| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 ,3 Department of Biological Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania 15260 USA; and Pymatuning Laboratory of Ecology, Linesville, Pennsylvania 16424 USA
Received for publication December 1, 1998. Accepted for publication June 8, 1999.
ABSTRACT
In this study we dissect the causes of variation in intra-inflorescence allocation in a sexually polymorphic species, Fragaria virginiana. We separated out the effects of resource competition during flowering from those of inflorescence architecture, as well as identified the effects of sex morph and genotype. We found position-based variation in petal length, ovule, pollen, and flower number to be influenced more by architecture than by our resource manipulations during flowering. We also found both genotype- and sex-specific intra-inflorescence patterns. Furthermore, our data indicate that the sex morph-specific intra-inflorescence patterns result from architectural modifications of the basic pattern. In fact, sex-differential intra-inflorescence patterns suggest that fitness through male and female function may be maximized by different resource distribution patterns within the inflorescence and may have been modified by past selection. Specifically, females invested heavily in ovules at positions where fruit set was most likely (primary and secondary), at the expense of flower number and allocation per flower at more distal positions. Whereas functional males invested minimally in ovules at all flower positions and produced the most abundantly flowered inflorescences, hermaphrodites, on the other hand, showed intermediate patterns, implying a compromise between sex functions. We suggest that consideration of intra-inflorescence allocation and inflorescence architecture may reveal the mechanism underlying sexual dimorphism in flower allocation and number.
Key Words: floral variation • Fragaria virginiana • genotypic variation • gynodioecy • inflorescence architecture • polygamodioecy • resource competition • sex allocation • sex-specific architecture
While evolutionary biologists have been compelled to explain the extraordinary diversity of flowers among species and among plants within a species, they have focused considerably less attention on the diversity of flowers within inflorescences. However, it has become evident that flowers of an inflorescence are not created equal. Intra-inflorescence variation in reproductive function has long been apparent in plants with specialized breeding systems, such as monoecy, where male and female functions are borne on separate flowers. Recent evidence suggests that even within hermaphrodite flowers there is considerable variation within an inflorescence in investment in attractive traits (e.g., petals, nectar), biomass distribution, and reproductive potential (reviewed in Stephenson, 1981
; Diggle, 1995
). Specifically, there is often a distinct trend from basal (earliest produced) to distal (latest produced) positions along an inflorescence: distal flowers are typically smaller, produce fewer, smaller ovules, seeds and fruits, but may have higher stamen or pollen production (reviewed in Herrera, 1991
; Ashman, 1992
; Ashman and Baker, 1992
; Wolfe, 1992
; Diggle, 1993, 1995, 1997
; Brunet and Charlesworth, 1995
; Guitián and Navarro, 1996
) relative to basal flowers. Acknowledgment of this widespread, position-based variation among flowers of an inflorescence has led to the recent formulation of a number of alternative (but not mutually exclusive) hypotheses.
One hypothesis to explain intra-inflorescence variation suggests that position-dependent variation results from competition for resources between basally and distally developing flowers and/or fruits. That is, since basal flowers often open first, they may have a competitive advantage over distal flowers for a limited supply of resources, and thus lead to a reduction in the size of distal flowers or their components (reviewed in Stephenson, 1981
; Wyatt, 1982
; Lee, 1988
; Diggle, 1995
). Results from studies that have manipulated resource conditions within the inflorescence have supported this hypothesis (Casper, 1984
; Solomon, 1988
; Herrera, 1991
; Guitián and Navarro, 1996
). However, this is clearly not the only contributing factor. The observation that strong intra-inflorescence patterns are maintained regardless of resource status (Wolfe, 1992
; Diggle, 1995, 1997
) led Diggle (1995)
to propose that variation among sequentially produced flowers is the unavoidable developmental consequence of inflorescence architecture, and as such, can constrain the evolution of alternative allocation patterns. Architectural effects may stem from inherent mechanical limitations in vasculature and/or support (Schoen and Dubuc, 1990
; Diggle, 1995
) imposed by inflorescence design. To date relatively few studies have separated architectural from resource competition effects on intra-inflorescence patterns (reviewed in Diggle, 1995, 1997
).
