| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
Department of Ecology, Evolution and Marine Biology, University of California, Santa Barbara, California 93106 USA
Received for publication August 7, 2003. Accepted for publication August 26, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Clarkia • development rate • life history • mating systems • Onagraceae • protandry • self-fertilization
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Wyatt, 1983
, 1986
, 1988
). Six mechanisms may contribute to this evolutionary transition. First, the automatic selection hypothesis proposes that alleles promoting self-fertilization will enjoy a transmission advantage whereby individuals that express such alleles will transmit to future generations more copies of their genome through male function than individuals that lack them (with no reduction in female function) (Fisher, 1941
; Jain, 1976
). If the negative effects of inbreeding depression do not outweigh this advantage, alleles promoting selfing will increase in frequency.
Second, in habitats where pollinators or potential mates are scarce or unreliable, genotypes whose flowers spontaneously self-pollinate will produce more offspring than those that do not (Darwin, 1876
; Baker, 1955
, 1967
; Jain, 1976
; Lloyd, 1979
, 1992
; Schemske, 1978
; Motton, 1982
; Piper et al., 1982; Holsinger, 1996
, 2000
; Schoen et al., 1996
; Fausto et al., 2001
; Culley, 2002
; Gomez, 2002
; Dubois et al., 2003
). The advantage of selfing under these conditions is termed reproductive assurance (Jain [1976]
and references therein), and should result in a prevalance of selfing populations in habitats with unreliable pollinators.
Third, self-fertilization may evolve as a mechanism to avoid interspecific pollen transfer in habitats where generalist pollinators do not reliably discriminate among plant species (Stucky, 1985
; Petanidou et al., 1998
; Fishman and Wyatt, 1999
). Where stigmas receive enough allospecific pollen to impede pollination or fertilization by conspecific pollen, self-fertilization provides reproductive assurance even where pollinators are abundant (Palmer et al., 1989
; Ehlers, 1999
). Fourth, self-fertilization may evolve where conspecific ecotypes adapted to distinct environments live within the range of gene flow (Cugen et al., 1989
; Wendt et al., 2002
). Genotypes that self-fertilize may be less likely than outcrossing genotypes to receive pollen produced by genotypes adapted to an alternative environment. Accordingly, the disadvantages of either inter- or intraspecific hybridization may generate selection favoring self-fertilization. Under these four scenarios, self-fertilization may evolve independently of whole-plant traits.
Two mechanisms for the evolution of selfing involve the joint evolution of floral and whole-plant traits. First, self-fertilizing flowers may evolve where selection favors the rapid completion of the life cycle, referred to here as "the accelerated life cycle hypothesis." For example, in populations that have colonized habitats with shorter growing seasons than those from which the colonizers migrated, natural selection may favor earlier and more rapid reproduction (Strid, 1969
; Guerrant, 1988
, 1989
; Andersson, 1997
; Eckhart and Geber, 1999; Morgan-Richards and Wolff, 1999
). This may be achieved by accelerating reproductive processes at both the individual and flower levels. At the whole-plant level, selection may favor genotypes that flower relatively early and that produce successive flowers at a rapid rate. At the level of individual flowers, selection may favor genotypes whose flowers progress rapidly through anther dehiscence, stigma receptivity, and petal loss. If the distance between anthers and stigmas increases with flower size and during development, then reductions in the floral size and lifespan may also result in reduced herkogamy and dichogamy, increasing the likelihood of selfing (cf. Sherry and Lord, 1996
; Holtsford and Ellstrand, 1989
, 1992
).
This hypothesis predicts that among populations that differ in the length of the growing season, there will be positive correlations across population means (within and among closely related taxa) between: (a) life history traits reflecting the whole-plant development rate, such as the age at first flower and the number of days between the production of successive flowers and (b) individual floral traits affecting the length of a reproductive episode, such as the mean number of days between bud break and the completion of floral development or the number of days of protandry (Fig. 1). Where mating system has co-evolved with these traits, selfing populations are predicted to have faster mean developmental rates than outcrossers for both types of traits. This hypothesis does not preclude, however, the independent evolution of whole-plant and floral traits among populations, depending on the strength of selection imposed. Moreover, within populations, we do not expect to observe strong or consistent genetic correlations among these traits, and the sign of any genetic correlations within populations need not be the same as the sign of the cross-population correlation coefficient (Fig. 1B and C).
|
; Conner, 2002
). Andersson (1997)
proposed such a mechanism as a contributing factor in the evolution of selfing taxa in Nigella (Ranunculaceae). We refer to this as "the correlated response to selection hypothesis." If development rates of individuals are determined by alleles that simultaneously affect both whole-plant and floral development, then genotypes that reproduce at an early age may also produce successive flowers at a rapid rate and produce flowers with little herkogamy or dichogamy, increasing the rate of self-fertilization. Strong pleiotropic effects could also cause floral traits to be highly integrated, constraining their independent evolution (Armbruster and Schwaegerle, 1996
).
