 3, Fig. 3).
|
|
|
Using a standardized measure of dimorphism that controls for the relative size of flowers between species [for a given trait, (MH – F)/MH], the frequency of females in the populations was not significantly correlated with the amount of dimorphism for any floral or inflorescence trait measured.
Correlations between traits
Correlations between male and female traits within species
Within hermaphroditic species, correlations between numbers of ovules and numbers of pollen grains and between ovary volume and anther volume were both positive (r = 0.81, P = 0.002 for the former, r = 0.66, P = 0.02 for the latter) rather than negative as predicted from sex allocation models. In dimorphic species, phenotypic correlations between ovary volume and anther volume were not significant within either females or MHs.
Correlations associated with wind pollination
If wind pollination is a strong selective factor, then in females, greater stigma length may be associated with shorter sepals, shorter nectary length, and fewer ovules, and greater number of flowers may be correlated with shorter inflorescences. Instead, we found that stigma length had a significant positive phenotypic correlation with sepal length in females (r = 0.78, P = 0.01) and in MH (r = 0.82, P = 0.007) and with nectary length (r = 0.70, P = 0.04). Sepal length was correlated with nectary length in both females (r = 0.77, P = 0.01) and MH (r = 0.78, P = 0.01). There was no correlation between stigma length and ovule number in females, or inflorescence length and number of flowers, or ovule number and ovule size in females or in MH. In MH of dimorphic species, filament length was correlated with sepal length (r = 0.68, P = 0.04). Filament length showed no significant correlation with pollen number, anther volume, or nectary length, and anther volume was not correlated with pollen number in MH.
There were no apparent packing constraints reflected in the phenotypic correlations. Species with smaller flowers did not have more condensed inflorescences; phenotypic correlations between the condensation index and sepal length (for hermaphroditic species, MH, or females) or filament length (for females or MH) were not significant. In hermaphroditic species, the condensation index and filament length were positively correlated (r = 0.80, P = 0.02).
Correlations of homologous structures between females and MH of dimorphic species
Sepal length had a significant phenotypic correlation between females and MH of dimorphic species (r = 0.93, P = 0.0003). Most other traits also had significant positive phenotypic correlations between homologous traits in females and MH of dimorphic species (ovary volume, r = 0.96, P < 0.0001; N ovules, r = 0.94, P = 0.0002; ovule width, r = 0.75, P = 0.05 with marginal significance; stigma length, r = 0.83, P = 0.006; total nectary length, r = 0.94, P = 0.0001; inflorescence length, r = 0.93, P = 0.002; N flowers, r = 0.93, P = 0.002). Neither anther volume nor filament length was significantly correlated between females and MH.
DISCUSSION
Trade-offs in sex allocation have been viewed widely as important factors promoting the evolution of breeding systems. In Schiedea, we find little evidence for such trade-offs. Instead, the evolution of wind pollination appears to have a more important role in the evolution of floral traits in dimorphic species. In this study, sexual dimorphism in the traits analyzed is related to wind pollination, but not to variation in female frequency.
Trade-offs between male and female allocation
There is little evidence that introduction of male sterility in populations of Schiedea results in major phenotypic trade-offs in sex allocation, when measured in terms of floral volume of ovaries and anthers or numbers of pollen grains and ovules. Although females have vestigial stamens and allocate very little to these structures, there is no compensatory gain relative to hermaphroditic species or MH of dimorphic species in any measured traits associated with female function at the individual flower level, including ovary volume, number, or width. The greater variation in investment in stamens relative to female investment is similar to patterns found in several hermaphroditic species (e.g., Mazer and Hultgard, 1993
; Parachnowitsch and Elle, 2004
). We also found no trade-offs in ovule number and ovule size within females or MH or between pollen number and anther size in MH.
To the extent that the positive phenotypic correlations between the sexes in ovary volume as well as ovule number reflect underlying genetic correlations, it may be difficult for the sexes to diverge in these traits (Lande, 1980
; Reeve and Fairbairn, 2001
; Ashman, 2003
; Delph et al., 2004
) and sex allocation trade-offs within a flower are not a driving force in further differentiation of the floral morphology of dimorphic species.
