| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
2Department of Biological Sciences, Idaho State University, Pocatello, Idaho 83209 USA; 3Department of Biological Sciences, Stanford University, Stanford, California 94305 USA; 4Department of Life Sciences, Indiana State University, Terre Haute, Indiana 47809 USA 5The Rocky Mountain Biological Laboratory, P.O. Box 519, Crested Butte, Colorado 81224 USA
Received for publication August 31, 2000. Accepted for publication February 13, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
= 0.055–0.126) than D. nuttallianum ( = 0.004–0.009) at spatial scales ranging from metres to 3.5 km. Spatial autocorrelation analysis also indicated more pronounced local genetic structure in D. barbeyi than D. nuttallianum populations. Fixation indices (FIS) of D. barbeyi adults were much lower than expected based on mating system equilibrium and suggest that differences in the degree of self-compatibility and/or the timing of postpollination selection/inbreeding depression between the two species further contribute to the genetic differences between them.
Key Words: breeding system • comparative study • floral display size • F statistics • geitonogamy • genetic structure • outcrossing rate; • pollinator behavior
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Richards, 1986
; Harder and Barrett, 1996
). In turn, the mating system is a primary determinant of plant population genetic structure. Inbreeding species are expected to have less genetic diversity and heterozygosity within populations and more genetic differentiation among populations than outcrossing species (Wright, 1921, 1951
; Allard, Jain, and Workman, 1968
; Jain, 1976
). Surveys comparing allozyme and quantitative genetic variation across a wide range of taxa generally support these theoretical predictions (Brown, 1979
; Hamrick, Linhart, and Mitton, 1979
; Schoen and Brown, 1991
; Charlesworth and Charlesworth, 1995
; Hamrick and Godt, 1996
). However, interspecific comparisons of many unrelated taxa may be confounded by the effects of correlated ecological traits or phylogeny in producing the observed pattern of genetic structure. In addition, comparisons of rare vs. widespread species indicate that measures of genetic diversity are often strongly correlated between congeneric species (Gitzendanner and Soltis, 2000
). Comparative studies of closely related taxa have the advantage of being able to better isolate the effects of variation in single traits. To this end, numerous intrageneric and intraspecific comparisons have examined the relationship between the mating system and allozyme diversity and heterozygosity within populations. Most of these studies support the expected reduction in diversity and heterozygosity of selfers relative to outcrossers (reviewed by Brown, 1979
; Schoen and Brown, 1991
; Charlesworth, Nordborg, and Charlesworth, 1997
).
In contrast, relatively few studies of closely related taxa have examined the effects of mating system variation on genetic differentiation among population subdivisions (Appendix). In most cases, these studies found that species with more highly selfing mating systems had greater inbreeding within subpopulations and more genetic differentiation among subpopulations. These studies mostly compared taxa with widely divergent mating systems (e.g., selfing vs. outcrossing or self-compatible vs. self-incompatible) and/or taxa that differ in flower traits affecting the breeding system (e.g., flower size, degree of protandry). Floral display, the number and arrangement of flowers on the plant, can also affect the mating system (Wyatt, 1982
; Geber, 1985
; Harder and Barrett, 1996
; Snow et al., 1996
). In particular, the number of open flowers on the plant should increase pollinator attraction, but is predicted to lead to more selfing through geitonogamy (the transfer of self pollen between flowers on the same plant) (Hessing, 1988
; de Jong, Waser, and Klinkhamer, 1993
; Harder and Barrett, 1996
; Snow et al., 1996
). These potentially conflicting effects of floral display size have been little studied in regard to their influence on the mating system and gene flow.
The purpose of our study was to investigate differences in floral display (plant size and number of open flowers) and the breeding system (autogamy and self-compatibility) and their effects on pollinator behavior, outcrossing rates, and population genetic structure in sympatric Delphinium barbeyi Huth and Delphinium nuttallianum Pritzel (= D. nelsonii Greene) (Ranunculaceae). These species share most life-history traits thought to affect genetic structure. Both are long-lived herbaceous perennials. They co-occur in the same subalpine meadows, with D. nuttallianum found on somewhat drier microsites. Delphinium barbeyi has limited clonal spread, forming distinct clumps, and D. nuttallianum has no vegetative reproduction. Both have gravity-dispersed seeds. The two species have morphologically similar, protandrous flowers (D. barbeyi flowers are slightly smaller). Although D. nuttallianum blooms earlier in the season, the species have similar flower visitors. The main pollinators of Delphinium barbeyi are hummingbirds (Selasphorus platycercus, S. rufus, and Stellula calliope), and queen and worker bumble bees (especially Bombus appositus and B. flavifrons). Delphinium nuttallianum is visited more by S. platycercus and queen bees (Waser, 1982
). Delphinium is a predominantly outcrossing but self-compatible genus (Epling and Lewis, 1952
), with widely varying degrees of self-incompatibility (0–50%) and autogamy (1–90%) reported among species (Macior, 1975
; Varney, 1979
; Powell and Jones, 1983
; Bosch, 1999
). Autogamy rates, partial self-incompatibility, and inbreeding depression have been well documented in D. nuttallianum (Waser, 1978
; Price and Waser, 1979
; Waser and Price, 1991, 1994
). Comparable data on the breeding system of D. barbeyi are presented here.