Architectural patterns of intra-inflorescence allocation have also been viewed, not so much as constraints, but as characteristics that are subject to natural selection and, thus, may reflect adaptations to specific mating opportunities and/or resource conditions (Wyatt, 1982
; Lloyd and Bawa, 1984
; Frank, 1987
; Brunet and Charlesworth, 1995
). To date there have been few tests of the adaptive hypothesis (see Brunet and Charlesworth, 1995
; Brunet, 1996
), and we know little about how conserved inflorescence design is within taxa (Wyatt, 1982
). Evidence of genetic variation in intra-inflorescence patterns would indicate the opportunity for response to selection. Moreover, sex-specific architectural patterns of intra-inflorescence allocation in sexually polymorphic species would provide evidence that such patterns can be subject to alteration (Ashman, 1992
) and could reflect different selective optima. Specifically, morphs that gain fitness through female function (females and hermaphrodites) may be selected to invest in a greater number of ovules at positions where resources for seed production are likely to be most abundant (basal positions), whereas morphs that do not (or rarely) gain fitness through female function (males, or "functional males" sensu Lloyd, 1980
) may be under selection to minimize the investment in these nonfunctional organs throughout the inflorescence. On the other hand, if fulfilling male function is limited more by pollinator visits than resources (e.g., Bell, 1985
), then plants with male function may invest maximally in attraction, and/or pollen production across all flowers on the inflorescence, and/or produce more flowers per inflorescence position to maximize opportunities for pollen export, whereas females may be selected to reduce allocation to attraction over the inflorescence (Ashman, 1992
). Furthermore, if large investment in floral organs early in the inflorescence reduces the number of flowers per inflorescence (i.e., if there is a size-number trade-off), and if flower number impacts male reproductive success more than female (e.g., Broyles and Wyatt, 1995
; Emms, Stratton, and Snow, 1997
), then the sex morphs may be under selection for different flower (or floral organ) size and number combinations.
Here, we describe a study that dissects the different causes of variation in intra-inflorescence allocation in a sexually polymorphic species, Fragaria virginiana. We separate out the effects of resource competition during flowering from those of inflorescence architecture via two forms of resource manipulation within the inflorescence (pollination and bud removal). Moreover, to shed light on the potential for adaptation in intra-inflorescence allocation we determine whether there are sex morph- or genotype-specific intra-inflorescence patterns.
METHODS
Study species
Fragaria virginiana (Rosaceae), the Virginian wild strawberry, is a creeping stoloniferous perennial herb that is native to North America (Staudt, 1989
). It has a gynodioecious (or polygamodioecious) breeding system, where females coexist with hermaphrodites, but some hermaphrodites never produce fruit and thus may be considered "functional males" (Valleau, 1918
; Staudt, 1989
; Stahler, Luby, and Ascher, 1991
; Ashman, 1999
). The frequency of females varies among populations (Stahler, Luby, and Ascher, 1991
; Ashman, 1999
), and populations with high frequencies of females have the highest frequency of nonfruiting hermaphrodites (Ashman, 1999
). Work to date suggests that sex determination is under nuclear control with male sterility (femaleness) dominant to male fertility (Valleau, 1923
; Staudt, 1967
; Ahmadi and Bringhurst, 1991
), but also indicates that modifying genes and the environment play a role in the regulation of sexual phenotype in Fragaria (Valleau, 1918, 1923
; Ahmadi and Bringhurst, 1991
; Ashman, 1999
).
Inflorescence design varies within Fragaria and closely related genera (e.g., Potentilla, Alchemilla [Morgan, Soltis, and Robertson, 1994
]) from single flowers to compound dichasia (Gleason and Cronquist, 1991
; Hitchens, personal observation). In particular, the inflorescence of Fragaria virginiana has been described as a corymb (Gleason and Cronquist, 1991
), or a di- or pleiochasial cyme (Valleau, 1918
), and there is some anecdotal evidence of intra-populational variation (Valleau, 1918
). The populations of wild strawberry we studied have flowers borne on di- or pleiochasial cymes (Fig. 1A, B). Within these inflorescences flowers at the primary position open first and are basal, whereas flowers at the quaternary position open last and are distal (Valleau, 1918
; Fig. 1C). Usually one flower per inflorescence is open at a time and individual flowers last about a day, so the number of flowers produced is correlated with the duration of flowering (r = 0.54; P < 0.0001; N = 153). Pollinators include small generalist bees, flies, and ants (Ashman, in press). Flowers of all sex morphs contain 20–30 stamens and a fleshy receptacle that supports numerous uniovulate pistils. Stamens of females are vestigial, white, and devoid of pollen (Valleau, 1918, 1923
; Ashman, personal observation).
|
for full descriptions). Female frequencies in these wild populations reflect the average (31%) of those studied in northwestern Pennsylvania (Ashman, 1999
). For this study, pollen-producing genotypes were classified as hermaphrodites or functional males based on their propensity to set fruit in previous experiments. Those that previously exhibited fruit set above 20% (average fruit set was 41%) were designated hermaphrodites, whereas those with fruit set of <5% (average fruit set 1%) were designated functional males. Using this scheme, we classify 40% as hermaphrodites and 30% as functional males. We randomly selected 21 genotypes (seven female, nine hermaphrodite, and five functional males) and attempted to produce eight clones per genotype by rooting plantlets (clones) into 440-cm3 pots containing FafardTM number 2 soil. Due to mortality, poor runnering, or nonflowering, the number of clones per genotype ranged from six to eight and averaged 7.6. During fall cultivation plants were subjected to natural day lengths, day/night temperatures of 18°/7°C, and moderate levels of fertilizer. Plants were chilled for 10 wk at 4°C, and then arranged in a growth chamber set for spring conditions (i.e., day/night photoperiod and temperature of 13 h/11 h and 18°/10°C). For the remainder of the experiment plants were supplied only water as needed.
