| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Natural History, Norwegian University of Science and Technology, N-7491 Trondheim, Norway; and 2 Department of Botany, Faculty of Chemistry and Biology, Norwegian University of Science and Technology, N-7491 Trondheim, Norway
Received for publication November 19, 1998. Accepted for publication April 13, 1999.
ABSTRACT
To study genetic adaptations in bryophytes on small ecological and spatial scales and to assess the adaptive significance of morphological trait variation, genotypes of Sphagnum angustifolium originating from habitats characterized by different pH and height above water table were clonally propagated and grown along the same gradients that exist in the field. Clones from ombrotrophic habitats grew consistently better ombrotrophically than clones from minerotrophic habitats and vice versa, suggesting that the genotypes were adapted to different pH levels. Genetic variation was found in several morphological traits, but habitat-specific genetic effects were detected only in length of spreading branches. Covariation between morphology and growth was generally environmentally induced. Positive and negative cross-environment genetic correlations suggested the presence of constraints on adaptive reaction norm evolution. The indications of small-scale genetic adaptations suggest either selective establishment of genotypes adapted to specific habitats, strong selective forces operating at the later stages of the life cycle, restricted gene flow over short distances, or a combination of these. In contrast to prevailing views, these results indicate that bryophytes are likely to respond genetically to small-scale environmental gradients.
Key Words: adaptation • bryophyte • multiple niche • phenotypic plasticity • reaction norm • Sphagnum.
Over the last 15 yr isozyme data have called into question the notion that mosses are genetically depauperate, as suggested by Crum (1972)
. Some bryophyte populations display levels of genetic variation equal to those found in highly outcrossed diploids (e.g., Wyatt, Odrzykoski, and Stoneburner, 1989
; Wyatt, Stoneburner, and Odrzykoski, 1989
). Thus, lack of sheltering of recessive alleles in the dominant haploid phase of the life cycle and the assumed high levels of asexual reproduction and selfing do not necessarily reduce the level of genetic variation in bryophytes. The degree to which these patterns reflect that populations consist of genotypes adaptated to different local niches vs. the selective neutrality of isozyme alleles has, however, been disputed (Cummins and Wyatt, 1981
; Yamazaki, 1981, 1984
; Wyatt, Odrzykoski, and Stoneburner, 1989
).
Few studies have demonstrated adaptive genetic variation within or between bryophyte populations. The broad environmental tolerances over extremely different habitats found in studies of metal ion tolerance (e.g., Shaw, Beer and Lutz, 1989
) and temperature optima (e.g., Longton, 1981
; Kallio and Saarnio, 1986
) do not support the hypothesis that genetic polymorphisms are maintained by selective responses to subtle microenvironmental variation (Shaw, 1991
). Instead, it is believed that most mosses have broad physiological optima and that genetically specialized ecotypes are uncommon (Shaw, 1991, 1992
). Yet, in some instances adaptive responses to metal ion concentrations have been found in populations from metal-contaminated sites (e.g., Shaw, 1988, 1990
; Jules and Shaw, 1994
). Experimental studies addressing genetic adaptations within populations are lacking in bryophytes.
High levels of electrophoretic variability have been demonstrated in the unisexual Sphagnum angustifolium (Russ.) C. Jens. (Stenøien and Såstad, 1999
) The distribution of this species spans gradients from hummocks to lawns, intermediate fens to ombrotrophic bogs, and forested wetlands to open mires. It displays considerable clinal variation in morphology along these gradients (Såstad and Flatberg, 1994
), which can cause problems in taxonomic delimitation from other closely related species. In typical mire ecosystems the different habitats often occur adjacent to each other in mosaics that cover large continuous areas. These mosaics constitute a suitable model ecosystem for studying the relative importance of genetic specialization and phenotypic plasticity across continuous gradients on a small spatial scale.
Adaptive significance of morphological variation has been suggested by field studies of mosses sampled along ecological gradients (e.g. Såstad and Flatberg, 1993
; Heegaard, 1997). Such studies may help us reveal general patterns of clinal variation in morphology and generate hypotheses regarding the ecological relevance of this variation. Such studies cannot, however, reveal the relative importance of genetic specialization and phenotypic plasticity, or genetic constraints on reaction norm evolution. To do this, experiments involving replicated genotypes exposed to a series of environments are required.
