Effects of progeny and size of the pollen load on progeny performance in Campanula americana.
That parentage plays a central role in determining the performance of offspring (survival, growth, and reproduction) is one of the most important tenets in evolutionary biology. The ability of a trait to respond to selection is dependent upon there being heritable additive genetic variation among the parents (Fisher, 1958; Falconer, 1981). In angiosperms, the processes of fertilization and seed development also permit factors other than the equal paternal and maternal nuclear gene contributions to have an effect on progeny performance. The development of seed requires the interaction of four genetically distinct tissues: the maternal plant (2n), the female gametophyte (n), the endosperm (usually 3n, and always maternally overrepresented), and the embryo (2n, n from each parent). Moreover, the maternal parent supplies most of the cytoplasm and all of the resources for the developing seed. This asymmetry in genetic composition, cytoplasm and resource contributions may elevate the maternal parent's role in the process of seed development relative to that of the paternal parent (Nakamura and Stanton, 1989). This can result in maternal effects that may extend through the life cycle of the progeny (e.g., Weis, 1982; Dolan, 1984; Stanton, 1984). In spite of a genetic underrepresentation in the developing seed, paternal effects on seed development are known, though they are of lesser magnitude at this stage than maternal influences (e.g., Nakamura and Stanton, 1989).
In addition to parental effects, recent studies have indicated that the number of pollen grains that are deposited onto the stigma can affect progeny performance (see Snow, 1986; Mulcahy and Mulcahy, 1987 for reviews). Several studies of cultivated species have shown that progeny produced by large pollen loads are more vigorous as seedlings and/or have greater reproductive output as adults than do progeny produced by smaller pollen loads (e.g., Mulcahy et al., 1975, 1978; Mulcahy and Mulcahy, 1975; Ottaviano et al., 1983; Stephenson et al., 1986; Davis et al., 1987; Winsor et al., 1987). The evidence for pollen load effects on progeny performance in noncultivated species is more equivocal. Progeny produced by large pollen loads outperformed progeny from smaller pollen loads in studies on Oenothera organensis (Fingerett, 1979), Turnera ulmifolia (McKenna, 1986) and Aureolaria flava (Ramstetter and Mulcahy, 1988). In Cassia fasciculata the size of the pollen load had a significant effect on progeny performance only under competitive conditions (Lee and Hartgerink, 1986), while in Campsis radicans the size of the pollen load had a significant effect on only one measure of progeny performance (percent germination) (Bertin, 1990). Other studies have failed to find an effect due to the size of the pollen load on progeny performance in Raphanus raphanistrum (Snow, 1990).
In this paper we report the results of a controlled crossing experiment designed to simultaneously determine the effects of parentage and size of the pollen load on progeny performance. The crosses were carried out in a natural population of Campanula americana L., while progeny performance was examined in controlled environment chambers and in an experimental garden. An earlier study has examined the effects of parentage and size of the pollen load on seed number per fruit and seed weight (Richardson and Stephenson, 1991).
Species Description and Study Site. -- Campanula americana is endemic to eastern North America and occurs in partial shade in moist thickets and floodplain woods, along streams, roads, and in drier upland woods (Gleason and Cronquist, 1963). The site for this study was an eight hectare oak, walnut, and hickory woodlot in Kalamazoo County, Michigan, approximately five km from the W. K. Kellogg Biological Station. The study site contained over 500 reproducing C. americana plants.
The life history of C. americana may be either that of a winter annual or a blennial herb depending upon the time of seed germination (Baskin and Baskin, 1984). Reproductive plants at our study site began to flower in late July. An individual plant of C. americana can produce more than 600 flowers borne in inflorescences of 1 to 10 flowers at each node. The pale blue, self-compatible flowers are pollinated primarily by bumblebees, and they are protandrous (a period of pollen presentation precedes the period of stigmatic receptivity). The fruit is a capsule containing 20 to 60 seeds. For further details on the floral biology of Campanula see Shetler (1979) and Richardson and Stephenson (1989).