Similar to the accelerated life cycle hypothesis, correlated responses to selection should result in early reproduction and rapid flower production and development in habitats with short growing seasons. Under both hypotheses, the evolution of selfing flowers is an example of heterochrony, in which the timing and/or the rate of developmental processes differs between an ancestor and its descendent (Diggle, 1992
; see also Lord and Hill, 1987
; Hill and Lord, 1990
; Johnston, 2000
) or between structures of different types (e.g., selfing vs. outcrossing flowers; cf. Gallardo et al., 1993
).
Both the accelerated life cycle and the correlated response to selection hypotheses predict the simultaneous evolution of multiple traits that shorten the life cycle. Such joint evolution should result in strong correlations across populations between whole-plant and floral traits, where selfing populations are represented by lower phenotypic means than outcrossers (Fig. 1A–F). The most important difference between these two hypotheses is that under the correlated response to selection hypothesis, selection cannot operate independently on the focal traits, so we do not expect their independent evolution. Under this hypothesis, where populations evolve to differ in one trait (e.g., the age of first flower), genetic correlations will necessarily drive evolutionary change in the others (e.g., the rate of floral development). In addition, given sufficiently strong pleiotropy or linkage, this hypothesis predicts a positive correlation between rates of whole-plant and floral development among genotypes within populations (given sufficient standing genetic variance in the focal traits; Fig. 1D and E). Genotypes with alleles coding for relatively fast developmental rates would necessarily flower earlier, produce successive flowers faster, and produce flowers with shorter lifespans than genotypes with slow developmental rates. Note that selfing populations may exhibit lower genetic variation than outcrossers for several reasons (e.g., founder effects, repeated inbreeding), so the magnitude of their genetic correlations may be lower than that of outcrossing populations (e.g., Fig. 1F).
In the current study, we aim to distinguish between the accelerated life history and the correlated response to selection hypotheses by comparing populations of outcrossing and selfing subspecies of Clarkia xantiana Gray (Onagraceae) with respect to both whole-plant and floral traits. Previous work on this species has detected correlations among population means between floral traits reflecting flower size, floral development rates, and the probability of selfing (Eckhart and Geber, 2000
) and has revealed differences between the subspecies in patterns of covariation among floral traits (Runions and Geber, 2000
); a detailed study of both floral and whole-plant development has not yet been reported.
Clarkia xantiana is a particularly appropriate taxon for such an investigation because field populations of the selfing and outcrossing subspecies are known to differ in life history and in floral traits; the predominantly selfing subspecies flowers earlier and produces shorter-lived flowers than the outcrossers (Moore and Lewis, 1965
; Eckhart and Geber, 2000
; Runions and Geber, 2000
). Here, we examined maternal family means to seek genetically based correlations among whole-plant and floral traits, and to determine whether selfing and outcrossing maternal genotypes fall on the same genetic trajectory. We also assessed whether, controlling for variation in whole-plant traits, floral traits evolved independently between populations or subspecies.
Specifically, we asked: (1) Do selfing populations exhibit more rapid development than outcrossing populations for both whole-plant traits and individual flowers? (2) Do phenotypic values of floral and whole-plant traits vary independently among maternal families within subspecies, or among populations, suggesting the potential for their independent evolution? (3) Alternatively, are floral and whole-plant traits correlated among families or populations? If so, do the bivariate distributions of these traits represent similar genetic trajectories in the two subspecies, as would be predicted by the correlated responses to selection hypothesis? In sum, we evaluate whether the patterns of variation and covariation observed provide stronger support for the accelerated life cycle or the correlated response to selection hypothesis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
, Moore and Lewis (1965)
, Runions and Geber (2000)
, and Eckhart and Geber (2000)
. The species is represented by two subspecies that differ in their floral biology and life history: Clarkia xantiana ssp. xantiana Gray (hereafter "xantiana") and C. xantiana ssp. parviflora (Eastw.) H. Lewis & Raven (hereafter "parviflora"). In the field and in the greenhouse, xantiana flowers have lower relative growth rates and longer development times, produce longer sepals and larger petals, and exhibit greater protandry than those of parviflora (Runions and Geber, 2000
). In addition, xantiana individuals have longer lifespans and initiate flowering at later dates than parviflora (Mazer and V. A. Delesalle, personal observation). Protandry and herkogamy in xantiana appear to contribute to high levels of outcrossing and strong dependence on pollinators for full seed set. In contrast, the anthers of parviflora flowers release their pollen onto the open stigma (often while still in bud), resulting in nearly 100% autogamous fruit and seed set in the absence of pollinators.