Here we: (1) compare genetic and environmental effects on growth and morphology of genotypes along two important ecological gradients in mires; (2) look for evidence of selection in heterogeneous environments in an electrophoretically variable species by including genotypes sampled from different habitats within the population; and (3) explore the adaptive significance of morphological trait variation between habitats.
MATERIALS AND METHODS
Experimental site
The experiment was carried out in 1997 in a vegetationally heterogeneous area that covered 40 x 80 m, with abundant S. angustifolium along the pH and water-level gradients. These gradients were defined in 1996 using floristic criteria, corroborated with measurements of ground-water pH and water-level measurements (K. Digre, Norwegian University of Science and Technology, unpublished data). As a basis for selection of experimental units, habitats were classified according to pH/nutrient richness (ombrotrophic-minerotrophic) and height above water table (hummock-lawn).
Experimental material
Gametophores were collected in the autumn of 1996 and clonally propagated by regeneration from fragments following Såstad, Bakken, and Pedersen (1998)
, assuring that all gametophores propagated from one sample were genetically identical. Plots selected for gametophore sampling were assigned the levels "high" or "low" for the two factors pH and height above water table, in a 2 x 2 factorial design. Three replicate plots (50 x 50 cm) were selected within each factor level combination. One gametophore was selected within each of three subplots (6.25 x 6.25 cm) within each plot in a nested design, reducing the chance of these three gametophores being from the same clone. Gametophores were kept in agar culture for ~5 mo and then moved to liquid culture for 1 mo before they were used in the field experiment. The clonally propagated gametophores had a normal gross morphology, including that of branch fascicles, but they were more slender and fragile and distinctly greener than typical gametophores encountered in the field.
Design of field experiment
The outcome of the clonal regeneration process influenced the design chosen for back-transplantation of gametophores to the field. About half of the sampled genotypes produced sufficient material for further experimentation. Of these, about half produced many gametophores, which allowed two experiments with slightly different goals to be performed.
Experiment 1 was performed as a 2 x 2 factorial design with the same factors and factor levels used for gametophore sampling. Clonally propagated material from two genotypes from each of two plots within each of the factor-level combinations was used (except for one of the plots in which material from only one gametophore could be obtained). This maintained the hierarchical sampling scheme of genotypes. One gametophore of each genotype was mounted together in an experimental unit (see below). Eight such units were used in the experiment (two replicates in each environment), giving a total of 120 gametophores. This experiment covered the extremes of the pH and water-level gradients and sought to reveal possible adaptive responses in the habitat-specific genotypes.
Experiment 2 contained the eight genotypes that produced the most new gametophores. They represented only a subset of habitats. These gametophores were replicated twice in 12 plots evenly distributed along the same gradients, giving a total of 192 gametophores. This provided an opportunity to study the effects of these gradients in more detail and to detect nonlinear responses.
The spatial distribution of experimental plots and plots used for collecting gametophores as well as their positions along the pH and water-level gradients is shown in Fig. 1. Mean water-level and pH in the experimental plots are shown in Fig. 2 for the year of the experiment and the year before. Because extremely low water-levels precluded measurements in the driest period in 1997 (i.e., the year of the experiment), that year appear to have the smallest mean distances to the water table. However, only the measurements from 1997 were used in the statistical analyses.
|
|
Character measurements
The experiments were initiated in mid-June and harvested in mid-October 1997. The same measurements were made in both experiments. Fresh masses of individual gametophores were measured before the experiment started. Before weighing, gametophores were kept in a water-saturated atmosphere at room temperature and were given a light pressure between two filter papers to allow external capillary water to be absorbed. In the same way, the gametophores were reweighed at the time of harvest. Also dry masses were measured at the end of the experiment, showing a strong linear relationship with fresh masses (r2 = 0.92) and implying that the method used for weighing gametophores adequately reflects the dry mass of the gametophores. Mean relative growth rate during the experimental period was calculated as
 |
Number of branch fascicles (with and without pendant branches) were counted using a section of stem located 1 cm below the capitulum, and the length of the longest spreading branch on any fascicle in this section was recorded. Breadth, length, and area of five stem leaves and five branch leaves (from one spreading branch) were measured. Microscopy of leaves was carried out following Såstad, Bakken, and Pedersen (1998)
.
