Do plants derived from seeds that readily germinate differ from plants derived from seeds that require forcing to germinate? A case study of the desert mustard Lesquerella fendleri.
In many environments, within-year seed dormancy delays germination until a time which favors successful germination and subsequent plant establishment and reproduction. The time that a seed germinates within this favorable period often is the single most important variable in explaining variation in subsequent plant performance (Weiner, 1988). Many studies have shown that seeds that germinate early gain a competitive advantage over relatively late-germinating seeds (e.g., Ross and Harper, 1972; Howell, 1981; Waller, 1985; Firbank and Watkinson, 1987; Miller, 1987). However, seeds that germinate too early in the season face the risk of unpredictable and often fatal environmental conditions (e.g., frosts, droughts and temperature extremes). Thus a trade-off exists between delaying germination until a stable favorable period, and initiating germination soon enough within a favorable period to gain a competitive advantage over later germinating seeds (Silvertown, 1988).
In addition to this within-year dormancy difference, seeds of many species also exhibit between-year seed dormancy, in which a fraction of the seeds within a given environment remain dormant throughout an entire season in which other conspecific seeds germinate. This type of dormancy has led to the creation of persistent soil seed banks in most of the world's major ecosystems (reviewed by Leck et al., 1989; Thompson, 1992). Numerous theoretical models have also suggested that this between-year dormancy may balance the risk of local extinction from germination in unfavorable years with the risk of missing good years by remaining dormant (e.g., Cohen, 1966; MacArthur, 1977; Venable and Lawlor, 1980; Brown and Venable, 1986; Venable and Brown, 1988).
Both within- and between-year seed dormancy are often broken by a particular combination of specific environmental cues [e.g., temperature, light and soil moisture; see reviews in Mayer and Poljakoff-Mayber (1975) and Baskin and Baskin (1989)]. Seeds of some species may have evolved the ability to "predict" favorable establishment periods by using dormancy-breaking environmental cues that are also correlated with subsequent conditions favorable for their growth and reproduction. Such seeds may thus be able to germinate in particular years and seasons within years that best match their particular ecophysiology; that is, over evolutionary time, germination and postgermination traits may have adaptively coevolved (Evans and Cabin, 1995). For example, in relatively dry years, individuals with more xerophytic traits (e.g., greater water-use efficiency) might germinate more readily than individuals with more mesophytic traits. There may thus be genetic and/or phenotypic differences between seeds that germinate in a particular season and seeds that remain dormant but viable under the same environmental conditions.
Direct investigation of this issue is difficult, because comparing the performance of seeds in the field that germinate in different seasons may be confounded by the different biotic and abiotic environments each cohort would encounter. We therefore used an experimental approach and compared the performance of plants originating from readily germinating seeds with the performance of plants obtained from dormant seeds forced to germinate by application of the dormancy-breaking plant hormone, gibberellic acid (GA), which required an extended germination period.
In this experiment, we used the desert mustard Lesquerella fendleri (Gray) S. Wats. This species is well-suited for this study because it maintains a considerable between-year seed bank, and the fitness of seeds germinating from this bank may be significantly affected by both spatial and temporal environmental variation (Cabin, 1995; Evans and Cabin, 1995). Evans and Cabin (1995) also found conditions that favor the coevolution of Lesquerella dormancy and postgermination traits. Thus it is conceivable that the Lesquerella seeds that remain dormant throughout a particular germination period may differ ecologically and evolutionarily from the seeds that do germinate.
In this study, we compare the performance of Lesquerella seeds that readily germinate ("natural" seeds) with dormant seeds that require forcing to germinate ("induced" seeds). To assess the generality of any potential differences between these groups, we used seeds collected from four different source populations, and transplanted plants from both seed groups into contrasting microenvironments in the field. Specifically, we asked (1) Are there differences in the survival and growth of natural and induced seeds? (2) Do natural and induced seeds respond differentially to the transplant microenvironments? (3) Do natural and induced seeds from the different source populations exhibit similar patterns of survival and growth?