In the greenhouse, we employed resource manipulations that reflect resource alterations typically experienced by wild strawberry inflorescences in nature: (1) pollination, (2) primary bud removal (in nature, this is affected by clipper weevils [Ashman, unpublished data]), and (3) nonpollination (as a baseline for architectural effects [see below]). Each plant within a given genotype was randomly assigned one of the three resource treatments before flower buds were visible. We hand-pollinated all flowers at each position on plants in the pollination treatment. Hand-pollination was conducted with a mixture of pollen from a set of several pollen donors that represented genotypes not used in the experiment. Pollen was stored at 4°C, and no pollen was stored for >3 d. Fruit set achieved by plants in the pollination treatment was characteristically high for females (~100%), moderate for hermaphrodites (~37%), and absent for functional males (0%). All flowers on plants in the nonpollination treatment were prevented from pollination, and in the case of male-fertile plants this involved emasculation just prior to anther dehiscence. At the first sign of bud formation, plants in the bud removal treatment had their first visible (but immature in size) bud removed from the inflorescence. All flowers on plants in this treatment were also prevented from pollination. We expected bud removal to increase resources available to the distal flowers relative to that available to distal flowers in the nonpollination treatment, and thus we expected a concomitant increase (or less of a decrease) in the size and/or number of distal flowers (or their organs) relative to the nonpollination treatment, whereas we expected the pollination treatment to reduce available resources to distal flowers, and thus we expected allocation to distal flowers to be lower than that observed in the nonpollination treatment. In addition, we reasoned that if fruit maturation represents a significant drain on resources, then this drain will vary with the different fruit-setting propensities of the sex morphs, and thus we expected the pollination treatment to have its greatest effect on females, with lesser effects on hermaphrodites and functional males. Such a result would be reflected in a significant flower position-by-sex-by-treatment effect in the ANOVA (see below).
Traits measured
For each plant we recorded number of flowers per position, and for a randomly chosen flower at each position we measured petal length and counted the number of ovules and pollen grains produced. Variation among positions is typically 1–4 times that within positions (Ashman and Hitchens, unpublished data). For hermaphrodite and functional male flowers, we collected anthers just before they dehisced and stored them in microcentrifuge tubes. Pollen samples were prepared by suspension and sonication in 10 mL of 1% NaCl, and then diluted to 10:1 prior to counting. The number of pollen grains in three 250 µl aliquots were counted on a Elzone particle counter (Elzone® 280PC; Norcross, Georgia, USA), and the average of these was used to calculate pollen production per flower. For all three sex morphs we measured petal length to the nearest 0.01 mm with electronic calipers on the second day of anthesis. Inflorescences were allowed to complete their flowering cycle fully before we preserved them in FAA. We counted ovule (or ovule and seed) number on preserved flowers with the aid of a dissecting microscope. We estimated first flower values for clones in the bud removal treatment as the mean of first flowers from other clones of that genotype. This approach provides a good estimate of first flowers because first flowers are not expected to be affected by the resource treatments employed and because there is a high clonal repeatability for the traits measured on the first flower (Ashman, 1999
). This approach is similar to that advocated by Diggle (1997)
.
Statistical analyses
We analyzed data on petal length, number of ovules per flower, number of pollen per flower, and number of flowers per position using mixed-model repeated-measures analyses of variance (PROC MIXED with RANDOM and REPEATED statements [SAS, 1989, 1996
; Littell et al., 1996
]). Treatment, sex morph, flower position, and their interactions were designated as fixed effects, whereas genotype (nested within sex) and its interactions with treatment and flower position were designated as random effects. In this analysis we included the estimated data for first flowers in the bud removal treatment. Additionally, we analyzed the data two other ways: (1) where the primary position was treated as missing data for the bud removal plants, and (2) where the primary position was eliminated for all plants (i.e., we only analyzed data from positions 2–4). We found the general pattern of results and our conclusions were the same regardless of the analytical approach, so we present the analyses, figures, and tables for the original data set for consistency and completeness.
Because the focus of this study was to determine the causes of variation in intra-inflorescence allocation, we were primarily interested in the effect of flower position and its interactions with other effects. In our analysis, a significant main effect of flower position indicates a significant overall position effect on floral allocation, whereas significant interactions of flower position with sex, treatment, or genotype suggest that these positional effects are not homogeneous across sex morphs, treatments, or genotypes. Specifically, a significant flower position-by-treatment effect indicates that treatment, and thus resource status during flowering, influences the pattern of floral allocation within the inflorescence. Likewise, a flower position-by-sex effect indicates that the sex morphs differ in their within-inflorescence patterns of allocation. Moreover, a flower position-by-genotype effect indicates that there is variation in intra-inflorescence allocation patterns among genotypes, and this may reflect genetic variation in positional pattern. Main effects of genotype, sex, treatment, and their interactions indicate that these factors influence average (over all positions) floral allocation. Following a significant effect, least square means of the dependent variable were calculated and compared, and significance levels per comparison were adjusted using the Dunn-Sidàk method (Sokal and Rohlf, 1981
). To further investigate the source of significant flower position-by-sex morph interactions, we subdivided the data based on three pairs of adjacent floral positions: (1) positions 1 and 2 only; (2) positions 2 and 3 only; and (3) positions 3 and 4 only. In this way we could determine where along the inflorescence the sex morphs were heterogeneous in their pattern of allocation.