These general results are consistent with more detailed studies of two gynodioecious species with contrasting female frequencies, S. salicaria (13% females) and S. adamantis (39% females). In both these species, females and hermaphrodites have similar female investment at the flower level (ovule number, ovary volume), although differences occur in allocation to fruit production. In S. salicaria, females and hermaphrodites have similar seed production (Weller and Sakai, 2005
), while in S. adamantis, females have much greater seed production relative to hermaphrodites because females have a greater number of viable seeds per capsule and a higher proportion of flowers forming fruits (Sakai et al., 1997a
). With male sterility, MH in dimorphic species of Schiedea continue to produce the same ovule number and ovary volume, and in females, the reduction in resources devoted to stamen production are not reallocated to female function within the flower, although changes do occur at the fruit level. In a meta-analysis of gynodioecious species, Shykoff et al. (2003)
found that in general, females produced more fruits rather than more seeds per fruit relative to hermaphrodites.
Floral size
In a meta-analysis of gynodioecious species, Shykoff et al. (2003)
also found a consistent pattern of females with smaller flowers than hermaphrodites. In Schiedea, flowers of dimorphic species are also smaller (shorter sepal length) than those of hermaphroditic species, and flowers of females are smaller than those of MH flowers in dimorphic species. Smaller flower size in females may reflect the effect of the male sterility mutation and may be unrelated to selection for wind pollination. In dimorphic species, sepal length has a positive phenotypic correlation with nectary length, stamen filament length, and stigma length. The size of nectaries, filaments, and stigmas may therefore result indirectly from correlations between traits affected by the male sterility mutation. Schultz (2003)
also found that sepal length and stamen size were correlated in Sidalcea hirtipes, a gynodioecious species. Alternatively, divergence in flower size between females and MHs in Schiedea may have resulted from selection for size differences related to wind pollination that counter the strong positive intersex correlations for flower size in dimorphic species (assuming that positive phenotypic correlations indicate the occurrence of underlying genetic correlations; e.g., Ashman, 2003
). Smaller nectaries in dimorphic flowers may also reflect direct selection for reduced nectary size with wind pollination. We did not find a significant negative correlation of longer filament length in MH with shorter sepals that might be expected to facilitate pollen dispersal in wind-pollinated plants (Faegri and van der Pijl, 1979
; Bertin, 1989
; Dafni, 1992
). Further studies of genetic correlations between these traits may be helpful in understanding the role of pleiotropy in the evolution of sexual dimorphism (Delph et al., 1996
; Schultz, 2003
; reviewed in Eckhart, 1999
).
Flower size tends to become smaller with increasing female frequency for both sexes of dimorphic species. None of these correlations is significant when only dimorphic species are included, but with the inclusion of hermaphroditic species (with a female frequency of zero) the pattern of decreasing size with increasing frequency of females is apparent (Fig. 4A). In wind-pollinated plants, sepals may become smaller because unimpeded dispersal and receipt of pollen into the wind is critical and because large perianths are not critical to attract biotic pollinators (Faegri and van der Pijl, 1979
; reviewed in Bertin, 1989
). Although differences between females and MH of dimorphic species differed in these traits, increasing frequency of females in Schiedea was not associated with increasing sexual dimorphism in flower size.
Although phylogenetic analyses of these traits awaits a more resolved phylogeny, it is most parsimonious to assume that the size dimorphism in flowers of dimorphic Schiedea is a result of the evolution of relatively smaller flower size in females, rather than evolution of larger size in male- hermaphrodites. These results are consistent with other studies of gynodioecious species that show a strong pattern of females with smaller flowers than hermaphrodites (Bell, 1985
; Schultz, 2003
; Shykoff et al., 2003
), although studies of monoecious and dioecious species show that the opposite pattern can also occur for some wind-pollinated species (Delph et al., 1996
; Eckhart, 1999
). In Schiedea, subdioecious and dioecious species generally show patterns similar to gynodioecious species.
