| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
Division of Biological Sciences, University of Missouri–Columbia, Columbia, Missouri 65211 USA
Received for publication March 26, 2005. Accepted for publication July 18, 2005.
ABSTRACT
Floral longevity, the time between corolla expansion and senescence, contributes directly and indirectly to a plant's overall fitness. Though mating opportunities for insect-pollinated species often differ among populations, few empirical studies have addressed whether floral longevity varies in a manner consistent with these differences. I conducted experiments at thermally distinct sites to examine whether the prevailing floral longevity model predicted such variation between a montane and an alpine population of Campanula rotundifolia. Staminate phase duration was significantly shorter for montane vs. alpine C. rotundifolia flowers in the presence of pollinators, but significantly longer when pollinators were excluded. Montane flowers had a significantly higher female fitness accrual rate, significantly shorter total longevities, and, unlike alpine flowers, were not pollen-limited. Delaying pollinator access to pistillate phase flowers significantly increased total longevity in alpine flowers only. Significant differences in total longevity between populations resulted from an extended pistillate phase in alpine flowers. Overall, the prevailing model accurately predicted the total floral longevity trends found here. However, I provide novel evidence for geographic and gender-specific plasticity in the floral longevity response to fitness accrual rate variation, suggesting C. rotundifolia populations in this study may be attuned to local schedules of pollinator activity.
Key Words: Campanulaceae • Colorado • conditional response • fitness accrual rate • floral longevity • geographic variation • pollination • pollen limitation
For plants dependent on external agents for pollen dissemination, floral longevity has indirect effects on pollen removal and deposition through its direct effect on the time available for pollinator visitation (Schoen and Ashman, 1995
; Bingham and Orthner, 1998
). Floral longevity, the period between anthesis and corolla senescence, varies among angiosperm species from several hours to several weeks (reviewed in van Doorn, 1997
) and is influenced by factors that include quantity and quality of pollinators, environmental conditions, and type of mating system (Primack 1985
; Bingham and Orthner, 1998
; Blionis and Vokou, 2001
; Rathcke, 2003
). Because individual populations of many plant species differ discreetly in their biotic and abiotic environments (Arroyo et al., 1985
; Galen, 1989
; Herrera, 1991
; Kearns and Inouye, 1994
; Bingham and Orthner, 1998
; Jonas and Geber, 1999
; Blionis and Vokou, 2001
; Caruso et al., 2003
), it is important to determine whether floral longevity varies in a manner consistent with these differences.
The optimization of floral longevity has most recently been modeled as a resource allocation tradeoff between the costs for constructing a new flower vs. those for maintaining an existing flower (Schoen and Ashman, 1995
; Charnov, 1996
). The Schoen and Ashman model utilizes fitness accrual rate through male and female function to determine the appropriate resource allocation strategy, and empirical studies on individual species have supported the model (Ashman and Schoen, 1994
, 1997
; Evanhoe and Galloway, 2003). However, the model's ability to predict geographic variation in floral longevity among populations of an individual species remains largely untested.
The senescence response of animal-pollinated flowers to pollen removal and pollen deposition has been studied and reviewed extensively (Devlin and Stephenson, 1984
; Richardson and Stephenson, 1989
; Stead, 1992
; Weiss, 1995
; Ashman and Schoen, 1997
; van Doorn, 1997
; Evanhoe and Galloway, 2002
). Recent studies have found that the rate of corolla senescence can vary within individual populations (Sargent and Roitberg, 2000
; Evanhoe and Galloway, 2002
; Rathcke, 2003
), demonstrating the importance of incorporating conditional responses to fitness accrual rate in the existing floral longevity model. What remains unclear is whether such conditional responses vary among natural populations that experience differences in fitness accrual rates and floral maintenance costs. For hermaphroditic flowers, an additional consideration is whether the functional gender phases (i.e., staminate, pistillate) respond differently to fitness accrual rate variation (Ashman, 2004
).
Campanula rotundifolia L. (Campanulaceae) is a protandrous, herbaceous perennial that spans large altitudinal gradients in western North America. Consequently, C. rotundifolia represents an excellent system for testing whether the Schoen and Ashman model accurately predicts patterns of floral longevity variation across populations of a single species. Moreover, the protandrous nature of C. rotundifolia flowers facilitates empirical tests of how total longevity is affected by sequential, gender-specific responses to fitness accrual rates.