The primary morphological difference between the two species is the size of their floral display. Delphinium barbeyi produces hundreds of flowers per genet, usually on multiple stalks, whereas D. nuttallianum produces only a few flowers on a single raceme (Waser, 1982
). Though pollinators fly similar distances between flowers in the two species (Waser, 1982
), the higher density of genetically distinct individuals in D. nuttallianum means that pollen travels to a greater number of genets. Conversely, D. barbeyi potentially receives more self pollen by virtue of the greater proportion of intraplant pollinator flights (D. barbeyi, 67.4%, N = 954 visits; D. nuttallianum, 44.8%, N = 2336 visits; 
2 = 138.77, df = 1, P < 0.0001; reanalyzed data of Waser, 1982
). These differences in pollen dispersal should lead to more inbreeding and less gene flow (Crawford, 1984
) in D. barbeyi than in D. nuttallianum. We therefore expect a lower outcrossing rate (t), a higher inbreeding coefficient (f or FIS), and a greater degree of population subdivision (higher 
or FST) in many-flowered D. barbeyi than in few-flowered D. nuttallianum. We report here comparisons of floral display size, pollinator foraging behavior, breeding system, outcrossing rates estimated from progeny arrays, and genetic structure over a variety of spatial scales for sympatric populations of D. barbeyi and D. nuttallianum.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Bosch and Waser, 1999
). Delphinium nuttallianum begins flowering shortly following snowmelt in late May or early June. Delphinium barbeyi typically begins blooming in late June or early July, at the end of the flowering period of D. nuttallianum (Waser, 1978
). All observations and experiments were carried out in Delphinium populations along a 3.5-km corridor of the East River valley and its tributary Copper Creek.
Floral display size
The number of flowers per inflorescence and number of inflorescences (flowering stems) per plant were examined on 61 D. barbeyi plants in the Gothic Research Meadow in 1990. Additionally, 17 plants in 1990 and 12 plants in 1992 were monitored throughout the flowering period to estimate the time of peak flowering and the mean number of flowers open per day. The number of flowers and inflorescences per plant were estimated for D. nuttallianum in four populations in 1991. The number of open flowers in the male phase (during anther dehiscence) and female phase (during stigma receptivity) were counted for each plant.
Pollinator behavior at Delphinium barbeyi
Paired plant experiment
We observed visitation by bumble bees to paired large and small plants in the Gothic Research Meadow. Per-plant and per-flower visit rates to two neighboring plants differing in size were recorded in seven 30-min trials between 1430 and 1730 MDT on 5–9 August 1990. Because the distributions of visitation measures were highly skewed, the central tendencies are presented as medians and midranges (middle 50% of observations) in this and the following analysis. Differences between large and small plants in visitation measures were analyzed using Wicoxon matched-pairs signed ranks tests (Sokal and Rohlf, 1981
).
Individual foraging bouts
Foraging bouts of individual bumble bees and hummingbirds were observed in a patch of 60 marked D. barbeyi plants between 28 July and 5 August 1990. We followed bumble bees foraging within the patch, placed a numbered flag by each plant visited, and recorded the number of flowers visited. We subsequently counted the number of open flowers on each visited plant. Hummingbirds were observed with binoculars from a distance of 10–25 m as they visited plants marked with numbered flags. For each bout we recorded the number of flowers visited on flagged, numbered plants. We counted the number of open flowers on marked plants each day. For bees and hummingbirds, we analyzed the relationship between plant size and flower visitation frequency using linear regression (Sokal and Rohlf, 1981
).
Inflorescence number
We also examined the effect of D. barbeyi plant size on bee foraging behavior by classifying plant size as the number of inflorescences with at least two open flowers. Plants were divided into 6 different size classes (1, 2–5, 6–10, 11–15, 16–24, and >24 inflorescences). Individual plants were observed for 10 min each, between 1100 and 1800 on 22–26 July 1994. No plant was observed more than twice. We recorded the number of bee visits to a plant per 10-min observation period and the number of different inflorescences visited during each bee's foraging bout. The relationships between the response variables (bee visits and number of inflorescences visited) and plant size were analyzed using one-way ANOVA (Sokal and Rohlf, 1981
). Plants were further grouped into small (
10 inflorescences) and large (>10 inflorescences) size classes and differences in foraging between these two groups were analyzed using ANOVA.
Breeding system
We investigated the breeding system of D. barbeyi using five different pollination treatments. Each treatment was performed on a separate inflorescence and replicated on eight different plants. At least nine flowers on each plant received each treatment. To test for autogamy, flowers were left unemasculated and the inflorescence covered with a fine mesh bag to exclude pollinators. To compare self and outcross pollination, inflorescences were kept covered and new male-phase flowers were emasculated each day. Pollen was applied daily to receptive stigmas using toothpicks. We hand-pollinated each flower for three consecutive days to assure female receptivity. For the hand-self treatment the pollen source was another flower from the same plant. For the hand-outcross treatment flowers from plants growing at least 10 m away served as donors. We marked two additional inflorescences on each plant, which were left open (unbagged) to pollinator visits. Flowers on one open inflorescence were unmanipulated and served as an open-pollinated control (natural treatment). Flowers on the other open-pollinated inflorescence were also hand-outcrossed daily with supplemental pollen to test for pollen limitation. We performed pollination treatments from 24 July to 6 August 1990 and collected fruits to assess seed set on 13 August. Because not all treatments were successful on each plant, we used paired t tests to compare seed set per flower between paired treatments on each plant. We compare these results to those published for D. nuttallianum (Waser, 1978
; Waser and Price, 1991
).