Traits potentially related to wind pollination
Although differences between hermaphroditic and sexually dimorphic species are not pronounced at the flower level, there are several ways that traits affecting wind pollination vary among sexually dimorphic species. These differences are especially evident for male traits. For both filament length and anther volume, the amount of sexual dimorphism depends upon the species. The amount of pollen is related to both anther volume and pollen size. Some species, such as S. globosa, produce large quantities of pollen because although the pollen grains are large, the anther volume is also large (Weller et al., 1998
). In contrast, S. ligustrina has substantial pollen production because although it has the smallest anthers of any sexually dimorphic species of Schiedea, it also has the smallest pollen grains.
Stigma length differences are also likely to be associated with the evolution of wind pollination. Within dimorphic species, the stigmas of female flowers are significantly longer than those of MH, but the amount of sexual dimorphism is not correlated with the frequency of females. The lack of correlation between the sexes for this trait in dimorphic species suggests that stigma length may not be constrained by underlying positive genetic correlations between the sexes.
Inflorescence architecture
Selection for wind pollination also has affected inflorescence architecture, and there is some evidence for sexual dimorphism in inflorescence structure. Hermaphroditic species have similar numbers of flowers per inflorescence but longer inflorescences with less inflorescence condensation than dimorphic species (Weller et al., 1998
). Within dimorphic species, our study shows that females and MH of dimorphic species do not differ in inflorescence length or inflorescence condensation, but both flower size and number of flowers are sexually dimorphic traits. Females have more flowers and smaller flowers than MH of dimorphic species, although we did not find a trade-off in flower size and number (i.e., no significant correlation of flower size and flower number) and the condensation index was not correlated with flower size. Delph et al. (2002)
also found sexual dimorphism in flower production, although in Silene latifolia, males had smaller flowers.
Niklas (1985)
has suggested that one way wind pollination can be enhanced is through the condensation of inflorescences. Flower size may affect pollen dispersal through its direct effect on individual flower architecture as well as an indirect effect on inflorescence structure (Freeman et al., 1997
). More condensed inflorescences may enhance pollen dispersal and receipt, but larger flowers may constrain the extent of inflorescence condensation. Our condensation index is a simple measure (flower number/length of the main axis), and other changes in the secondary and tertiary branching of the inflorescence may need to be considered to examine changes in inflorescence architecture with wind pollination. Quantitative genetic studies of inflorescence architecture of S. salicaria (S. Weller et al., unpublished manuscript) suggest that females and hermaphrodites may respond differently to selection for wind pollination.
Relationship of frequency of females and sexual dimorphism
Many of the traits measured show sexual dimorphism, and there is a consistent trend for smaller flowers and flower parts in both sexes with greater female frequencies. When the difference between the sexes within a species is standardized for flower size, the amount of dimorphism in these traits is not correlated with the frequency of females across species. Several factors may affect female frequency, including the underlying genetics of male sterility, as well as environmental factors that influence the relative seed fitness of females and hermaphrodites (Delph and Carroll, 2001
). In Schiedea, a system with nuclear inheritance of male sterility, the frequency of females may not correspond with the amount of sexual dimorphism in species because of many factors. For example, it may be difficult to estimate female frequency through space or time, phenotypic measures of traits may not reflect underlying genetic relationships because of variation in resources or other factors, genetic correlations may constrain responses to selection, or genetic responses to factors such as pollen availability and inbreeding depression may show lag times.
Conclusions
With the exception of four presumably bird-pollinated species formerly placed in Alsinidendron, flowers of Schiedea are very similar across the 34 species in the genus (Wagner et al., 2005
). Dimorphic and hermaphroditic species differ in several floral and inflorescence traits consistent with wind pollination, but there is little indication of trade-offs in allocation to male and female function consistent with sex allocation models. Females of dimorphic species have vestigial stamens, but there is no evidence of greater investment in female floral parts related to loss of male function. Several changes in dimorphic species are consistent with wind pollination (smaller flower size, greater inflorescence condensation, longer stamen filaments in MH than females of dimorphic species, and longer stigmas in females relative to MH of dimorphic species). Species with low frequencies of females (e.g., S. salicaria) have less sexual dimorphism than species with higher frequencies, but female frequency in general is not a good indicator of sexual dimorphism. How rapidly specialization in separate sexes will occur depends not only upon the strength of selection and heritabilities of the traits, but also on genetic correlations among traits. Further work examining the heritabilities and genetic correlations of these traits may reveal the genetic potential of these species to change with wind pollination or other selective factors.