Data analyses
For experiment 1, analyses of variance (ANOVA) of growth rate and morphology were carried out according to the following model:
klmn is the clonal-level genetic effect (nested within plots). Ppik, Wwjl, Pwil, and Wpjk are interactions between the environmental and genetic main effects, and 
ijklmno is the residual error. All effects were treated as fixed except plot and clone, which were considered random.
Analyses of covariance (ANCOVA) were used for data from experiment 2 according to the following model:
k is the effect of the kth clone (treated as a fixed effect), and ß1pik and ß2wjk is the genotype interaction with pH and water level, respectively.
Homoscedasticity was tested using the Levene test of homogeneity. Length of spreading branches, stem leaf area, breadth, and length, and branch leaf area and breadth were ln-transformed to meet with this assumption. ANOVA and ANCOVA were carried out using the general linear modeling option in SPSS 7.5 for Windows (SPSS, 1997
).
Cross-environment genetic correlations
To assess possible genetic constraints on evolution of reaction norms in the measured traits, cross-environment genetic correlations (Via, 1994
) between all pairs of habitats (in experiment 1) were calculated. The genetic variance–covariance matrix was estimated as the difference between the phenotypic and environmental variance–covariance matrices. These were obtained from a multivariate ANOVA entering the trait values in the four different habitats as separate variables and specifying no effects and genetic effects in the model, respectively.
Correlations between growth rate and morphological traits
Phenotypic and environmental correlations were calculated within the four habitats in experiment 1. The environmental variance–covariance matrix was obtained from a multivariate ANOVA, specifying only genetic effects in the model. The genetic variance–covariance matrix was estimated as the difference between the phenotypic and environmental variance–covariance matrices (cf. Gebhardt and Stearns, 1993
). Regressions of genotypic means of growth rate (as a measure of fitness) on the means and plasticities in the other measured traits were performed to assess their adaptive significance (cf. Lacey et al., 1983
; Schlichting, 1986
; Scheiner, 1993
). Plasticities were calculated as the variance of the genotypic means of the pH and water-level gradients, respectively, for separate evaluation of each gradient.
RESULTS
Of the experimental gametophores, 11.9% could not be recovered at the end of the experiment. One unit (eight gametophores) in experiment 2 was lost completely, probably due to moose urination. Occasional shoots also had either lost the capitulum or been completely detached from the grid, most likely during harvest.
During the experiment the plants experienced the driest and warmest July and August in the area encountered in over 150 yr; the situation normalized in the autumn. The dry and warm summer seemed to hamper the growth of the gametophores, and some clones encountered negative growth rates in some environments. The gametophores retained the long hemi-isophyllous leaves and poorly developed pendant branches characteristic of agar-cultivated material through the end of the experiment.
Growth rates
A significant interaction between pH and water level was observed for growth rate in experiment 1 (Table 1). The highest growth rates were generally found in the wet habitat with high pH and in the dry habitat with low pH. Clones that originated from habitats with high and low pH showed opposite responses according to the pH gradient (Fig. 3). Clones from the low pH habitats grew better at low pH than clones from high pH habitats, and vice versa. The difference was particularly pronounced in high-pH habitats.
|
|
|
|
|
|
|
|
|
|
The relationships between growth rate and morphology were weak across habitats (Table 5). Moreover, the relatively low number of genotypes yielded low statistical power in the analyses. Plasticities of stem leaf size and branch fascicle density affected growth rate negatively along the pH gradient, whereas a positive response was found along the water-level gradient. Examination of the impact of plasticity of spreading branch lengths on clones originating from wet and dry habitat separately yielded distinctly different relationships between plasticities of this trait and growth rate ("dry" genotypes ß = 0.69; "wet" genotypes ß = -0.64).