MATERIALS AND METHODS
In mid July 1988, just prior to the start of flowering, 55 plants at the study site were tagged. The 55 plants were haphazardly chosen from the entire range of this population. To have a sufficient number of flowers for the experimental crosses, 8 of the largest of these plants were designated as pollen recipients, and the remaining 47 as pollen donors. The donor plants were monitored daily for flower production. Flower buds likely to open within 24 hours were covered with a cheesecloth bag to prevent pollinators from removing pollen and/or depositing pollen from other plants. The pollen from open flowers was collected and used for hand-pollinations (see below). The recipient plants were also monitored daily. Since C. americana is self-compatible (Kalisz, unpubl. data), it was necessary to emasculate each flower bud to prevent self-fertilization. Following the controlled pollination (see below), a sleeve of glassine photographic paper was placed over the ovary and remained there until either the flower was dropped or the fruit had developed for approximately two weeks, at which time the sleeve was removed and a tag was attached to the fruit. We recorded the pollination treatment (see below), the pollen donor, and the date of the pollination. Five days after pollination each flower was checked to determine if the ovary had enlarged. Whenever possible, any failed pollinations were repeated on later opening flowers.
Pollination Treatments. -- Each recipient flower was hand-pollinated using pollen from a single donor plant. Pollen was applied by lightly touching a freshly collected style covered with pollen against each of the three stigmatic surfaces. The pollen of C. americana is approximately 40 microns in diameter and is purple which makes it visible against the white stigmatic surface. This gave us a measure of control over the amount of pollen deposited. In addition, the hairs that hold the pollen onto the style prevent large clumps of pollen from being deposited with a single touch.
To investigate the possible effects of pollination intensity on fruit and seed production, two pollination levels were used for each donor and recipient combination. Each of the donors was used to make a "high pollen load" and a "low pollen load" pollination on each of the eight recipient plants. High pollen loads were achieved by touching each of the stigmatic surfaces two to three times with a pollen covered style from an individual donor. Low pollen loads were achieved by touching the stigmatic surfaces once with a pollen covered style from an individual donor. The ability to see the amount of pollen deposited allowed us to maintain a qualitative difference between the high and low pollen load treatments.
To estimate the number of grains actually deposited by the two pollination treatments, 42 flowers from nonexperimental plants in the population were hand-pollinated as described above, and collected. Twenty-one of these flowers received a high pollen load and 21 received a low pollen load. Immediately following pollination, each pistil was removed and placed in a scintillation vial filled with FAA to prevent pollen germination and tube growth. In the laboratory the FAA was evaporated, and five ml of 1% NaCl was added to each vial. The vials were then sonicated for five minutes to dislodge the pollen from the stigma and evenly distribute the pollen throughout the salt solution. Ten samples of five [Mu]l each were then taken from the solution, placed on a microscope slide, and the number of pollen grains in each sample was counted at 40 x. The sum of the number of the grains in the 10 samples was multiplied by 10 to get an estimate of the total number of pollen grains deposited. We determined that high pollen loads averaged 215 [+ or -] 79 pollen grains [Chi] [bar] [+ or -] SD, N = 21),and low pollen loads 77 [+ or -] 65 pollen grains [Chi] [bar] [+ or -] SD, N = 21).
From August to October, we collected the mature fruits prior to dehiscence and stored them individually. Later, the number of seeds per fruit and the total seed mass per fruit (to the nearest 0.1 mg) were determined. We then calculated the mean seed mass for each fruit as the total seed mass for that fruit divided by the seed number for that fruit.
Measures o Progeny Performance. -- We eliminated one pollen recipient from the progeny screening tests because the plant senesced early and aborted many of the developing fruits. We then selected the nine pollen donors that were best represented on the remaining seven recipients (i.e., both the low and high pollen loads produced a mature fruit containing at least 10 seeds). From fruits with more than 10 seeds we randomly selected 10 seeds. All groups of 10 seeds were weighed to the nearest [Mu]g. The individual seeds were randomly assigned to wells in a germination flat. The flats were placed in a growth chamber on a 14 hr light/10 hr dark cycle with a fluctuating temperature of 15 [degrees] C during the light cycle and 5 [degrees] C during the dark cycle and misted on alternate days to keep the soil moist (see Baskin and Baskin, 1984).