FOOTNOTES
1 The authors thank Em Arpawong for help with floral measurements and reviewers for constructive criticism. This work was supported by funds from an NSF Research Experience for Undergraduates supplement for A.M.G. and Em Arpawong and National Science Foundation grants (DEB-9207724 and DEB 9815878) to A.K.S. and S.G.W. 
2 Current address: Department of Biology, Duke University, P.O. Box 90338, Durham, North Carolina 27708-0338 USA
3 Author for correspondence (e-mail: aksakai{at}uci.edu
)
LITERATURE CITED
Ashman T. L. 1999 Determinants of sex allocation in a gynodioecious wild strawberry: implications for the evolution of dioecy and sexual dimorphism. Journal of Evolutionary Biology 12: 648-661[CrossRef][ISI]
Ashman T. L. 2003 Constraints on the evolution of males and sexual dimorphism: field estimates of genetic architecture of reproductive traits in three populations of gynodioecious Fragaria virginiana. Evolution 57: 2012-2025[CrossRef][ISI][Medline]
Ashman T. L. C. Diefenderfer 2001 Sex ratio represents a unique context for selection on attractive traits: consequences for the evolution of sexual dimorphism. American Naturalist 157: 334-347[CrossRef][ISI]
Bell G. 1985 On the function of flowers. Proceedings of the Royal Society of London, B 224: 223-265
Bertin R. I. 1989 Pollination biology. In W. G. Abrahamson [ed.], Plant– animal interactions, 23–86. McGraw-Hill, New York, New York, USA
Campbell D. R. 2000 Experimental tests of sex-allocation theory in plants. Trends in Ecology and Evolution 15: 227-232
Case A. S. C. H. Barrett 2004 Floral biology of gender monomorphism and dimorphism in Wurmbea dioica (Colchicaceae) in western Australia. International Journal of Plant Sciences 165: 289-301[CrossRef]
Charlesworth B. D. Charlesworth 1978 A model for the evolution of dioecy and gynodioecy. American Naturalist 112: 975-997[CrossRef][ISI]
Charlesworth D. 1999 Theories of the evolution of dioecy. In M. A. Geber, T. E. Dawson, and L. F. Delph [eds.], Gender and sexual dimorphism in flowering plants, 33–60. Springer-Verlag, Berlin, Germany
Costich D. E. T. R. Meagher 2001 Impacts of floral gender and whole-plant gender on floral evolution in Echballium elaterium (Cucurbitaceae). Biological Journal of the Linnean Society 74: 475-487
Culley T. M. S. G. Weller A. K. Sakai 2002 The evolution of wind pollination in angiosperms. Trends in Ecology and Evolution 17: 361-369, plus corrigendum [CrossRef]
Dafni A. 1992 Pollination ecology: a practical approach. Oxford University Press, Oxford, UK
Delph L. F. S. B. Carroll 2001 Factors affecting relative seed fitness and female frequency in a gynodioecious species, Silene acaulis. Evolutionary Ecology Research 3: 487-505[ISI]
Delph L. F. L. F. Galloway M. L. Stanton 1996 Sexual dimorphism in flower size. American Naturalist 148: 299-320[CrossRef][ISI]
Delph L. F. J. L. Gehring F. M. Frey A. M. Arntz M. Levri 2004 Genetic constraints on floral evolution in a sexually dimorphic plant revealed by artificial selection. Evolution 58: 1936-1946[CrossRef][ISI][Medline]
Delph L. F. F. N. Knapczyk D. R. Taylor 2002 Among-population variation and correlations in sexually dimorphic traits of Silene latifolia. Journal of Evolutionary Biology 15: 1011-1020[CrossRef][ISI]
Eckhart V. M. 1999 Sexual dimorphism in flowers and inflorescences. In M. A. Geber, T. E. Dawson, and L. F. Delph [eds.], Gender and sexual dimorphism in flowering plants, 123–148. Springer-Verlag, Berlin, Germany