Genetic techniques
We used allozymes as genetic markers to estimate the mating system and genetic structure for both species. Horizontal starch-gel electrophoresis was performed on both leaf material and seeds. The electrophoretic techniques and grinding and running buffers for leaves and seeds of both species were as described in Williams and Waser (1999)
for D. nuttallianum leaves. In D. barbeyi, we resolved three polymorphic isozymes in both leaf and seed material: phosphoglucose isomerase (Pgi-2; three alleles), phosphoglucomutase (Pgm-1; three alleles), and acid phosphatase (Acp-2; four alleles). Allozyme results from leaf material used for analysis of genetic structure in D. nuttallianum are reported in Williams and Waser (1999)
. In D. nuttallianum we resolved four polymorphic isozymes from seeds: malate dehydrogenase (Mdh-1, two alleles; Mdh-2, three alleles), phosphoglucomutase (Pgm-1; three alleles), and phosphoglucose isomerase (Pgi-2, three alleles). These loci yielded clear banding patterns consistent with their previously reported quaternary structure (Kephart, 1990
) and segregated in expected Mendelian fashion in progeny arrays.
Outcrossing rates
We estimated the mating system of each species from allozyme genotypes of seeds in progeny arrays. Fruits of D. nuttallianum were collected from two populations at each of two sites (populations KPA, KPB, JFA, and JFB of Williams and Waser, 1999
) in 1991. Fruits were air dried and stored at room temperature in coin envelopes for 12–18 mo before using seeds for electrophoresis. Delphinium barbeyi seed collected for mating system analysis in 1995–1996 were lost because of insect predation. We successfully collected mature infructescences from 22 D. barbeyi plants in the KPB population in late August 1997. Infructescences were first oven dried at 50°C for several days to arrest fungal growth and destroy the eggs and larvae of a dipteran seed predator. We stored fruits at room temperature with a small amount of para-dichlorobenzene until using the seeds for electrophoresis in January 1998.
We estimated the outcrossing rates of both Delphinium species using the multilocus maximum-likelihood, mixed-mating procedure of Ritland and Jain (1981)
with the computer program MLT written by K. Ritland. All seeds from a single maternal plant (from 1 to 5 fruits) were treated as separate families or progeny arrays. The maternal genotypes of most D. nuttallianum plants were known from separate analysis of leaf material. We inferred maternal genotypes of all D. barbeyi and a few D. nuttallianum plants from progeny arrays using the procedure of Brown and Allard (1970)
, implemented in MLT. In the analysis, pollen and ovule genotype frequencies were estimated separately. The standard errors (SE) of multilocus outcrossing rates were calculated from the distribution of 500 bootstrap estimates, where the unit of resampling was the family (progeny array of a maternal plant).
Spatial genetic structure
Sampling design
We used a nested, hierarchical sampling design to analyze the genetic structure of both species. In 1995 Delphinium barbeyi were sampled at four sites, each site containing three populations, and each population made up of three subpopulations. These four sites (KP, RM, CC, and BP) correspond to sympatric D. nuttallianum sites (Kettle Ponds, Research Meadow, Copper Creek, and 401 Trail, respectively) sampled in an earlier study (Williams and Waser, 1999
). The four sites were isolated by forest or other habitat barriers unsuitable for Delphinium and were separated by 
1 km. We located three semi-isolated populations, 40–120 m apart (e.g., populations KPA, KPB, and KPC at site KP), which formed distinct patches within the larger metapopulation at each site. In each population we mapped all plants within 10 m of a transect through its center. Transects were further subdivided along their length in approximate thirds to create three subpopulations per population. Sample sizes per population ranged from 25 to 125 plants (total N = 725 mapped plants). A small amount of leaf material was collected from each mapped plant and refrigerated until used for electrophoresis, usually within 24 h.
We sampled D. nuttallianum at the same four sites used for D. barbeyi, plus two additional sites within the East River valley. Each of the six sites contained two (three at one site) populations, and each population was composed of four subpopulations. Sample sizes ranged from 40 to 340 individuals per population (total N = 1620 plants). Further details of the sampling design and a map of sampled sites for D. nuttallianum are described elsewhere (Williams and Waser, 1999
).