We monitored the trays twice daily (8:00 am and 8:00 pm) and recorded the date of emergence, and the dates on which each seedling produced the first and second fully expanded leaves. Ten days after each seedling produced its second leaf (approximately 60 days after emergence) the seedling was transplanted into another tray and moved to a cold room (5 [degrees] C with an 8 hr light/ 1 6 hr dark cycle) for vernalization (light intensity 13 to 15 [Mu][Em.sup.-2][s.sup.-1]). The seedlings were fertilized every two weeks (1/2 strength Peters 15-16-17 NPK Peat Lite Special, Robert B. Peters Co., Allentown, PA). Campanula americana seedlings require approximately 1,000 hr of 0 to 10 [degrees] C vernalization to bolt and flower (Baskin and Baskin, 1984), however, longer periods do not appear to have any deleterious effects, as the seedlings grow only slightly during the vernalization period. Therefore, by transferring the seedlings to the cold room at a uniform size we eliminated most of the size variation resulting from variation in emergence time. When the seedlings were removed from the cold room there was little variation in size, and later measures of growth and reproductive output were not merely continued effects of speed of seedling emergence. In early May, after all seedlings had spent at least 1,000 hr in the cold room, we recorded the number of fully expanded leaves on each seedling and moved the trays outside under a 75% shade cloth near University Park, PA. One month later, each seedling was transferred into a one gallon plastic pot containing a standard potting mix and returned to the shade enclosure. Throughout the summer and early autumn the pots were watered daily (unless there was adequate rainfall) and fertilized with 1/2 strength Peters solution every two weeks. We monitored the plants daily and recorded the number of days from vernalization to the first flower for each plant. The plants were naturally pollinated, primarily by bumblebees. In the autumn we collected every mature fruit and recorded the height and number of lateral branches for each plant. In the laboratory, we counted the number of mature fruits, separated the seeds from the chaff, and determined the total seed mass of each plant. Seed mass per fruit was calculated by dividing the total seed mass of each plant by the number of fruits produced.
Data Analyses. -- We analyzed our data on progeny performance using a fixed effects analysis of covariance (PROC GLM, SAS, 1985) to detect effects due to maternal parent, paternal parent, pollination intensity, and the maternal * paternal interaction. In all analyses the initial 10 seed sample weight was used as a covariate. In addition, two multivariate ANCOVAs (PROC GLM MANCOVA, SAS, 1985) were performed to detect overall effects on progeny performance. One MANCOVA was performed for the early traits: days to emergence, days to first leaf, days to second leaf, and number of leaves at the end of vernalization, and a second MANCOVA was performed for all subsequent measures. Separate analyses were performed because (a) our vernalization protocol served as a way to reduce the variation among the seedlings prior to transferring them outside and effectively splits our performance measures into early seedling growth and adult growth and reproduction and (b) because our sample size was reduced due to a Fusarium infection after transplanting to the field site.
Seed Production of Crossing Experiment. -- For the crossing experiment as a whole, we found significant effects of the maternal parent and the maternal by paternal parent interaction on seed number per fruit, and significant maternal and paternal main effects on mean seed mass. The size of the pollen load did not affect either seed number per fruit or mean seed mass in our whole sample of fruit (Richardson and Stephenson, 1991). However, for the sample of fruits from which seeds were selected for the progeny performance tests reported here, there were significant differences between the pollination treatments for both seed number per fruit and mean seed mass. Fruits we selected from high pollen loads contained an average of 44.7 [+ or -] 2.5 seeds (N = 63; [Chi] [bar] [+ or -] SE) as compared to 35.4 [+ or -] 2.1 (N = 63; [Chi] [bar] [+ or -] SE) seeds for fruits produced by low pollen loads (t-test; P < 0.05). In addition, the mean seed sample weights (10 seeds) for high pollen load fruits was 1.5 [+ or -] 0.007 mg (least square mean [+ or -] SE) versus 1.6 [+ or -] 0.007 mg (least square mean [+ or -] SE) for low pollen load fruits (LSMEANS/ PDIFF, SAS, 1985; P < 0.0001).
