| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ecology |
Department of Biology, Queen's University, Kingston, Ontario, Canada K7L 3N6
Received for publication June 11, 2003. Accepted for publication October 16, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: allometry • gender allocation • plasticity • pollen-dispersal • ragweed • size advantage • time limitation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Freeman et al., 1981
; Brunet, 1992
). Patterns of gender allocation in plants have been most commonly associated with variation in plant size. Size-dependent (allometric) gender allocation occurs when investment in male and/or female function changes disproportionately with increasing plant size and is predicted when male and/or female fitness returns differ between small and large plants (Ghiselin, 1969
; Klinkhamer et al., 1997
). Because within-species size variation in plants is commonly a product of environmental variation (e.g., in light, soil resources, local proximity of neighbors) and because plants have no control over the environments that they experience, theory predicts that size-dependent gender allocation (through effects of environment on plant size) should have adaptive significance in heterogeneous environments (Charnov and Bull, 1977
; Freeman et al., 1981
; Lloyd and Bawa, 1984
; Solomon, 1985
; Delph and Lloyd, 1991
; Korpelainen, 1998
; Barrett et al., 1999
).
According to established theory, plant size can influence the relative fitness costs and benefits of male and female function in two qualitatively different ways. The first is based on the assumption that resources or time for reproduction are limited and that this results, therefore, in a trade-off between male and female function. Relative to male function (pollen production and dispersal), female function in flowering plants demands a higher energetic investment (greater resource requirement) (Bierzychudek, 1984
; Lloyd and Bawa, 1984
; de Jong and Klinkhamer, 1994
) and also a longer time commitment for the production of mature seed and fruit (Day and Aarssen, 1997
). Smaller plants are presumed to have both smaller energetic reserves and a higher risk of mortality and hence, an abbreviated life span. Thus, according to the "size-advantage" hypothesis (Ghiselin, 1969
; Charnov, 1982
; Lloyd and Bawa, 1984
) and the "time-commitment" hypothesis (Day and Aarssen, 1997
), a negative correlation between maleness and plant size evolves by natural selection because the energetic and temporal limitations (respectively) associated with relatively small plant size imposes less severe limitation for male reproductive success than for female reproductive success. Accordingly, allocation to female function is expected to increase with increasing size faster than allocation to male function. Although the time-commitment hypothesis has never been tested directly, support for the size-advantage hypothesis has been interpreted from the results of several empirical studies (Freeman et al., 1981
; Bierzychudek, 1982
, 1984
; Lovett Doust and Cavers, 1982
; Policansky, 1987
; Schlessman, 1988
, 1991
; de Jong and Klinkhamer, 1989
; Zimmerman, 1991
; Pickering and Ash, 1993
; Klinkhamer et al., 1997
; Korpelainen, 1998
; Wright and Barrett, 1999
; Koelewijn and Hunscheid, 2000
).
Plant size, however, can also influence fitness returns directly. When pollination occurs by wind, the prediction for allometric gender allocation may be the converse of the size-advantage and time-commitment hypotheses, that is, maleness may be expected to increase with increasing plant size and, in particular, increasing plant height. The predicted advantage to male success here comes from the direct positive effect of plant height on pollen flight distances, thus increasing access to mates and reducing local mate competition among sib pollen. According to the "pollen-dispersal" hypothesis (Burd and Allen, 1987), therefore, a positive correlation between maleness and plant size evolves by natural selection in wind-pollinated species because success as an outcrossing donor of wind-dispersed pollen increases with increasing height, especially relative height within a crowded population (Lundholm and Aarssen, 1994
), whereas success as a recipient of wind-dispersed pollen is independent of height or increases with decreasing height. Support for the pollen-dispersal hypothesis has been demonstrated for a variety of wind-pollinated herbs, shrubs, and trees (McKone and Tonkyn, 1986
; Solomon, 1989
; Ackerly and Jasienski, 1990
; Traveset, 1992
; Fox, 1993
; Lundholm and Aarssen, 1994
; Pannell, 1997
; see Klinkhamer et al., 1997
for a review).