F statistics
We estimated hierarchical F statistics using the approach of Weir and Cockerham (1984)
, in which total genetic variation, FIT, is apportioned into components due to differentiation among nested population subdivisions, [equivalent to Wright's (1951)
FST] and that due to fixation within the smallest subdivisions, FIS. We calculated genotype frequencies and estimated genetic differentiation among three nested levels of population subdivision. These three estimators are: (1) SUBPOP, differentiation among subpopulations within populations, (2) POP, differentiation among populations within sites, and (3) SITE, differentiation among sites within the total sample. Comparable hierarchical results for D. nuttallianum are summarized from an earlier paper (Williams and Waser, 1999
). The 95% confidence intervals (CIs) of the multilocus estimates of each F statistic were calculated from the distribution of 1500 bootstraps over loci (Weir and Cockerham, 1984
). We tested deviations of from zero and differences between species' estimates by nonoverlap of the 95% confidence intervals of multilocus estimates.
Spatial autocorrelation
Analyses of spatial autocorrelation between individual plants on single-locus allele frequency data were performed using a computer program provided by J. S. Heywood. The autocorrelation coefficient, Moran's I, was calculated for each allele between all nearest-neighbor pairs of individuals (INN) in each of five D. barbeyi populations containing >75 mapped plants and in four mapped D. nuttallianum populations (Williams and Waser, 1999
). Autocorrelation coefficients were also calculated among all pairs of individuals in 1-m distance intervals over 0–20 m in these populations. We considered single allele estimates of autocorrelation coefficients significantly different from random expectations at P < 0.05 (two-tailed) if the standard normal deviate (SND) of the estimate was 1.96, where SND = [I – µ]/
(Heywood, 1991
). We calculated the mean autocorrelation coefficients for nearest neighbors and at 1-m intervals for each population by averaging single-allele estimates in each distance class. The least common allele at each locus was omitted from the averages to maintain statistical independence of samples.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
170 flowers per plant were open at peak bloom (Table 1).
|
20–30 d (F. Saavedra and C. Williams, unpublished data), beginning soon after snowmelt in late May or early June. In 1991 the total number of flowers per plant averaged 4.7 flowers in four populations. The average number of open flowers per plant was 3.1, 2.1 in male phase and 1.0 in female phase (Table 1).
Pollinator behavior
Paired plant experiment
A total of 2965 visits by bumble bees to D. barbeyi flowers were recorded in seven 30-min observations of paired large and small plants in 1990. The larger plants had significantly more bumble bee approaches, a significantly greater median number of flowers visited, and significantly higher per flower visitation rates than the small plants (Table 2). Bees often visited a relatively large number of flowers per plant, especially on larger plants (Table 2). Workers of Bombus appositus and B. flavifrons accounted for 80 and 19%, respectively, of flower visits, with the remainder by queen B. appositus.
|
|
10 inflorescences), the number of bee visits per 10 min was significantly greater for large than for small plants (F1,59 = 15.56, P = 0.0002; Table 4). Bees also visited multiple inflorescences on the same D. barbeyi plant during a foraging bout. Although there were no significant differences in number of inflorescences visited among the six size classes (F5,81 = 1.71, P = 0.142), when plants are grouped into large and small size classes as above, bees visited significantly more inflorescences per bout on large plants than small plants (F1,85 = 6.82, P = 0.011; Table 4).
|
|
) showed that autogamy was rare, seed set of bagged flowers averaging only 2.2% of open-pollinated flowers (Table 5). Likewise, three previous experiments (Waser and Price, 1991
) demonstrated that self-pollinated flowers produced an average of only 59.4% as many seeds as outcrossed matings between plants growing 10 m apart (Table 5).
Outcrossing rates
The multilocus outcrossing rate estimate (tm) for D. barbeyi was significantly lower than those of D. nuttallianum (Table 6). Multilocus estimates were significantly <1.0 in D. barbeyi and D. nuttallianum populations KPA and KPB. The selfing rate (1 – tm) in D. barbeyi was almost 45%, indicating a mixed-mating system. In contrast, the selfing rate averaged only 7% in the four D. nuttallianum populations. Differences between average single-locus (ts) and multilocus (tm) outcrossing rates, an estimate of the contribution of biparental inbreeding to estimated selfing, are small and not significant.
|
were >10 times larger, suggesting much more restricted gene flow in D. barbeyi. There was significantly greater genetic differentiation at the two larger scales than at the smallest scale (among subpopulations within populations) for D. barbeyi. In D. nuttallianum there were no significant differences in the degree of genetic differentiation among different levels of the spatial hierarchy. Both 
IT and IS differed significantly between the two species. IT was greater in D. barbeyi as expected. However, IS was significantly smaller in D. barbeyi than in D. nuttallianum and significantly <0, indicating an unexpected heterozygote excess in adults of the apparently more inbred species.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
found that pollinators (mostly bees) visit an average of only 2.5 flowers per plant in D. nuttallianum. Furthermore, bumble bee movement among the strongly protandrous flowers of D. nuttallianum is characterized by bees arriving at older female-phase flowers low on the inflorescence and moving up, subsequently visiting younger male-phase flowers, promoting outcrossing (Pyke, 1978
). There thus appears to be only limited opportunity for self-pollination, either by autogamy or geitonogamy, in this species.