Study system. - Lesquerella fendleri (Gray) S. Wats. (Brassicaceae) is a self-incompatible, short-lived perennial native to southwestern North America (Rollins and Shaw, 1973). Our study site is located within the Five Points region of the Sevilleta National Wildlife Refuge, 80 km S of Albuquerque, New Mexico. The Sevilleta Refuge is one of the few large areas in the Southwest protected from human disturbances; cattle grazing and human uses other than research have been prohibited for over 20 yr. Within this area, Lesquerella occurs in ecologically diverse habitats, including creosote bush (Larrea tridentata) shrublands, grasslands and open sandy washes. Our study site consists mostly of creosote bush shrubs and open, largely barren patches in the spaces between the shrubs. Lesquerella flowering at this site usually occurs between March and April, as well as following suitable precipitation in the autumn. Lesquerella seedlings may emerge in the autumn and/or spring, depending on the amount and timing of rainfall (R. Cabin, pers. observ.).
Seed sources. - To investigate potential differences between induced and naturally germinating seeds, we used seeds collected from four different sources. The first source was a bulk collection of seeds produced by hand crosses performed on a set of 30 Lesquerella plants in a greenhouse in the spring of 1992 ("greenhouse" seed source). The parent plants had been transplanted from the Sevilleta Refuge in the preceding year. The remaining three seed sources were in fruits matured on wild plants at the Sevilleta Refuge during the summer of 1992. Two bulk collections were made from plants at the Five Points study site; one from plants growing directly beneath creosote bush shrubs ("undershrub" seed source), and the other from plants growing in the creosote shrub interspaces ("intershrub" seed source). The last source ("transition" seed source) was in fruits matured 3 km E of the Five Points study site, in a transition zone between creosote bush shrublands and short desert grassland habitats.
Seed germination. - On 9 September 1994, approximately 35 seeds from each seed source were sown on the surface of each of 16 plastic pots filled with sand, for a total of 64 pots and 560 seeds per seed source. Each pot was then bottom-watered until the sand was completely saturated, removed and allowed to drain. All pots were placed into trays in a randomized block design so that each tray contained four pots from each seed source. The trays were placed in a growth chamber set for 12 h of light and 12 h of darkness, and maintained at room temperature (ca. 22 c).
Five days after sowing, a 100 ppm gibberellic acid (GA) solution was sprayed onto the soil surface of all pots in the growth chamber. This technique is highly effective in forcing viable but dormant Lesquerella seeds to germinate (Sharir and Gelmond, 1971; Evans and Cabin, 1995; Evans et al., 1996). Because previous trials showed that Lesquerella seeds require over 2 days to germinate once the seeds have begun to imbibe water, we considered all seedlings that emerged between 4 and 7 days after sowing (no seedlings emerged before 4 days after sowing) to have initiated germination before the application of GA (i.e., if a seed was dormant until it received GA, the earliest its seedling could have emerged would have been 8 days after sowing, because GA was applied after the seeds had been in the growth chamber for 5 days), and thus hereafter we refer to plants derived from this group as "natural" plants. Seeds that emerged 9 or more days after sowing were considered to be either dormant seeds that were forced to germinate by the GA and/or seeds that required an extended period of moist soil and moderate temperature seldom encountered at the Sevilleta Refuge [previous experiments under conditions similar to this study found that application of GA resulted in a 26% increase in Lesquerella germination relative to untreated control seeds; see Evans et al. (1996)]. Plants from these groups will thus hereafter be referred to as "induced" plants. Seeds that germinated 8 days after sowing were not considered as being in either group and were removed and discarded. We only used induced plants that came from seeds that germinated between 9 and 17 days after planting [ILLUSTRATION FOR FIGURE 1 OMITTED], as only 10 seedlings emerged after day 17.
Emerging seedlings were transplanted daily into individual 6.5 x 6.5 x 9 cm deep pots by carefully removing each seedling with a forceps and laying it on the soil surface of the new pot, then covering the exposed roots with sand. Each transplanted seedling was then randomly assigned to a plastic tray and placed in a greenhouse receiving ambient light (ca. 12:12 day:night) and 21 C day and 10 C night temperatures. To allow the transplanted seedlings to recover from transplant shock, and to minimize any potential head-start advantage acquired by the early germinating seeds, all seedlings were kept in the greenhouse for 24 days after the last seedling had been transplanted out of the growth chamber.