Correlations between Seed Weight and Progeny Performance. -- The seed sample weight was significantly correlated with three of our nine measures of progeny performance (Table 1). For each of the three traits (days to first leaf, days to second leaf, and the number of leaves after vernalization) the sign of the correlation indicates that seedlings from larger 10 seed sample weights were more vigorous than seedlings from seeds of lower weights. That is, seedlings from larger seeds produced their first and second leaf faster and had more leaves after vernalization than seedlings from smaller seeds. [TABULAR DATA 1 OMITTED]
Analyses of Progeny Performance. -- The fixed effects ANCOVA revealed significant maternal effects on days to emergence (Table 2B), final plant height, total seed mass, and seed mass per fruit (Table 3B). In addition there was an indication of a maternal effect on days to second leaf (P = 0.08). The MANCOVA for early traits indicated no overall maternal effect on progeny performance (Wilks' Criterion, Table 2B); however, the analysis for later traits revealed a significant overall maternal effect on progeny performance (Wilks' Criterion, Table 3B). [TABULAR DATA 2 OMITTED]
Significant paternal effects were found for days to emergence, days to first leaf (Table 2B), and seed mass per fruit (Table 3B), and there was also an indication of a paternal effect on days to second leaf (P = 0.09). The MANCOVA for early traits again revealed no significant overall effect on progeny performance due to the pollen parent (Wilks' Criterion, Table 2B), while the analysis for later traits revealed a significant overall paternal effect on progeny performance (Table 3B). The specific interaction of the parents had a significant effect on two early traits: days to emergence, and days to first leaf (Table 2B), but neither of the multivariate analyses detected an overall effect of this interaction on progeny performance (Tables 2B and 3B).
The analyses of covariance also revealed a significant effect due to the size of the pollen load on days to first leaf, days to second leaf, and the number of leaves after vernalization (Table 2B), and an indication of an effect on days to first flower (P = 0.08). Tables 2A and 3A contain the least square means [+ or -] SE) for our nine measures of progeny performance for seedlings derived from high and low pollen loads. For seven of the nine traits the progeny from high pollen load fruits outperformed the progeny from low pollen load fruits. Moreover, in each case where there is a significant (P < 0.05), or marginally significant (P < 0.09) difference between the means for the two pollination treatments, progeny from high pollen loads outperformed the progeny from the low pollen load fruits (Tables 2A and 3A). The multivariate analysis of early traits revealed an overall pollen load effect (Table 2B), whereas the analysis on the later traits gave no indication of an overall pollen load effect (Table 3B). [TABULAR DATA 3 OMITTED]
Seed Weight. -- Studies of many species have found that seed size has a positive effect on germination, early growth and/or seedling establishment but that the effects of seed size on progeny performance are transient especially under growth chamber, greenhouse or experimental garden conditions (e.g., Marshall, 1986; Wulff, 1986). For example, in Oenothera biennis, seed size was positively correlated with seedling size for the first five weeks after germination but was unrelated to adult size or reproductive output (Kromer and Gross, 1987). However, under competitive conditions other studies have found that seed size is associated not only with larger seedlings but also with increased reproductive output (e.g., Stanton, 1984; Waller, 1984; Wulff, 1986). In this study of C. americana we found that mean sample weight of the seeds had a significant effect on three measures of seedling vigor but no significant effect on later measures of vegetative vigor or reproductive output under noncompetitive garden conditions. In natural populations of C. americana, however, these early traits may be important in seedling establishment where seedling densities can average 100/[m.sub.2] and seedling mortality is high (Kalisz, pers. comm.).