Delphinium barbeyi on the other hand has an average of 170 flowers on multiple inflorescences open per day. Consequently, pollinators often visit many flowers per plant, and move between several inflorescences on the same plant during a foraging bout. Such interinflorescence movements are the most likely mechanism leading to self-pollination in D. barbeyi. More haphazard interflower and interplant movements by hummingbirds, which do not adhere as closely to the "bottom-up" foraging pattern of bumble bees (N. Waser, personal communication), may further contribute to selfing in both species. Although the pollinator taxa are qualitatively similar between the two species, quantitative differences in pollinator abundance, pollen carryover, and seasonal variation in interspecific plant movements may also affect differences in the mating system between these Delphinium species. Because plant size increases attractiveness to pollinators in both species (e.g., Schulke, 1999
for D. nuttallianum), more detailed investigations of the effects of intraspecific variation in floral display on pollinator behavior and the mating system are needed.
Outcrossing rates and breeding systems
As predicted because of its larger floral display, D. barbeyi had a significantly greater selfing rate than D. nuttallianum. Pollinator foraging behavior in response to plant size and inflorescence number may in part explain differences in the two species' outcrossing rates. Comparable outcrossing rate estimates are available for the narrow endemic Delphinium viridescens (Richter, 1993
). Bumble bee-pollinated D. viridescens is intermediate in size between D. nuttallianum and D. barbeyi with an average of 46 flowers on a single stalk (Varney, 1979
). Outcrossing rates estimated from progeny arrays in five populations of D. viridescens averaged 0.717 (range 0.641–0.820; Richter, 1993
), also intermediate between the two Delphinium species examined in this study. Outcrossing rates estimated for the above and a number of other Delphinium species in North America (N = 1; Dodd, 1997
) and the Meditteranean (N = 8; Bosch, 1999
) vary widely (0.62–1.0) and do not show an association with floral display size. However, most of these mating system estimates are equilibrium expectations derived from adult fixation indices (t = (1 – F)/(1 + F); Haldane, 1924)
and may not reflect the true mating system (see below).
Postpollination events may further act to modify the mating system. Self-incompatibility, autogamy, and inbreeding depression may all influence the success of particular matings, and hence the genetic estimates of outcrossing made at the seed and other stages of the life cycle. Delphinium barbeyi produced more seed by autogamy (9%) than did D. nuttallianum (2%; Waser, 1978
). Delphinium barbeyi also appears to be self-compatible, in contrast to its congener, which sets 40% fewer seed in hand-self vs. hand-outcross pollinations (Waser and Price, 1991
). For these reasons differences in the mating system between these two species associated with differences in floral display size will be magnified by differences in their breeding systems.
If inbreeding from geitonogamous pollination has a detrimental effect on fitness, then a greater degree of self-incompatibility may evolve in species with larger floral displays (McDade, 1985
). This is clearly not the case with D. barbeyi and D. nuttallianum, which show the opposite pattern. Likewise, there is no clear relationship between floral display size and the degree of self-compatibility among these and several other Delphinium species reported in the literature (N = 10). There is a similar lack of association between autogamy rates and floral display size (N = 19 species). These aspects of the breeding system appear to be relatively independent of floral display, although annuals have somewhat higher autogamy rates and are more self-compatible than perennials (Macior, 1975
; Varney, 1979
; Powell and Jones, 1983
; Bosch, 1999
).
Population genetic structure
We predicted that more intraplant pollen tranfer in D. barbeyi would result in reduced gene flow and greater genetic differentiation among population subdivisions compared to D. nuttallianum. This prediction was strongly supported by the >10 times higher estimates of at all spatial scales in the hierarchical F statistics analysis and more pronounced local genetic structure indicated by spatial autocorrelation for D. barbeyi. Reviews of genetic structure studies based on allozyme variation indicate that the mating system is one of the strongest determinants of genetic differentiation across a wide range of plant taxa (Heywood, 1991
; Hamrick and Godt, 1996
). Likewise, almost all previous interspecific comparisons of congeneric species and intraspecific comparisons of populations differing in their breeding systems have found greater genetic differentiation among populations of the more inbred taxa (Appendix). The magnitude of genetic differentiation () observed in these two sympatric Delphinium species is overall lower, and differences between species somewhat less, than seen in the species and populations compared in the Appendix. This may be due in part to the fact that Delphinium populations were sympatric and sampled over a much more limited geographic area than previous studies. It may also reflect that differences in geitonogamy and the breeding system between these two Delphinium species reduce gene flow less strongly than the somewhat different factors affecting the mating system and genetic structure in the other species.
Two comparable studies, one from 24 populations of Delphinium variegatum from San Clemente Island (Dodd, 1997
) and the other from 17 populations of the narrow endemic D. viridescens (Richter, 1993
; Richter, Soltis, and Soltis, 1994
), further support the relationship between floral display size and spatial genetic structure. Delphinium variegatum is of similar size to D. nuttallianum and has a similar low estimate of FIS (0.067) and a somewhat higher FST estimate (0.047). The geographic range of sampling for this species was 25 km. Delphinium viridescens has a single inflorescence and is intermediate in total floral display size (46 flowers) between D. nuttallianum and D. barbeyi. It has somewhat higher genetic differentiation among populations (ST = 0.209) than does D. barbeyi (SITE = 0.126), although populations were more widely dispersed (1–30 km) than in our study (1–3 km). For these four genetically well-characterized Delphinium species the predicted relationships among floral display size, mating systems, and population genetic structure appear to hold. However, there is no significant association between floral display and genetic differentiation among populations when five additional Delphinium species studied by Bosch (1999)
are considered.