Wind-pollinated species are particularly interesting here because they have potential to incur fitness benefits from size-dependent gender allocation associated with not only the pollen-dispersal hypothesis but also the size-advantage and/or time-commitment hypotheses, simultaneously. How do the selection mechanisms of these hypotheses interact in wind-pollinated species? If they operate simultaneously, then, because they involve conflicting predictions, we might expect their effects on gender allocation to be confounded, with no consistent allometric pattern evident at all. Indeed, Bickel and Freeman (1993)
analyzed patterns of floral sex allocation with respect to size for 22 species (involving 12 families) of monoecious plants. They found that femaleness increased with size in the eight species pollinated by animals, whereas maleness increased with size in only eight of the 14 wind-pollinated species.
Alternatively, is it possible that in wind-pollinated species, the pattern and/or degree of allometry in gender allocation might display adaptive plasticity depending on the main environmental causes of plant size variation? This question was explored in the present study using the monoecious, wind-pollinated annual Ambrosia artemisiifolia (common ragweed). Monoecious plants are convenient subjects for testing theories of gender allocation because gender can be continuously modified (and easily recorded) through the relative production of separate male and female flowers. Previous research indicates support for the pollen-dispersal hypothesis in Ambrosia artemisiifola; a positive correlation between maleness and plant size has been reported in both natural populations and greenhouse studies (McKone and Tonkyn, 1986
; Ackerly and Jasienski, 1990
; Traveset, 1992
; Lundholm and Aarssen, 1994
). However, Lundholm and Aarssen (1994)
also found that plants that grew in solitude from the seedling stage had greater relative female allocation than smaller plants that had grown with close neighbors, suggesting that these different hypotheses for allometric gender allocation may involve interacting mechanisms. Hence, in the present study, parallel experiments in the field and the greenhouse were designed to vary and control the effects of neighbor proximity, light availability, and soil nutrient availability, separately and in combination. A wide range of treatment levels was used to generate a range of plant sizes, making it possible to compare patterns of plastic allometry in gender allocation when the main environmental cause of size variation was associated with variation in neighbor (including shading) effects vs. variation in soil nutrient levels only. We predicted that variation in size resulting from variation in shading effects (including from neighbors) will generate allometric gender allocation consistent with the pollen-dispersal hypothesis, whereas variation in size resulting from variation in soil nutrient levels only (without shading effects) will generate allometric gender allocation consistent with the size-advantage and time-commitment hypotheses.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Ragweed was chosen as the study species because it is monoecious with separate, relatively long-lived, and easily distinguishable male and female flowers on the same individual. This makes it relatively easy to score individual plants for gender allocation. Being monocarpic, it is also possible to record final ("end of life") gender allocation for each individual. Finally, ragweed is economically important as one of the most noxious annual weeds of disturbed ground, and its pollen is the main cause of allergic "hayfever" reactions in humans (Bassett and Crompton, 1975
). Understanding patterns of gender allocation in ragweed, therefore, may have implications for ragweed control, ensuing from the direct relationship between "maleness" and pollen production. Plant material for this study was collected as seeds, pooled from approximately 100 parent plants selected randomly from a small (approximately 100 m2) local ragweed population in the late summer of 1998. The seeds were stored at 4°C until use.
Study site
Parallel experiments in the field and the greenhouse were conducted to examine the effects of neighbors and soil resources on allometric gender allocation. A flat, uniform field plot (50 x 100 m) that is tilled each autumn and a similar plot within an adjacent abandoned sand quarry were used in the field portion of the study, located on the property of Queen's University Biological Station (QUBS) at Lake Opinicon (Storrington District, South Frontenac Township), 50 km north of Kingston, Ontario, Canada (44°34'02'' N, 760°21'52'' W). Several weedy species, consisting mostly of Ambrosia artemisiifolia, emerged naturally from the seed bank at this site and served as the "neigbors" in the field experiments described below. The greenhouse portion of the study was conducted at the Queen's University Phytotron.
