| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
2Division of Biological Sciences, 105 Tucker Hall, University of Missouri, Columbia, Missouri 65211 USA; and 3Department of Forestry, 203 ABNR Bldg, University of Missouri, Columbia, Missouri 65211 USA
Received for publication March 30, 2000. Accepted for publication June 2, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
l) and floral traits were monitored over a 12-d period of soil moisture depletion. Soil moisture depletion induced drought stress over time, as revealed by significant treatment x day interactions for predawn and midday l. Nectar volume and flower size showed significant negative responses to drought stress, but nectar sugar concentration did not vary between treatments. Floral traits were more buffered from drought than leaf water potentials. We used path analysis to examine direct and indirect effects of l on floral traits for plants in well-watered (control) vs. drought treatments. According to the best-fit path models, midday l has significant positive effects on flower size and nectar volume in both environments. However, for controls midday l also had a significant negative effect on nectar sugar concentration. Results indicate that traits influencing floral attractiveness to pollinators in E. angustifolium vary with plant water status, such that pollinator-mediated selection could indirectly target physiological or biochemical controls on l. Moreover, under mesic conditions selection for greater nectar sugar reward may be constrained by the antagonistic effects of plant water status on nectar volume and sugar concentration.
Key Words: drought stress • Epilobium angustifolium • flower size • leaf water potential • nectar production • nectar sugar concentration • Onagraceae • phenotypic plasticity • pollinator-mediated selection
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) and high nectar sugar production (Pyke, 1978
; Best and Bierzychudek, 1982
; Galen and Plowright, 1985
; Thomson, Maddison, and Plowright, 1989
; Cresswell, 1990
; Hodges, 1995
). Plants exhibiting these traits disperse and receive more pollen, increasing male fitness, and, potentially, female fitness (but see Hodges, 1995
). Despite strong selection from pollinators, natural plant populations display significant variation in flower size and nectar production rate. Some of the variation in floral traits is caused by heritable differences among individuals (Mitchell and Shaw, 1993
; Andersson and Widen, 1993
; Fenster and Ritland, 1994
; Young et al., 1994
; Campbell, 1996
; Galen, 1996
). However, the remainder reflects phenotypic plasticity in response to the local environment (Mazer and Schick, 1991
; Stratton, 1992
; Holtsford and Ellstrand, 1992
). Through plasticity, the abiotic environment may not only limit plant attractiveness to pollinators, but may alter the nature of pollinator-mediated selection on flower traits.
The significance of floral trait responses to the abiotic environment is twofold. First, evolutionary responses of floral traits to selection from the abiotic environment may conflict with pollinator-mediated selection. For example, despite directional selection by pollinators for large flowers, smaller flowered genotypes persist in at least two well-characterized plant taxa, Polemonium viscosum and Ipomopsis aggregata (Campbell, 1991
; Galen, 1999
). The persistence of less attractive flowers may be partially explained by selective pressures of the abiotic environment on floral traits (Campbell, 1996
; Galen, 1999
). Second, plastic responses of floral traits to the abiotic environment may influence how pollinators perceive and service plants. If more attractive flowers are produced under one set of abiotic conditions than under another, plant–pollinator interactions may differ between the two environments. Potentially, differences in the quality of pollinator service between environments could ensue. Furthermore, if genotypes vary in their sensitivity to the abiotic gradient distinguishing environments, such variation in reaction norms or in the physiological mechanisms underlying them could itself be subject to pollinator-mediated selection.
Abiotic heterogeneity can cause variation among individuals in the capacity to invest in sexual reproduction. In animal- pollinated species, those individuals that are capable of investing disproportionately greater resources in floral display should achieve higher rates of visitation by pollinators and, for that reason, higher rates of pollen transfer (Kaul, 1979
; Bell, 1985
; Sutherland, 1986
; Schmid-Hempel and Speiser, 1988
). Such variation should be especially pronounced when environments vary in resources that are essential for pollinator advertisement and reward. One resource of this kind is water (Nobel, 1977
; Galen, Sherry, and Carroll, 1999
). Water is used throughout the lifetime of a flower. An influx of water from vegetative portions of the plant is required for floral bud expansion, flower opening, nectar production, and turgor maintenance in floral organs under evapotranspirational demand (Mohan Ram and Rao, 1984
). Although the water costs of floral display can be substantial (Nobel, 1977
; Whiley, Chapman, and Saranah, 1988
), studies addressing the effects of resource availability on flowering in wild plants have largely disregarded this phenomenon. In the present study, we address the hypothesis that the abiotic environment indirectly influences floral advertisements and rewards through its effects on plant water status.