Previous research on C. rotundifolia demonstrated that stigma receptivity, a component of total floral longevity, was 33% shorter in montane vs. alpine populations (Bingham and Orthner, 1998
). Here, I manipulated fitness accrual rates in populations of C. rotundifolia from geographically isolated, thermally distinct sites to address the following questions: (1) Do staminate and pistillate longevities reflect site-specific rates of pollen removal and deposition? (2) Do flowers at both sites respond similarly to fitness accrual rate manipulation? (3) If total floral longevity differs between populations, what is the relative contribution of each gender phase to this difference?
MATERIALS AND METHODS
Study plant and site
Campanula rotundifolia flowers are insect-pollinated and self-incompatible (Shetler, 1982
; Nyman, 1992
). They achieve secondary pollen presentation through introrse anther dehiscence prior to anthesis that deposits the pollen upon hairs lining the style. Pollen removal by floral visitors causes the hairs to retract, which in turn induces stigmatic lobe expansion and initiation of the flower's pistillate phase (Nyman, 1993a
).
Two field sites in Colorado (CO), USA separated by 120 km were selected for conducting this study in the summer of 1999: Aiken Canyon Preserve (montane habitat, 38°37' N, 104°54' W, 1900 m a.s.l.), 20 km southwest of Colorado Springs, and Pennsylvania Mountain (alpine habitat, 39°15' N, 106°07' W, 3400 m a.s.l.), 10 km west of Fairplay, CO. The flowering period at Aiken Canyon is late May through June, and on Pennsylvania Mountain from late July through August (Giblin, 2001
).
Prior to my study, I compiled mean daily temperature and monthly precipitation data collected over a 20-yr-period (1978–1998) at weather stations in Colorado Springs, CO (montane) and Leadville, CO (alpine) to estimate climatic differences between the study sites during the respective flowering periods. Mean values for daily temperature and monthly precipitation were higher at the montane station (18.3 ± 0.7°C; 55.6 ± 8.6 mm; here and elsewhere all values are mean ± 1 SE unless otherwise noted) than at the alpine (11.7 ± 0.3°C; 36.2 ± 6.1 mm) (Colorado Climate Center, 1999
). To verify these trends at each site over the duration of my experiments, I measured ambient temperature at 15-min intervals with Hobo Temp data loggers (H01-001-01; Onset, Bourne, Massachusetts) positioned 30 cm above the ground surface (approximate inflorescence height). Sensors were spaced regularly throughout sampling locations (blocks) at each site. Daily precipitation was recorded with one Tru-Chek rain gauge per site. Mean daily ambient temperature at Aiken Canyon (1–30 June; 18.4 ± 0.1°C;) was significantly greater than at Pennsylvania Mountain (2–31 August; 8.5 ± 0.01°C; F1,4 = 418, P = 0.0001), while Pennsylvania Mountain (88 mm) received greater total precipitation than Aiken Canyon (44 mm).
Experimental design
Three blocks separated by a minimum of 200 m were located at each site to span potential within-site microclimatic variation. One hundred plants were haphazardly selected within each block, and for all experiments I sampled the first flower on each plant to avoid potentially confounding ontogenetic effects (Diggle, 1994
; Vogler et al., 1999
). Longevity of the first and subsequent flowers was highly correlated under field conditions (Aiken Canyon: R = 0.67, N = 19; P = 0.0007; Pennsylvania Mountain: R = 0.78, N = 9, P = 0.01). Flower gender was checked twice daily (0800– 0900 hours and 1500–1600 hours).
Staminate longevity and male fitness accrual rate
The staminate phase of C. rotundifolia extends from initial opening of the corolla until stigmatic lobe expansion. Staminate longevity was recorded on 50 plants per block that were evenly and randomly assigned between two treatments: (1) complete pollinator exclusion, (2) open pollination. Pollinator exclusion was achieved by placing mesh netting over the entire inflorescence of each plant. I scored staminate longevity to the nearest 0.5 d for all flowers.