A puzzling result was the high heterozygosity observed in adult D. barbeyi despite low outcrossing rates estimated from seeds. This result also appears at odds with other intrageneric and intraspecific studies in which the more inbred taxa have higher estimates of FIS (Appendix). This suggests that significant changes in the genetic composition of the populations occur between the seed and adult stages of the life cycle of D. barbeyi, but not D. nuttallianum. An increase in heterozygote frequency throughout the life cycle has been found in numerous studies of genetic demography (e.g., Kahler, Clegg, and Allard, 1975
; Tonsor et al., 1993
; Hossaert-McKey et al., 1996
) and suggests selection for heterozygosity. Schoen (1982)
, Glover and Barrett (1987)
, and Holtsford and Ellstrand (1989)
also found a pattern of excess heterozygosity in adults relative to inbreeding equilibrium expectations in selfing populations, while outcrossing populations showed slight heterozygote deficits. A similar pattern is seen in the comparison of highly outcrossing Plantago lanceolata and mixed-mating P. coronopus (Wolff, 1991
; Appendix).
Selection for particular heterozygous genotypes in D. barbeyi appears unlikely as the cause of its adult heterozygote excesses because similar patterns are found at all three polymorphic loci studied. Also, population subdivisions differed sufficiently in their genotype and allele frequencies at various spatial scales to produce significant genetic differentiation among them. Gitzendanner and Soltis (2000)
point out that genetic variability measures are often correlated among rare and widespread species within a genus. If levels of heterozygosity were conserved among Delphinium species, then deviations from outcrossing equilibrium estimates such as seen in these two species might result. However, the excess heterozygosity seen in D. barbeyi appears to reflect high overall levels of adult heterozygosity (HO = 0.54; this study) compared to those reported for other 15 Delphinium species (HO = 0.17, range = 0.07–0.30; Richter, 1993
; Dodd, 1997
; Bosch, 1999
), including D. nuttallianum (HO = 0.16; this study).
Greater demographic change in heterozygosity in D. barbeyi relative to D. nuttallianum may reflect differences in the magnitude and timing of inbreeding depression in the two species (Husband and Schemske, 1996
). Partial self-incompatibility, either from inbreeding depression or maternal control (Waser et al., 1987
), strongly reduces seed set in D. nuttallianum but not D. barbeyi. Early-acting inbreeding depression (on seed set) is expected to be more prevalent in outcrossing populations, while selfing populations, which may already be purged of some of their genetic load, are predicted to exhibit greater late-acting inbreeding depression (on survival and reproduction) (Husband and Schemske, 1996
). Because the mating system is estimated from surviving seed, early selection against selfing may reduce the differences in fixation indices estimated from seed and adults in D. nuttallianum more than D. barbeyi. Alternatively, changes in the mating system over generations or year-to-year variation in the outcrossing rate could explain the discrepency between selfing rate and adult heterozygosity in D. barbeyi.
Conclusions
This interspecific comparison of two Delphinium species has demonstrated that floral display is associated with predicted changes in pollinator foraging behavior, the mating system, and population genetic structure. Comparable studies of two other North American Delphinium further suggest that species with larger floral displays are more self-pollinated and have reduced gene flow. However, these results should be interpreted with caution. Differences in the breeding systems (autogamy and self-compatibility) of these species may exaggerate the influence of floral display on the estimated genetic parameters. A brief survey of Delphinium species suggest that breeding systems vary widely and are not strongly associated with floral display size. More detailed intraspecific studies of variation in plant size and its effects on pollinator behavior and outcrossing rates are needed to understand the trade-offs in the evolution of floral display size. The wide variation in floral display size and long-lived perennial habit of Delphinium barbeyi provides an excellent system in which to test these relationships. Likewise, further interspecific studies of floral display size, breeding systems, outcrossing rates, and spatial genetic structure are necessary to gain a better understanding of their relationships. The genus Delphinium provides ample opportunities for such comparisons. Over 60 species of Delphinium occur in North America, of which D. nuttallianum and D. barbeyi represent the extremes of floral display size. These two species are placed in different subsections in the most recent taxonomic treatment of North American Delphinium (Warnock, 1997
), which may introduce historical factors as a potential cause of their breeding system and genetic differences. Future studies of the relationships between floral display, the breeding system, outcrossing rates, and genetic structure should capitalize on the wide range of variation in Delphinium, especially if combined with phylogenetic information allowing sister taxa to be compared.