The two subspecies differ in their habitat preferences or tolerances, although they may occur sympatrically. Where xantiana occurs in the absence of parviflora, it is visited by seven species of pollinators; where parviflora appears in the absence of xantiana, only two pollinator species have been observed visiting its flowers (Fausto et al., 2001
). Lone populations of xantiana exhibit higher population densities, higher floral densities, and higher pollinator visitation rates than lone populations of parviflora (Fausto et al., 2001
). Parviflora tends to occupy habitats in which both pollinators and potential mates are relatively scarce, consistent with the reproductive assurance hypothesis. Lone populations of xantiana occupy the relatively mesic western edge of the species range, while allopatric populations of parviflora are restricted to its xeric eastern edge.
We collected seeds from four wild populations of Clarkia xantiana: two populations represent xantiana (Saw Mill Road and Camp 3) and two populations represent parviflora (Saw Mill Road and Long Valley Road; Table 1). The xantiana and parviflora populations on Saw Mill Road are about one km apart, while the Camp 3 and Long Valley Road sites are inhabited by lone populations of xantiana and parviflora, respectively. In June 2001, mature seeds were collected from maternal plants in the field; maternal families were stored separately in paper envelopes at room temperature until prepared for germination.
|
Cultivation methods
For each trial, 12–15 seeds of each maternal family were placed in a 3-cm diam plastic petri dish filled with an agar medium produced by dissolving 2 g of agar and 0.125 g of MES (2-[N-Morpholino]ethanesulfonic acid; SIGMA M-2933) in 250 mL of heated distilled water. Seeds were placed on the agar surface; the petri dish was then covered, wrapped in aluminum foil, and refrigerated for 1 wk. The moist seed coats were then nicked with a razor blade to facilitate imbibition, and seeds were placed in a growth chamber at temperatures of 70° F day/55° F night and a fluorescent/incandescent light regime of 16 h light/8 h dark. Germination occurred within one to seven days. One week following placement on the agar, seedlings were transplanted to plastic tubes (25 cm long x 4 cm diameter, with drainage holes at the base) filled with UC Soil Mix and transferred to a greenhouse.
Data recorded
The following data were recorded: (a) the date on which the first flower opened on the primary stem; (b) the date on which the sixth flower opened on the primary stem. In addition, each flower on the primary stem was observed on a daily basis (monitoring occurred between 15:00–17: 00 h) and the following data recorded: (c) the date on which the bud opened; (d) the date on which all eight anthers had completely dehisced lengthwise along their locules; and (e) the date on which the stigma fully opened, with the stigmatic branches oriented perpendicular to the style and the papillae well developed. For each flower, these traits were used to determine: (f) the number of days between bud opening and complete anther dehiscence, (hereafter referred to as days to anthesis); (g) the number of days between bud opening and stigma receptivity (hereafter referred to as days to stigma receptivity); and (h) the number of days of protandry (g–f). Individual phenotypic means were estimated using the values recorded for the first six flowers produced by the primary stem.
The data were truncated to include only the first six flowers because some individuals produced no more than six flowers on the primary stem and we aimed to create a balanced data set to control for potential effects of flower position on floral development. Accounting for some plants that produced fewer than six flowers on the primary stem and for some missing data, a total of 624 flowers was observed in the two trials (2 trials x 2 populations/trial x 10 maternal families/population x 3 offspring/maternal family x 6 flowers/ offspring). Maternal family means for each trait were then estimated from the phenotypic means of the three siblings representing each maternal family. These maternal family means were used as data points in the statistical analyses described below.
Our statistical analyses targeted three traits reflecting the rate of floral development and two traits reflecting whole plant development rates. Floral development traits included: (a) the number of days to anthesis; (b) the number of days to stigma receptivity; and (c) the number of days of protandry. Whole-plant development traits included: (d) the age at first flower production (AFF), measured as the number of days between when newly germinated seeds were transplanted into soil and the first flower opened on the main branch; and (e) the number of days elapsed between the first and the sixth flowers opening on the primary branch (ND1–6), which is a proxy for the rate of successive flower development.
Statistical analyses
Detecting differences between selfing and outcrossing taxa in floral and whole-plant development
We used the analysis of covariance (ANCOVA) to detect differences between subspecies in floral traits (the dependent variables) while controlling for variation in whole-plant traits (AFF and ND1–6) and to detect covariation between floral and whole-plant traits. We conducted two sets of ANCOVAs, each with a different covariate (AFF vs. ND1–6). First, to detect sources of variance in floral traits, and to detect covariance between floral traits and AFF, we conducted an ANCOVA based on type III sums of squares and including the terms: trial, subspecies, AFF (included as a covariate), trial x AFF, subspecies x AFF, subspecies x trial, and subspecies x AFF x trial. With this model we tested: (1) whether mean values of floral traits differed between subspecies (detected as a significant effect of subspecies); (2) whether there is a general correlation between each floral trait and AFF across subspecies (detected as a significant effect of AFF); (3) whether slopes of the relationships between each floral trait and AFF (pooling trials) differed between subspecies (subspecies x AFF); (4) whether the mean values of subspecies differed similarly between trials (subspecies x Trial), and (5) whether within each experimental trial, slopes of the relationship between each floral trait and AFF differ between subspecies (detected as a significant subspecies x trial x AFF interaction).
This model therefore served to seek evidence of a relationship between floral traits and AFF at several ecological levels: within each population, within each subspecies (pooling populations), and across all genotypes. We started by testing the highest order interaction and removed interaction terms that appeared to be non-significant. After identifying the non-significant interactions (see Results), we designed and used a reduced model, described below. These analyses were repeated using ND1–6 as the covariate instead of AFF. All analyses were conducted using JMP 5.01 (SAS Institute, Inc.).
Pearson correlation coefficients were estimated among maternal family means within each subspecies to evaluate whether the individual floral traits are strongly correlated with each other, that is, whether the floral traits are highly integrated.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
Design of reduced ANCOVA
Given these non-significant interactions, the two trials were then pooled to test whether the slopes of the relationships between each floral trait and AFF (and ND1–6), or the mean values of floral traits, differed between subspecies. Two sets of reduced ANCOVAs (from which the three-way interactions were excluded) were then conducted using each whole-plant trait as the covariate. These reduced ANCOVAs included the terms: trial (as a block), subspecies, covariate (AFF or ND1–6) and the subspecies x covariate interaction. The subspecies x trial term was excluded from these models because a large proportion of variation in AFF and ND1–6 was accounted for by trial (Fig. 2), and thus the subspecies x AFF (or ND1–6) and subspecies x trial interactions were somewhat redundant. These ANCOVAs were used and interpreted assuming that the high variance observed in whole-plant traits was likely due to different growing conditions between trials.
Main effects of trial and whole-plant traits on floral traits
The reduced ANCOVA detected no significant differences between trials in the expression of floral traits independently of either AFF or ND1–6 (Table 3). Similarly, no significant covariance was detected between floral traits and AFF or ND1–6 independently of trial.
|
Interactions between subspecies and whole-plant development in selfing and outcrossing taxa
The reduced ANCOVAs detected significant subspecies x AFF and subspecies x ND1–6 interactions for days to anthesis and for days to stigma receptivity but not for days of protandry (Table 3). The significant subspecies x AFF interaction terms reflect the fact that xantiana exhibited positive relationships between days to anthesis and AFF, and between days to stigma receptivity and AFF, while in parviflora these floral traits varied negatively with, or independently of AFF (Fig. 2A–B). The significant subspecies x ND1–6 interaction terms reflect the same differences between subspecies in their slopes as for AFF (Fig. 2D–E).
Variation within maternal families
Variation among siblings in the age of first flower production was low relative to the phenotypic mean of each population (the mean coefficient of variation [CV] among siblings was less than 10%; Table 4). For the rate of successive flower production (ND1–6) and for the floral traits, however, the mean CV among siblings was much higher (12.5%–98.3%, depending on the trait and population), and there was substantial variation among maternal families in the magnitude of the CV among the siblings within them (compare the minimum vs. maximum CV among siblings for each trait; Table 4).
|
Among parviflora families, the number of days to anthesis was correlated with neither the number of days to stigma receptivity (r = 0.266; P > 0.258; df = 19) nor with the duration of protandry (r = 0.318; P > 0.172). Similar to xantiana, the number of days to stigma receptivity was positively correlated with the duration of protandry (r = 0.816; P < 0.0001) in parviflora.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Runions and Geber (2000)
also found that the number of days to anthesis is lower, and the rate of successive flower production is higher, in parviflora than xantiana. These investigators sought evidence for correlated evolution between floral parts (sepals, petals, filaments, styles, and ovaries) by examining covariation among population means within each of the two subspecies. Based on the absence of strong among-population correlations, they concluded that these floral parts may well have evolved independently during the transition between outcrossing and selfing in response to selection favoring rapid floral maturation in the arid habitats tolerated by parviflora. The current study begins to extend earlier work in two ways. First, we focused on the correlated evolution between floral and whole-plant traits rather than on the relationships among multiple floral traits. Our aim was to evaluate whether the evolution of rapid floral development in parviflora may be interpreted simply as a correlated response to selection on rates of whole-plant maturation or whether floral and whole-plant developmental traits may have evolved independently in the transition from outcrossing to selfing in C. xantiana. Second, we examined maternally inherited (and presumably genetically based) covariation within populations and subspecies to assess whether pleiotropy or linkage may generate correlated responses to selection among floral and whole-plant traits.
Distinguishing between the accelerated life cycle and the correlated response to selection hypotheses
The correlated response to selection hypothesis for the joint evolution of floral and whole-plant traits may be rejected under three conditions: (1) when no evidence of a positive correlation between floral and whole-plant rates of development is observed at any ecological level; (2) where opposing relationships between floral and whole-plant traits are observed (e.g., where subspecies differ qualitatively in the sign of the genetic correlations between traits [Fig. 1B] or where the correlation across population means differ in sign from the within-population correlations [Fig. 1C]); and (3) where populations differ greatly in one kind of trait (e.g., the age of first reproduction) but not in the other (e.g., the rate of floral development), that is, where the two types of traits differentiate independently among populations. In the current study, opposing relationships are observed between subspecies (Fig. 2A and D) and parviflora populations differ greatly in the age at first flower but not in rates of floral development (Fig. 2A–C).
Developmental changes associated with the evolution of selfing
As observed previously (Moore and Lewis, 1965
; Runions and Geber, 2000
), our greenhouse-raised populations of parviflora flowered earlier and produced faster developing flowers with less protandry than xantiana. This pattern is consistent with the view that self-fertilization evolved as a correlated response to selection favoring rapid completion of the life cycle in the arid habitats to which much of parviflora is restricted (Runions and Geber, 2000
). However, we found no correlation between the duration of protandry and whole-plant development rates within either species (Fig. 2), implying that these traits may evolve independently.
Within both subspecies, the timing of stigma receptivity influences the duration of protandry more than the time of anthesis; among maternal families of both subspecies, early stigma receptivity results in a shorter period of protandry. The subspecies appear to differ in the degree of floral trait integration. In xantiana, the timing of anthesis and stigma receptivity were strongly positively correlated. By contrast, these traits were not significantly correlated among maternal families of parviflora, in part reflecting the low among-family variation observed in these traits in this subspecies. Other studies of insect-pollinated species have found strong correlations among floral traits (Armbruster, 1985
, 1988
; Herrera, 2001
), suggesting that successful pollination requires accurate positioning of male and female parts and that selection has favored a high degree of floral integration. This should be true as well for autogamously self-fertilizing species, in which herkogamy or dichogamy may prevent successful pollination. It is important to note that genetically based correlations among floral traits can be sensitive to environmental conditions, so a realistic estimate of the correlations reported in this study would best be obtained in the field (Conner et al., 2003
).
Within subspecies, population differentiation in floral traits appears more constrained in parviflora than in xantiana. This may simply reflect the fact that selfing taxa must exhibit highly synchronous anthesis and stigma receptivity, which may be achieved most efficiently by the early maturation of both organ types. By contrast, outcrossing individuals and populations may exhibit a range of variation in the duration of, and overlap between, male and female phases while still acheiving high levels of seed production. Directional selection on floral traits influencing the degree of herkogamy and dichogamy is likely to be much stronger in obligate selfers than in facultative outcrossers, but this interpretation is tentative due to the small number of populations sampled here.
The joint change in whole-plant and individual floral traits
We examined the joint change of floral and whole-plant developmental traits at three levels: between subspecies, between populations within subspecies, and within populations. Between subspecies, parviflora exhibited faster floral and whole-plant development than xantiana. Between populations within subspecies, we found evidence of joint change between floral and whole-plant development, but the pattern of change differed between the subspecies. Xantiana populations tended to covary positively between floral and whole plant traits (Fig. 2), while parviflora exhibited negative or independent relationships between population means. Within populations, we found no evidence that genetic correlations currently contribute to the correlated evolution of floral and whole-plant traits, but the low number of families represented may have limited our ability to detect significant correlations. An absence of significant genetic correlations between flowering time and five of six floral traits was likewise found in Raphanus raphanistrum (Conner and Via, 1992). Similarly, in a comparison of genetic correlations among floral traits, among vegetative traits, and across the two types of traits in Brassica napus, Brassica nigra, Raphanus raphanistrum, Hesperis matronalis, and Phlox divaricata, the correlations between floral and vegetative traits were uniformly weak (Conner and Sterling, 1996
). Similar patterns were found by Armbruster et al. (1999)
in a study of nine species. In sum, to date there is no strong evidence for pleiotropic effects influencing both whole-plant and floral development rates.
The positive associations between floral and whole-plant development traits exhibited by xantiana (but not parviflora) populations implies either that in xantiana the joint evolution of floral and whole-plant traits has a genetic basis or that xantiana exhibits more floral trait plasticity (in response to conditions that differed between Trials) than parviflora. In either case, the selfing populations are not simply accelerated versions of the outcrossers. If floral and whole-plant development traits were subject to strong linkage or pleiotropy, we would not see opposing changes in floral and whole-plant traits between the subspecies (as seen in Fig. 2A and D). Whether through differences in genetic or environmental influences, the relationship between floral and whole-plant trait expression differs between the subspecies (Fig. 2; Table 3). Correlated responses to selection cannot alone account for the differences observed here between xantiana and parviflora, but the generality of these patterns must be tempered by the fact that only two populations were sampled per subspecies.
Genetic covariance vs. concerted evolution
The correlated response to selection hypothesis predicts that parviflora genotypes will exhibit low phenotypic values of floral and whole-plant traits that overlap with the general genetic trajectory represented by xantiana genotypes. None of the traits observed exhibited this pattern. The relationships between floral and whole-plant development traits (Fig. 2) correspond most closely to the patterns shown in Fig. 1A and B. Both subspecies exhibit large inter-population differences in whole-plant development, but only xantiana exhibits consistent variation between populations in floral traits. The joint evolution of these traits in the outcrossing-to-selfing transition in C. xantiana conforms better to the accelerated life history hypothesis than to the correlated response to selection hypothesis. A larger sample of populations would be necessary to corroborate this finding more rigorously. This potential for the independent evolution of floral and whole-plant traits was also found in a greenhouse study of 16 populations of the predominantly outcrossing C. unguiculata, in which the age at first flowering evolves independently of a variety of floral traits influencing the likelihood of autogamous self-fertilization (Jonas and Geber, 1999).
Alternative hypotheses for the evolution of selfing
This study was not designed to identify the selective forces generating the divergence in mating system between xantiana and parviflora. Nevertheless, the low population densities and the lack of pollinator visitors across the eastern portion of the species' range support the view that reproductive assurance has favored the evolution of autogamy in parviflora (cf. Fausto et al., 2001
), as has been proposed for other species (Ramsey, 1996
; Donnelly et al., 1998
; Goodwillie, 1999
, 2001
; Kalisz et al., 1999
; Kephart et al., 1999
; Culley, 2002
; Gomez, 2002
; Herlihy and Eckert, 2002
; Dubois et al., 2003
). The reproductive assurance hypothesis does not predict that parviflora would necessarily exhibit either earlier reproduction or more rapid successive flower development. The data presented here support the view that floral and whole-plant development rates may evolve independently in the two subspecies but that strong selection favoring both early and rapid reproduction in arid habitats with sparse pollinators led to the combination of life history and floral developmental traits that distinguish parviflora from xantiana.
|
FOOTNOTES |
|---|
2 Reprint requests: mazer{at}lifesci.ucsb.edu
3 Current address: Department of Ecology, Evolution and Marine Biology, University of California, Santa Barbara, California, 93106; paz{at}lifesci.ucsb.edu
4 Permanent address as of September 2003: Centro de Investigaciones en Ecosistemas, Universidad Nacional Autónoma de México, Apartado Postal 27-3 (Xangari), 58089, Morelia, Michoacán, México; hpaz{at}ate.oikos.unam.mx
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Armbruster W. S. 1985 Patterns of character divergence and the evolution of reproductive ecotypes of Dalechampia scandens (Euphorbiaceae). Evolution 39: 733-752[CrossRef][ISI]
Armbruster W. S. 1988 Multilevel comparative analysis of the morphology, function, and evolution of Dalechampia blossoms. Ecology 69: 1746-1761[CrossRef][ISI]
Armbruster W. S. V. S. Di Stilio J. D. Tuxill T. C. Flores J. L. Velasquez Runk 1999 Covariance and decoupling of floral and vegetative traits in nine neotropical plants: a re-evaluation of Berg's correlation-pleides concept. American Journal of Botany 86: 39-55
Armbruster W. S. K. E. Schwaegerle 1996 Causes of covariation of phenotypic traits among populations. Journal of Evolutionary Biology 9: 261-276[CrossRef][ISI]
Baker H. G. 1955 Self-compatibility and establishment after ‘long-distance’ dispersal. Evolution 9: 347-348[CrossRef][ISI]
Baker H. G. 1967 Support for Baker's Law as a rule. Evolution 21: 853-856[CrossRef][ISI]
Conner J. S. Via 1993 Patterns of phenotypic and genetic correlations among morphological and life-history traits in wild radish, Raphanus raphanistrum. Evolution 47: 704-711[CrossRef][ISI]
Conner J. K. 2002 Genetic mechanisms of floral trait correlations in a natural population. Nature (London) 420: 407-410[CrossRef][Medline]
Conner J. K. R. Franks C. Stewart 2003 Expression of additive genetic variances and covariances for wild radish floral traits: comparison between field and greenhouse environments. Evolution 57: 487-495[ISI][Medline]
Conner J. K. A. Sterling 1996 Selection for independence of floral and vegetative traits: evidence from correlation patterns in five species. Canadian Journal of Botany 74: 642-644[ISI]
Cugen J. M. Acheroy A. L. Loutfi D. Petit P. Vernet 1989 Breeding system differentiation in Arrhenatherum elatius populations: evolution toward selfing?. Evolutionary Trends in Plants 3: 17-24
Culley T. M. 2002 Reproductive biology and delayed selfing in Viola pubescens (Violaceae), an understory herb with chasmogamous and cleistogamous flowers. International Journal of Plant Sciences 163: 113-122[CrossRef]
Darwin C. 1876 The effects of cross- and self-fertilization in the vegetable kingdom. John Murray, London, UK
Diggle P. K. 1992 Development and the evolution of plant reproductive characters. In R. Wyatt [ed.], Ecology and evolution of plant reproduction: new approaches, 326–355. Chapman and Hall, New York, New York, USA
Donnelly S. E. C. J. Lortie L. W. Aarssen 1998 Pollination in Verbascum thapsus (Scrophulariacae): the advantage of being tall. American Journal of Botany 85: 1618-1625
Dubois S. P.-O. Cheptou C. Petit P. Meerts M. Poncelet X. Vekemans C. Lefebvre J. Escarre 2003 Genetic structure and mating systems of metallicolous and nonmetallicolous populations of Thlaspi caerulescens. New Phytologist 157: 633-641[CrossRef][ISI]
Eckhart V. M. M. A. Geber 2000 Character variation and geographic range in Clarkia xantiana (Onagraceae): breeding system and phenology distinguish two common subspecies. Madroño 46: 117-125
Ehlers B. K. 1999 Variation in fruit set within and among natural populations of the self-incompatible herb Centaurea scabiosa (Asteraceae). Nordic Journal of Botany 19: 653-663[ISI]
Fausto J. A., Jr. V. M. Eckhart M. A. Geber 2001 Reproductive assurance and the evolutionary ecoogy of self-pollination in Clarkia xantiana (Onagraceae). American Journal of Botany 88: 1794-1800
Fisher R. A. 1941 Average excess and average effect of a gene substitution. Annals of Eugenics 11: 53-63[ISI]
Fishman L. R. Wyatt 1999 Pollinator-mediated competition, reproductive character displacement, and the evolution of selfing in Arenaria uniflora (Caryophyllaceae). Evolution 53: 1723-1733[CrossRef][ISI]
Gallardo R. E. Dominguez J. M. Muñoz 1993 The heterochronic origin of the cleistogamous flower in Astragalus cymbicarpos (Fabaceae). American Journal of Botany 80: 814-823[CrossRef][ISI]
Gomez J. M. 2002 Self-pollination in Euphrasia willkommii Freyn (Scrophulariaceae), an endemic species from the alpine of the Sierra Nevada (Spain). Plant Systematics & Evolution 232: 63-71[CrossRef]
Goodwillie C. 1999 Wind pollination and reproductive assurance inLinanthus parviflorus (Polemoniaceae), a self-incompatible annual. American Journal of Botany 86: 948-954
Goodwillie C. 2001 Pollen limitation and the evolution of self-compatibility inLinanthus (Polemoniaceae). International Journal of Plant Sciences 162: 1283-1292[CrossRef][ISI]
Guerrant E. O. 1988 Heterochrony in plants: the intersection of evolution, ecology, and ontogeny. In M. L. McKinney [ed.], Heterochrony in evolution, 111–133. Plenum Press, New York, New York, USA
Guerrant E. O. 1989 Early maturity, small flowers and autogamy: a developmental connection?. In J. H. Bock and Y. B. Linhart [eds.], The evolutionary biology of plants, 61–84. Westview Press, Boulder, Colorado, USA
Herlihy C. R. C. G. Eckert 2002 Genetic cost of reproductive assurance in a self-fertilizing plant. Nature (London) 416: 320-323[CrossRef][Medline]
Herrera J. 2001 The variability of organs differentially involved in pollination, and correlations of traits in Genisteae (Leguminosae: Papilionoideae). Annals of Botany 88: 1027-1037
Hill J. P. E. M. Lord 1990 The role of developmental timing in the evolution of floral form. Developmental Biology 1: 281-287
Holsinger K. E. 1996 Pollination biology and the evolution of mating systems in flowering plants. Evolutionary Biology 29: 107-149[ISI]
Holsinger K. E. 2000 Reproductive systems and evolution in vascular plants. Proceedings of the National Academy of Sciences, USA 97: 7037-7042
Holtsford T. P. N. C. Ellstrand 1989 Variation in outcrossing rate and population genetic structure of Clarkia tembloriensis (Onagraceae). Theoretical and Applied Genetics 78: 480-488[CrossRef][ISI]
Holtsford T. P. N. C. Ellstrand 1992 Genetic and environmental variation in floral traits affecting outcrossing rate in Clarkia tembloriensis (Onagraceae). Evolution 46: 216-225[CrossRef][ISI]
Jain S. K. 1976 The evolution of inbreeding in plants. Annual Review of Ecology and Systematics 7: 469-495
Johnston M. O. 2000 Heterochrony in plant evolutionary studies through the twentieth century. Botanical Review 66: 57-88[ISI]