Results of the few studies that have addressed the effects of water on floral display in plants are equivocal. The single component of flowering that has been studied extensively in relation to water availability is nectar secretion. Supplemental watering increased nectar volume in Delphinium nelsonii (Zimmerman, 1983
), Polemonium foliosissimum (Zimmerman and Pyke, 1988
), and Asclepias syriaca (Wyatt, Broyles, and Derda, 1992
) and increased nectar sucrose content in Asclepias syriaca (Wyatt, Broyles, and Derda, 1992
). Comparisons between growing seasons that differ in precipitation also suggest that nectar production is greater in wet years than dry years (e.g., Prosopis glandulosa—Lee and Felker, 1992
; Ipomopsis aggregata—Campbell, 1996
). In addition, Boose (1997)
found that clones of Epilobium canum produced less nectar when watered every other day with approximately half the amount of water that control plants received daily. Except for Lee and Felker (1992)
, none of these workers used an index of plant water status to determine the extent of stress experienced by the plant, so the exact nature of the relationship between water availability and nectar secretion is still unclear.
In this paper, we report on an experiment in which we artificially induced drought stress during flowering of Epilobium angustifolium (fireweed), an animal-pollinated plant. We then monitored plant water status and floral trait plasticity after 12- d of soil moisture depletion. The goal of our study was to assess whether drought stress, by limiting plant water status, indirectly alters three floral traits that influence the outcome of plant-pollinator interactions: flower size, lifetime nectar volume and nectar sugar concentration. Specifically we asked the following questions: (1) Do floral traits show plastic responses to drought stress? (2) Do floral traits and vegetative traits differ in the degree of plasticity exhibited under drought? Specifically, are flowers more buffered from drought stress than vegetative organs? (3) Does among-plant variation in the physiological response to watering treatments (l) affect floral trait expression within wet and dry environments?
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Plants flower for 3–5 wk with individual flowers open 4–5 d (Mosquin, 1966
). Nectar production begins at the onset of anthesis and peaks at the time of stigma receptivity (Galen and Plowright, 1985
). Although fireweed is self-compatible, spatial and temporal (protandry) separation of anthers and stigma prevent self- pollination within individual flowers (Mosquin, 1966
). As a result, E. angustifolium plants depend on animal-mediated pollination for sexual reproduction. Flowers are pollinated by various Hymenoptera, Lepidoptera, and, occasionally, hummingbirds (Percival, 1965
; Heinrich, 1976
; Zimmerman, 1988
; Carroll, personal observation).
Experimental design
Seeds obtained in bulk (Western Native Seeds, Salida, Colorado, USA) were germinated and the plants grown to flowering from December to April 1998, in the greenhouse at the University of Missouri- Columbia (Boone County, Missouri; 38°57'30'' N, 92°20'00'' W). Seeds were scattered on the surface of a peat-based growing medium (Pro-Mix BX Professional General Purpose Growing Medium; Premier Horticulture, Inc. Red Hill, Pennsylvania, USA) under natural light at a constant temperature of 20°C. Thirty healthy, vigorous seedlings were transplanted three times to successively larger pots, ending with a 3.78-L pot in early March. Pots were distributed randomly across a greenhouse bench, and position was re-randomized at each successive transplant. At the final transplant, plants were moved to a separate greenhouse maintained at a night temperature of 16°C and a day temperature ranging from 16° to 25°C. Plants were watered as needed and fertilized every 2 wk until watering treatments began (Peters Professional All Purpose Plant Food [20-10-20]; United Industries Corp., St. Louis, Missouri, USA). On initiation of floral buds, plants were matched by size and each matched pair was randomly assigned to a block on the greenhouse bench.
As seeds were obtained from Colorado (A. Tonnesson, personal communication), watering levels were based on average precipitation for central Colorado during July and August, the time of flowering for natural populations of E. angustifolium (Zimmerman, 1988
). Over the past century, mean precipitation ranged from 
2 to 10 cm/mo in this region (National Climate Data Center, NOAA, 1998
), which is the equivalent to 150–715 mL/wk for a 3.78- L pot. Plants in each pair were randomly assigned between the well-watered control group and the drought-stressed group. Controls received 500 mL of water every other day, while drought-stressed plants received 500 mL of water once on the first day of a 12-d period of soil moisture depletion. For each pair of plants, watering treatments were initiated when both individuals had at least two flowers open on the central axis of the raceme. Treatments began on four dates to accommodate the range of flowering schedules in the study population: 5, 12, 19, and 26 April 1998. The last cohort completed the treatment period on 8 May. Cohorts were composed of four pairs each except on 5 April, when only three pairs were included. Responses of plants to watering regime were monitored 24 h before the initiation of the watering treatments or pretreatment (day 0), and on day 6 and day 12 of soil moisture depletion.