Experimental design
The seeds collected in 1998 were germinated in the greenhouse in May 1999. Seeds were planted individually in 5-cm pots using a standard greenhouse potting mixture containing no added nutrients. At 2 wk of age, seedlings of uniform size were transplanted both to the field site directly into the soil and, in the greenhouse, into 20-cm pots filled with turface (Profile Products LLC, Buffalo Grove, Illinois, USA). Turface, an inert growth medium consisting of coarse inorganic particles, was used as the growth medium because it facilitates careful control over nutrient availability, it is well drained thereby minimizing the occurrence of pests and disease, and its roots are relatively easily washed off at harvest, thus allowing root biomass to be measured.
Plants in the field were spaced 2 m apart and watered daily for 1 wk after planting to ensure successful establishment before beginning the treatments. Plants in the greenhouse were spaced 1 m apart, watered daily using an automated sprinkler, and maintained in a 16-h day/8-h night cycle throughout the study using supplemental greenhouse lighting. To minimize the problems associated with a heterogeneous environment, randomized complete blocks (with replication) were used in the design of the experiments. The field plots were partitioned into four blocks, and the treatments were randomly allocated within each of the plots. The greenhouse benches were organized into four blocks from front to back of the greenhouse, and the treatments were randomly assigned to benches within each of the blocks.
Five individual experiments, three in the greenhouse and two in the field, were designed to test predictions for allometric gender allocation in ragweed. All of the experiments had a similar general design but differed in the number of treatment levels. The main objective of varying treatment levels was to generate a gradient in plant size, not primarily to test the effect of treatment levels per se (although treatment effects were also analyzed). Two predictions were tested, associated with two different potential causes of plant size variation: variation in shading effects and variation in soil nutrient levels.
Variation in size resulting from variation in effects from shading will generate allometric gender allocation consistent with the "pollen-dispersal" hypothesis
In the field neighbor experiment, a range of plants sizes was achieved by growing the plants under a range of neighbor densities. In June 1999, 30 ragweed seedlings were randomly assigned to each of three neighbor density treatments: low, all naturally occurring neighbors within a 50-cm radius around the target plant were removed at ground level and kept cleared (by hand weeding) for the duration of the experiment; medium, a 20-cm neighbor removal radius was maintained for the duration of the experiment; and high, no vegetation was removed around the target plant. The plants were grown under natural field soil nutrient levels and were watered periodically and evenly throughout the growing season to minimize variation in local soil moisture levels.
In the greenhouse "light" experiment, aboveground neighbor effects were simulated by a gradient of light treatments using a green filter (Lee, Andover, UK; number 121 Lee green) and various neutral density shade cloths (Plant Products, Brampton, Ontario, Canada) supported on circular wire cages completely surrounding each potted plant. A 30-cm square panel covered with the appropriate filter/shade cloth combination was placed on top of the cage to provide shade from directly above the plant. To maintain airflow around the plants, a gap of 5 cm was left at both the bottom of the cage and between the top and side shading layers of the cage. The green filter was used because it produces a change in the spectral quality of light that is similar to the change resulting from light being filtered through the leaves of neighboring plants (Bonser and Aarssen, 2003
). The filter reduced light transmittance by 36%. Further shading, without further changing the spectral quality of the light, was achieved by adding one of four neutral density shade cloths that reduced light transmittance by an additional 30, 51, 63, and 80%, respectively. Hence, six levels of shading were created: natural light intensity, filter only, filter plus 30% shade cloth, filter plus 51% shade cloth, filter plus 63% shade cloth, and filter plus 80% shade cloth. These treatments represented 100, 64, 45, 32, 24, and 13% of incident light reaching the target plant, corresponding respectively to a range of light intensities from 700 to 167 µmol · m–2 · s–1 (based on an average of three measurements taken at mid-day on three separate cloudless days using a LI-COR light meter, model lL-189; LI-COR, Lincoln, Nebraska, USA). Ten plants were randomly assigned to each light intensity treatment for a total of 60 plants. All of the plants were treated weekly with 100% of full strength ("very high") nutrient solution (see next section).