On 17 October, between 53-64 naturally germinating seeds per seed source and 29-35 induced seeds per seed source were randomly selected for transplantation to the Sevilleta field site on the following morning (because of the germination timing, there were many more naturally germinating seeds than there were induced seeds; [ILLUSTRATION FOR FIGURE 1 OMITTED]). To test for initial differences among these groups before transplanting, the number of true leaves was counted and the maximum plant diameter was recorded using a digital calipers.
Field transplanting. - Previous work at the Five Points study site has shown that Lesquerella germination, establishment and reproduction tend to be much greater beneath creosote bush canopies (especially under the N side of the shrubs) than in the adjacent intershrub areas (Evans and Cabin, 1995). This may be because the lower light intensity and soil temperature, combined with greater surface organic matter, create a wetter, more benign microenvironment in this otherwise arid ecosystem (Evans and Cabin, 1995). To create two contrasting transplant environments, we established 0.5 x 1 m plots on the N side of 10 creosote shrubs, and paired each of these plots with a 0.5 x 1 m plot located approximately 1 m away in an adjacent, open intershrub area. Within each plot, four rows of four 25 x 12.5 cm cells were established, with each cell receiving natural or induced plants from one of the four seed sources in a randomized block design. Since we had a limited number of Lesquerella seedlings to transplant and assumed that neither natural nor induced plants could survive in the intershrub plots for more than a few weeks (Evans and Cabin, 1995), we transplanted more plants into the undershrub plots so that overwinter survival of the two germination treatments could be compared. Each cell within an undershrub plot containing natural plants received three Lesquerella plants, and each cell containing induced plants received two plants. The remaining natural and induced plants were divided equally among the intershrub plots, with half the cells receiving 2-3 natural plants, and half the cells receiving 1-2 induced plants.
Plots were watered until the soil was completely saturated every 2-3 days for the 1st 2 wk after transplanting, and once or twice per week for the following 3 wk. Each Lesquerella plant was scored as alive or dead at weekly intervals for 5 wk after transplanting. The maximum height and diameter, and the number of leaves of each surviving plant was recorded 5 wk after transplanting. A period of cold weather and frosts after this time precluded any additional data collection, as even small Lesquerella plants that die back following autumn frosts can successfully re-establish in the following spring (R. Cabin, pers. observ.). A final census of the Lesquerella transplants was performed on 10 March 1995, 5 mo after transplanting. Each remaining plant was scored as alive or dead; however, significant overwinter mortality [ILLUSTRATION FOR FIGURE 3 OMITTED] precluded meaningful statistical analysis of the growth and morphology of the relatively few plants surviving to this date.
Data analysis. - To test for initial differences among the Lesquerella plants before transplantation in the field, we used three separate ANOVAs. First, a two-way ANOVA was employed using germination treatment (natural vs. induced plants) and seed source as main effects, and plant diameter and leaf number as dependent variables. Two additional, separate ANOVAs were then used to test for differences among the four seed sources within the natural and induced plant groups. In these and all subsequent ANOVAs, we used the GLM procedure of SAS, considered all effects to be fixed, and used Type III Sums of Squares.
To compare the survivorship curves of the natural and induced plants in the field, we used Peto and Peto's (1972) logrank test as discussed in Peto and Pyke (1973) and Pyke and Thompson (1986). To compare the proportion of natural and induced plants surviving to 5 mo after transplantation, we used a standard chi-square contingency table analysis (Pyke and Thompson, 1986).
To analyze the field transplant data, we considered each cell within a plot to be the experimental unit (i.e., the set of 1-3 transplanted seedlings), and calculated average values per cell for each dependent variable. Although statistical results were very similar when data from each individual within a cell were used, using the cell averages produced more conservative results and allowed us to treat the number of leaves as a continuous variable. Seed source, germination treatment and transplant habitat (undershrub vs. intershrub plot) were used as independent variables, and height, diameter and number of leaves were used as dependent variables. Tukey's Studentized Range Test was used to test for pairwise differences among groups defined by the independent variables.
Growth chamber and greenhouse. - The fraction of seeds emerging within the growth chamber ranged from 66 to 85% among the four seed sources [ILLUSTRATION FOR FIGURE 1 OMITTED]. Of the seeds that germinated, 63% emerged within 7 days after planting, and thus were considered "natural" seeds. Twenty-two percent of all germinating seeds emerged between 9 and 17 days after planting, and were classified as "induced" seeds.
There were highly significant differences in plant diameter and number of leaves of natural and induced plants on the day before transplanting from the greenhouse to the field (Table 1). The diameter of induced plants was significantly greater than natural plants (mean diam = 9.80 and 8.83 mm, respectively), whereas induced plants had significantly fewer leaves than natural plants (mean leaf number = 3.13 and 3.69, respectively). There [TABULAR DATA FOR TABLE 1 OMITTED] were also significant differences in these variables for seed source, and significant seed source by treatment interactions. Interestingly, in both of these variables, the induced plants from the intershrub seed source showed an atypical response [ILLUSTRATION FOR FIGURE 2A AND B OMITTED] that was primarily responsible for the significant seed source by treatment interaction (Table 1). Because of these significant interactions, we also analyzed the natural and induced plants separately. Among the natural plants, there were no differences among the four seed sources in diameter (P [greater than] 0.26) or number of leaves (P [greater than] 0.20), but among the induced plants, there was a significant effect for both these variables (P [less than] 0.001 and 0.03, respectively; see Figs. 2a and b).
Field transplants. - There were highly significant differences in the survivorship curves of the natural and induced plants transplanted into the field [Peto and Peto's (1972) logrank test: [X.sup.2] = 17.0, df = 1, P [less than] 0.001], primarily due to the higher mortality of induced plants 1 wk after transplanting 25 October and at the end of the experiment in the following spring [ILLUSTRATION FOR FIGURE 3 OMITTED]. There were also highly significant differences in the proportion of natural and induced plants surviving to the end of the study ([X.sup.2] = 18.0, df = 1, P [less than] 0.001), with more than twice the proportion of natural plants (43.3%) as induced plants (21.3%) still alive [ILLUSTRATION FOR FIGURE 3 OMITTED]. Surprisingly, however, the final survival of Lesquerella plants transplanted into the undershrub and interspace plots was identical (32.3% survival in each). There were also no significant differences in the proportion of plants from the four different seed populations surviving to the end of the study ([X.sup.2] = 5.6, df = 3, P [greater than] 0.2).
At 5 wk after transplanting there were no significant differences between natural and induced plants in height and diameter, although natural plants had marginally more leaves (Table 2; mean leaf number for natural and induced plants = 4.61 and 4.07, respectively). However, there was a highly significant effect of seed source on the height of plants, with the greenhouse and undershrub seed sources producing plants significantly taller than plants originating from the intershrub seeds (mean height = 6.07, 6.07, and 4.35 mm, respectively, P [less than] 0.05 for each comparison, Tukey's Studentized Range Test). Lesquerella plants transplanted in the undershrub plots were also significantly taller than the transplants in the intershrub plots (mean height = 5.89 and 5.11 mm, respectively).
Beneath the shrubs, the transplanted natural Lesquerella plants had larger diameters than the induced plants (mean diam [+ or -] SE were 7.0 [+ or -] 0.42 and 6.5 [+ or -] 0.48 cm for the natural and induced plants, respectively), while the converse was true in the shrub interspace plots (mean diam [+ or -] SE were 5.5 [+ or -] 0.45 and 7.4 [+ or -] 0.67 cm for the natural and induced plants, respectively; see Table 2 for statistics). There was also a significant seed source by transplant habitat interaction for leaf number (Table 2; [ILLUSTRATION FOR FIGURE 4 OMITTED]). Interestingly, the plants grown from the undershrub seed source produced over 50% more leaves in the undershrub plots than in the intershrub plots, whereas seeds from the other three sources produced as many or more leaves in the intershrub plots as from beneath the shrubs. Finally, there was a significant three-way interaction of transplant habitat by seed source by germination treatment for leaf number, and this same interaction was marginally significant for plant diameter (Table 2).
Differences between induced and naturally germinating seeds. - Our results demonstrate that Lesquerella plants derived from naturally germinating seeds differ from plants originating from seeds induced to germinate. Before transplanting to the field, there were initial differences in the diameter and number of leaves produced by natural and induced seedlings in the greenhouse. In the field, more than twice as many natural plants survived to 5 mo after transplanting.
Why might plants produced by seeds that readily germinate differ from plants derived from dormant seeds that were forced to germinate? We can think of four plausible, nonmutually exclusive explanations for these differences: (1) A "side effect" of gibberellic acid; (2) a "head-start" advantage of the naturally germinating seeds; (3) initial seed mass differences between induced and naturally germinating seeds; and/or (4) genetic differences between induced and naturally germinating seeds. Although we cannot entirely discount the first three explanations, we argue below that the results of the present study, combined with evidence from previous studies, suggest that the last explanation may be at least partially [TABULAR DATA FOR TABLE 2 OMITTED] responsible for the observed differences between induced and naturally germinating seeds.
In other experiments (Evans et al., 1996), we found that Lesquerella seeds that receive GA develop into taller seedlings with fewer but longer leaves than seeds that do not receive GA. However, it is unlikely that this phenotypic effect of GA led to size differences between the natural and induced plants in this experiment for two reasons: (1) Most of the plants derived from naturally germinating seeds actually received GA, since GA was sprayed on all of the pots in the growth chamber 5 days after sowing (both ungerminated, dormant seeds, and newly emerging "natural" seedlings that had not yet been transplanted out of the growth chamber and into the greenhouse), and most of the naturally emerging seedlings were still in the growth chamber at this time [ILLUSTRATION FOR FIGURE 1 OMITTED]; and (2) there were no significant differences in the height, number of leaves, and diameters of natural and induced plants in the field. This result also suggests that although the natural plants by definition germinated earlier and thus had an additional 1-2 wk to grow in the greenhouse, this initial "head-start" did not result in any significant size differences between natural and induced plants 5 wk after transplantation to the field.
The third explanation for the observed differences between natural and induced plants is that seeds from these two groups differed in mass and/or size. Numerous studies have demonstrated intraspecific differences in germination behavior that are correlated with seed size, with relatively large, heavy seeds usually germinating sooner than relatively small, lighter seeds (e.g., Silvertown, 1984; Waller, 1985; Winn, 1985; Kalisz, 1989; Zammitt and Zedler, 1990; but see Rabinowitz, 1978; Hendrix, 1984). However, it seems unlikely that seed size differences were largely responsible for the differences found in this study. In previous germination trials, we found that relatively large Lesquerella seeds (mean [+ or -] SE = 0.79 [+ or -] 0.01 mg) germinated only 0.6 days sooner than relatively small seeds (mean [+ or -] SE = 0.32 [+ or -] 0.01 mg), a difference that was not statistically significant (n = 107 seeds, P [greater than] 0.14). In addition, the two seed sources with the heaviest seeds in the present study (greenhouse and undershrub seeds; mean mass of five batches of 50 seeds = 0.029 and 0.028 g, respectively) actually germinated more slowly than the lighter seeds (intershrub and transition seeds; means = 0.026 and 0.025 g, respectively; see [ILLUSTRATION FOR FIGURE 1 OMITTED]). However, it is plausible that other maternal effects besides seed mass [e.g., position of seeds within fruits or fruits within inflorescences; see Roach and Wulff (1987) for review] may have contributed to the variation in Lesquerella germination speeds observed in this study.
Although we did not investigate whether there were genetic differences between induced and naturally germinating seeds in the present study, other studies have detected genetic differences between Lesquerella seeds that germinate at different times and under different environmental conditions. For example, there were significant differences in allozyme frequencies between Lesquerella seeds that germinated in the field and other viable seeds remaining dormant in the soil (Evans and Cabin, 1995; Cabin, 1996). Previous experiments also found significant allozyme differences between Lesquerella plants derived from early vs. late-germinating seeds and between seeds germinating under relatively wet vs. dry soil water conditions. These kinds of genetic differences between seeds with apparently different germination speeds and environmental requirements may reflect the evolution of adaptive syndromes of germination and postgermination traits (Evans and Cabin, 1995). For instance, the seeds that did not readily germinate in the present study may have had genotypes that were adapted for germination and growth in a particular environment that differs from the optimal environment of the seeds that readily germinated. While this idea remains to be tested directly, the differential responses of the induced and naturally germinating seeds to the two contrasting transplant habitats used in this study are consistent with this hypothesis.
Natural and induced plant responses to transplant habitat. - While the natural plants had much larger diameters beneath the shrubs than in the interspaces between the shrubs, induced plants had the largest diameters in the shrub interspaces. It is not clear why these dormant seeds were able to grow more effectively in the relatively stressful intershrub environment. One possibility is that induced plants were less plastic than natural plants, and thus were either well-adapted to the interspace microenvironment, or were poorly adapted and simply died. There is some support for this hypothesis from the different (but not statistically significant due to low sample sizes) survival patterns of natural and induced plants in the two transplant habitats. Survival of natural plants 5 wk after transplanting was nearly identical beneath the shrubs and in the interspaces (mean survival = 56.7 and 55.8%, respectively), and the overwinter survival of these remaining plants was nearly the same (76.0 and 77.6%, respectively). In contrast, survival of the induced plants 5 wk after transplanting was 68% greater beneath the shrubs (means = 50.0 and 33.8%), but overwinter survival of these remaining plants was greater in the shrub interspaces (42.6 and 63.0%, respectively). Thus it is plausible that the induced plants that survived in the shrub interspaces contain genotypes that adapt them to this high stress desert environment.
Effect of seed source on natural and induced plants. - The initial size and number of leaves of natural and induced seedlings in the greenhouse were significantly affected by the seed source. For both of these variables, there was considerably more variation among the induced plants than among the natural plants, primarily because of the performance of seeds matured in the shrub interspaces. In the field, the only direct effect of seed source was that the plants derived from the creosote and greenhouse seeds grew taller than the plants produced by the intershrub seeds. However, there were several significant or marginally significant two-way and three-way interactions among the seed sources, germination treatments and transplant habitats. Thus the relative performance of induced and naturally germinating plants may be determined by where the seeds come from and what kind of habitat they are placed into.
Many studies have found intraspecific population variation in germination requirements and subsequent performance (e.g., Cavers and Harper, 1966; Baskin and Baskin, 1973; Winn, 1985; Kalisz, 1986, 1991). In some cases, these differing population germination requirements may reflect local adaptation to particular environments and selection regimes (reviewed by Quinn and Colosi, 1977). In general, we found inconsistent evidence of microsite adaptation among the seed populations used in this study. For example, when transplanted beneath the creosote bush shrubs, plants derived from seeds matured under these conditions did not show greater relative survival than plants derived from the other seed sources. However, plants produced from the undershurb seeds did produce more leaves when beneath the shrubs, whereas plants from the other three seed sources all produced as many or more leaves when transplanted into the shrub interspaces.
Implications of this study. - We have argued that the phenotypic differences between the Lesquerella plants produced by induced and naturally germinating seeds may be at least partially due to genetic differences between these two groups. The existence of such differences may indicate that particular seed genotypes differentially germinate or persist in the soil under specific environmental conditions. This result would have many important consequences for our understanding of the demographic and genetic structure of natural plant populations. For example, seed banks could function as a type of sieve, screening when and under what circumstances particular seed genotypes germinate from the soil (Cabin, 1996). Differential seed genotype success could also affect traits expressed during later life-history stages (Evans and Cabin, 1995), and may help explain the creation and maintenance of nonrandom patterns of genetic variation found in many plant populations (reviewed by Nevo, 1988; Hamrick, 1987, 1989; Mitton, 1989). Unfortunately, these ideas are still largely speculative, as very few empirical studies address these issues. We therefore encourage other researchers studying different systems to explore these kinds of questions.
Acknowledgments. - We thank Joy Avritt for considerable help in the greenhouse, and Diane Marshall and John Heywood for helpful comments on an earlier version of the manuscript. This research was supported by National Science Foundation Grant DEB-9318433 to ASE.
BASKIN, J. M. AND C. C. BASKIN. 1973. Plant population differences in dormancy and germination characteristics of seeds: Heredity or environment? Am. Midl. Nat., 90: 493-498.
----- AND -----. 1989. Physiology of dormancy and germination in relation to seed bank ecology, p. 53-66. In: M. A. Leck, V. T. Parker and R. L. Simpson. (eds.). Ecology of soil seed banks. Academic Press, Inc. San Diego, Calif.
BROWN, J. S. AND I. L. VENABLE. 1986. Evolutionary ecology of seed bank annuals in temporally varying environments. Am. Nat., 127: 31-47.
CABIN, R.J. 1995. An examination of the ecological and evolutionary relationship between the seed bank and the surface plant population of the desert mustard, Lesquerella fendleri. Ph.D. Diss., p. 223, University of New Mexico, Albuquerque.
-----. 1996. Genetic comparisons of seed bank and seedling populations of the desert mustard, Lesquerella fendleri. Evolution. 50: 1830-1841.
CAVERS, P. B. AND J. L. HARPER. 1966. Germination polymorphism in Rumex crispus and Rumex obtusifolius. J. Ecol., 54: 367-382.
COHEN, D. 1966. Optimizing reproduction in a randomly varying environment. J. Theor. Biol., 12:119-129.
EVANS, A. S. AND R. J. CABIN. 1995. Can dormancy affect the evolution of post-germination traits? An assessment of the desert mustard Lesquerella fendleri. Ecology, 76: 344-356.
-----, R.J. MITCHELL, AND R.J. CABIN. 1990. Morphological side effects of using gibberellic acid to induce germination: consequences for the study of seed dormancy. Am. J. Bot. 83: 543-549.
FIRBANK, L. G. AND A. R. WATKINSON. 1987. On the analysis of competition at the level of the individual. Oecologia, 71: 308-317.
HAMRICK, J. L. 1987. Gene flow and distribution of genetic variation in plant populations, p. 53-68. In: K. M. Urbanska (ed.). Differentiation patterns in higher plants. Academic Press, London.
-----. 1989. Isozymes and the analysis of genetic structure in plant populations, p. 87-105. In: D. Soltis and P. Soltis (eds.). Isozymes in plant biology. Dioscorides Press, Portland, Oregon.
HENDRIX, S. D. 1984. Variation in seed weight and its effects on germination in Pasticaca sativa L. (Umbelliferae). Am. J. Bot., 71: 795-802.
HOWELL, N. 1981. The effect of seed size and relative emergence time on fitness in a natural population of Impatiens capensis Meerb. (Balsaminaceae). Am. Midl. Nat., 105: 312-320.
KALISZ, S. 1986. Variable selection on the timing of germination in Collinsia verna (Scrophulariaceae). Evolution, 40: 479-491.
----- 1989. Fitness consequences of mating system, seed weight and emergence date in a winter annual, Collinsia verna. Evolution, 43: 1263-1272.
----- 1991. Experimental determination of seed bank age structure in the winter annual Collinsia verna. Ecology, 72: 575-585.
LECK, M. A., V. T. PARKER AND R. L. SIMPSON (EDS.). 1989. Ecology of soil seed banks. Academic Press, Inc. San Diego, Calif. 462 p.
MACARTHUR, R. H. 1977. Geographical ecology: patterns in the distribution of species. Harper & Row, New York. 269 p.
MAYER, A.M. AND A. POLJAKOFF-MAYBER. 1975. The germination of seeds, 3rd ed. Pergamon, London, U.K. 236 p.
MILLER, T. E. 1987. Effects of emergence time on survival and growth in an early old-field plant community. Oecologia, 72: 272-278.
MITTON, J. B. 1989. Physiological and demographic variation associated with allozyme variation, p. 127-145. In: D. Soltis and P. Soltis (eds.). Isozymes in plant biology. Dioscorides Press, Portland, Oregon.
NEVO, E. A. 1988. Genetic diversity in nature: patterns and theory. Evol. Biol., 23: 217-246.
PETO, R. AND J. PETO. 1972. Asymptotically efficient rank invariant procedures. J. R. Statist. Soc. Ser. A, 135: 185-207.
----- AND M. C. PIKE. 1973. Conservation of the approximation [Sigma][(C - E).sup.2]/E in the logrank test for survival data or tumor incidence data. Biometrics, 29: 579-584.
PYKE, D. A. AND J. N. THOMPSON. 1986. Statistical analysis of survival and removal rate experiments. Ecology, 67: 240-245.
QUINN, J. A. AND J. C. COLOSI. 1977. Separating genotype from environment in germination ecology studies. Am Midl. Nat., 97: 484-489.
RABINOWITZ, D. 1978. Abundance and diaspore weight in rare and common prairie grasses. Oecologia, 37: 213-219.
ROACH, D. A. AND R. I. WULFF. 1987. Maternal effects in plants. Annu. Rev. Ecol. Syst., 18: 209-235.
ROLLINS, R. C. AND E. A. SHAW. 1973. The genus Lesquerella (Cruciferae) in North America. Harvard University Press, Cambridge. Mass. 300 p.
ROSS, M. A. AND I. L. HARPER. 1972. Occupation of biological space during seedling establishment. J. Ecol., 60: 77-88.
SHARIR, A. AND H. GELMOND. 1971. Germination studies of Lesquerella fendleri and Lesquerella gordonii, with reference to their cultivation. Econ. Bot., 25: 55-59.
SILVERTOWN, J. 1988. The demographic and evolutionary consequences of seed dormancy, p. 205-219. In: A. J. Davy, M.J. Hutchings and A. R. Watkinson (eds.). Plant population ecology. Blackwell, Oxford.
-----. 1984. Phenotypic variety in seed germination behavior: The ontogeny and evolution of somatic polymorphism in seeds. Am. Nat., 124: 1-16.
THOMPSON, K. 1992. The functional ecology of seed bands, p. 231-258. In: M. Fenner, (ed.). Seeds: the ecology of regeneration in plant communities. C.A.B International, U.K.
VENABLE, D. L. AND J. S. BROWN. 1988. The selective interactions of dispersal, dormancy, and seed size as adaptations for reducing risk in variable environments. Am. Nat., 131: 360-384.
----- AND L. LAWLOR. 1980. Delayed germination and dispersal in desert annuals: Escape in space and time. Oecologia, 46: 272-282.
WALLER, D. M. 1985. The genesis of size hierarchies in seedling populations of Impatiens capensis Meerb. New Phytol., 100: 243-260.
WEINER, J. 1988. Variation in the performance of individuals in plant populations, p. 59-81. In: A.J. Davy, M.J. Hutchings and A. R. Watkinson (eds.). Plant population ecology. Blackwell, Oxford.
WINN, A. A. 1985. Effects of seed size and microsite on seedling emergence of Prunella vulgaris in four habitats. J. Ecol., 73: 831-840.
ZAMMITT, C. AND P. H. ZEDLER. 1990. Seed yield, seed size and germination behaviour in the annual Pogogyne abramsii. Oecologia, 84: 24-28.
|Printer friendly Cite/link Email Feedback|
|Author:||Cabin, Robert J.; Evans, Ann S.; Mitchell, Randall J.|
|Publication:||The American Midland Naturalist|
|Date:||Jul 1, 1997|
|Previous Article:||Demography and life history characteristics of the rare Kachina daisy (Erigeron kachinensis, Asteraceae).|
|Next Article:||Polymorphic floral traits in Linaria canadensis (Scrophulariaceae).|