|
|
FOOTNOTES |
|---|
6 Author for reprint requests (willcha2{at}isu.edu
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Allard R. W. S. K. Jain P. L. Workman 1968 The genetics of inbreeding populations. Advances in Genetics 14: 55-131
Bosch M. 1999 Biologia de la reproducció de la tribu Delphinieae a la Mediterrània occidental. Arxius de la Secció de Ciències, 120. Institut d'Estudis Catalans, Barcelona, Spain
———, and N. M. Waser 1999 Effects of local density on pollination and reproduction in Delphinium nuttallianum and Aconitum columbianum (Ranunculaceae). American Journal of Botany 86: 871-879
Brown A. H. D. 1979 Enzyme polymorphism in plant populations. Theoretical Population Biology 15: 1-42
———, and R. W. Allard 1970 Estimation of the mating system in open-pollinated maize using isozyme polymorphisms. Genetics 66: 133-145
Charlesworth B. M. Nordborg D. Charlesworth 1997 The effects of local selection, balanced polymorphism and background selection on equilibrium patterns of genetic diversity in subdivided populations. Genetical Research 70: 155-174[CrossRef][ISI][Medline]
Charlesworth D. B. Charlesworth 1995 Quantitative genetics in plants: the effect of breeding system on genetic variability. Evolution 49: 911-920[CrossRef][ISI]
———, and Z. Yang 1998 Allozyme diversity in Leavenworthia populations with different inbreeding levels. Heredity 81: 453-461
Crawford T. J. 1984 What is a population?. In B. Shorrocks [ed.], Evolutionary ecology, 135–173. Blackwell, Oxford, UK
de Jong T. J. N. M. Waser P. G. L. Klinkhamer 1993 Geitonogamy: the neglected side of selfing. Trends in Ecology and Evolution 8: 321-325
Dodd S. C. 1997 Genetic diversity in Delphinium variegatum (Ranunculaceae): a comparison of two island endemic subspecies and their widespread mainland relative. M.S. thesis, San Diego State University, San Diego, California, USA
Ellstrand N. C. D. A. Levin 1980 Recombination system and population structure in Oenothera. Evolution 34: 923-933[CrossRef][ISI]
Epling C. H. Lewis 1952 Increase in the adaptive range of the genus Delphinium. Evolution 6: 253-267
Fenster C. B. K. Ritland 1992 Chloroplast DNA and isozyme diversity in two Mimulus species (Scrophulariaceae) with contrasting breeding systems. American Journal of Botany 79: 1440-1447[CrossRef][ISI]
Geber M. A. 1985 The relationship of plant size to self-pollination in Mertensia ciliata. Ecology 66: 762-777[CrossRef][ISI]
Gitzendanner M. A. P. S. Soltis 2000 Patterns of genetic variation in rare and widespread plant congeners. American Journal of Botany 87: 783-792
Glover D. E. S. C. H. Barrett 1987 Genetic variation in continental and island populations of Eichornia paniculata (Pontederiaceae). Heredity 59: 7-17[ISI]
Haldane J. B. S. 1924 A mathematical theory of natural and artificial selection. Part II. The influence of partial self-fertilization, inbreeding, assortative mating, and selective fertilization on the composition of Mendelian populations and on natural selection. Proceedings of the Cambridge Philosophical Society, Biological Sciences 1: 158-163
Hamrick J. L. M. J. W. Godt 1996 Effects of life history traits on genetic diversity in plant species. Philosophical Transactions of the Royal Society of London B 351: 1291-1298[CrossRef]
———, Y. B. Linhart J. B. Mitton 1979 Relationships between life history characteristics and electrophoretically detectable genetic variation in plants. Annual Review of Ecology and Systematics 10: 173-200
Harder L. D. S. C. H. Barrett 1996 Pollen dispersal and mating patterns in animal pollinated plants. In D. G. Lloyd and S. C. H. Barrett [eds.], Floral biology: studies on floral evolution in animal-pollinated plants, 140–190. Chapman and Hall, New York, New York, USA
Hessing M. B. 1988 Geitonogamous pollination and its consequences in Geranium caespitosum. American Journal of Botany 75: 1324-1333[CrossRef][ISI]
Heywood J. S. 1991 Spatial analysis of genetic variation in plant populations. Annual Review of Ecology and Systematics 22: 335-355[CrossRef][ISI]
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]
Hossaert-McKey M. M. Valero D. Magda M. Jarry J. Cuguen P. Vernet 1996 The evolving genetic history of a population of Lathyrus sylvestris: evidence from temporal and spatial genetic structure. Evolution 50: 1808-1821[CrossRef][ISI]
Husband B. C. D. W. Schemske 1996 Evolution of the magnitude and timing of inbreeding depression in plants. Evolution 50: 54-70
Inoue K. T. Kawahara 1990 Allozyme differentiation and genetic structure in island and mainland Japanese populations of Campanula punctata (Campanulaceae). American Journal of Botany 77: 1440-1448[CrossRef][ISI]
Jain S. K. 1976 The evolution of inbreeding in plants. Annual Review of Ecology and Systematics 7: 469-495
Kahler A. L. M. T. Clegg R. W. Allard 1975 Evolutionary changes in the mating system of an experimental population of barley (Hordeum vulgare L.). Proceedings of the National Academy of Sciences, USA 72: 943-946