Variation in size resulting from variation only in soil nutrient levels will generate allometric gender allocation consistent with the "size-advantage"/"time-commitment" hypotheses
In the field nutrient experiment, a range of plant sizes was achieved by varying soil nutrients available for growth. Twenty ragweed seedlings were randomly assigned to each of five nutrient levels: VH (very high), treated weekly with 100% of full strength nutrient solution; H (high), treated weekly with 10% of full strength nutrient solution; M (medium), treated weekly with water only; L (low), no nutrients or water; and VL (very low), planted in adjacent abandoned sand quarry (VH, H, M, and L treatments were in loam soil, which has a higher nutrient- and water-holding capacity). The nutrient solution was prepared from a 20 : 20 : 20 (N : P : K) fertilizer and mixed at 2.00 g/L to obtain the full strength solution. For each nutrient level, 250 mL of the appropriate dilution was applied to each plant. All vegetation within 50 cm of each target plant was removed at ground level and kept cleared for the duration of the experiment.
In the greenhouse nutrient experiment, 10 ragweed seedlings were randomly assigned to each of six nutrient treatments: water only, 0.1, 1, 10, 50, and 100% of full strength nutrient solution. For each nutrient level, 50 mL of the appropriate dilution was applied to each plant on a weekly basis. Each plant was grown in full (100% of incident) light.
The pattern and strength of allometric gender allocation may not be evident if the selection mechanisms associated with the pollen-dispersal hypothesis and size-advantage hypothesis (variation in neighbor effects vs. variation in soil resources) operate simultaneously. Moreover, neighbor effects are likely to involve both light and nutrient deprivation simultaneously. To explore this effect, greenhouse light and nutrient treatments were varied simultaneously (i.e., not factorially), ranging from resource rich soil with no shading effects to impoverished soil with strong shading effects (called here the greenhouse "light/nutrient" experiment). Ten plants were randomly assigned to each of six treatments representing a gradient of combined resources declining from 100% incident light and 100% full strength nutrient solution, respectively (i.e., 100%/100%, 64%/50%, 45%/10%, 31%/1%, 24%/0.1%, and 13%/0.0%).
Harvesting and data collection
Field plants were harvested in mid-September when they had stopped flowering but had not yet dispersed their seeds. This ensured that all of the plants had reached their final, "end-of-life" gender allocation. Field plants were harvested at ground level and stored in airtight bags at 4°C until they could be scored for gender allocation in the laboratory. The total number of male inflorescences (capitula) and the total number of seeds (corresponding to the number of female flowers; i.e., fruit set = 1) were recorded for each individual plant. Some of the male capitula had fallen off by the time of harvest, but these could still be accounted for by counting the persistent stalks to which these capitula were attached. Each plant was then measured for height (to the highest point on the plant), oven-dried (at 90°C for 48 h), and then weighed. The same procedure was repeated for the greenhouse plants, except that these were harvested with their roots intact after rinsing away the turface with tap water.
Statistical analyses
All of the variables measured were log transformed to achieve normality and homogeneity of variances. For each of the experiments, except the field nutrient experiment, a two-way mixed model ANOVA was used with blocks as the random effect and nutrient level, light level, or neighbor density as fixed effects. The field nutrient experiment was an incomplete block design (the very low treatment was present in the sand quarry only); hence, a one-way ANOVA model was used here to detect significant treatment effects. Multiple comparison tests of treatment means were carried out using the Tukey-Kramer method.