Parental Effects. -- From previous analyses, we know that mean seed weight in the entire set of field crosses (8 recipients, 47 donors) was significantly influenced by both the maternal parent and the paternal parent, with the maternal parent explaining a larger percentage of the variance in seed weight (Richardson and Stephenson, 1991). Given the correlation between seed weight and progeny performance in this study (Table 1), it is reasonable to conclude that the maternal and paternal parents have an indirect effect on progeny performance through their effects on seed weight. Moreover, it is reasonable to assume that the seed size mediated effects of the maternal parent on progeny performance exceed those of the paternal parent.
In the analyses of progeny performance presented here, seed weight is used as a covariate. Consequently, the maternal and paternal effects in these analyses represent the total effects, both genetic and environmental (predominantly maternal), that are independent of seed weight. Our analyses reveal that several measures of progeny performance are significantly affected by the maternal and paternal parents. Moreover, these effects include early and late measures of vegetative vigor and measures of reproductive performance.
Although most studies of maternal effects on progeny performance do not separate the effects of seed size from nonseed size effects (see Roach and Wulff, 1987), the few that do make this separation have found that the maternal effects on progeny performance extend throughout the life cycle even when the progeny are not grown under competitive conditions (e.g., Rocha and Stephenson, 1990, 1991; Schlichting et al., 1990). Consequently, we suspect that maternal effects that are independent of seed weight are more likely to reflect genetic differences among maternal parents than maternal effects that include seed size.
Finally, our earlier analyses revealed no significant interaction of the maternal and paternal parents on seed size (Richardson and Stephenson, 1991) and here we found that only two early measures of progeny performance and no overall effects were significantly influenced by the interaction of the parents. The lack of specific parental combinations whose progeny performed exceptionally poorly or exceptionally well indicate that, among the parents we selected for this study, inbreeding or outbreeding depression was not a major factor affecting progeny performance (Charlesworth and Charlesworth, 1987; Price and Waser, 1979).
Size of Pollen Load Effects. -- This study clearly reveals an effect of the size of the pollen load on the vigor of C. americana seedlings (Tables 2B, 3B). For each significant trait, the progeny from high pollen loads outperformed the progeny from low pollen loads under controlled environmental conditions (Tables 2A, 3A). Although these effects of pollen load size are independent of seed weight, it should be noted that even when seed weight is eliminated from the ANOVA model, the seedlings from high pollen loads still significantly outperform seedlings from low pollen loads for two of our measures of performance (days to first leaf, and number of leaves after vernalization). That is, even with a 7% size advantage (on average), the progeny from low pollen loads are not as vigorous as the progeny from the high pollen loads. With the exception of days to first flower however, the effects of pollen load size do not extend to later measures of vegetative vigor or reproductive output. This could be an artifact of the experimental design in which all plants were transferred into the cold room for vernalization at the same size (not the same age). Consequently, the slower growing seedlings remained in the growth chamber for a longer period of time and were exposed to the vernalization conditions for a shorter period than the faster growing seedlings. Under natural conditions, inclement weather would preserve size differences. Moreover, in several species of winter annuals, rosette size is correlated with survivorship over the winter and occasionally with later measures of vigor (e.g., Werner, 1975).
The size of the pollen load is known to affect progeny performance in both cultivated and noncultivated species (e.g., Mulcahy and Mulcahy, 1975; Mulcahy et al., 1975; Lee and Hartgerink, 1986; McKenna, 1986; Stephenson et al., 1986; Davis et al., 1987; Winsor et al., 1987; Ramstetter and Mulcahy, 1988; Bertin, 1990; but see Snow, 1990). However, the causes of these pollen load effects have been the subject of much debate (e.g., Mulcahy, 1979; Charlesworth, 1988; Stephenson et al., 1988; Snow and Mazer, 1988; Schlichting et al., 1990). Both prezygotic hypotheses, such as nonrandom fertilization on the basis of pollen genotype and/or pollen-pistil interactions, and postzygotic hypotheses, such as nongenetic maternal effects and nonrandom seed abortion, have been advanced to explain the relationship between the size of the pollen load and progeny vigor (for a detailed discussion, see Schlichting et al., 1990).