Male fitness accrual is the amount of pollen removed from a flower and entering the pool of conspecific pollen competing for unfertilized ovules (Ashman and Schoen, 1994
). Though problems exist with equating male fitness accrual with pollen removal (Galen, 1992
; Stanton et al., 1992
; Snow and Lewis, 1993
), the two are correlated in some systems (Ashman, 1998
).
I estimated male fitness accrual rate over the first 2 d of the staminate phase by counting the amount of pollen present in flowers of 30 haphazardly selected plants per block at each site. Plants were evenly and randomly assigned among three age classes: (1) newly opened, (2) 1 d old, and (3) 2 d old. The style and anthers of each sample were placed into an individual 1.5-mL Eppendorf tube containing 70% ethanol, and stored at room temperature prior to counting of pollen grains. All sampling occurred over a 3-day period at each site.
The Eppendorf tubes were sonicated in the lab for 10 s to dislodge all pollen from the anthers and style, and these floral parts were then discarded. The samples were centrifuged, the ethanol pipetted off, and the pollen suspended in 1 mL Isoton (Coulter Electronics, Hialeah, Florida, USA) before transfer to a 15-ml test tube. Each Eppendorf tube was rinsed with 1 mL Isoton to capture residual pollen, and the rinse added to the 1 mL Isoton suspension. The total pollen amount in the 2 mL suspension was quantified using a Coulter counter (model ZBI, Coulter Electronics, Hialeah, Florida, USA). Pollen quantities recorded were square-root transformed prior to analysis to satisfy assumptions of normality (Zar, 1999
).
Pistillate longevity and female fitness accrual rate
Female fitness accrual is defined as the amount of conspecific pollen deposited on a flower's stigmatic surface (Ashman and Schoen, 1994
). Residual staminate phase pollen can adhere to the unreceptive underside of C. rotundifolia stigmatic lobes (D. Giblin, personal observation). Microscopy could not be used to differentiate whether stained pollen grains arrived through pollination or contamination during staining. To avoid potentially misleading results, I used seed production per flower as a surrogate for female fitness accrual rate.
To assess geographic differences in pistillate phase longevity and female fitness accrual rate, 75 plants per block were evenly and randomly assigned among three flower age classes at each site: (1) excluding pollinators after 1 d in the pistillate phase, (2) excluding pollinators after 2 d in the pistillate phase, and (3) open pollination. The 25 plants per block receiving complete pollinator exclusion during the staminate longevity study constituted a fourth age class and were similarly denied exposure to pollinators. Age classes were achieved by placing netting over each plant's entire inflorescence. Pistillate longevity of each flower was measured to the nearest 0.5 d. Fruits matured approximately 3 weeks after corolla senescence and were then placed in individual envelopes upon harvest. The number of seeds per fruit were counted in the laboratory using a 10x Zeiss dissecting microscope. I calculated the daily fitness accrual rate by counting the number of seeds per fruit for each treatment at each site.
Enhancing female fitness accrual rate
I randomly selected 10 additional open-pollinated plants per block at each site to receive supplemental pollen. The first open flower of each plant received cross-pollen within 12 h of pistillate phase inception and afterwards remained available to pollinators. Donor pollen was collected from within a 10-m radius of the recipient flower, unique donors were used for each flower, and pollen was rubbed over the stigmatic lobes until they formed a layer visible to the naked eye. Pollinations were repeated once daily until corolla senescence. Pistillate phase duration was scored to the nearest 0.5 d.
Partitioning total longevity by gender
Total longevity for each flower was calculated as the number of days (to the nearest 0.5) between corolla expansion and senescence. I divided the number of days pistillate by total longevity, and these percentage values were angular-transformed prior to analysis to satisfy assumptions of normality (Zar, 1999
).
Data analyses
All analyses were performed with the Statistical Analysis System (SAS) version 6.12 using the general linear models procedure (PROC GLM; SAS Institute, 2000
). Site and treatment were fixed effects, block was random, and block was nested within site. I used the TEST option in all analyses to test fixed effects against the appropriate error terms specified in the models (Zar, 1999
). Nonsignificant (P > 0.30) block, block x site, and block x treatment interaction terms were eliminated from the models, with main effects then tested over the residual error term (Underwood, 1997
).
I performed mixed-model analysis of covariance (ANCOVA) on least square mean values to test for site and treatment effects on staminate and total longevity, and the percentage of total time spent in the pistillate phase (PROC GLM; SAS Institute, 2000
). Corolla width served as a covariate in the staminate longevity analysis to control for potential initial differences in pollen quantity due to flower size. The number of flowers per inflorescence served as a covariate in the total longevity analysis to control for size variation among sampled plants. Flower number and plant size were highly correlated among plants in this study (R2 = 0.36, N = 277, P < 0.0001). I used corolla depth as a covariate in the percentage pistillate analysis to control for variation in the stylar distance traveled by pollen tubes. To estimate depth, I hand-measured to the nearest millimeter the distance between each corolla's distal lobe and base. Nonsignificant covariate x site interactions were detected in all analyses.
I performed mixed-model analysis of variance (ANOVA) on least square mean values to test for site and treatment effects on pollen removal, pistillate longevity, and seed production per flower. All analyses were followed by post-hoc Tukey's honest significant difference tests for multiple comparisons where necessary (PROC GLM; SAS Institute, 2000
).
RESULTS
Staminate longevity and male fitness accrual rate
Mean staminate longevity across treatments averaged 3.2 (±0.09; N = 150) d at Aiken Canyon and 2.5 (±0.09; N = 141) d at Pennsylvania Mountain. The significant site x treatment interaction indicates site-specific responses to the pollen removal treatments (Table 1A). Post-hoc pairwise comparisons detected significant differences (P = 0.012) between treatments at Aiken Canyon only (Fig. 1). Corolla width approached significance as a covariate, and a significant block effect suggests that spatial variation in environmental factors within sites contributed to staminate longevity variation.
|
|
Pistillate longevity and female fitness accrual rate
Across treatments, mean pistillate longevity of Aiken Canyon flowers (2.4 ± 0.1 d; N = 327) was nearly three times less than those at Pennsylvania Mountain (6.5 ± 0.1 d; N = 315). The site x treatment interaction was highly significant, indicating site-specific responses to the female fitness accrual rate treatments (Table 2A). Complete pollinator exclusion significantly increased pistillate longevity relative to all other treatments at each site. Pistillate longevity was significantly different between 2 d and pollen-supplemented flowers at Aiken Canyon, while 1 d flowers at Pennsylvania Mountain had significantly greater longevities relative to open-pollinated and pollen-supplemented flowers (Fig. 3).
|
|
The block and block x treatment interaction terms were highly significant, indicating spatial heterogeneity in the response to female fitness accrual rate variation (Table 2B).
Partitioning total longevity by gender
Across treatments, a nearly two-fold difference was found in total longevity between Aiken Canyon (4.9 ± 0.1 d; N = 299) and Pennsylvania Mountain (9.2 ± 0.1 d; N = 285) flowers. The response of total floral longevity to pollinator access treatment was site specific (Table 3A). Complete screening of pollinators for the duration of flowering significantly increased total longevity relative to all other treatments at each site (Fig. 5). Incremental manipulations of pollinator access to pistillate flowers significantly increased total longevity relative to open-pollinated flowers at Pennsylvania Mountain only. Aiken Canyon and Pennsylvania Mountain flowers were pistillate for approximately 50% and 75% of their total longevities, respectively, and this difference was significant (Table 3B; Fig. 6).
|
|
A common element among widely distributed plant species is their ability to survive and reproduce under a range of environmental conditions. For insect-pollinated species, this includes exploiting the services of pollinator communities that typically change in both quantity and quality across a species's distribution range (Kearns and Inouye, 1994
; Bingham and Orthner, 1998
). Schoen and Ashman (1995)
constructed a model that demonstrated the contribution of floral longevity to a plant's reproductive ecology across floral maintenance cost and fitness accrual rate gradients. This model serves as a conceptual framework for understanding why floral longevity varies among species. In this study, I used the Schoen and Ashman model in a novel context by exploring floral longevity variation among populations of an individual species.
For species with protandrous flowers, the Schoen and Ashman model predicts that relative male and female fitness accrual rates affect a flower's relative staminate and pistillate longevities and that cumulatively these gender-specific responses impact a flower's total longevity (Schoen and Ashman, 1995
; Ashman, 2004
). I found that variation in male and female fitness accrual rates between C. rotundifolia flowers at Aiken Canyon and Pennsylvania Mountain produced parallel between-site differences in staminate, pistillate, and total longevity. These results validate the model's ability to predict floral longevity trends not only in protandrous flowers, but also between populations of an individual species that experience fitness accrual rate differences.
For C. rotundifolia flowers in this study, both gender phases varied geographically in their conditional response to fitness accrual rate variation, a result not predicted by the Schoen and Ashman model. Previous studies have shown that geographic differences in fitness accrual rates can lead to overall floral longevities differences (Primack, 1985
; Ashman and Schoen, 1994
; Bingham and Orthner, 1998
; Blionis and Vokou, 2001
), but these studies did not address how each gender phase contributed to total longevity differences. Here, I provide novel evidence of site-specific response patterns to fitness accrual rate variation by each gender phase: staminate longevity was significantly increased by the lack of pollen removal (male fitness accrual) at Aiken Canyon only, while pistillate longevity was reduced incrementally by increased pollen deposition (female fitness accrual) at Pennsylvania Mountain only. These results demonstrate how population-level differences in the conditional response to fitness accrual can contribute to geographic differences in total longevity and suggest that floral longevity among plants in this study may be adapted to local schedules of pollen exchange.
Variation in staminate longevity
In previous studies on Campanulaceae taxa, pollen removal reduced staminate longevity (Devlin and Stephenson, 1984
; Richardson and Stephenson, 1989
; Koptur et al., 1990
; Nyman, 1993b
; Evanhoe and Galloway, 2002
), but many of these studies did not compare this response across sites or populations (but see Nyman, 1993b
). Here, pollen removal significantly reduced staminate longevity in C. rotundifolia flowers at Aiken Canyon only.
On average, C. rotundifolia flowers at Pennsylvania Mountain transitioned to the pistillate phase after dispersing only 50% of their total pollen, and seed set in these flowers was pollen-limited. For C. rotundifolia plants at this site, it is potentially maladaptive to have a flower's staminate phase duration dependent upon complete pollen dispersal, since mounting floral maintenance costs could result in corolla senescence prior to pistillate phase initiation. In this context, the geographic differences observed in the staminate longevity response to fitness accrual rate variation may have been shaped by historic mating opportunities for C. rotundifolia flowers at these sites.
Only a modest between-site difference was observed in staminate longevity of open-pollinated flowers, and no significant difference existed between sites for the amount of pollen removed from sampled flowers. Solitary bees were the primary visitors to staminate flowers at Aiken Canyon, while dipteran taxa (Muscidae) dominated such visits at Pennsylvania Mountain (D. Giblin, personal observation). Flower-visiting Muscidae are typically pollen foragers (Proctor et al., 1996
), a behavior observed in other C. rotundifolia studies (Nyman, 1993b
; Blionis and Vokou, 2001
). Stylar hair stimulation by foraging Muscidae at Pennsylvania Mountain may have triggered stylar hair retraction despite an overall smaller percentage of pollen removal during the sample period. Nevertheless, pollen removal did not appear to be the primary cue for gender phase transition in Pennsylvania Mountain flowers, since there was no significant staminate longevity difference between pollinator excluded and open-pollinated flowers at this site. Similar results from subsequent field and common garden greenhouse studies on these populations provide additional support for the potential adaptation of staminate longevity in C. rotundifolia flowers to local schedules of pollinator activity (Giblin, 2001
).
Variation in pistillate phase longevity
Patterns of pistillate longevity in this study were consistent with Schoen and Ashman model predictions (Ashman and Schoen, 1994
). The daily fitness accrual rate of open-pollinated flowers at Aiken Canyon was six times greater and pistillate longevity four times less than that of similar flowers at Pennsylvania Mountain. Pistillate longevities at Aiken Canyon of 1-d pistillate, open-pollinated and pollen-supplemented flowers were similar, indicating saturation of female function after 1 day of pollinator visitation. At Pennsylvania Mountain, only pollen-supplemented flowers were significantly shorter-lived relative to all other pollination treatments, resulting in the significant site x treatment interaction.