In addition, to test specifically for architectural and resource effects we employed the comparisons advocated by Diggle (1997)
. To test for a resource effect we focused comparisons on distal flowers (at the quaternary position) and used ANOVA to compare distal flowers of the resource manipulation treatments (bud removal or pollination) to those on plants in the nonpollination treatment. To test for architectural effects (and sex-specific architectural effects) we performed a separate repeated-measures ANOVA on data from plants in the nonpollination treatment only (i.e., plants for which resource competition with developing fruit has been eliminated as a cause [Diggle, 1997
]). In this analysis a significant position effect reflects an architectural one, and a significant position-by-sex effect reflects a significant sex-specific architectural effect. When appropriate, we followed with comparisons between flowers at the primary position vs. flowers at the most distal position (quaternary). We also compared primary flowers to those at intervening positions (secondary, tertiary) so that we could detect architectural effects that were nonlinear.
RESULTS
The effect of resource manipulations on intra-inflorescence floral allocation
There were strong, highly significant (P < 0.0001) positional effects for all traits measured. Primary flowers produced more ovules and pollen and had larger petals than distal flowers (Fig. 2A–C). This was the case regardless of resource treatment (there was no significant flower position-by-treatment effect for any trait [Table 1]), suggesting that there was little effect of resource manipulation on position-based floral allocation (Table 1). Moreover, all comparisons of quaternary flowers in the nonpollination treatment to those in either the bud removal or pollination treatment were nonsignificant (Table 2A, Fig. 2A–C). The only trait to show an effect of treatment was number of flowers per position: plants in the bud removal treatment produced an average of 20% more flowers per position than plants in either of the other treatments (mean ± 1 SE flower per position: bud removal—2.06 + 0.12 flowers; nonpollination—1.72 + 0.10 flowers; pollination—1.74 + 0.10 flowers). This translated into nearly an extra flower per plant for plants in the bud removal treatment.
|
|
|
). Overall, traits of flowers at the primary position were significantly larger that those at the quaternary position and all intervening positions (Table 2B).
The effects of sexual identity on intra-inflorescence floral allocation
In addition to strong effects of sexual identity on average floral trait values (Table 3), there were significant sex-specific positional effects for ovule number, petal length, and flowers per position (all P < 0.01; Table 1, Fig. 2A–D). The decline in ovule number between the primary and secondary positions was greatest for females and hermaphrodites (22 and 20%, respectively), and lowest for functional males (13%) (P < 0.03). Declines in ovule number between secondary and tertiary flowers were similar among the sex morphs (all 17%; P > 0.30), whereas declines between tertiary and quaternary flowers were highest for hermaphrodites (30%) compared to either female or functional males (11 and 10%, respectively; P < 0.01). These patterns resulted in less distinctive sex-specific ovule numbers per flower in the later positions (tertiary and quaternary) than in the earlier positions (primary and secondary) (Fig. 2A). The sex-specific pattern of petal length decrement was most pronounced between the primary and secondary flowers, with functional males exhibiting a lower reduction (3.5%) compared to either hermaphrodites (11%) or females (7.5%) (P < 0.02). Petal length declined similarly for all three sex morphs between secondary and tertiary positions (P > 0.8), and between tertiary and quaternary positions (P > 0.15) (Fig. 2C).
|
2 = 10.94, P < 0.005; 2 = 4.24, P < 0.05, respectively).
|
Genotypic variation in intra-inflorescence allocation
There was significant genotypic variation for average allocation to all floral traits (Table 1). But significant genotype-specific patterns of intra-inflorescence allocation were seen only for pollen number per flower (genotype-by-flower position interaction, P < 0.01). Three genotypes increased or maintained pollen production from primary to quaternary positions, while others decreased pollen production along the inflorescence. There was also a suggestion of genotypic variation in the pattern of ovule allocation over the inflorescence (Table 1). Limited sample size precluded testing for an architectural cause of genotype-specific positional effects.
DISCUSSION
In this study we have shown that pronounced intra-inflorescence variation among Fragaria virginiana flowers is due more to architecture than to our resource manipulations during flowering. However, because we also found strong evidence for sex-specific architectural patterns and genotype-specific positional variation in one trait, the effect of architecture cannot be viewed purely as a constraint on the evolution of new intra-inflorescence patterns.