Although our data do not allow us to differentiate among these competing hypotheses, we should note several factors in our experimental design that are relevant to the hypotheses. First, we performed only single donor pollinations. Consequently, our low and high pollen loads varied only the size of the sample of haploid genomes derived from a single pollen producing plant. To the extent that pollen performance is genetically determined, it is reasonable to assume that the variation in pollen performance within a single donor is less than the variation between donors in most plant populations. Second, in the low pollen loads, approximately 77 pollen grains produced 35.4 seeds (2.2 pollen grains per seed) whereas in the high pollen loads, approximately 215 pollen grains produced 44.7 seeds (4.8 pollen grains per seed). That is, a three fold increase in pollen deposition resulted in only a 25% increase in seed production. Consequently, the potential for microgametophyte selection (nonrandom fertilization based on pollen genotype) was far greater in the high pollen loads. Moreover, because the low pollen loads produced fewer seeds than the high pollen loads, we can assume that 77 pollen grains is insufficient (for the specific plants used in this study) to produce a full complement of seeds in each fruit. Consequently, seeds from the low pollen loads were produced under conditions of little or no pollen competition. Finally, our measures of progeny performance were taken under noncompetitive growth conditions which may fail to disclose subtle differences in progeny vigor (e.g., Lee and Hartgerink, 1986).
In conclusion, these results demonstrate the importance of the maternal and paternal parents and the size of the pollen load on progeny performance throughout the life cycle of the native plant, Campanula americana. In addition to significant parental effects on individual traits, we found a significant overall effect of both parents on a suite of traits related to reproductive output, and an overall effect of the size of the pollen load on a suite of early seedling traits which are likely to be important during seedling establishment in natural populations. These results suggest that pollen load effects can certainly be important in noncultivated plants, especially when one considers the variation in pollen deposition rates in natural populations (Snow, 1986; Galen and Newport, 1988; Levin, 1990), and the high frequency of pollen carryover which results in multidonor pollen loads in zoophilous plants (e.g., Schaal, 1980; Handel, 1982; Waser and Price, 1984; Ellstrand and Marshall, 1986).
We wish to thank B. Devlin and C. Schlichting for advice and discussion throughout the project, and S. Kalisz for allowing us to work at the Cheff Center site. C. Boyer and P. Grun made useful comments on a previous version of the manuscript. The W. K. Kellogg Biological Station provided housing and logistical support during the Michigan portion of the project. Field and laboratory assistance were provided by A. Hrincevich, C. Keifman, T.-C. Lau, H.-C. Lin, A. McNall, B. Naugle, T. O'Malley, M. Quesada, O. Rocha, and M. Zaldivar. This work was supported in part by a Henry W. Popp Fellowship and a Hill-Hill award from Pennsylvania State University to T.E.R., and NSF Grant BSR-8818189 to A.G.S.
Baskin, J. M., and C. C. Baskin. 1984. The ecological life cycle of Campanula americana in northcentral Kentucky. Bull. Torrey Bot. Club 111:329-337. Bertin, R. I. 1990. Effects of pollination intensity in Campsis radicans. Am. J. Bot. 77:178-187. Charlesworth, B., and D. Charlesworth. 1987. Inbreeding and its evolutionary consequences. Annu. Rev. Ecol. Syst. 18:237-268. Charlesworth, D. 1988. A comment on the evidence for pollen competition in plants and its relationship to progeny fitness. Am. Nat. 132:298-302. Davis, L. E., A. G. Stephenson, and J. A. Winsor. 1987. Pollen competition improves performance and reproductive output of the common zucchini squash under field conditions. J. Am. Soc. Hortic. Sci. 112:711-716. Dolan, R. W. 1984. The effect of seed size and maternal source on individual size and a population of Ludwigia leptocarpa (Onagraceae). Am. J. Bot. 71:1302-1307. Ellstrand, N. C., and D. L. Marshall. 1986. Patterns of multiple paternity in populations of Raphanus sativus. Evolution 40:837-842. Falconer, D. L. 1981. Introduction to Quantitative Genetics, 2nd ed. Longman, London, UK and N.Y., USA. Fingerett, E. R. 1979. Pollen competition in a species of evening primrose, Oenothera organensis Munz. Master's Thesis. Washington St. Univ., Pullman, WA USA. Fisher, R.A. 1958. The Genetical Theory of Natural Selection, 2nd ed. Dover, UK and N.Y., USA. Galen, C., and M. E. Newport. 1988. Pollination quality, seed set and flower traits in Polemonium viscosum: Complementary effects of variation in floral scent and size. Am. J. Bot. 75:900-905. Gleason, H. A., and A. Cronquist. 1963. Manual of the Vascular Plants of Northeastern United States and Adjacent Canada. Van Nostrand, N.Y., USA. Handel, S. N. 1982. Dynamics of gene flow in an experimental population of Cucumis meto (Cucurbitaceae), Am. J. Bot. 69:1538-1546. Kromer, M., AND K. L. Gross. 1987. Seed mass, genotype, and density effects on growth and yield of Oenothera biennis L. Oecologia 73:207-212. Lee, T. D., and A. P. Hartgerink. 1986. Pollination intensity, fruit maturation pattern, and offspring quality in Cassiafasciculata (Leguminosae), pp. 417-422. In D. L. Mulcahy, G. B. Mulcahy, and E. Ottaviano (eds.), Biotechnology and Ecology of Pollen. Springer-Verlag, N.Y., USA. Levin, D. A. 1990. Size of natural microgametophyte populations in pistils of Phlox drummondii. Am. J. Bot. 77:356-363. Marshall, D. L. 1986. Effect of seed size on seedling success in three species of Sesbania (Fabaceae). Am. J. Bot. 73:457-464. McKenna, M. A. 1986. Heterostyly and microgametophytic selection: The effect of pollen competition on sporophytic vigor in two distylous species, pp. 443-448. In D. L. Mulcahy, G. B. Mulcahy, and E. Ottaviano (eds.), Biotechnology and Ecology of Pollen. Springer-Verlag, N.Y. Mulchany, D. L. 1979. The rise of the Angiosperms: A genecological factor. Science 206:20-23. Mulchany, D. L., and G. B. Mulchany. 1975. The influence of gametophytic competition on sporophyte quality in Dianthus chinensis. Theor. Appi. Genet. 46:277-280. --. 1987. The effects of pollen competition. Am. Sci. 75:44-50. Mulchany, D. L., G. B. Mulchany, and E. M. Ottaviano. 1975. Sporophytic expression of gametophytic competition in Petunia hybrida, pp. 227-232. In D. L. Mulcahy (ed.), Gamete Competition in Plants and Animals. North Holland Pub). Co., Amsterdam, The Netherlands. --. 1978. Further evidence that gametophytic selection modifies the genetic quality of the sporophyte. Soc. Bot. Fr. Actualities Bot. 1:57-60. Nakamura, R. R., and M, L. Stanton. 1989. Embryo growth and seed size in Raphanus sativus: Maternal and paternal effects in vivo and in vitro. Evolution 43:1435-1443. Ottaviano, E. M., M. Sari-Gorla, and I. Arenary. 1983. Male gametophyte competitive ability in maize: Selection and implications with regard to breeding, pp. 367-374. In D. L. Mulcahy and E. M. Ottaviano (eds.), Pollen: Biology and Implications for Plant Breeding. Elsevier, N.Y., USA. Price, M. V., and N. M. Waser. 