Water potential measurements
A pressure chamber (Plant Water Status Console-Model 3005-1412; Soilmoisture Equipment Corp., Goleta, California, USA; Pearcy et al., 1989
) was used to measure plant water status. Water potential (l) was assessed for healthy mature leaves between 0500 and 0630 (predawn) and 1200 and 1330 (midday). Predawn l estimates soil water potential, while midday l estimates the maximal water stress experienced by a plant in a diurnal period (Lambers, Chapin, and Poris, 1998
). All l measurements were taken immediately after leaves were excised from the stem. A random order of sampling within each pair and among pairs was followed for measurements in each cohort. After selection of the first stem leaf, each subsequent measurement was taken on the leaf at the next highest node.
Floral trait measurements
All measurements were taken from flowers in the female phase (2–3 d old). The most basal flower on the central axis of the raceme was selected for the first observation (day 0). To assure that subsequently measured flowers opened after the onset of soil moisture depletion, we marked the pedicel of the most distal open flower at this time with a dot of enamel paint. The most basal flower above the mark was measured on days 6 and 12. On average, 15 nodes separated subsequently measured flowers. Nectar measurements were taken from 1100 to 1200 and 1330 to 1600. Nectar sampling was interrupted to accommodate midday measurements of leaf water potential. Plants in each block were sampled during the same time interval on each day of measurement. Sampling order on each day was determined at random. Nectar volume per flower was measured with a calibrated 5-µL disposable micropipet and nectar sugar concentration (mg sucrose equivalents/ mg solution) was measured with a hand-held refractometer (Bellingham and Stanley, Inc., Lawrenceville, Georgia, USA). Flowers were excised from the stem and pressed after nectar collection so that surface area could be measured using a leaf area meter (Model 201; CID, Inc., Vancouver, Washington, USA).
Statistical analysis
All analyses were performed using the Statistical Analysis System (SAS System, version 6.12; SAS Institute, Inc., Cary, North Carolina, USA). To test whether watering treatment caused a change in leaf water status over time, leaf water potential was subjected to repeated-measures analysis of variance (RM-ANOVA). Time (predawn or midday) and day of the water depletion period (day 0, 6, or 12) were specified as repeated factors (within-subject) and treatment and cohort as fixed effects (between-subjects). Block was dropped from the analysis as it did not explain a significant portion of the variance. Because the RM-ANOVA showed significant differences in l between treatments, separate one-way RM-ANOVAs were conducted for predawn l and midday l, with treatment and cohort as the fixed main effects. Each RM-ANOVA was followed by planned contrasts to evaluate a priori comparisons between treatment means on each day. Similarly, separate one- way RM-ANOVAs and planned contrasts were used to test whether treatments caused significant changes in floral traits over time. The validity of a within- subjects test depends on the equality of covariances. We calculated a measure of deviation that addresses this assumption (Huynh-Feldt statistic) and report adjusted P values (H-F P) (Deutschman, Levin, and Pacala, 1999
). Nectar sugar concentration was angular transformed prior to analysis to satisfy the assumption of normality.
The extent of phenotypic change over the course of treatment was compared among physiological and floral traits within each environment. Here, controls show changes in trait values over time due to plant development or age, while drought-treated plants show changes that also reflect a progressively restricted water supply. As the difference between treatments in predawn and midday l was greatest on day 12 (see Fig. 1), day-0 and day-12 trait values were used in this analysis. The extent to which diverse traits vary in response to an environmental signal can be obscured by differences among traits in units of measurement. To resolve this problem, standardized scores were obtained for all floral traits as well as both measures of leaf water potential. For each trait, individual observations were divided by the grand mean across day 0 and day 12 (pooled) to yield standardized scores. Means and standard errors of the standardized scores are unitless, allowing for meaningful comparisons among traits. To test whether traits changed by varying magnitudes over time in each environment, final scores on day 12 were subtracted from scores on day 1 and the absolute value of the resulting difference was obtained for each plant. We used separate mixed-model analyses of variance followed by a posteriori (Tukey's) contrasts to compare the magnitude of change exhibited among traits in each environment, with plant as a random factor and kind of trait as a fixed effect.
|
l on day 12 (Fig. 1), we used data collected at this time to characterize the relationships between plant water status and floral traits under well-watered vs. drought environments. Path analysis was used to determine direct and indirect effects of l on floral traits. Path analysis is a form of multiple regression that allows for nonindependence between steps of a modeled causal scheme such that relationships among independent and dependent variables are partitioned into direct and indirect effects (Wright, 1921, 1934
; Mitchell, 1993
). Based on the course along which water moves through a plant we proposed an a priori path model (Fig. 2). We hypothesized that predawn l, which is indicative of soil water status, first affects midday l, which in turn influences flower size, nectar volume, and nectar sugar concentration. Separate path diagrams were constructed for plants in each treatment. First we calculated observed and expected correlations between l and floral traits. We next estimated direct, indirect, and total effects for the model using SAS PROC CALIS (Mitchell, 1993
). The total explained variance for dependent variables (R2) was used to calculate the total unexplained variance (U = 
) (Mitchell, 1993
). Alternative models were analyzed in the same manner. A goodness-of-fit test allowed for comparison of explanatory power among the alternative models (Mitchell, 1993
). The model with the lowest 
2 value has the best fit, and the model with the fewest estimated paths has greatest parsimony.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1 under drought stressl (P = 0.0001), showing that the change in leaf water potential over the 12 d treatment interval depended on watering schedule. Univariate analyses indicated that treatments significantly affected both predawn l and midday l (Table 1; day x treatment interaction P = 0.0001 for both). Although plants in both treatments had similar water status before soil moisture depletion, planned contrasts revealed that treatments differed significantly (P < 0.01) in predawn l and midday l by day 6, with differences sustained through day 12 (Fig. 1). On both days, l was less negative for controls than for drought-treated plants. Flowering cohorts responded similarly to progressive drought stress in both predawn l and midday l (for the day by treatment by cohort interaction, P > 0.71 and P > 0.26, respectively). Results show that treatments had marked effects on plant water status over time.
|
|
|
|
1 and floral traitsl varied independently of predawn l (Table 3, model 7, 2 = 3.61, P = 0.3074). According to this model, midday 
l had a positive direct effect (DE) on both flower size and nectar volume (DE = 0.60, df = 1, P = 0.0226, and DE = 0.63, df = 1, P = 0.0127, respectively). Conversely, midday l had a negative direct effect on nectar sugar concentration (DE = 0.80, df = 1, P = 0.0003; Table 4, Fig. 5). For drought-stressed plants, the best-fit path model differed in two respects from that of controls (Table 3, model 4, 2 = 2.55, P = 0.8632). First, the a priori hypothesis of a positive relationship between predawn l and midday l was supported under drought but not under control conditions (DE = 0.93, df = 1, P = 0.0001; Table 4, Fig. 5). Next, in drought-stressed plants midday l had no effect on nectar sugar concentration. As in controls, midday l in drought-stressed plants had significant positive direct effects on nectar volume and flower size (DE = 0.8002, df = 1, P = 0.0003 and DE = 0.8522, df = 1, P = 0.0001, respectively). Path analysis suggests that with soil moisture depletion, the extent to which plants avoided drought stress, as indexed by predawn l, affected floral traits indirectly through midday l. Specifically, predawn l had positive indirect effects (IE) on nectar volume (IE = 0.7181) and flower size (IE = 0.7751, Fig. 5).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
l relative to controls. Drought also led to a threefold reduction in nectar volume and a 33% decrease in flower size relative to controls. Nectar sugar concentration was largely unaffected by water deficit. Comparison of the phenotypic response to treatments among traits showed that traits varied significantly in the degree of change over time in relation to both plant age and drought. However, variation in plasticity among traits was most pronounced under drought. Floral traits of drought-stressed plants showed significantly less plasticity than leaf water potentials. Plants also varied significantly in average plasticity under drought, hinting at underlying variation in sensitivity to water availability. Variation among plants in water status affected floral traits differently under wet and dry conditions. The most pronounced difference between environments concerned the relationship between midday l and nectar sugar concentration. Under drought stress no relationship was observed, while in controls midday l negatively affected nectar sugar concentration.
While levels of drought stress are likely to differ between the greenhouse and field, treatments in this study were effective in simulating natural drought stress for E. angustifolium. Chapin (1995)
found that mean l of fireweed declined from -0.20 MPa at predawn to -1.20 MPa at midday at Mount St. Helens National Monument (Skamania County, Washington, USA). Similarly, in the Rocky Mountains (Park County, Colorado, USA) l in flowering plants of E. angustifolium (N = 15) averaged -0.39 Mpa before dawn (range: -0.25 to -0.53 MPa) and -0.99 MPa at midday (range: -0.71 to -1.36 MPa) (Carroll, 2000)
. As values of l generated in the greenhouse lie within the range exhibited in nature, findings of our study are likely to apply to natural populations of E. angustifolium.
Subjecting plants of E. angustifolium to drought during flowering significantly reduced predawn and midday l, as well as nectar volume and flower size. Other field and greenhouse experiments have also shown that nectar production varies in relation to water supply, although concomitant variation in plant water status was not measured (Zimmerman, 1983
; Zimmerman and Pyke, 1988
; Wyatt, Broyles, and Derda, 1992
; Boose, 1997
). The trend for reduced flower size under drought stress is mirrored in populations of Clarkia unguiculata distributed along a natural moisture gradient (Jonas and Geber, 1999
). Water availability had little effect on nectar sugar concentration in our study or others (Wyatt, Broyles, and Derda, 1992
; Campbell, 1996
). Variation among traits in plasticity, as observed in the present study, may limit the efficacy of pollinators in exerting selection on pollination "syndromes" in heterogeneous habitats. Plastic traits should exhibit lower adaptive responses to pollinator-mediated selection than more canalized ones, causing divergent rates of evolutionary change.
Under control conditions, l changed little over 12 d of flowering. In contrast, nectar volume showed a moderate increase. Nectar production also depends on plant age in natural populations of Lobelia cardinalis (Devlin, Horton, and Stephenson, 1987
) and Ipomopsis aggregata (Pleasants, 1983
). However, in both species nectar production decreases over time, most likely due to resource limitation late in the flowering period. By experimentally manipulating resource (water) availability in the present study, we could tease apart ontogenetic and plastic sources of trait variation that are often confounded in nature. Depletion of soil moisture over time substantially decreased predawn and midday l. While significant responses of floral traits to drought were also observed, even the most plastic floral trait, lifetime nectar volume, was buffered compared to leaf water potential. Flower size is also buffered from resource limitation in Hydrophyllum appendiculatum (Wolfe, 1992
). The finding of significant among-plant variation in the degree to which traits change under drought stress suggests that trait plasticity in E. angustifolium may have a genetic basis (Carroll, 2000)
.
Previous studies examining how stress affects floral display have focused primarily on herbivory (Hendrix, 1988
; Marquis, 1992
; Frazee and Marquis, 1994
; Quesada, Bollman, and Stephenson, 1995
; Strauss, Conner, and Rush, 1996
; Lehtila and Strauss, 1997, 1999
). Simulated herbivory reduces mean petal size, nectar secretion and pollen production in flowers of Raphanus raphanistrum (Strauss, Conner, and Rush, 1996
). Plastic responses of floral traits to herbivory and drought are likely to weaken the role of pollinators in driving trait evolution. In general, animal pollinators prefer large flowers with high nectar production rates over smaller, less rewarding flowers (e.g., Best and Bierzychudek, 1982
; Thomson, Maddison, and Plowright, 1989
; Cresswell, 1990
; Cresswell and Galen, 1991
; Hodges, 1995
; Johnson, Delph, and Elderkin, 1995
; Conner and Rush, 1996
; Lehtila and Strauss, 1997
). If floral trait expression varies with abiotic or biotic stress, patterns of pollinator visitation may depend on microhabitat quality rather than on genotypic differences among plants. In Raphanus raphanistrum, pollinators prefer plants undamaged by herbivory (Strauss, Conner, and Rush, 1996
; Lehtila and Strauss, 1997
). A similar pattern of pollinator behavior would be expected for plants of E. angustifolium under drought stress, especially because bumble bees preferentially visit and remain longer at fireweed flowers with enriched nectar volume (Galen and Plowright, 1985
). Nonetheless, if drought stress has strong negative effects on plant survival or reproductive success, its indirect effect on pollinator behavior could have little bearing on fitness.
Path models for relationships between plant water status and floral traits in control and drought treatments indicated that plant water status has positive effects on flower size and nectar volume across environments. Flower size is positively correlated with nectar production in a number of species (reviewed in Cresswell and Galen, 1991
; see also Stanton and Young, 1994
). The phenotypic association between advertisement and reward may reflect selection from pollinators (Cresswell and Galen, 1991
), genetic constraint (Murrell, Tomes, and Shuel, 1982
; Teuber, Barnes, and Rincker, 1983
) or environmental correlation. The parallel responses of flower size and nectar volume to plant water status in control and drought-stressed E. angustifolium suggest that the two traits could covary in nature due to a common response to the environment.
Two paths distinguished best-fit models for control and drought-stressed individuals. First, predawn l had a direct positive effect on midday l under drought, indicating that external water supply limited midday l. However, under control conditions predawn and midday l were uncoupled, suggesting that unmeasured facators limited midday l. Phenotypic variation in midday l of controls may have been due to underlying variation in plant hydraulic characteristics, photosynthetic rate, stomatal conductance, or stomata number and distribution. Stomatal conductance and photosynthetic rate are genetically variable in the herbaceous annual, Polygonum arenastrum (Geber and Dawson, 1990
). We plan to investigate further whether phenotypic variation in midday l also has a genetic basis in E. angustifolium. Next, the relationship between midday l and nectar sugar concentration differed for controls and drought-stressed plants. Specifically, midday l had a negative effect on nectar sugar concentration in controls, and no effect on nectar sugar concentration under drought. Control plants were well watered and fertilized (N-P-K), eliminating resource limitation of photosynthesis as a main cause of low nectar sugar concentration. Rather, because nectar represents sap from phloem, plants capable of maintaining high l under wet conditions can sustain higher nectar secretion rates, which in turn dilute sugar concentration (Salisbury and Ross, 1978
). In mesic habitats, these antagonistic effects of plant water status on nectar volume and sugar concentration could constrain pollinator selection for high sugar reward.
Because changes in floral traits affect pollinator taxa differentially (Thomson, Maddison, and Plowright, 1982
; Conner and Rush, 1996
; Johnson, Delph, and Elderkin, 1995
; Strauss, Conner, and Rush, 1996
), our results suggest that variation in water availability will influence the pollinator community visiting E. angustifolium in the wild. If flowers are more attractive to specific pollinators under wet conditions than under dry conditions, then pollinator assemblages servicing fireweed may differ between environments. For example, hummingbird flowers often have more dilute nectar than bumble bee flowers (reviewed in Pyke and Waser, 1981
). It follows that wet environments in which dilute nectar is prevalent may indirectly promote hummingbird visitation of fireweed (Tamm and Gass, 1986
). Similarly, plants capable of maintaining high l, by producing copious amounts of dilute nectar may gain an advantage in hummingbird-mediated pollination in dry habitats. In this way, variation in the capacity to maintain l itself could be subject to indirect selection from pollinators.
In conclusion, by affecting plant water status, drought indirectly influences floral traits that function as pollinator advertisements and rewards in E. angustifolium. Variation in floral traits with abiotic heterogeneity will likely cause differences in the quality of pollinator service between environments. Our results suggest that in heterogeneous environments selection by pollinators is intimately linked to the physiology of resource acquisition in their host plants.
|
FOOTNOTES |
|---|
4 Current address: Boreal Ecology Cooperative Research Unit, University of Alaska, P.O. Box 756780, Fairbanks, Alaska 99775-6780 USA.
5 Author for correspondence (e-mail: galenc{at}missouri.edu
)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Bell, G. 1985 On the function of flowers. Proceedings of the Royal Society of London 224: 223–265
Best, L. S., and P. Bierzychudek. 1982 Pollinator foraging of Foxglove (Digitalis purpurea): a test of a new model. Evolution 36: 70–79[CrossRef][ISI]
Boose, D. L. 1997 Sources of variation in floral nectar production rate in Epilobium canum (Onagraceae): implications for natural selection. Oecologia 110: 493–500[CrossRef][ISI]
Campbell, D. R. 1991 Effects of floral traits on sequential components of fitness in Ipomopsis aggregata. American Naturalist 137: 713–737[CrossRef][ISI]
———. 1996 Evolution of floral traits in a hermaphroditic plant: field measurements of heritabilities and genetic correlations. Evolution 50: 1442– 1453[CrossRef][ISI]
Carroll, A. B. 2000 The relationship between plant water status and floral trait expression in fireweed, Epilobium angustifolium L. Master's thesis, University of Missouri, Columbia, Missouri, USA
Chapin, D. M. 1995 Physiological and morphological attributes of two colonizing plant species on Mount St. Helens. American Midland Naturalist 133: 76–87[CrossRef][ISI]
Conner, J. K., and S. Rush. 1996 Effects of flower size and number on pollinator visitation to wild radish, Raphanus raphanistrum. Oecologia 105: 509–516[CrossRef][ISI]
Cresswell, J. E. 1990 How and why do nectar-foraging bumblebees initiate movements between inflorescences of wild bergamont (Monarda fistulosa) (Lamiaceae). Oecologia 82: 450–460[CrossRef][ISI]
———, and C. Galen. 1991 Frequency-dependent selection and adaptive surfaces for floral character combinations: the pollination of Polemonium viscosum. American Naturalist 138: 1342–1353[CrossRef][ISI]
Devlin, B., J. B. Horton, and A. G. Stephenson. 1987 Patterns of nectar production of Lobelia cardinalis. American Midland Naturalist 117: 289–295[CrossRef][ISI]
Deutschman, D. H., S. A. Levin, and S. W. Pacala. 1999 Error propagation in a forest succession model: the role of fine-scale heterogeneity in light. Ecology 80: 1927–1943[CrossRef][ISI]
Fenster, C. B., and K. Ritland. 1994 Quantitative genetics of mating system divergence in the yellow monkey flower species complex. Heredity 73: 422–435[ISI]
Frazee, J. E., and R. J. Marquis. 1994 Environmental contribution to floral trait variation in Chamaecrista fasciculata (Fabaceae: Caesalpinoideae). American Journal of Botany 81: 206–215[CrossRef][ISI]
Galen, C. 1996 Rates of floral evolution: adaptation to bumblebee pollination in the alpine wildflower, Polemonium viscosum. Evolution 50: 120–125[CrossRef][ISI]
———. 1999 Why do flowers vary? The functional ecology of variation in flower size and form within natural plant populations. BioScience 49: 631–640[CrossRef][ISI]
———, and R. C. Plowright. 1985 The effects of nectar level and flower development on pollen carry-over in inflorescences of fireweed (Epilobium angustifolium) (Onagraceae). Canadian Journal of Botany 63: 488– 491
———, R. A. Sherry, and A. B. Carroll. 1999 Are flowers physiological sinks or faucets? Costs and correlates of water use by flowers of Polemonium viscosum. Oecologia 118: 461–470[CrossRef][ISI]
Geber, M. A., and T. E. Dawson. 1990 Genetic variation in and covariation between leaf gas exchange, morphology, and development in Polygonum arenastrum, an annual plant. Oecologia 85: 153–158[CrossRef][ISI]
Heinrich, B. 1976 Resource partioning among some eusocial insects: bumblebees. Ecology 57: 874–889[CrossRef][ISI]
Hendrix, S. D. 1988 Herbivory and its impact on plant reproduction. In J. D. Lovett Doust and L. L. Lovett Doust [eds.], Plant reproductive ecology: patterns and strategies, 246–263. Oxford University Press, New York, New York, USA
Hodges, S. 1995 The influence of nectar production on hawkmoth behavior, self pollination, and seed production in Mirabilis multiflora (Nyctaginaceae). American Journal of Botany 82: 197–204[CrossRef][ISI]
Holtsford, T. P., and N. C. Ellstrand. 1992 Genetic and environmental variation in floral traits affecting outcrossing rate in Clarkia tembloriensis (Onagraceae). Evolution 46: 216–225[CrossRef][ISI]
Johnson, S. G., L. F. Delph, and 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., and M. A. Geber. 1999 Variation among populations of Clarkia unguiculata (Onagraceae) along altitudinal and latitudinal gradients. American Journal of Botany 86: 333–343
Kaul, R. B. 1979 Inflorescence architecture and flower sex ratios in Sagittaria brevirostra (Alismataceae). American Journal of Botany 66: 1062–1066[CrossRef][ISI]
Lambers, H., F. S. Chapin, and T. L. Poris. 1998 Plant physiological ecology. Springer-Verlag, New York, New York, USA
Lee, S. G., and P. Felker. 1992 Influence of water/heat stress on flowering and fruiting of mesquite (Prosopis glandulosa var. glandulosa). Journal of Arid Environments 23: 309–319[ISI]
Lehtila, K., and S. Y. Strauss. 1997 Leaf damage by herbivores affects attractiveness to pollinators in wild radish, Raphanus raphanistrum. Oecologia 111: 396–403[CrossRef][ISI]
———, and ———. 1999 Effects of foliar herbivory on male and female reproductive traits of wild radish, Raphanus raphanistrum. Ecology 80: 116–124[CrossRef][ISI]
Marquis, R. J. 1992 The selective impact of herbivores. In R. S. Fritz and E. L. Simms [eds.], Plant resistance to herbivores and pathogens: ecology, evolution, and genetics, 301–325. University of Chicago Press, Chicago, Illinois, USA
Mazer, S. J., and C. T. Schick. 1991 Constancy of population parameters for life-history and floral traits in Raphanus sativus L. II. Effects of planting density on phenotype and heritablity estimates. Evolution 45: 1888–1907[CrossRef][ISI]
Mitchell, R. J. 1993 Path analysis: pollination. In S. M. Scheiner and J. Gurevitch [eds.], Design and analysis of ecological experiments, 211– 231. Chapman and Hall, New York, New York, USA
———, and R. G. Shaw. 1993 Heritability of floral traits for the perennial wild flower Penstemon centranthifolius (Scrophulariaceae): clones and crosses. Heredity 71: 185–192[ISI]
Mohan Ram, H. Y., and I. V. Rao. 1984 Physiology of flower bud growth and opening. Proceedings of the Indian Academy of Natural Sciences 93: 253–274
Mosquin, T. 1966 A new taxonomy for Epilobium angustifolium L. (Onagraceae). Brittonia 18: 167–188[CrossRef][ISI]
Murrell, D. C., D. T. Tomes, and R. W. Shuel. 1982 Inheritance of nectar production in birdsfoot trefoil. Canadian Journal of Plant Sciences 62: 101–105
National Climate Data Center (NOAA). 1998 Monthly precipitation records for Division 4: Platte River drainage basin. http://www.ncdc.noaa.gov
Nobel, P. S. 1977 Water relations of flowering Agave deserti. Botanical Gazette 138: 1–6
Pearcy, R. W., J. Ehleringer, H. A. Mooney, and R. W. Rundel. 1989 Plant physiological ecology. Chapman and Hall, New York, New York, USA
Percival, M. S. 1965 Floral biology. Pergamon Press, New York, New York, USA
Pleasants, J. M. 1983 Nectar production patterns in Ipomopsis aggregata (Polemoniaceae). American Journal of Botany 70: 1468–1475[CrossRef][ISI]
Pyke, G. H. 1978 Optimal foraging in hummingbirds: testing the marginal value theorem. American Zoologist 18: 739–752