Pollen deposition can induce floral senescence, but the rate of this response varies among taxa (Weiss, 1995
; van Doorn, 1997
). Here I demonstrate geographic variation in the conditional response by C. rotundifolia flowers to female fitness accrual rate variation. Mean pistillate longevity of pollen-supplemented flowers at Pennsylvania Mountain was twice that of similarly treated flowers at Aiken Canyon. Plumb (1999)
found no evidence of a Q10 response (the rate at one temperature divided by the rate at a temperature 10°C lower) in floral metabolism between montane and alpine C. rotundifolia populations at study sites near mine. These data discount the interpretation that increased pistillate longevity in Pennsylvania Mountain flowers was due to a rate reduction in post-pollination metabolic processes (e.g., pollen tube growth) resulting from lower mean daily temperatures at that site.
One potential reason for geographic variation in the conditional response of C. rotundifolia flowers to female fitness accrual rate can be seen in the fitness consequences experienced by Pennsylvania Mountain flowers if corolla senescence was induced upon initial pollen deposition. One-day pistillate flowers at Pennsylvania Mountain experienced a five-fold reduction in per flower seed production relative to open-pollinated flowers, suggesting a severe reproductive cost for premature senescence.
An alternative conclusion to the response pattern observed in the supplemental pollen study is that intrinsic differences exist between populations in the regulation of pistillate longevity. This interpretation is supported by the fact that pollinator-excluded flowers at Pennsylvania Mountain were also pistillate twice as long as similarly treated flowers at Aiken Canyon. However, a subsequent common garden greenhouse study on offspring from these populations found no evidence to support this interpretation (Giblin, 2001
).
In addition to maximizing pollen deposition opportunities, extended pistillate longevity in Pennsylvania Mountain flowers may serve as a competitive mechanism that enhances pollinator attraction through increased floral display size (Klinkhamer and de Jong, 1990
; Ishii and Sakai, 2001
; Galloway et al., 2002
; Rathcke, 2003
). The pollen-limited per flower seed production found at Pennsylvania Mountain suggests competition for pollinator services existed among pistillate C. rotundifolia flowers at that site.
Partitioning total longevity by gender
Campanula rotundifolia flowers at Pennsylvania Mountain spent a significantly greater percentage of their total longevity in the pistillate phase relative to Aiken Canyon flowers. This between-site lability in relative resource allocation to gender function is consistent with the findings of Johnson et al. (1995)
for C. americana. In that study, pollinator quality determined whether male or female function received the greater benefit from attractive corollas. Pollinator quantity is more likely responsible for the differences observed here, because Pennsylvania Mountain flowers required significantly greater time for saturation of female function. I did not quantify between-site differences in pollinator abundance in this study. However, lower pollinator visitation rates are typically found at alpine vs. montane sites (Arroyo et al., 1985
; Kearns and Inouye, 1994
; Utelli and Roy, 2000
), and this pattern was reported for other C. rotundifolia populations in relative proximity to my sites (Bingham and Orthner, 1998
).
Conclusions
The general floral longevity trends in my study were consistent with predictions of the Schoen and Ashman model: Campanula rotundifolia flowers had greater longevity when fitness accrual rates were lower and floral maintenance costs were likely less. However, I found that geographic variation in total floral longevity was mediated through gender-specific differences in the conditional response to fitness accrual rates, a result not predicted by the model. The pattern of these differences correlated with local pollination schedules and is consistent with theoretical expectations for gender-specific reproductive limitations due to competition. Results from this study emphasize the value of including flowers' conditional responses to fitness accrual rates in refining future floral longevity models (Yasaka et al., 1998
; Evanhoe and Galloway, 2002
).
FOOTNOTES
1 Discussions with and comments from C. Galen greatly improved this manuscript, as did comments from two anonymous reviewers. The author thanks M. Brock, L. Dudley, and J. Farrell for field assistance; The Nature Conservancy for providing access to Aiken Canyon Preserve; and the University of Colorado at Colorado Springs for providing access to Pennsylvania Mountain. This research was supported by a research grant to D.G. from the Conservation Biology Program at the University of Missouri-Columbia. 
2 Author for correspondence (e-mail: dgiblin{at}u.washington.edu
) current address: University of Washington Herbarium, University of Washington, Box 355325, Seattle, Washington 98195-5325 USA
LITERATURE CITED
Arroyo M. T. K. J. Armesto R. B. Primack 1985 Community studies in pollination ecology in the high temperate Andes of central Chile. II. Effects of temperature on visitation rates and pollination possibilities. Plant Systematics and Evolution 149: 187-203[CrossRef][ISI]
Ashman T. L. 2004 Flower longevity. In L. D. Nooden [ed.], Plant cell death processes, 349–362. Elsevier, San Diego, California, USA
Ashman T. L. 1998 Is relative pollen production or removal a good predictor of relative male fitness? An experimental exploration with a wild strawberry (Fragaria virginiana, Rosaceae). American Journal of Botany 85: 1166-1171[Abstract]
Ashman T. L. D. J. Schoen 1994 How long should flowers live?. Nature 371: 788-790[CrossRef]
Ashman T. L. D. J. Schoen 1997 The cost of floral longevity in Clarkia tembloriensis: an experimental investigation. Evolutionary Ecology 11: 289-300
Bingham R. A. A. R. Orthner 1998 Efficient pollination of plants. Nature 391: 238-239[CrossRef]
Blionis G. J. D. Vokou 2001 Pollination ecology of Campanula species on Mt. Olympus, Greece. Ecography 24: 287-297
Caruso C. M. S. B. Peterson C. E. Ridley 2003 Natural selection on floral traits of Lobelia (Lobeliaceae): spatial and temporal variation. American Journal of Botany 90: 1333-1340
Charnov E. L. 1996 Optimal floral lifetimes. Evolutionary Ecology 10: 245-248[CrossRef][ISI]
Colorado Climate Center. 1999 Summaries for selected stations. Colorado State University, Fort Collins, Colorado, USA. [online, accessed on 12 July 2005], http://ulysses.atmos.colostate.edu/sum_form.html
Devlin B. A. G. Stephenson 1984 Factors that influence the duration of the staminate and pistillate phases of Lobelia cardinalis flowers. Botanical Gazette 145: 323-328[CrossRef]
Diggle P. K. 1994 The expression of andromonoecy in Solanum hirsutum (Solanaceae): phenotypic plasticity and ontogenetic contingency. American Journal of Botany 81: 1354-1365[CrossRef][ISI]
Evanhoe L. L. F. Galloway 2002 Floral longevity in Campanula americana (Campanulaceae): a comparison of morphological and functional gender phases. American Journal of Botany 89: 587-591
Galen C. 1989 Measuring pollinator-mediated selection on morphometric floral traits: bumblebees and the alpine sky pilot Polemonium viscosum. Evolution 43: 882-890[CrossRef][ISI]
Galen C. 1992 Pollen dispersal dynamics in an alpine wildflower, Polemonium viscosum. Evolution 46: 1043-1051[CrossRef][ISI]
Galloway L. F. T. Cirigliana K. Gremski 2002 The contribution of display size and dichogamy to potential geitonogamy in Campanula americana. International Journal of Plant Sciences 163: 133-139[CrossRef][ISI]
Giblin D. E. 2001 Ecological genetics of floral longevity in Campanula rotundifolia, the alpine harebell. Ph.D. dissertation, University of Missouri, Columbia, Missouri, USA
Herrera C. M. 1991 Dissecting factors responsible for individual variation in plant fecundity. Ecology 72: 1436-1448[CrossRef][ISI]
Ishii H. S. S. Sakai 2001 Effects of display size and position on individual floral longevity in racemes of Narthecium asiaticum (Liliaceae). Functional Ecology 15: 396-405[CrossRef][ISI]
Johnson S. G. L. F. Delph C. L. Elderkin 1995 The effect of petal-size manipulation on pollen removal, seed set, and insect-visitor behavior in Campanula americana. Oecologia 102: 174-179[CrossRef][ISI]
Jonas C. S. M. A. Geber 1999 Variation among populations of Clarkia unguiculata (Onagraceae) along altitudinal and latitudinal gradients. American Journal of Botany 86: 333-343
Kearns C. A. D. W. Inouye 1994 Fly pollination of Linum lewisii (Linaceae). American Journal of Botany 81: 1091-1095[CrossRef][ISI]