Faegri K. L. van der Pijl 1979 The principles of pollination ecology. Pergamon Press, New York, New York, USA
Freeman D. C. J. L. Doust A. El-Keblawy K. J. Miglia E. D. McArthur 1997 Sexual specialization and inbreeding avoidance in the evolution of dioecy. Botanical Review 63: 66-92
Geber M. A. T. E. Dawson L. F. Delph [eds.] 1999 Gender and sexual dimorphism in flowering plants. Springer-Verlag, Berlin, Germany
Kearns C. A. D. W. Inouye 1993 Techniques for pollination biologists. University of Colorado Press, Niwot, Colorado, USA
Lande R. 1980 Sexual dimorphism, sexual selection, and adaptation in polygenic characters. Evolution 34: 292-305[CrossRef][ISI]
Lloyd D. G. 1975 The maintenance of gynodioecy and androdioecy in angiosperms. Genetica 45: 325-339[CrossRef][ISI]
Lloyd D. G. 1979 Evolution towards dioecy in heterostylous populations. Plant Systematics and Evolution 131: 71-80[CrossRef][ISI]
Lloyd D. G. C. J. Webb 1977 Secondary sex characters in plants. Botanical Review 43: 177-216
Mazer S. J. U.-M. Hultgard 1993 Variation and covariation among floral traits within and among four species of northern European Primula (Primulaceae). American Journal of Botany 80: 474-485[CrossRef][ISI]
Meagher T. R. 1992 The quantitative genetics of sexual dimorphism in Silene latifolia (Caryophyllaceae). I. Genetic variation. Evolution 46: 445-457[CrossRef][ISI]
Meagher T. R. 1994 The quantitative genetics of sexual dimorphism in Silene latifolia (Caryophyllaceae) II. Response to sex-specific selection. Evolution 48: 939-951[CrossRef][ISI]
Miller J. S. D. L. Venable 2003 Floral morphometrics and the evolution of sexual dimorphism in Lycium (Solanaceae). Evolution 57: 74-86[ISI][Medline]
Nepokroeff M. W. L. Wagner P. S. Soltis S. G. Weller D. E. Soltis A. K. Sakai E. A. Zimmer 2005 Phylogeny of Schiedea. In W. L. Wagner, S. G. Weller, and A. K. Sakai. Monograph of Schiedea (Caryophyllaceae-Alsinoideae), 13–20 Systematic Botany Monographs 72: 1-169
Niklas K. J. 1985 The aerodynamics of wind pollination. Botanical Review 51: 328-386[ISI]
Parachnowitsch A. L. E. Elle 2004 Variation in sex allocation and male–female trade-offs in six populations of Collinsia parviflora (Scrophulariaceae s.l). American Journal of Botany 91: 1200-1207
Ramula S. P. Mutikainen 2003 Sex allocation of females and hermaphrodites in the gynodioecious Geranium sylvaticum. Annals of Botany 92: 207-213
Rankin A. E. S. G. Weller A. K. Sakai 2002 Mating system instability in Schiedea menziesii (Caryophyllaceae). Evolution 56: 1574-1585[CrossRef][ISI][Medline]
Reeve J. P. D. J. Fairbairn 2001 Predicting the evolution of sexual size dimorphism. Journal of Evolutionary Biology 14: 244-254
Renner S. S. R. E. Ricklefs 1995 Dioecy and its correlates in the flowering plants. American Journal of Botany 82: 596-606[CrossRef][ISI]
Sakai A. K. K. Karoly S. G. Weller 1989 Inbreeding depression in Schiedea globosa and S. salicaria (Caryophyllaceae), subdioecious and gynodioecious Hawaiian species. American Journal of Botany 76: 437-444[CrossRef][ISI]
Sakai A. K. S. G. Weller M.-L. Chen S.-Y. Chou C. Tasanont 1997a Evolution of gynodioecy and maintenance of females: the role of inbreeding depression, outcrossing rates, and resource allocation in Schiedea adamantis (Caryophyllaceae). Evolution 51: 724-736[CrossRef][ISI]