———, and N. M. Waser. 1981 The production of dilute nectar by hummingbird and honeyeater flowers. Biotropica 13: 26–270
Quesada, M., K. Bollman, and A. G. Stephenson. 1995 Leaf damage decreases pollen production and hinders pollen performance in Cucurbita texana. Ecology 76: 437–443[CrossRef][ISI]
Salisbury, F. B., and C. W. Ross. 1978 Plant physiology. Wadsworth, Belmont, California, USA
Schmid-Hempel, P., and B. Speiser. 1988 Effects of inflorescence size on pollination in Epilobium angustifolium. Oikos 53: 98–104[CrossRef][ISI]
Stanton, M. L., and H. J. Young. 1994 Selecting for floral character associations in wild radish, Raphanus sativus L. Journal of Evolutionary Biology 7: 271–285[CrossRef][ISI]
Stratton, D. A. 1992 Life-cycle components of selection in Erigeron annuus: I. Genetic variation. Evolution 46: 107–120[CrossRef][ISI]
Strauss, S. Y., J. K. Conner, and S. L. Rush. 1996 Foliar herbivory affects floral characters and plant attractiveness to pollinators: implications for male and female fitness plant fitness. American Naturalist 147: 1098–1107[CrossRef][ISI]
Sutherland, S. 1986 Floral sex ratios, fruit set, and resource allocation in plants. Ecology 67: 991–1001[CrossRef][ISI]
Tamm, S., and C. L. Gass. 1986 Energy intake rates and nectar concentration preferences by hummingbirds. Oecologia 70: 20–23[CrossRef][ISI]
Teuber, L. R., D. K. Barnes, and C. M. Rincker. 1983 Effectiveness of selection for nectar volume, receptacle diameter, and seed yield characteristics in alfalfa. Crop Science 23: 283–289
Thomson, J. D., W. P. Maddison, and R. C. Plowright. 1982 Behavior of bumblebee pollinators of Aralia hispida Vent. (Araliaceae). Evolutionary Ecology 2: 65–76; Ecology 70: 1061–1068[CrossRef]
Whiley, A. W., K. R. Chapman, and J. B. Saranah. 1988 Water loss by floral structures of avocado (Persea americana cv. Fuerte) during flowering. Australian Journal of Agricultural Research 39: 457–467[CrossRef][ISI]
Wolfe, L. M. 1992 Why does the size of reproductive structures decline through time in Hydrophyllum appendiculatum (Hydrophyllaceae)?: developmental constraints vs. resource limitation. American Journal of Botany 79: 1286–1290[CrossRef][ISI]
Wright, S. 1921 Correlation and causation. Journal of Agricultural Research 20: 557–558
———. 1934 The method of path coefficients. Annals of Mathematics and Statistics 5: 161–215[CrossRef]
Wyatt, R., S. B. Broyles, and G. S. Derda. 1992 Environmental influences on nectar production in milkweeds (Ascelpias syriaca and A. exaltata). American Journal of Botany 79: 636–642[CrossRef][ISI]
Young, H. J., M. L. Stanton, N. C. Ellstrand, and J. M. Clegg. 1994 Temporal and spatial variation in heritability and genetic correlations among floral traits in Raphanus sativus, wild radish. Heredity 73: 298– 308[ISI]
Zimmerman, M. 1983 Plant reproduction and optimal foraging: experimental nectar manipulations in Delphinium nelsonii. Oikos 41: 57–63[CrossRef][ISI]
———. 1988 Pollination biology of montane plants: relationship between rate of nectar production and standing crop. American Midland Naturalist 120: 50–57[CrossRef][ISI]
———, and G. H. Pyke. 1988 Experimental manipulations of Polemonium foliosissimum: effects on subsequent nectar production, seed production and growth. Journal of Ecology 76: 777–789[CrossRef]
This article has been cited by other articles:
|
|
 |
|
  U. Effmert, C. Dinse, and B. Piechulla Influence of Green Leaf Herbivory by Manduca sexta on Floral Volatile Emission by Nicotiana suaveolens Plant Physiology, April 1, 2008; 146(4): 1996 - 2007. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. B. Fenster, G. Cheely, M. R. Dudash, and R. J. Reynolds Nectar reward and advertisement in hummingbird-pollinated Silene virginica (Caryophyllaceae) Am. J. Botany, December 1, 2006; 93(12): 1800 - 1807. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. ACKERMANN and M. WEIGEND Nectar, Floral Morphology and Pollination Syndrome in Loasaceae subfam. Loasoideae (Cornales) Ann. Bot., September 1, 2006; 98(3): 503 - 514. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Caruso Plasticity of inflorescence traits in Lobelia siphilitica (Lobeliaceae) in response to soil water availability Am. J. Botany, April 1, 2006; 93(4): 531 - 538. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. HERRERA Flower Size Variation in Rosmarinus officinalis: Individuals, Populations and Habitats Ann. Bot., February 1, 2005; 95(3): 431 - 437. |
} = 1.0760 for predawn {