|
Genetic specialization and reaction norm evolution
Restricted gene flow can generally promote local adaptation and genetic divergence between different microhabitats (cf. Via and Lande, 1985
). The magnitude of gene flow between the habitats in the current study is difficult to assess. Gamete dispersal distances are considered highly restricted in bryophytes (Wyatt and Anderson, 1984
). Mean fertilization distance in another unisexual Sphagnum (S. subtile) has been estimated as 2.2 cm (McQueen, 1985
) from the spatial distribution of male and female gametophores within a single colony. This makes recombination between individuals from the same habitat patch much more likely than recombination between habitat patches. Fertilization and formation of sporophytes do not seem to be abundant at the site (sporulating individuals were found in seven of 32 patches analyzed in 1996), but this may vary considerably among years. Spore dispersal distances are probably orders of magnitude higher than gamete dispersal distances. Maximum spore dispersal distance was estimated as 0.75 m in Sphagnum subtile (McQueen, 1985
), but it is probably higher as some spores are likely to become airborne and not detected by the methods used in McQueen's study. Gametophytic fragmentation is an additional potentially effective dispersal mechanism, and there may be no strong limits to the dispersal of propagules between the habitats in the current study. Whether this dispersal actually effects gene flow between habitats depends on the degree to which new gametophores are established from spores or vegetative fragments. Studies of clonal structure have revealed many different genotypes intermixed at small spatial scales both in S. capillifolium (Cronberg, 1996
) and in S. angustifolium (Stenøien and Såstad, 1999
), suggesting active recruitment to the populations. Clones have also been shown to cover large areas, however, suggesting vegetative growth of clones for many years (Cronberg, 1996
). Given active recruitment to the population, the evolution of ecologically specialized genotypes will depend on the strength of the selective forces, and whether selective establishment of genotypes adapted to specific habitats reduces the effective gene flow between habitats (van Tienderen, 1992
). Given large differences in the frequency of different habitats, specialization may occur to the most frequent habitat type even when gene flow is high, because the performance of a genotype in less frequent habitats is less important (van Tienderen, 1992
). In the case of negative genetic correlations among traits, the evolutionary dynamics may be dominated by the common habitat, which can actually cause the character states expressed in the rare environments to evolve away from their optima (Via, 1987
). In the present study, dry habitats with low pH and wet habitats with high pH were by far the most common. The comparatively high growth rates found in these habitats, compared to the less common habitats, might be explained in terms of such constraints.
Patterns of cross-environment genetic correlations may present possible constraints on adaptive reaction norm evolution (Via and Lande, 1985
; Via, 1987
; van Tienderen, 1991
; Gomulkiewicz and Kirkpatrick, 1992
). Whether genetic correlations will constrain evolution depends on whether selection acts to change the phenotypes towards the same or different optima in different habitats (Via, 1994
). For growth rate, negative correlations among habitats may constrain evolution towards a similar optimum in all habitats. On the other hand, positive correlations may constrain the evolution of an adaptive plastic response. As discussed below, this may be the case regarding spreading branch length or branch fascicle density.
Temporal and spatial variability of environments
Differences in pH remain relatively stable across environments, even though periods of high precipitation tend to decrease differences in pH values. The pH values in a single patch are mostly either above or below 5 throughout the two seasons. Under these conditions, different selection pressures at different pH levels seem to promote locally adapted genotypes. Specialization is also demonstrated in length of branches across the water-level gradient. In contrast to pH, water levels display considerable within- and between-year variation. In long periods of drought that may be encountered in mid-summer, even the wettest parts of the area will dry out, whereas in periods of high rainfall the water table may exceed the lawns, leaving most mosses submerged or nearly submerged. Thus, with respect to water level, all patches experience conditions close to both extremes within a single season. Reaction norm evolution in environments with temporal variation within generations depends on whether the character is labile (subject to change during lifetime) or nonlabile (unable to be affected by environmental changes after some critical period) (Gomulkiewicz and Kirkpatrick, 1992
; Via, 1994
). Spreading branches develop in the gametophore capitulum and, although they may have the potential for continuous growth, they are not labile in the sense of having the ability to change as fast as the environment (cf. Scheiner, 1993
). Nonlabile traits can evolve an optimal "compromise" in the absence of genetic constraints, reflecting the duration and intensity of within-generation fluctuations in selection (Gomulkiewicz and Kirkpatrick, 1992
).