Both least squares (Type I) and reduced major axis (Type II) regression analyses were performed to detect relationships between gender allocation and size based on the allometric model: log Y = log a + b log X (Klinkhamer and de Jong, 1997). The Type II regression coefficient b was tested for significant departure from 1.0 indicating a significant departure from isometry. A regression coefficient smaller than 1 indicates that an increase in X is accompanied by a less than proportional increase in Y, whereas a regression coefficient greater than 1 indicates a greater than proportional increase in Y. Tests for departure from isometry and for significant differences between slopes were carried out using a t test or directional t test (Rice and Gaines, 1994
). All analyses were done using the JMP statistical package (version 3.02, SAS Institute, 1994).
Relationship between plant size and pollen production per capitulum
To determine if relative pollen production on plants of different sizes could be accurately estimated from a count of the male capitula, pollen production per capitulum was recorded for a subset of plants (n = 20) selected from the field experiments ranging in size from 192 to 686 mm tall, from 0.13 g to 10.24 g dry mass, from 32 to 2291 male capitula, and from 11 to 1310 seeds.
To account for possible within-plant variation in capitulum size, male capitula were collected from three locations on each plant; the highest, middle, and lowest branches (with three replicates per location, per plant). However, in very small plants, capitula were collected from one location only. The capitula were preserved in 0.3 mL of 70% ethanol and pollen was counted using a hemacytometer. Before the counts, the anthers were crushed with a blunt metal rod to separate the pollen grains, stained with one drop of 1% (m/v) aqueous aniline blue and vortexed for 30 s. Using a glass pipette, a small subsample was placed on the grid of the hemacytometer for counting. Six replicate counts were performed for each capitulum (Kearns and Inouye, 1993
). ANCOVA was used to compare pollen production per capitulum between locations within plants using plant dry mass and plant height (in separate analyses), as the covariate.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Bassett and Crompton, 1975; McKone and Tonkyn, 1986
). As these unisexual genotypes are incapable of environmental gender modification and hence, incapable of responding to the experimental treatments, they introduce a potentially confounding source of variation in the data. Accordingly, to minimize this effect, prior to statistical analyses, these all-female plants were omitted from data sets involving treatments with the highest light level, and the largest 10% of the all-female plants (corresponding to the presumed percentage of genetically determined all-females) were eliminated from the data sets involving treatments with lower light levels.
|
|
|
Gender allocation along a plant size gradient
With increasing reproductive output (associated with increasing plant size), the number of male capitula increased more rapidly than did the number of seeds in the greenhouse light experiment (P = 0.05; Fig. 2a), in the field neighbor experiment (P = 0.04; Fig. 2b), and in the greenhouse light/nutrient experiment (P = 0.04; Fig. 2e). These results all represent a significant departure from isometry, i.e., the regression slope is greater than 1.0 indicating positive allometry for all three regressions (Fig. 2). In the field nutrient experiment, however, there was a negative allometric relationship, i.e., the number of male capitula increased less rapidly than the number of seeds with increasing reproductive output (plant size), with a slope <1.0 (P = 0.05) (Fig. 2d). The slope for the field nutrient experiment (Fig. 2d) was also significantly less than the slope for the field neighbor experiment (Fig. 2b) (t = 2.37, Pdir < 0.01). The slope for the greenhouse nutrient experiment (Fig. 2c) was also less than the slope for the greenhouse light experiment (Fig. 2a), although not significantly so (t = 1.35, Pdir = 0.11). The regression slope for the greenhouse nutrient treatment (Fig. 2c) did not show a significant departure from isometry.
|
1). However, the high frequency of the all-female condition (with zero male capitula) among the smallest plants associated with low light/strong neighbor effects (Fig. 1) violates the assumption of normality in regression analysis. Thus, while the slope for the greenhouse light experiment (Fig. 2a), the field neighbor experiment (Fig. 2b), and greenhouse light/nutrient experiment (Fig. 2e) are significantly (P < 0.05) greater than 1.0, this disproportionately greater increase (with increasing plant size) in male reproductive output appears to be driven, to some extent at least, by the presence of several all-female plants (P > 0.05 if these plants are excluded from analysis).