Architecture and resource effects on intra-inflorescence variation
In contrast to other studies that found significant effects of resource competition on allocation to individual flowers within an inflorescence (Casper, 1984
; Solomon, 1988
; Macnair and Cumbes, 1989
; Herrera, 1991
; Ashman, 1992
; Wolfe, 1992
; Diggle, 1995, 1997
; Guitián and Navarro, 1996
), we found little general impact of our resource manipulations during flowering on intra-inflorescence patterns. Rather, we found a resource effect at the level of the whole inflorescence. Specifically, the removal of primary buds resulted in an increase in the total number of flowers produced per inflorescence, but did not alter allocation to pollen, ovules, or petal length within flowers. This suggests that the pattern of floral organ size and number throughout the inflorescence may be determined prior to flowering (possibly in the previous fall; Darrow, 1966
) and is inalterable by resources status during flowering, or at least, is inalterable by the removal of the primary bud prior to flowering. (We cannot, however, rule out the possibility that removal of more buds might result in a greater response.) However, the number of flower buds that develop into flowers appears to be subject to resource release provided by the level of bud removal we employed. While bud removal elicited a response, pollination treatment did not, not even in females where fruit production was highest. This suggests that maturing fruit did not represent a significant competitor for resources with flowers on the same inflorescence. Strawberry fruits forming at primary and secondary positions may not have competed for resources with distal flowers because inflorescences were relatively few flowered (averaging 7.3 flowers) and the temporal overlap between flower development and fruit maturation was brief (flowering lasted, on average, 2.5 wk).
Nonetheless, we found strong effects of architecture on all traits (Table 2B), and this agrees with the findings of several studies, which despite finding resource effects, conclude that architectural effects were at least as important (Diggle, 1995, 1997
; Wolfe, 1992
). For example, our work with Fragaria virginiana and that of Wolfe (1992) working with Hydrophyllum appendiculatum revealed only an architectural effect on petal length, whereas Diggle (1997)
and Ashman (1992)
found both architecture and resource effects on petal length in Arabidopsis thaliana and in gynodioecious Sidalcea oregana spicata, respectively.
Given the relatively small number of studies that have specifically endeavored to separate architectural vs. resource causes of intra-inflorescence allocation thus far, it is difficult to uncover a predictive pattern regarding the predominance of architectural or resource effects. However, it does appear that the relative importance of architecture and resources can be trait specific and phenologically variable. Some traits may become fixed earlier in development than others. It would seem that organ number (ovule, stamen, pollen) might be influenced more by architecture than by resources during flowering because primordia, or mother cell numbers, may be determined early in bud formation. In contrast, size or biomass of organs may be more plastic in response to resource conditions during flowering. Our results for pollen and ovule number are consistent with this argument, however those for petal length are not. Nevertheless, such a pattern does emerge in other studies. For example, Diggle (1995)
found that ovule number declined over the inflorescence in 15 liliaceous species that have preformed flowers, despite the fact that the direction of flower opening and, thus, the competition-based resource gradient, differed among the species, whereas most size and biomass traits responded to resource manipulations in Mimulus guttatus (Macnair and Cumbes, 1989
; Mossop, McNair, and Robertson, 1994
), and Sidalcea oregana spicata (Ashman, 1992
). However, this does not exclude the possibility that plants will prioritize certain organs over others when faced with resource constraints, and size may be retained in primary sexual organs at the expense of secondary organs. For example, Ashman (1992)
found biomass allocated to sepals, androecium, and petals declined in pollinated plants faster than in unpollinated plants, but allocation to pollen and ovules did not. Likewise, Diggle (1995)
found that corolla and ovary length declined in response to resources, while anther length did not in Solanum hirtum. Clearly, more studies that measure both size and number characters simultaneously and studies that perform manipulations at various developmental stages are needed to reveal the general theme of resource vs. architectural predominance in intra-inflorescence variation in floral organ formation, if one exists.