Adaptive significance of morphological trait variation
Our results suggest that the bulk of genetic variation in the morphological traits is not correlated with genotypes from particular habitats. The exception is the habitat-specific genetic response in length of spreading branches to the water-level gradient. The genotypes from dry habitats respond according to what normally occurs (i.e., spreading branches become shorter in drier habitats and are presumably less vulnerable to desiccation). Moreover, the gametophores growing best were those with the longest branches when growing wet and the shortest branches when growing dry. The branches of genotypes from wet habitats were shorter when growing wet than when growing dry, and growth rate decreased with increasing branch length. Genotypes from wet habitats seem to invest more resources in elongation growth of the shoot under favorable conditions compared to genotypes from dry habitats. The results suggest that horizontal expansion (e.g., by having longer spreading branches) to monopolize resources may be restricted both by vulnerability to desiccation in dry habitats and by the need to maintain vertical growth to keep up with surrounding gametophores in wet habitats. Thus, despite the fact that genotypes from wet and dry habitats experience a comparable range of environmental conditions, it may seem that selection pressures are in fact very different, because of varying durations of the different environmental conditions. This can explain why patterns of genetic specialization are found in these habitats. Moreover, the adaptive significance of the plastic response seemed to vary between habitats. Variable dry-habitat genotypes grew better overall, suggesting that this plastic response is adaptive, whereas for wet-habitat genotypes the most variable genotypes showed lowest overall growth, suggesting a developmental constraint.
The positive correlation between growth rate and branch fascicle density in dry habitats may be explained in terms of improved capillary movement of water and better water retention ability in dense gametophores (Clymo and Hayward, 1982
; Rydin, 1985
). The fact that the number of branch fascicles decreased in the dry habitats may be a consequence of lower growth overall. Rydin and Mcdonald (1985)
concluded that the ability of S. fuscum to grow in hummocks was related more to its greater water transport to living tissue than to its ability to alter photosynthetic rate in response to water content. In wet habitats, in which capillary movement of water is less important, high branch fascicle density is more likely to hamper effective photosynthesis by reducing light availability. The positive correlation between phenotypic plasticity of branch fascicle density and growth rate in our study indicates that this constitutes an adaptive response to different water levels. However, the negative genetic correlations between fascicle density and growth rate within three of four environments may constrain the evolution of this response.
The pH gradient seems to be the most important variable with regard to plastic responses in branch and stem leaf morphology. The frequently significant quadratic relationships might explain why these patterns were largely undetected in the purely linear model. Assuming linearity by involving only two environments to estimate reaction norms may be misleading when relationships between morphology and environment are not monotonic. Broad stem leaves may have a function in capillary water movement similar to that of branch fascicles, which may explain the positive relationships between stem leaf breadth plasticity and growth rate along the water-level gradient. The corresponding negative relationship along the pH gradient is more difficult to interpret, but may imply that nutrient levels constrain stem leaf development.
The retention of hemi-isophyllous leaves and poorly developed pendant branches may imply that the duration of the experiment was too short to yield significant environmental responses of morphology. If the amount of morphological change during the experiment varies mainly in response to how much different gametophores have grown at field conditions, a correlation between morphology and growth is to be expected. There was no overall phenotypic correlation between growth rate and any morphological trait (results not shown), which indicates that the observed differentiation in morphological characters was not merely a consequence of differences in growth. Rather the altered patterns of correlation between morphology and growth in the various habitats suggest that changes in growth habit between these habitats reflect responses to the environment.
There is no direct apriori model to describe how growth rate contributes to future representation in the population; thus, its adequacy as a measure of fitness might be questioned. High growth in length in relation to neighbors will increase the risk of more frequent desiccation (Hayward and Clymo, 1983) because the water retention capacity will be better in a dense carpet of gametophores. However, if length growth equals that of neighbors, high relative growth rate may increase fitness as a large capitulum increases surface area, reduces risk of overgrowth, and increases the possibility of vegetative reproduction through dispersal of fragments and branching (Rydin, 1993
). In this study gametophores may be viewed as being in a phase of establishment, which probably increases the relevance of growth rate as a measure of fitness. The gametophores did not produce sporophytes; thus, the impact of sexual reproduction on fitness and possible trade-offs between this and growth could not be evaluated.
Concluding remarks
The present study demonstrates genetic specialization to habitat within local populations of a haploid bryophyte. Multiple niche selection has been suggested as a potential process maintaining high electrophoretic variability in bryophyte populations (Wyatt, Odrzykoski, and Stoneburner, 1989
). Even if tests of selective neutrality indicate that most isozymes act as if they are selectively neutral (Stenøien, 1999
; Stenøien and Såstad, 1999
), patterns revealed here fit a model of local adaptation to multiple niches within a single population. In contrast to the former beliefs that the broad tolerances observed for some physiological traits make genetic adaptation to habitats with relatively minor differences in ecological conditions unlikely among bryophytes (e.g., Shaw, 1991
), these results suggest that such adaptation may exist.
FOOTNOTES
1 The authors thank Kjell I. Flatberg for assistance during fieldwork; Carolyn Baggerud for help with inoculations, and Line Bretten for help with morphometry; Hans Stenøien, Nils Cronberg, Håkan Rydin, Jon Shaw, and Robert Wyatt for valuable comments on the manuscript. This work was supported by grant 107627/420 from the Norwegian Research Council. 
LITERATURE CITED
Clymo, R. S., and P. M. Hayward. 1982 The ecology of Sphagnum. In A. J. E. Smith [ed.], Bryophyte ecology, 229–289. Chapman and Hall, London.
Cronberg, N. 1996 Clonal structure and fertility in a sympatric population of the peat mosses Sphagnum rubellum and Sphagnum capillifolium. Canadian Journal of Botany 74: 1375–1385.
Crum, H. 1972 The geographic origins of the mosses of North America's eastern deciduous forests. Journal of the Hattori Botanical Laboratory 35: 269–298.
Cummins, H., and R. Wyatt. 1981 Genetic variability in natural populations of the moss Atrichum angustatum. Bryologist 84: 30–38.
Gebhardt, M. D., and S. C. Stearns. 1993 Phenotypic plasticity for life history traits in Drosophila melanogaster: I. Effect on phenotypic and environmental correlations. Journal of Evolutionary Biology 6: 1–16.
Gomulkiewicz, R., and M. Kirkpatrick. 1992 Quantitative genetics and the evolution of reaction norms. Evolution 46: 390–411. [CrossRef][ISI]
Hayward, P. M., and R. S. Clymo. 1983 The growth of Sphagnum: experiments on, and simulation of, some effects of light flux and water table depth. Journal of Ecology 71: 845–863. [CrossRef]
Heegaard, E. 1997 Morphological variation within Andreaea blyttii in relation to the environment on Hardangervidda, western Norway: a quantitative analysis. Bryologist 100: 308–323. [ISI]
Jules, E. S., and A. J. Shaw. 1994 Adaptation to metal-contaminated soils in populations of the moss, Ceratodon purpureus: vegetative growth and reproductive expression. American Journal of Botany 81: 791–797. [CrossRef][ISI]
Kallio, P., and E. Saarnio. 1986 The effect on mosses of transplantation to different latitudes. Journal of Bryology 14: 159–178. [ISI]
Lacey, E. P., L. Real, J. Antonovics, and D. G. Heckel. 1983 Variance models in the study of life histories. American Naturalist 122: 114–131. [CrossRef][ISI]
Longton, R. E. 1981 Inter-population variation in morphology and physiology in the cosmopolitan moss Bryum argenteum Hedw. Journal of Bryology 11: 501–520. [ISI]
McQueen, C. B. 1985 Spatial pattern and gene flow distances in Sphagnum subtile. Bryologist 88: 333–336.
Rydin, H. 1985 Effects of water level on desiccation of Sphagnum in relation to surrounding Sphagna. Oikos 45: 374–379. [CrossRef][ISI]
———. 1993 Mechanisms of interactions among Sphagnum species along water level gradients. Advances in Bryology 5: 153–185.