Klinkhamer P. G. L. T. J. de Jong 1990 Effects of plant size, plant density, and sex differential nectar reward on pollinator visitation in protandrous Echium vulgare (Boraginaceae). Oikos 57: 399-405[CrossRef][ISI]
Koptur S. E. N. Davila D. R. Gordon B. J. Davis MacPhail C. G. Murphy J. B. Slowinski 1990 The effect of pollen removal on the duration of the staminate phase of Centropogon talamancensis. Brenesia 33: 15-18
Nyman Y. 1992 Pollination mechanisms in six Campanula species (Campanulaceae). Plant Systematics and Evolution 181: 97-108[CrossRef][ISI]
Nyman Y. 1993a The pollen-collecting hairs of Campanula (Campanulaceae). I. Morphological variation and the retractive mechanism. American Journal of Botany 80: 1427-1436[CrossRef][ISI]
Nyman Y. 1993b The pollen-collecting hairs of Campanula (Campanulaceae). II. Function and adaptive significance in relation to pollination. American Journal of Botany 80: 1437-1443[CrossRef][ISI]
Plumb C. L. 1999 Spatial and temporal variation of respiration in natural populations of plants (Bistorta bistortoides, Campanula rotundifolia). Ph.D. dissertation, University of Colorado, Boulder, Colorado, USA
Primack R. 1985 Longevity of individual flowers. Annual Review of Ecology and Systematics 16: 15-37[CrossRef][ISI]
Proctor M. P. Yeo A. Lack 1996 The natural history of pollination. Timber Press, Portland, Oregon, USA
Rathcke B. J. 2003 Floral longevity and reproductive assurance: seasonal patterns and an experimental test with Kalmia latifolia (Ericaceae). American Journal of Botany 90: 1328-1332
Richardson T. E. A. G. Stephenson 1989 Pollen removal and pollen deposition affect the duration of the staminate and pistillate phases in Campanula rapunculoides. American Journal of Botany 76: 532-538[CrossRef][ISI]
Sargent R. D. B. D. Roitberg 2000 Seasonal decline in male-phase duration in a protandrous plant: a response to increased mating opportunities?. Functional Ecology 14: 484-489[CrossRef][ISI]
SAS Institute. 2000 SAS/STAT, version 6.12. SAS Institute, Cary, North Carolina, USA
Schoen D. J. T. L. Ashman 1995 The evolution of floral longevity: resource allocation to maintenance vs. construction of repeated parts in modular organisms. Evolution 49: 131-139
Shetler S. G. 1982 Variation and evolution of the Nearctic harebells (Campanula subsect. Heterophylla) (Phanerogamarum Monographiae XI). J. Cramer, Vaduz, Liechtenstein
Snow A. A. P. O. Lewis 1993 Reproductive traits and male fertility in plants: empirical approaches. Annual Review of Ecology and Systematics 24: 331-351[ISI]
Stanton M. L. T. L. Ashman L. F. Galloway H. J. Young 1992 Estimating male fitness of plants in natural populations. In R. Wyatt [ed.], Ecology and evolution of plant reproduction. Chapman and Hall, New York, New York, USA
Stead A. D. 1992 Pollination-induced floral senescence: a review. Plant Growth Regulation 11: 13-20
Underwood A. J. 1997 Experiments in ecology. Cambridge University Press, Cambridge, UK
Utelli A. B. B. A. Roy 2000 Pollinator abundance and behavior on Aconitum lycoctonum (Ranunculaceae): an analysis of the quantity and quality components of pollination. Oikos 89: 461-470[CrossRef][ISI]
Van Doorn W. G. 1997 Effects of pollination on floral attraction and longevity. Journal of Experimental Botany 48: 1615-1615
Vogler D. W. S. Peretz A. G. Stephenson 1999 Floral plasticity in an iteroparous plant: the interactive effects of genotype, environment, and ontogeny in Campanula rapunculoides (Campanulaceae). American Journal of Botany 86: 482-494
Weiss M. R. 1995 Floral color change: a widespread functional convergence. American Journal of Botany 82: 167-185[CrossRef][ISI]
Yasaka M. Y. Nishiwaki Y. Konno 1998 Plasticity of flower longevity in Corydalis ambigua. Ecological Research 13: 211-216
Zar J. H. 1999 Biostatistical analysis, 4th ed. Prentice-Hall, Upper Saddle River, New Jersey, USA
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
0.05)