Sex-specific and genotypic effects on intra-inflorescence variation
We found evidence for genotype-specific intra-inflorescence patterns in one trait, a hint of genotypic patterns in one other trait, and sex-specific patterns in several traits, suggesting that intra-inflorescence patterns can be genetically variable in this species. Specifically, pollen number per flower showed a genotype-specific intra-inflorescence pattern. The three genotypes with the lowest pollen number in the primary flower increased or maintained pollen number over the inflorescence, whereas among the remaining genotypes there was a positive relationship between pollen number in the primary flower and the rate of decline in pollen over the inflorescence (r = 0.66; P = 0.004; N = 11). Few studies have explored genotype-specific inflorescence allocation patterns. However, some have shown genotypes to respond differently to resource manipulations (Diggle 1993, 1997
), and several (Pelofske and Lawrence, 1984
; Diggle, 1997
) have also revealed genetic variation in patterns of investment. Diggle (1997)
found different architectural effects on stigma length in two ecotypes of Arabidopsis thaliana; stigma length in the ecotype with the longer stigmas (Landsberg) declined three times faster over the inflorescence than the ecotype with the shorter stigmas (Columbia). Likewise, Pelofske and Lawrence (1984)
have shown that the size ratio of primary to secondary fruits is heritable in cultivars of hermaphrodite strawberries and suggested that the larger the primary fruit the faster the decline in fruit size over the inflorescence. In our study, we also find a positive correlation between investment in ovule number at the primary position and the rate of decline in ovule number over the inflorescence across all genotypes (r = 0.61; P < 0.005; N = 21). This relationship is maintained even when only nonpollination plants are considered (r = 0.49; P < 0.05; N = 18), suggesting that it has, in part, an architectural basis. Taken together, these studies suggest that genotypes with large investments in early flowers suffer more dramatic declines over the inflorescence, whereas those with lesser investments in primary flowers have more uniformity over the inflorescence. More studies are needed to determine whether genotypic variation in architectural effects is common, as well as, to relate this variation to fitness, if we are to understand the adaptive potential of intra-inflorescence variation.
The sex morphs showed the most pronounced differences in intra-inflorescence patterns of allocation with respect to ovule number and flowers per position. Females and hermaphrodites invest most heavily in ovules in the primary position, but their investment in ovules declines rapidly, and the decline of hermaphrodites surpasses that of females at the quaternary position (Fig. 2B). The primary and secondary positions are the most likely to set fruit in both females and hermaphrodites, whereas quaternary positions are the least likely to set fruit and this probability is even lower in hermaphrodites than in females (Valleau, 1918
; Ashman, personal observation). Thus, in F. virginiana the flowers with the greatest probability of setting fruit have been provisioned with the most ovules, and this is most pronounced in females. Heavily provisioning the primary flower, however, appears to come at a cost in terms of flower number: the sex morph that invests most heavily in the primary position (females) also produces the least branching and fewest flowered inflorescences of the sex morphs (Figs. 2D, 3A). This architectural pattern may be favored by selection through female function if vasculature to more distal flowers is insufficient to adequately supply developing fruit (reviewed in Diggle, 1995
), and/or if structural tissue is inadequate (or too costly) to support developing fruit at the more distal positions (Schoen and Dubuc, 1990
). In contrast, functional males invested the least in ovules at the primary position and exhibited the lowest decrements over the inflorescence, but also showed the greatest increase in the number of flowers between primary and secondary positions and maintained higher flower numbers at the quaternary position. As a consequence, functional males had the most branching inflorescences of the sex morphs (Fig. 3C). It appears that functional males maintain production of as many distal flowers as possible and thus increase their opportunities for pollen export. Because pollen maturation is brief and requires less of an investment relative to fruit production, less vasculature or structural support at distal positions is not likely to significantly affect success through male function. Hermaphrodites show patterns intermediate between females and functional males, indicating that their allocation pattern may reflect a balance between the sex functions.
When these data are considered in light of the common observation that males produce more flowers per inflorescence than females in dioecious species (Lloyd and Webb, 1977
; Delph, 1996
), they suggest that flower number may be reduced in females as a consequence of selection for enhanced allocation to ovules and fruit at positions most favorable for their maturation. In contrast, no such architectural bias would be favored in males, rather selection is likely to favor spreading pollen out over as many flowers on an inflorescence as possible. More studies of intra-inflorescence variation in sexually polymorphic species are needed to determine the universality of this pattern. Moreover, experiments that link intra-inflorescence allocation patterns to fitness through male and female reproductive success achieved under natural conditions are needed to test this hypothesis. Evidence of disruptive selection with respect to inflorescence size and intra-inflorescence patterns would be support for this hypothesis.
Conclusions
We have shown that intra-inflorescence variation among Fragaria virginiana flowers has a strong architectural component, but one that varies among the sex morphs. This variation suggests that increased investment in early flowers leads to an increase in the rate of intra-inflorescence decline in floral allocation. Sex-differential intra-inflorescence patterns suggest that fitness through male and female function may be maximized by different resource distribution patterns within the inflorescence. More studies are needed to determine how common sex-differential architectural effects are and to relate this variation to fitness, if we are to understand the adaptive potential of intra-inflorescence variation.
FOOTNOTES
1 The authors thank Jim Swetz, Sandy Shivitz, Jason Javitz, and Ellen York for greenhouse and laboratory assistance; Susan Kalisz, Donna Vogler, Shannon Bliss, Tiffany Knight, and anonymous reviewers for comments on the manuscript. This research was supported by grants from NSF (DEB-9508635 and 9707247), University of Pittsburgh Faculty of Arts and Sciences, Central Research and Development Fund, and the Department of Biological Sciences. This is Contribution number 107 to the Pymatuning Laboratory of Ecology. 
2 Author for correspondence (e-mail: TIA1+{at}PITT.EDU
).