———, and A. J. S. Mcdonald. 1985 Photosynthesis in Sphagnum at different water contents. Journal of Bryology 13: 579–584.
Såstad, S. M., and K. I. Flatberg. 1993 Leaf morphology of Sphagnum strictum in Norway, related to habitat characteristics. Lindbergia 18: 71–77.
———, and ———. 1994 Leaf size and shape in the Sphagnum recurvum complex: taxonomic significance and habitat variation. Journal of Bryology 18: 261–275.
———, S. Bakken, and B. Pedersen. 1998 Propagation of Sphagnum in axenic culture—a method for obtaining large numbers of cloned gametophores. Lindbergia 23: 65–73.
Scheiner, S. M. 1993 Genetics and evolution of phenotypic plasticity. Annual Review of Ecology and Systematics 24: 35–68. [CrossRef][ISI]
Schlichting, C. D. 1986 The evolution of phenotypic plasticity in plants. Annual Review of Ecology and Systematics 17: 667–693. [CrossRef][ISI]
Shaw, J. 1988 Genetic variation for tolerance to copper and zinc within and among populations of the moss, Funaria hygrometrica Hedw. New Phytologist 109: 211–222. [CrossRef][ISI]
———. 1990 Metal tolerances and cotolerances in the moss Funaria hygrometrica. Canadian Journal of Botany 68: 2275–2282.
———. 1991 Ecological genetics, evolutionary constraints, and the systematics of bryophytes. Advances in Bryology 4: 29–74.
———. 1992 The evolutionary capacity of bryophytes and lichens. In J. W. Bates and A. M. Farmer [eds.], Bryophytes and lichens in a changing environment, 362–380. Clarendon Press, Oxford.
———, S. C. Beer, and J. Lutz. 1989 Potential for the evolution of heavy metal tolerance in Bryum argenteum, a moss. I. Variation within and among populations. Bryologist 92: 73–80. [CrossRef][ISI]
SPSS. 1997 SPSS advanced statistics 7.5. SPSS, Chicago, IL.
StenØien, H. K. 1999 Are enzyme loci selectively neutral in haploid populations of nonvascular plants? Evolution 53: 1050–1059. [CrossRef][ISI]
———, and S.M. Såstad. 1999 Population structure in three species of peat mosses (Sphagnum). Heredity 82: 391–400.
van Tienderen, P. 1991 Evolution of generalists and specialists in spatially heterogeneous environments. Evolution 42: 1342–1347. [CrossRef]
———. 1992 Variation in a population of Plantago lanceolata along a topographical gradient. Oikos 64: 560–572. [CrossRef][ISI]
Via, S. 1987 Genetic constraints on the evolution of phenotypic plasticity. In V. Loeschcke [ed.], Genetic constraints on adaptive evolution, 47–71. Springer-Verlag, Berlin.
———. 1994 The evolution of phenotypic plasticity: what do we really know? In L. A. Real [ed.], Ecological genetics, 35–57. Princeton University Press, Princeton, NJ.
———, and R. Lande. 1985 Genotype-environment interaction and the evolution of phenotypic plasticity. Evolution 39: 505–522. [CrossRef][ISI]
Wyatt, R., and L. E. Anderson. 1984 Breeding systems in bryophytes. In A. F. Dyer and J. G. Duckett [eds.], The experimental biology of bryophytes, 39–64. Academic Press, London.
———, I. J. Odrzykoski, and A. Stoneburner. 1989 High levels of genetic variability in the haploid moss Plagiomnium ciliare. Evolution 43: 1085–1096.
———, A. Stoneburner, and I. J. Odrzykoski. 1989 Bryophyte isozymes: Systematic and evolutionary implications. In D. E. Soltis and P. S. Soltis [eds.], Isozymes in plant biology, 221–240. Dioscorides Press, Portland, OR.
Yamazaki, T. 1981 Genetic variabilities in natural population of haploid plant, Conocephalum conicum. I. The amount of heterozygosity. Japanese Journal of Genetics 56: 373–383. [CrossRef][ISI]
———. 1984 The amount of polymorphism and genetic differentiation in natural populations of the haploid liverwort, Conocephalum conicum. Japanese Journal of Genetics 59: 133–139.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