1979. Pollen dispersal and optimal outcrossing in Delphinium nelsonii. Nature 277:294-296. Ramstetter, J., and D. L. Mulchny. 1988. Consequences of pollen competition for Aureolaria flava seedlings. Bull. Ecol. Soc. Amer. Suppl. 69:269-270. Richardson, T. E., and A. G. Stephenson. 1989. Pollen removal and pollen deposition affect the duration of the staminate and pistillate phases in Campanula rapunculoides. Am. J. Bot. 76:532-538. --. 1991. Effects of parentage, prior fruit set and pollen load on fruit and seed production in Campanula americana L. Oecologia 87:80-85. Roach, D. A., and R. D. Wulff. 1987. Maternal effects in plants. Annu. Rev. Ecol. Syst. 18:209-235. Rocha, R. J., and A. G. Stephenson. 1990. Effect of ovule position on seed production, seed weight, and progeny performance in Phaseolus coccineus L. (Leguminosae). Am. J. Bot. 77:1320-1329. --. 1991. Effects on non-random seed abortion on progeny performance in Phaseolus coccineus L. Evolution 45:1198-1208. SAS. 1985. SAS User's Guide: Statistics. SAS Institute, Cary, NC USA. Schaal, B. A. 1980. Reproductive capacity and seed size in Lupinus texensis. Am. J. Bot. 67:703-709. Schlinching, C. D., A. G. Stephenson, L. E. Small, and J. A. Winsor. 1990. Pollen loads and progeny vigor in Cucurbita pepo: The next generation. Evolution 44:1358-1372. Shetler, S. G. 1979. Pollen-collecting hairs of Campanula (Campanulaceae), 1: Historical review. Taxon 28:205-215. Snow, A. A. 1986. Evidence for and against pollen tube competition in natural plant populations, pp. 405-410. In D. L. Mulcahy, G. B. Mulcahy, and E, Ottaviano (eds.), Biotechnology and Ecology of Pollen. Springer-Verlag, N.Y., USA. --. 1990. Effects of pollen load size and number of donors on sporophyte fitness in wild radish (Raphanus raphanistrum). Am. Nat. 136:742-458. Snow, A. A., and S. J. Mazer. 1988. Gametophytic selection in Raphanus raphanistrum: A test for heritable variation in pollen competition ability. Evolution 42:1065-1075. Stanton, M. L. 1984. Seed variation in wild radish: Effect of seed size on components of seedling and adult fitness. Ecology 65:1105-1112. Stephenson, A. G., J. A. Winsor, and L. E. Davis. 1986. Effects of pollen load size on fruit maturation and sporophyte quality in zucchini, pp. 429-434. In D. L. Mulcahy, G. B. Mulcahy, and E. Ottaviano (eds.), Biotechnology and Ecology of Pollen. Springer-Verlag, N.Y., USA. Stephenson, A. G., J. A. Winsor, C. D. Schlichting, AND L. E. DAVIS. 1988. Pollen competition, nonrandom fertilization, and progeny fitness: A reply to Charlesworth. Am. Nat. 132:303-308. Waller, D. M., 1984. Differences in fitness between seedlings derived from cleistogamous and chasmogamous flowers in Impatiens capensis. Evolution 38:427-440. Waser, N. M., and M. V. PRICE. 1984. Experimental studies of pollen carryover: Effects of floral variability in Ipomopsis aggregata. Oecologia 62:262-268. Weis, I. M. 1982. The effects of propagule size in germination and seedling growth in Mirabilis hirsuta. Can. J. Bot. 60:1868-1874. Werner , P. A. 1975. Predictions of fate from rosette size in teasel (Dipsacus fullonum L.). Oecologia 20: 197-201. Winsor, J. A., L. E. Davis, and A. G. Stephenson. 1987. The relationship between pollen load and fruit maturation and the effect of the pollen load on offspring vigor in Cucurbita pepo. Am. Nat. 129: 643-656. Wulff, R. 1986. Seed size variation in Desmodium paniculatum 111. Effects on reproductive yield and competitive ability. J. Ecol. 74:115-121.
|Printer friendly Cite/link Email Feedback|
|Author:||Richardson, Thomas E.; Stephenson, Andrew G.|
|Date:||Dec 1, 1992|
|Previous Article:||Sex ratio selection and the evolution of environmental sex determination in laboratory populations of Menidia menidia.|
|Next Article:||Sexual selection and fitness variation in a population of small mouth bass, Micropterus dolomieui (Pisces: Centrarchidae).|