LITERATURE CITED
Ahmadi, H., and R. S. Bringhurst. 1991 Genetics of sex expression in Fragaria species. American Journal of Botany 78: 504–514. [CrossRef][ISI]
Ashman, T.-L. 1992 Indirect costs of seed production within and between seasons in a gynodioecious species. Oecologia 92: 266–272. [CrossRef][ISI]
———. 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.
———. In press. Pollinator selectivity and its implications for the evolution of dioecy and sexual dimorphism. Ecology.
———, and I. Baker. 1992 Variation in floral sex allocation with time of season and currency. Ecology 73: 1237–1243. [CrossRef][ISI]
Bell, G. 1985 On the function of flowers. Proceedings of the Royal Society of London 224: 223–265.
Broyles, S. B., and W. Wyatt. 1995 A reexamination of the pollen-donation hypothesis in an experimental population of Asclepias exaltata. Evolution 49: 89–99.
Brunet, J. 1996 Male reproductive success and variation in fruit and seed set in Aquilegia caerulea (Ranunculaceae). Ecology 77: 2458–2471. [CrossRef][ISI]
———, and D. Charlesworth. 1995 Floral sex allocation in sequentially blooming plants. Evolution 49: 70–79. [CrossRef][ISI]
Casper, B. B. 1984 On the evolution of embryo abortion in the herbaceous perennial Cryptantha flava. Evolution 38: 1337–1349.
Darrow, G. M. 1966 The strawberry: history, breeding, and physiology. Holt, Rinehart, and Winston, New York, New York, USA.
Delph, L. F. 1996 Flower size dimorphism in plants with unisexual flowers. In D. G. Lloyd and S. C. H. Barrett [eds.], Floral biology: studies on floral evolution in animal pollinated Plants, 217–237. Chapman and Hall, New York, New York, USA.
Diggle, P. K. 1993 Developmental plasticity, genetic variation, and the evolution of andromonoecy in Solanum hirtum (Solanaceae). American Journal of Botany 80: 967–973. [CrossRef][ISI]
———. 1995 Architectural effects and the interpretation of patterns of fruit and seed development. Annual Review of Ecology and Systematics 26: 531–552. [CrossRef][ISI]
———. 1997 Ontogenetic contingency and floral morphology: The effects of architecture and resource limitation. International Journal of Plant Sciences 158 supplement: s99–s107.
Emms, S. K., D. A. Stratton, and A. A. Snow. 1997 The effect of inflorescence size on male fitness: Experimental tests in the andromonoecious lily, Zigadenus paniculatus. Evolution 51: 1481–1489.
Frank, S. A. 1987 Individual and population sex allocation patterns. Theoretical Population Biology 31: 47–74. [CrossRef][ISI][Medline]
Gleason, H. A., and A. Cronquist. 1991 Manual of vascular plants of northeastern United States and adjacent Canada. New York Botanical Garden, Bronx, New York, USA.
Guitián, J., and L. Navarro. 1996 Allocation of reproductive resources within the inflorescences of Petrocoptis grandiflora (Caryophyllaceae). Canadian Journal of Botany 74: 1482–1486.
Herrera, J. 1991 Allocation of reproductive resources within and among inflorescences of Lavandula stoechas (Lamiaceae). American Journal of Botany 78: 789–794. [CrossRef][ISI]
Lee, T. D. 1988 Patterns of fruit and seed production. In J. Lovett-Doust and L. Lovett-Doust [eds.], Plant reproductive ecology: patterns and strategies, 179–202. Oxford, University Press, New York, New York, USA.
Littell, R. C., G. A. Milliken, W. W. Stroup, and R. D. Wolfinger. 1996 SAS system for mixed models. SAS Institute, Cary, North Carolina, USA.
Lloyd, D. G. 1980 Sexual strategies in plants. III. A quantitative method for describing the gender of plants. New Zealand Journal of Botany 18: 103–108.
———, and K. S. Bawa. 1984 Modification of the gender of seed plants in varying conditions. Evolutionary Biology 17: 255–338. [ISI]
———, and C. J. Webb. 1977 Secondary sex characters in plants. Botanical Review 43: 177–216.
Macnair, M. R., and Q. J. Cumbes. 1989 The genetic architecture of interspecific variation in Mimulus. Genetics 122: 211–222.
Morgan, D. R., D. E. Soltis, and R. K. Robertson. 1994 Systematic and evolutionary implications of rbcL sequence variation in Rosaceae. American Journal of Botany 81: 890–903. [CrossRef][ISI]
Mossop, R., M. R. McNair, and A. W. Robertson. 1994 Within-population variation in sexual resource allocation in Mimulus guttatus. Functional Ecology 8: 410–418. [CrossRef][ISI]
Pelofske, P. J., and F. J. Lawrence. 1984 Inheritance of size relationship of primary and secondary berries of strawberry. HortScience 19: 641–642. [ISI]
SAS. 1989 SAS/STAT user's guide, version 6, 4th ed., vol. 1. SAS Institute, Cary, North Carolina, USA.