Maleness, femaleness, and plant mass
In the field neighbor experiment, there was a positive allometric increase in maleness (male capitula production) with increasing plant mass, i.e., with a regression slope (b = 1.28) significantly greater than 1.0 (P = 0.05) (Fig. 3a), and this slope was also significantly greater than the slope for femaleness (seed production) vs. plant mass (b = 0.96; Fig. 3b) (t = 1.85, Pdir = 0.043). In contrast, maleness was isometric with plant mass in the field nutrient experiment (Fig. 3c), but this slope (b = 0.89) was significantly less than the slope for maleness vs. plant mass in the field neighbor experiment (b = 1.28; Fig. 3a) (t = 2.03, Pdir = 0.028). There was no significant departure from isometry (slope = 1.0) for femaleness vs. plant mass in either the field neighbor (Fig. 3b) or field nutrient (Fig. 3d) experiments.
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
If a plant's relative height is more important than its absolute height in affecting its relative success as a pollen donor (Lundholm and Aarssen, 1994
), then the pollen dispersal hypothesis may not apply when there are no close neighbors and hence, no shading effects at all. If, instead of neighbor effects, size variation is a consequence of only variation in soil nutrient level, then we predicted that A. artemissiifolia would display allometric gender allocation patterns consistent with the size-advantage/time-commitment hypotheses. Significant regression results were obtained that support this prediction: in the field nutrient experiment, greater reproductive output (associated with larger plant size) was associated with increased femaleness (i.e., relative to maleness, with b < 1.0; Fig. 2d). Also, the rate of increase in female reproductive output per unit plant mass in the greenhouse was significantly greater when size variation was affected by soil nutrient level only (Fig. 4d) than when size variation was due to effects that included variation in shade (Fig. 4b and f).
While the increase in total reproductive output (male and female) was mostly isometric (i.e., showing proportional increase) with increasing plant mass (Figs. 3 and 4), total reproductive output was allometric in terms of its relationship with plant height; i.e., taller plants, regardless of the environmental source of height variation, produced disproportionately more pollen and disproportionately more seeds than shorter plants (i.e., with b > 1.0) in both the greenhouse (Fig. 5) and in the field (Fig. 6). This can be explained by the negative allometric relationship between plant mass and plant height; i.e., for approximately every 104-fold increase in plant mass, there was only about a 10-fold increase in plant height in both the greenhouse experiments (compare abscissa scales in Fig. 4 vs. Fig. 5) and field experiments (compare abscissa scales in Figs. 3 vs. 6). This, in turn, is accounted for by the fact that, as plants increase in size, they add mass not only in terms of increased height but also (and more extensively) in terms of increased branching.
Significant relationships involving plant height, however, also support the pollen-dispersal hypothesis for the response in maleness to shading effects and for the response in femaleness to soil nutrient level effects: The increase in maleness per unit height was significantly greater when height variation was caused by variation in field neighbor proximity (Fig. 6a) than when height variation was caused by variation in field soil nutrient levels (Fig. 6c). Conversely, in the greenhouse, the increase in femaleness per unit plant height was significantly greater when height variation was caused by variation in soil nutrient levels only (Fig. 5d) than when height variation was caused by combined variation in nutrient and shading levels (Fig. 5f).
Several previous studies have reported size-dependent gender allocation in A. artemesiifolia (McKone and Tonkyn, 1986
; Ackerly and Jasienski, 1990
; Traveset, 1992
; Lundholm and Aarssen, 1994
). Most of these studies support the pollen-dispersal hypothesis. However, the independent control and manipulation of both shading effects and soil nutrient availability in this study have never been used in previous studies of gender allocation. The use of light filters and shade cloths to control the quantity and quality of light reaching target plants allowed direct and precise manipulation of shading effects, simulating competition for light without any effects of competition for belowground resources. Our results confirm that variation in shading effects alone (i.e., even when belowground competition for soil resources is absent) is the cue for the shift in gender allocation in A. artemissifolia, favoring femaleness when small and maleness when large, consistent with the pollen-dispersal hypothesis (Figs. 1a, 2a). However, when plant size variation is due to variation in soil resources only (without variation in shading), A. artemissiifolia is also capable of displaying size-dependent gender consistent with the size-advantage/time limitation hypotheses. To the best of our knowledge, these results represent the first report of allometric gender allocation in which the pattern of allometry displays adaptive plasticity consistent with the interpretations of established theory.
|
FOOTNOTES |
|---|
2 aarssenl{at}biology.queensu.ca
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Barrett S. C. H. A. L. Case G. B. Peters 1999 Gender modification and resource allocation in subdioecious Wurmbea Dioica (Colchicaceae). Journal of Ecology 87: 123-137[CrossRef][ISI]
Bassett I. J. C. W. Crompton 1975 The biology of Canadian weeds 11. Ambrosia artemisiifolia L. and A. psilostachya D.C. Canadian Journal of Plant Science 55: 463-476[ISI]
Bickel A. M. D. C. Freeman 1993 Effects of pollen vector and plant geometry on floral sex ratio in monoecious plants. American Midland Naturalist 130: 239-247[CrossRef][ISI]
Bierzychudek P. 1982 The demography of Jack-in-the-pulpit, a forest perrenial that changes sex. Ecological Monographs 52: 335-352[CrossRef][ISI]
Bierzychudek P. 1984 Determinants of gender in Jack-in-the-pulpit: the influence of plant size and reproductive history. Oecologia 65: 14-18[CrossRef][ISI]
Bonser S. P. L. W. Aarssen 2003 Allometry and development in herbaceous plants: functional responses of meristem allocation to light and nutrient availability. American Journal of Botany 90: 404-412
Brunet J. 1992 Sex allocation in hermaphroditic plants. Trends in Ecology and Evolution 7: 79-84[CrossRef]
Burd M. T. F. H. Allen 1987 Sexual allocation in wind-pollinated plants. Evolution 42: 403-407[CrossRef][ISI]
Charnov E. 1982 The theory of sex allocation. Princeton University Press, Princeton, New Jersey, USA
Charnov E. J. Bull 1977 When is sex environmentally determined?. Nature 266: 828-830[CrossRef][Medline]
Day T. L. W. Aarssen 1997 A time-commitment hypothesis for size-dependent gender allocation. Evolution 51: 988-993[CrossRef][ISI]
De Jong T. J. P. G. L. Klinkhamer 1989 Size-dependency of sex-allocation in hermaphroditic, monocarpic plants. Functional Ecology 3: 201-206
De Jong T. J. P. G. L. Klinkhamer 1994 Plant size and reproductive success through female and male function. Ecology 82: 399-402[CrossRef]
Delph L. F. D. G. Lloyd 1991 Environmental and genetic control of gender in the dimorphic shrub Hebe Subalpina. Evolution 45: 1957-1964[CrossRef][ISI]
Fox J. F. 1993 Size and sex allocation in monoecious woody plants. Oceologia 94: 110-113[CrossRef]