Kalisz S. D. Vogler B. Fails M. Finer E. Shepard T. Herman R. Gonzales 1999 The mechanism of delayed selfing in Collinsia verna (Scrophulariaceae). American Journal of Botany 86: 1239-1247
Kephart S. R. E. Brown J. Hall 1999 Inbreeding depression and partial selfing: evolutionary implications of mixed-mating in a coastal endemic, Silene douglasii var. oraria (Caryophyllaceae). Heredity 82: 543-554
Lewis H. M. E. Lewis 1955 The genus Clarkia. University of California Publications in Botany 20: 241-392
Lloyd D. G. 1979 Some reproductive factors affecting the selection of self-fertilization in plants. American Naturalist 113: 67-79[CrossRef][ISI]
Lloyd D. G. 1992 Self- and cross-fertilization in plants. II. The selection of self-fertilization. International Journal of Plant Sciences 153: 370-380[CrossRef]
Lord E. M. J. P. Hill 1987 Evidence for heterochrony in the evolution of plant form. In R. A. Raff and E. C. Raff [eds.], Development as an evolutionary process, 47–70. Alan R. Liss, New York, New York, USA
Mitchell-Olds T. 1996 Pleiotropy causes long-term genetic constraints on life-history evolution in Brassica rapa. Evolution 50: 1849-1858[CrossRef][ISI]
Moore D. M. H. Lewis 1965 The evolution of self-pollination in Clarkia xantiana. Evolution 19: 104-114[CrossRef][ISI]
Morgan-Richards M. K. Wolff 1999 Genetic structure and differentiation of Plantago major reveals a pair of sympatric sister species. Molecular Ecology 8: 1027-1036
Motton A. F. 1982 Autogamy and competition for pollinators in Hepatica americana (Ranunculceae). American Journal of Botany 69: 1296-1305[CrossRef][ISI]
Palmer M. J. Travis J. Antonovics 1989 Temporal mechanisms influencing gender expression and pollen flow within a self-incompatible perennial, Amianthum muscaetoxicum (Liliaceae). Oecologia 78: 231-236[CrossRef][ISI]
Petanidou T. A. C. Ellis-Adam J. C. M. Den Nijs J. G. B. Oostermeijer 1998 Pollination ecology of Gentianella uliginosa, a rare annual of the Dutch coastal dunes. Nordic Journal of Botany 18: 537-548[ISI]
Piper J. B. Charlesworth D. Charlesworth 1986 Breeding system evolution in Primula vulgaris and the role of reproductive assurance. Heredity 56: 207-217[ISI]
Ramsey M. 1996 Inbreeding depression and pollinator availability in a partially self-fertile perennial herb Blandfordia grandiflora (Liliaceae). Oikos 76: 465-474[CrossRef][ISI]
Runions C. J. M. A. Geber 2000 Evolution of the self-pollinating flower in Clarkia xantiana (Onagraceae). Size and development of floral organs. American Journal of Botany 87: 1439-1451
Schemske D. W. 1978 Sexual reproduction in an Illinois population of Sanguinaria canadensis L. American Midland Naturalist 100: 261-268[CrossRef][ISI]
Schoen D. J. M. T. Morgan T. Bataillon 1996 How does self-pollination evolve? Inferences from floral ecology and molecular genetic variation. Philosophical Transactions of the Royal Society of London, B, Biological Sciences 351: 1281-1290[CrossRef]
Sherry R. A. E. M. Lord 1996 Developmental stability in flowers of Clarkia tembloriensis (Onagraceae). Journal of Evolutionary Biology 9: 911-930[CrossRef][ISI]
Stebbins G. L. 1970 Adaptive radiation in angiosperms. I. Pollination mechanisms. Annual Review of Ecology and Systematics 1: 307-326
Strid A. 1969 Evolutionary trends in the breeding system of Nigella (Ranunculaceae). Botaniska Notiser 122: 380-396
Stucky J. M. 1985 Pollination systems of sympatric Ipomoea hederacea and Ipomoea purpurea and the significance of interspecific pollen flow. American Journal of Botany 72: 32-43[CrossRef][ISI]
Wendt T. M. B. F. Canela D. E. Klein R. I. Rios 2002 Selfing facilitates reproductive isolation among three sympatric species of Pitcairnia (Bromeliaceae). Plant Systematics & Evolution 232: 201-212[CrossRef]
Wyatt R. 1983 Pollinator-plant interactions and the evolution of breeding systems. In L. Real [ed.], Pollination biology, 51–95. Academic Press, Orlando, Forida, USA
Wyatt R. 1986 The ecology and evolution of self-pollination in Arenaria uniflora (Caryophyllaceae). Journal of Ecology 74: 403-418[CrossRef][ISI]
Wyatt R. 1988 Phylogenetic aspects of the evolution of self-pollination. In L. D. Gottlieb and S. K. Jain [eds.], Plant evolutionary biology, 109– 131. Chapman and Hall, London, UK
This article has been cited by other articles:
|
|
 |
|
  I. A. Anderson and J. W. Busch Relaxed pollinator-mediated selection weakens floral integration in self-compatible taxa of Leavenworthia (Brassicaceae) Am. J. Botany, June 1, 2006; 93(6): 860 - 867. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||