———. 1996 SAS/STAT software: changes and enhancements through release 6.11. SAS Institute, Cary, North Carolina, USA.
Schoen, D. J., and M. Dubuc. 1990 The evolution of inflorescence size and number: a gamete- packaging strategy in plants. American Naturalist 135: 841–857. [CrossRef][ISI]
Sokal, R. R., and F. J. Rohlf. 1981 Biometry. W. H. Freeman, New York, New York, USA.
Solomon, B. P. 1988 Patterns of pre- and post-fertilization resource allocation within an inflorescence: evidence for interovary competition. American Journal of Botany 75: 1074–1079. [CrossRef][ISI]
Stahler, M. M., J. J. Luby, and P. D. Ascher. 1991 Comparative yield of female and hermaphroditic Fragaria virginiana germplasm collected in Minnesota and Wisconsin. In A. Dale and J. J. Luby [eds.], The strawberry into the 21st century, 104–105. Timber Press, Portland, Oregon, USA.
Staudt, G. 1967 The genetics and evolution of heteroecy in the genus Fragaria. I. Investigations on Fragaria orientalis. Zeitschrift Pflanzenzuchtung 58: 245–277.
———. 1989 The species of Fragaria, their taxonomic and geographical distribution. Acta Horticulturae 265: 23–33.
Stephenson, A. G. 1981 Flower and fruit abortion: Proximate causes and ultimate functions. Annual Review of Ecology and Systematics 12: 253–279.
Valleau, W. D. 1918 Sterility in the strawberry. Journal of Agricultural Research 12: 613–669.
———. 1923 The inheritance of flower types an fertility in the strawberry. American Journal of Botany 10: 259–274. [CrossRef][ISI]
Wolfe, L. M. 1992 Why does the size of reproductive structures decline through time in Hydrophyllum appendiculatum (Hydrophyllaceae)? Developmental constraints vs. resource limitation. American Journal of Botany 79: 1286–1290. [CrossRef][ISI]
Wyatt, R. 1982 Inflorescence architecture: how flower number, arrangement, and phenology affect pollination and fruit-set. American Journal of Botany 69: 585–594. [CrossRef][ISI]
This article has been cited by other articles:
|
|
 |
|
  M{a} L. Buide Disentangling the causes of intrainflorescence variation in floral traits and fecundity in the hermaphrodite Silene acutifolia Am. J. Botany, April 1, 2008; 95(4): 490 - 497. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Narbona, P. L. Ortiz, and M. Arista Sexual Dimorphism in the Andromonoecious Euphorbia nicaeensis: Effects of Gender and Inflorescence Development Ann. Bot., April 1, 2008; 101(5): 717 - 726. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Abdala-Roberts, V. Parra-Tabla, and J. Navarro Is Floral Longevity Influenced by Reproductive Costs and Pollination Success in Cohniella ascendens (Orchidaceae)? Ann. Bot., November 1, 2007; 100(6): 1367 - 1371. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
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]
|
|
|
|
|
|
T. Itagaki and S. Sakai Relationship between floral longevity and sex allocation among flowers within inflorescences in Aquilegia buergeriana var. oxysepala (Ranunculaceae) Am. J. Botany, September 1, 2006; 93(9): 1320 - 1327. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L Marr and M. L Marshall The role of fungal pathogens in flower size and seed mass variation in three species of Hydrophyllum (Hydrophyllaceae) Am. J. Botany, March 1, 2006; 93(3): 389 - 398. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. H. Cole and T.-L. Ashman Sexes show differential tolerance to spittlebug damage and consequences of damage for multi-species interactions Am. J. Botany, October 1, 2005; 92(10): 1708 - 1713. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. NARBONA, P. L. ORTIZ, and M. ARISTA Dichogamy and Sexual Dimorphism in Floral Traits in the Andromonoecious Euphorbia boetica Ann. Bot., April 1, 2005; 95(5): 779 - 787. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. GUITIAN, M. MEDRANO, and J. E. OTI Variation in Floral Sex Allocation in Polygonatum odoratum (Liliaceae) Ann. Bot., September 1, 2004; 94(3): 433 - 440. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. BUIDE Intra-inflorescence Variation in Floral Traits and Reproductive Success of the Hermaphrodite Silene acutifolia Ann. Bot., September 1, 2004; 94(3): 441 - 448. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. S. Ishii and S. Sakai Temporal variation in floral display size and individual floral sex allocation in racemes of Narthecium asiaticum (Liliaceae) Am. J. Botany, March 1, 2002; 89(3): 441 - 446. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. Mazer and K. A. Dawson Size-dependent sex allocation within flowers of the annual herb Clarkia unguiculata (Onagraceae): ontogenetic and among-plant variation Am. J. Botany, May 1, 2001; 88(5): 819 - 831. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||
0.05 < P < 0.1, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001