Kephart S. R. 1990 Starch gel electrophoresis of plant isozymes: a comparative analysis of techniques. American Journal of Botany 77: 693-712[CrossRef][ISI]
Langenheim J. H. 1962 Vegetation and environmental patterns in the Crested Butte area, Gunnison County, Colorado. Ecological Monographs 32: 249-285[CrossRef][ISI]
Layton C. R. F. R. Ganders 1984 The genetic consequences of contrasting breeding systems in Plectritis (Valerianaceae). Evolution 38: 1308-1325[CrossRef][ISI]
Levin D. A. 1978 Genetic variation in annual Phlox: self-compatible versus self-incompatible species. Evolution 32: 245-263[CrossRef][ISI]
Loos B. P. 1993 Allozyme variation within and between populations in Lolium (Poaceae). Plant Systematics and Evolution 188: 101-113[ISI]
Macior L. W. 1975 The pollination ecology of Delphinium tricorne (Ranunculaceae). American Journal of Botany 62: 1009-1016[CrossRef][ISI]
McDade L. A. 1985 Breeding systems of Central American Aphelandra (Acanthaceae). American Journal of Botany 72: 1515-1521[CrossRef][ISI]
Powell E. A. C. E. Jones 1983 Floral mutualism in Lupinus benthamii (Fabaceae) and Delphinium parryi (Ranunculaceae). In C. E. Jones and R. J. Little [eds.], Handbook of experimental pollination biology, 310–329. Van Nostrand Reinhold, New York, New York, USA
Price M. V. N. M. Waser 1979 Pollen dispersal and optimal outcrossing in Delphinium nelsonii. Nature 277: 294-297[CrossRef]
Pyke G. H. 1978 Optimal foraging in bumblebees and coevolution with their plants. Oecologia 36: 282-293
Richards A. J. 1986 Plant breeding systems. Allen and Unwin, London, UK
Richter T. S. 1993 Conservation genetics and outcrossing rates of the narrow endemic, Delphinium viridescens. M.S. thesis, Washington State University, Pullman, Washington, USA
———, P. S. Soltis D. E. Soltis 1994 Genetic variation within and among populations of the narrow endemic, Delphinium viridescens (Ranunculaceae). American Journal of Botany 8: 1070-1076
Ritland K. S. K. Jain 1981 A model for the estimation of outcrossing rates and gene frequencies using n independent loci. Heredity 47: 35-52[ISI]
Schoen D. J. 1982 Genetic variation and the breeding system of Gilia achilleifolia. Evolution 36: 361-370[CrossRef][ISI]
———, and A. H. D. Brown 1991 Intraspecific variation in population gene diversity and effective population size correlates with the mating system in plants. Proceedings of the National Academy of Sciences, USA 88: 4494-4497
Schulke B. M. 1999 Experimental studies of long distance pollen-mediated gene flow in Delphinium nelsonii. M.S. thesis, University of California, Riverside, California, USA
Snow A. A. T. P. Spira R. Simpson R. A. Klips 1996 The ecology of geitonogamous pollination. In D. G. Lloyd and S. C. H. Barrett [eds.], Floral biology: studies on floral evolution in animal-pollinated plants, 191–216. Chapman and Hall, New York, New York, USA
Sokal R. R. F. J. Rohlf 1981 Biometry, 2nd ed. W. H. Freeman, New York, New York, USA
Tonsor S. J. S. Kalisz J. Fisher T. P. Holtsford 1993 A life-history based study of population genetic structure: seed bank to adults in Plantago lanceolata. Evolution 47: 833-843[CrossRef][ISI]
Van Dijk H. K. Wolff A. De Vries 1988 Genetic variability in Plantago species in relation to their ecology. 3. Genetic structure of populations of P. major, P. lanceolata and P. coronopus. Theoretical and Applied Genetics 75: 518-528[CrossRef][ISI]
Varney D. M. 1979 Reproductive biology of four species of Delphinium endemic to the Wenatchee Mountains. M.S. thesis, University of Washington, Seattle, Washington, USA
Warnock M. J. 1997 Delphinium. In Flora of North America Editorial Committee [eds.], Flora of North America north of Mexico, vol. 3, 196–240. Oxford University Press, New York, New York, USA
Waser N. M. 1978 Competition for hummingbird pollination and sequential flowering in two Colorado wildflowers. Ecology 59: 934-944[CrossRef][ISI]
———. 1982 A comparison of distances flown by different visitors to flowers of the same species. Oecologia 55: 251-257[CrossRef][ISI]
———, and M. V. Price 1991 Outcrossing distance effects in Delphinium nelsonii: pollen load, pollen tubes, and seed set. Ecology 72: 171-179[CrossRef][ISI]
———, and ———. 1994 Crossing distance effects in Delphinium nelsonii: outbreeding and inbreeding depression in progeny fitness. Evolution 48: 842-852[CrossRef][ISI]
———, ———, A. M. Montalvo R. N. Gray 1987 Female mate choice in a perennial herbaceous wildflower, Delphinium nelsonii. Evolutionary Trends in Plants 1: 29-33
Weir B. S. C. C. Cockerham 1984 Estimating F-statistics for the analysis of population structure. Evolution 38: 1358-1370[CrossRef][ISI]
Williams C. F. N. M. Waser 1999 Spatial genetic structure of Delphinium nuttallianum populations: inferences about gene flow. Heredity 83: 541-550
Wolff K. 1991 Analysis of allozyme variability in three Plantago species and a comparison to morphological variability. Theoretical and Applied Genetics 81: 119-126[ISI]