Freeman D. C. E. D. McArthur K. T. Harper A. C. Blauer 1981 Influence of environment on on the floral sex ratio of monoecious plants. Evolution 35: 194-197[CrossRef][ISI]
Ghiselin M. T. 1969 The evolution of hermaphroditism among animals. Quarterly Review of Biology 44: 189-208[CrossRef][Medline]
Jones K. L. 1936 Studies on Ambrosia. I. The inheritance of floral types in the ragweed Ambrosia elaitior. American Midland Naturalist 17: 673-699[CrossRef]
Kearns C. A. D. W. Inouye 1993 Techniques for pollination biologists. University Press of Colorado, Niwot, Colorada, USA
Klinkhamer P. G. L. T. J. De Jong 1997 Size-dependent allocation to male and female reproduction. In F. A. Bazzaz and J. Grace [eds.], Plant resource allocation, 211–229. Academic Press, San Diego, California, USA
Klinkhamer P. G. L. T. J. De Jong H. Metz 1997 Sex and size in cosexual plants. Trends in Ecology and Evolution 12: 260-265[CrossRef]
Koelewijn H. P. M. P. H. Hunscheid 2000 Intraspecific variation in sex allocation in Plantago coronopus. Journal of Evolutionary Biology 13: 302-315[CrossRef][ISI]
Korpelainen H. 1998 Labile sex expression in plants. Biological Review 73: 157-180[CrossRef]
Lloyd D. G. 1979 Parental strategies of angiosperms. New Zealand Journal of Botany 17: 595-606[ISI]
Lloyd D. G. K. S. Bawa 1984 Modification of the gender of seed plants in varying conditions. Evolutionary Biology 17: 225-338
Lovett Doust J. L. P. B. Cavers 1982 Sex and gender dynamics in Jack-in-the pulpit, Arisaema triphylum (Araceae). Ecology 63: 797-808[CrossRef][ISI]
Lundholm J. T. L. W. Aarssen 1994 Neighbor effects on gender variation in Ambrosia artemisiifolia. Canadian Journal of Botany 72: 94-800
McKone M. J. D. W. Tonkyn 1986 Intrapopulation gender variation in common ragweed (Asteraceae: Ambrosia artemisiifolia L.), a monoecious, annual herb. Oecologia 70: 3-67
Pannell J. 1997 Variation in sex ratios and sex allocation in andromonoecious Mercurialis annua. Journal of Ecology 85: 57-69[CrossRef]
Pickering C. M. J. E. Ash 1993 Gender variation in hermaphrodite plants: evidence from five species of alpine Ranunculus. Oikos 68: 539-548[CrossRef][ISI]
Policansky D. 1987 Sex choice and reproductive costs in Jack-in-the-pulpit: size determines a plant's sexual state. Bioscience 37: 476-481[CrossRef][ISI]
Rice W. R. S. D. Gaines 1994 Heads I win, tails you lose: testing directional alternative hypotheses in ecological and evolutionary research. Trends in Ecology and Evolution 9: 235-237[CrossRef]
SAS Institute. 1994. SAS user's guide. SAS Institute, Cary, North Carolina, USA
Schlessman M. A. 1988 Gender modification in North American ginsengs. Bioscience 37: 469-475[CrossRef][ISI]
Schlessman M. A. 1991 Size, gender, and sex change in dwarf ginseng, Panax trifolium (Araliacea). Oceologia 87: 588-595[CrossRef]
Solomon B. P. 1985 Environmentally influenced changes in sex expression in an andromonoecious plant. Ecology 66: 1321-1332[CrossRef][ISI]
Solomon B. P. 1989 Size-dependent sex ratios in the monoecious, wind-pollinated annual, Xanthium struarium. American Midland Naturalist 121: 209-218[CrossRef][ISI]
Traveset A. 1992 Sex expression in a natural population of the monoecious annual, Ambrosia artemisiifolia (Asteraceae). American Midland Naturalist 127: 309-315[CrossRef][ISI]
Wright S. I. S. C. H. Barrett 1999 Size-dependent gender modification in a hermaphroditic perennial herb. Proceedings of the Royal Society of London B 266: 225-232[CrossRef]
Zimmerman J. K. 1991 Ecological correlates of labile sex expression in the orchid Catasetum viridiflavum. Ecology 72: 597-608[CrossRef][ISI]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