Sakai A. K. S. G. Weller W. L. Wagner P. S. Soltis D. E. Soltis 1997b Phylogenetic perspectives on the evolution of dioecy: adaptive radiation in the endemic Hawaiian genera Schiedea and Alsinidendron (Caryophyllaceae: Alsinoideae). In T. Givnish and K. Sytsma [eds.], Molecular evolution and adaptive radiation, 455–474. Cambridge University Press, New York, New York, USA
SAS Institute. 2001 SAS version 8.02. SAS Institute, Cary, North Carolina, USA
Schultz S. T. 2003 Sexual dimorphism in gynodioecious Sidalcea hirtipes (Malvaceae). II. Floral traits. International Journal of Plant Sciences 164: 175-180[CrossRef]
Shykoff J. A. S.-O. Kolokotronis C. L. Collin M. Lopez-Villavicencio 2003 Effects of male sterility on reproductive traits in gynodioecious plants: a meta-analysis. Oecologia 135: 1-9[ISI][Medline]
Soltis P. E. D. E. Soltis S. G. Weller A. K. Sakai W. L. Wagner 1996 Molecular phylogenetic analysis of the Hawaiian endemics Schiedea and Alsinidendron (Caryophyllaceae). Systematic Botany 21: 365-379[CrossRef][ISI]
Steven J. C. D. M. Waller 2004 Reproductive alternatives to insect pollination in four species of Thalictrum (Ranunculaceae). Plant Species Biology 19: 73-80
Thomson J. D. J. Brunet 1990 Hypotheses for the evolution of dioecy in seed plants. Trends in Ecology and Evolution 5: 11-16
Vaughton G. M. Ramsey 2002 Evidence of gynodioecy and sex ratio variation in Wurmbea biglandulosa (Colchicaceae). Plant Systematics and Evolution 232: 167-179[CrossRef][ISI]
Wagner W. L. S. G. Weller A. K. Sakai 1995 Phylogeny and biogeography in Schiedea and Alsinidendron (Caryophyllaceae). In W. L. Wagner and V. A. Funk [eds.], Hawaiian biogeography: evolution on a hot spot archipelago, 221–258. Smithsonian Press, Washington, D.C., USA
Wagner W. L. S. G. Weller A. K. Sakai 2005 Monograph of Schiedea (Caryophyllaceae-Alsinoideae). Systematic Botany Monographs 72: 1-169
Weller S. G. A. K. Sakai 1990 The evolution of dicliny in Schiedea (Caryophyllaceae), an endemic Hawaiian genus. Plant Species Biology 5: 83-95[CrossRef]
Weller S. G. A. K. Sakai 1991 The genetic basis of male sterility in Schiedea (Caryophyllaceae), an endemic Hawaiian genus. Heredity 67: 265-273[ISI]
Weller S. G. A. K. Sakai 2005 Selfing and resource allocation in Schiedea salicaria (Caryophyllaceae), a gynodioecious species. Journal of Evolutionary Biology 18: 301-308[CrossRef][ISI][Medline]
Weller S. G. A. K. Sakai A. E. Rankin A. Golonka B. Kutcher K. E. Ashby 1998 Dioecy and the evolution of pollination systems in Schiedea and Alsinidendron (Caryophyllaceae: Alsinoideae) in the Hawaiian Islands. American Journal of Botany 85: 1377-1388
Weller S. G. W. L. Wagner A. K. Sakai 1995 A phylogenetic analysis of Schiedea and Alsinidendron (Caryophyllaceae: Alsinoideae): implications for the evolution of breeding systems. Systematic Botany 20: 315-337[CrossRef][ISI]
This article has been cited by other articles:
|
|
 |
|
  S. G. Weller, A. K. Sakai, T. M. Culley, D. R. Campbell, P. Ngo, and A. K. Dunbar-Wallis Sexually dimorphic inflorescence traits in a wind-pollinated species: heritabilities and genetic correlations in Schiedea adamantis (Caryophyllaceae) Am. J. Botany, October 1, 2007; 94(10): 1716 - 1725. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Greenway and L. D. Harder Variation in ovule and seed size and associated size-number trade-offs in angiosperms Am. J. Botany, May 1, 2007; 94(5): 840 - 846. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS |
