| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ecology |
2Department of Botany 3Department of Ecology & Evolutionary Biology, University of Tennessee, Knoxville, Tennessee 37996 USA
Received for publication March 1, 2001. Accepted for publication November 6, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Brassica • Brassicaceae • character correlations • hybridization • phenotypic integration • phylogeny • rapid-cycling Brassica
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Schlichting and Pigliucci, 1998
; Wagner and Schwenk, 2000
). As it is obvious that a single trait can have a direct effect on the fitness of an organism, the covariation among multiple phenotypic characters arising from these types of relationships (Stearns, de Jong, and Newman, 1991
) may also strongly affect fitness, potentially making trait–trait interactions adaptive (Schlichting, 1989a, b
).
Environments can also affect correlations among characters, although underlying mechanisms may change to maintain adaptive correlations. Across light levels, Polygonum persicaria is successful at producing seeds by changing resource allocation to different traits within a functionally related set of characters (Sultan, 1995
). The relationships among functionally or developmentally related characters and their changes across environments have been considered important components of phenotypic evolution (e.g., Cheverud, 1984
; Zelditch, Bookstein, and Lundrigan, 1992
; Klingenberg and Zaklan, 2000
).
Interest in multivariate analyses of animal and plant phenotypes dates back to the early part of the twentieth century, most notably the descriptive work done by D'Arcy Thompson (1917)
on fossil skulls and Clausen, Keck, and Heisey's (Clausen, Keck, and Hiesey, 1940
; Clausen and Heisey, 1958
) long-term studies on several plant species in the Sierra Nevada. These authors hypothesized that species distinctions are due in large part to differences in patterns of character coherence, which they considered to have an underlying genetic basis. Olson and Miller (1958)
formalized the definition of morphological integration as the interdependence of morphological traits that produces an organized, functional organism. Berg (1960)
independently developed ideas similar to Olson and Miller (1958)
following the work of Terentjev (1931)
. Berg (1960)
developed and tested specific functional hypotheses of correlation patterns of plant reproductive characters. In her studies, she expected to find more correlations within floral characters than between floral and vegetative modules. This hypothesis was supported for plants with species-specific pollinators, but correlations between "modules" (i.e., a functionally related set of traits) were greater for generalist-pollinated species. In contrast to ecologically based comparisons, a few recent studies have adopted a phylogenetic perspective to address the question of the evolution of phenotypic integration (e.g., Fink and Zelditch, 1996
; Shubin and Wake, 1996
; Steppan, 1997
). Combining ecological and phylogenetic approaches may get us closer to understanding the evolution of correlations as traits.
The relationship between hybrids and their parental taxa have been examined as a specific type of phylogenetic relationship and to understand the evolution of phenotypic integration. Grant (1979)
examined correlation structures of a number of hybrid swarms. In particular, he examined a population that included hybrids (between Iris fulva and Iris hexagona) and a parental taxon (Iris hexagona). In this population, the hybrids maintained only seven of the ten significant correlation coefficients in comparison with the parentals. In contrast, in the genus Aquilegia, Grant (1979)
found that the magnitudes of the correlations were reduced for the hybrids, in comparison with their parents, but all pairwise correlations remained significant. Clausen and Heisey (1958)
examined the correlation structure of hybrid crosses of several distinct pairs of ecotypes of Potentilla glandulosa and how offspring of intra-specific hybrid crosses differed from the parental ecotypes. They found that the hybrids of the ecotypes were more integrated than the parents, perhaps due to blending of genetic variation from the parental ecotypes.
We chose the Brassica complex as a suitable system to determine whether ecological or genetic history influence patterns of phenotypic integration, as hybrid species and their parents are characterized by variability in their ecologies (Arnold, 1997
). We studied the evolution of phenotypic integration using six members of the genus Brassica and Raphanus sativus, for which a robust phylogenetic hypothesis is available (Williams, 1989
). We investigated three types of modules: life-history traits, leaf/resource gathering traits, and architectural traits. First, we asked (1) are there differences in trait means in morphological or life-history characters among closely related species of Brassica? As trait means and correlations may evolve independently, we next asked (2) do patterns of integration vary among species? Specifically, are hybrids more/less integrated than their parents? Additionally, are there greater numbers of correlations within our hypothesized modules (architectural, life-history, and leaf/light gathering) than across modules and how do these patterns vary? Then, we went on to examine relationships across the entire clade, asking (3) how do correlation patterns relate to overall similarities in the plants' cytogenetics? We tested the hypothesis that species with similar cytogenetic relationships would have more similar patterns of phenotypic correlations than species with distinct cytogenetics. Alternatively, species with shared ecologies may share phenotypic integration patterns regardless of the phylogenetic relationships among the species. Finally, we asked (4) how are our hypothesized modules associated with fitness and are these patterns consistent across taxa? This leads to the question of whether these units function together as a compound trait and whether selection can operate on the multitrait correlations.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). It is thought that R. sativus is derived from a bridge species that was a hybrid between B. nigra and another species in the B. oleracea lineage (Song, Osborn, and Williams, 1990
; Song, Tang, and Osborn, 1995
). Therefore, although R. sativus is in a separate genus, it shares a relatively recent common ancestor with B. oleracea and a more distant one with B. nigra. Seeds were obtained from Wisconsin fast plants (Carolina Biological Supply Company, Burlington, North Carolina, USA).
|
We took data on traits in three potential functional modules. The first module consisted of life-history characters: days to first leaf, days to first flower, and days to senescence. Life-history characters were measured throughout ontogeny. Date of onset of senescence was determined as the time when the last flower's pistil had elongated, at which point the plants were harvested. At senescence, plants were harvested, and we measured traits of the remaining two modules. In the next module of leaf/light gathering traits, we measured leaf length, leaf width, petiole length, and counted total leaf number. We also counted fruit number—our measure of fitness. Following these measurements, plants were dried in an oven at 32.2°C, and biomass of stems and leaves were recorded. The architectural module included total plant height, stem biomass, and leaf biomass.
Data analysis
We employed analyses of variance (ANOVAs using SYSTAT, version 8.0; SPSS Science, Chicago, Illinois, USA, 1998) to examine whether individual characters differed across species and across experimental blocks (tiers on the light rack). We applied a sequential Bonferroni correction to account for multiple simultaneous tests. To meet the assumptions of normality for the ANOVAs, we log-transformed number of days to first leaf, number of days to flower, number of fruits, height, and leaf length, and square-root-transformed number of days to onset of senescence, stem and leaf biomass. We then specified a priori contrasts in SYSTAT to compare a hybrid to the mean of its parental species. Matrices of pairwise Pearson correlations between characters were constructed for each species independently. We created a set of diagrams to depict the network of correlations among traits for each species to illustrate the degree and pattern of phenotypic integration (Fig. 2). In these diagrams, a line connecting a pair of traits represents a significant correlation after a matrix-wide Bonferroni correction for multiple tests. Numbers of significant correlations were used as a measure of the amount of phenotypic integration.
|
Second, we generated a matrix of among-species similarities in their phenotypic trait correlation patterns. This phenotypic integration matrix was constructed in a stepwise fashion by calculating a correlation coefficient between phenotypic correlation matrices for each pair of taxa. Therefore, the rows and columns of this matrix are the seven species; the entries in this matrix are correlation coefficients from the pairwise comparisons of the Pearson correlation matrix of the phenotypic traits for every species–species combination (calculations were carried out using NTSYS, version 2.0; Rohlf, 1998
).
A third matrix was constructed using ecological data of the habitat types and the geographic distribution of the species as reported in a number of floras from Europe (these data have been archived at the Botanical Society of America supplemental data website: http://ajbsupp.botany.org/v89/murren.doc). We began with a matrix of ecological attributes (in columns) for each of seven species (in rows). We calculated Jaccard coefficients (Jaccard, 1908
) as our measure of ecological similarity for each species pair. The Jaccard coefficient is a similarity that provides a standardized measure of the number of matches in a set. To build the ecological matrix, the rows and columns were the Jaccard coefficients (Jaccard, 1908
) between pairwise combinations of the seven species. The Jaccard coefficients were calculated in NTSYS.
Because the relationships among all pairwise species comparisons are not independent, the value of the correlation between matrices is expected to be greater than zero even if the null hypothesis of no correlation between matrices is correct. We therefore used a Mantel test (Cheverud, Wagner, and Dow, 1989
; implemented in NTSYS) to compare the phenotypic integration matrix to the genomic matrix and then to the ecological matrix. This test identifies whether the architecture of any given matrix is predicted by a hypothesis matrix by randomizing the entries in one matrix, recalculating the matrix correlation coefficient, and determining an empirical distribution of the matrix correlation coefficients (for further details of the method see: Manly, 1986
; Smouse, Long, and Sokal, 1986
; Cheverud, Wagner, and Dow, 1989
; Cheverud, 1995
). The actual value of the inter-matrix correlation is then compared with the empirical distribution to determine the true probability of rejection of the null hypothesis of no correlation. Other recently proposed methods of matrix comparison such as common principal components analyses (Phillips and Arnold, 1999
) are not appropriate in this case because the principal components of our matrices are not biological homologues of each other. An unweighted pair group method with arithmetic mean (UPGMA) diagram was constructed in NTSYS to display graphically the similarities among species in their phenotypic integration (Cheverud, Wagner, and Dow, 1989
; Waldmann and Andersson, 2000
). This diagram reflects the overall pattern of correlations, regardless of their statistical significance.
To understand how sets of life-history, vegetative, or architectural traits were related to fitness, we used set correlations in SYSTAT. Set correlation analysis is a generalized form of single and multiple correlations and is related to general linear models. Thus, it provides a broad framework for the study of the evolution of sets of correlations. We examine both the individual regressions of fitness on each trait in the module and the multivariate regression of fitness on all characters in the module simultaneously.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
|
The visual inspection of the similarities among correlation networks was substantiated by the cluster analysis grouping species in proportion to the overall similarity in their correlation patterns (UPGMA diagram; Fig. 3). Brassica napus and R. sativus shared 79% of their correlation structure. Brassica juncea shared about 72% of its phenotypic integration pattern with B. napus and R. sativus. Brassica carinata joined the juncea-napus-sativus group, but with only about 55% similarity. Notice that three of these species (B. carinata, B. juncea, and B. napus) are hybrids. Brassica rapa then joined the main cluster with only 50% similarity. Brassica nigra and B. oleracea showed 70% commonality in their integration patterns, but fell into a distinct group from all other study species (sharing only 42% similarity with the rest).
|
We used set correlation analysis to test whether our hypothesized modules functioned as a single trait. We found that there was heterogeneity across species in the relationship of these modules and fitness (Table 4a–c). For example, in the life-history module, the overall model, which explains the relationship of all three characters and fitness, was significant for five out of seven taxa (Table 4a). However, across all species, days to first leaf did not have a significant relationship with fitness. Days to first flower and days to harvest were significantly related to fitness for B. juncea, B. napus, B. nigra, and Raphanus sativus. Days to harvest alone was significantly related to fitness in B. carinata and B. rapa. Brassica oleracea had no component of module A significantly related to fitness. Together, the architectural characteristics for Brassica juncea (Table 4b) clearly defined a module (the overall model is significant with an R2 value = 0.9). In this case each of the components of the module were significantly positively associated with fitness. Across all taxa the overall models for life-history traits were highly significant, but which particular trait was related to fitness was highly heterogeneous across taxa. In module C (leaf characteristics), leaf number was associated with fitness for two taxa (Table 4c). For five species, the overall model was significant, suggesting that these traits do comprise a functional module. However, different traits were significantly associated with fitness across the species.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Clausen and Hiesey, 1960
; Schlichting, 1989a, b
), and recently it has received increasing attention both from a theoretical (Wagner, 1995
; Wagner and Schwenk, 2000
) and an empirical (Stirling, Roff, and Fairbairn, 1999
; Camara, Ancell, and Pigliucci, 2000
; Klingenberg and Zaklan, 2000
; Mezey, Cheverud, and Wagner, 2000
) standpoint. Integration has been viewed primarily as a constraint to phenotypic change since developmental and functional processes are thought to change in a coordinated manner across a phylogeny (Fink and Zeldich, 1996
). In this scenario, one would expect closely related taxa to share similar patterns of integration. Alternatively, others have hypothesized that similar selection pressures could result in similar patterns of character coherence particularly for hybrid taxa (Grant, 1979
). Our goal in this paper was to use the multiple hybrid/parental species of the Brassica complex (Song, Osborn, and Williams, 1990
; Galloway, Malmberg, and Price, 1998
) to test the effects of shared genetic or ecological history of these species on the evolution of phenotypic integration.
Hybrids and their parental taxa
The patterns of integration varied widely among these seven closely related species, from many (B. juncea) to few (B. rapa) significant pairwise correlations. However, for all species examined in this clade, traits were always positively correlated. In contrast to our expectation that hybrid species would have levels of integration intermediate between their parental species, hybrids had greater numbers of significant correlations (i.e., they were more integrated) than either of their parents in all three cases. Although this is an interesting result, we interpret these patterns of integration with caution, as we only examined three hybrid taxa and their parental species, and we encourage further investigation into other hybrid/parental taxa.
In the case of among-species comparisons, we found that hybrids were more integrated than their parents. However, we suggest (following Grant, 1979
) that there is likely to be a variety of possible patterns when comparing the correlation structure between parental taxa and hybrids. The resulting pattern may be influenced by both the amount of recombination and the maintenance of functionally coherent units by selection.
Patterns of correlations were more similar among the hybrids than they were to their parents. In contrast, character means of the hybrid species were generally similar to the mean trait of the two parents (intermediate phenotype). We found that there were significant differences among species for all character means. Therefore, our data suggest that the relationships of correlations and means may follow distinct patterns. One may choose to examine trait means and trait correlations as distinct aspects of phenotypic evolution (Schlichting and Pigliucci, 1998
: 191–226).
Our hypothesized functional modules held up to statistical investigation for some species, but they had variable success in predicting evolutionarily important relationships among traits and fitness. Overall, the architectural module (comprised of leaf and stem biomass and plant height) was significantly associated with potential reproductive fitness; however, which particular traits were significantly related to fitness within this module varied among species. This may be an indication of differential selective pressures on patterns of correlations across species during their evolutionary history (Cheverud, 1984
). In Berg's (1960)
classic study, she found more correlations within a module (floral or vegetative) than between modules. In contrast, we found a large number of correlations across modules. Although biologically reasonable (Pigliucci et al., 1991
; Pigliucci, 1992
), the modules as we defined them may not correspond to functional groups within some of the species of Brassica. Moreover, functional modules may differ across species. Cross-module correlations, particularly in the hybrid taxa, may be due to genetic linkage or to traits with shared developmental trajectories. Understanding these mechanisms warrants further investigation, but is distinct from a study documenting the patterns of phenotypic correlations and their evolutionary fate (Clausen and Hiesey, 1958, 1960
).
Shared ecology and cytogenetics
We did not find a significant relationship between cytogenetic similarity and patterns of phenotypic integration. Nor did we find a relationship between gross ecological characteristics (which included biogeography) of the species we examined and patterns of phenotypic integration. Our analyses therefore may suggest that differences in correlation patterns are most likely due to evolution of specific components of the genetic architecture rather than to shared cytogenetics or ecology. It is of course possible that more refined measures of genetic distance or ecological similarity for these same species might yield different results.
The lack of significant relationships between the matrix of integration and the matrices of cytogenetics or ecology may have several interpretations. The number of pairwise correlations may not reflect the quantitative pattern of integration as expressed by the strength of the correlations (but this does not always have to be the case). A set of taxa may have the same number of significant correlations, but which particular correlations are significant may be unique for each taxon. As the number of significant pairwise correlations increases, the pattern of similarity may also increase simply because there are not many alternative ways to get a large number of significant correlations. As hybrids were characterized by a greater number of significant correlations overall than their parents, this mechanism may lead to more similarity among hybrids, regardless of their genetic relationships. Moreover, examination of different modules than those studied here (e.g., floral traits) may lead to different patterns of similarity among taxa.
A few studies have examined whether patterns of phenotypic integration evolve across a clade or across environments (e.g., Dudley and Schmitt, 1996
; Ackerly and Donoghue, 1998
). Ackerly and Donoghue (1998)
used a phylogenetic approach to investigate 17 species of Acer. They found two sets of coevolving characters that remained integrated during evolution, suggesting that adaptive evolution was not significantly affected by historical contingencies. Callahan and Waller (2000)
used a path analysis approach to examine the relative importance of habitat similarity or genetic relatedness. They examined plasticity of individual traits and the plasticity of integration in two plant varieties of the genus Amphicarpaea. Surprisingly, the sun-adapted variety and the populations from the low light of another variety were most similar. This suggests that for these particular varieties, neither habitat similarity nor genetic relatedness were good predictors of the patterns of integration. Nicotra, Chazdon, and Schlichting (1997)
examined two closely related species of Piper, one classified as shade tolerant and one classified as a pioneer species. Unlike Callahan and Waller's work, Nicotra, Chazdon, and Schlichting (1997)
found similar patterns of plasticity and phenotypic integration for morphological and physiological characters in the two species. On the other hand, Riska (1985)
neglected to find an association between phenotypic correlation matrices and geography in populations of alates and stem mothers of aphids, Pemphigus populicaulis. He hypothesized that this pattern was caused by chance and by differences in genetic variation, in combination with ecological variation among localities. These results and ours suggest that the evolution of phenotypic integration may be influenced by shared ecology or shared evolutionary history, but their effect on integration may be quite distinct among taxa.
Interestingly, we found a significant negative relationship between the ecological and the cytogenetic similarity matrices of these species. This might demonstrate the opposite of what Westoby, Leishman, and Lord (1995)
referred to as "phylogenetic niche conservatism." According to Westoby, Leishman, and Lord (1995)
taxa that share a recent common ancestor also tend to occupy similar niches. If our results hold when using more sophisticated indicators of phylogenetic relationship and ecological similarity, this would suggest that Brassicas have experienced higher levels of ecological diversification among closely related species, possibly related to their ubiquitous episodes of hybridization.
The relationship between ecological diversification and phenotypic integration is a topic that deserves additional investigation. Berg's classic study (1960)
on phenotypic correlations among flower traits and how they correspond to the different pollinator guilds sheds some light on this issue (see also Armbruster et al., 1999
). These studies and ours suggest that our focus should be on how specific subsets of characters relate in a functional manner. Additionally, we should test the validity of these definitions of functional modules across taxa. The consensus seems to be that phenotypic integration is important for our understanding of how complex phenotypes evolve (Pigliucci et al., 1996
; Phillips and Arnold, 1999
; Roff, 2000
; Wagner, 2001
).
|
|
FOOTNOTES |
|---|
4 Current address: Department of Biology, H.J. Patterson Hall, University of Maryland, College Park, Maryland 20742 USA
5 Author for reprint requests (phone: 301-405-1640; fax: 301-314-9081; cmurren{at}wam.umd.edu
; URL: www.genotype-environment.org)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Armbruster W. S. V. S. D. Stilio J. D. Tuxill T. C. Flores J. L. V. Runk 1999 Covariance and decoupling of floral and vegetative traits in nine Neotropical plants: a re-evaluation of Berg's correlation-pleiades concept. American Journal of Botany 86: 39-55
Arnold M. L. 1997 Natural hybridization and evolution. Oxford University Press, New York, New York, USA
Berg R. L. 1960 The ecological significance of correlation pleiades. Evolution 14: 171-180[CrossRef][ISI]
Callahan H. S. D. M. Waller 2000 Phenotypic integration and the plasticity of integration in an amphicarpic annual. International Journal of Plant Sciences 161: 89-98[CrossRef][ISI][Medline]
Camara M. D. C. A. Ancell M. Pigliucci 2000 Induced mutations: a novel tool to study phenotypic integration and evolutionary constraints in Arabidopsis thaliana. Evolutionary Ecology Research 2: 1009-1029[ISI]
Cheverud J. M. 1984 Quantitative genetics and developmental constraints on evolution by selection. Journal of Theoretical Biology 110: 155-171[ISI][Medline]
Cheverud J. M. 1995 Morphological integration in the saddle-back tamarin (Saguinus fuscicollis) cranium. American Naturalist 145: 63-89[CrossRef][ISI]
Cheverud J. M. G. P. Wagner M. M. Dow 1989 Methods for the comparative analysis of variation patterns. Systematic Zoology 38: 201-213[CrossRef]
Clausen J. W. M. Hiesey 1958 Experimental studies on the nature of plant species. IV. Genetic structure of ecological races. Publication 615, Carnegie Institution of Washington, Washington, D.C., USA
Clausen J. W. M. Hiesey 1960 The balance between coherence and variation in evolution. Proceedings of the National Academy of Sciences, USA 46: 494-506
Clausen J. D. Keck W. M. Hiesey 1940 Experimental studies on the nature of plant species. I. Effect of varied environment on western North American plants. Publication 520, Carnegie Institution of Washington, Washington, D.C., USA
Dudley S. A. J. Schmitt 1996 Testing the adaptive plasticity hypothesis: density-dependent selection on manipulated stem length in Impatiens capensis. American Naturalist 147: 445-465[CrossRef][ISI]
Fink W. L. M. L. Zelditch 1996 Historical patterns of developmental integration in Piranhas. American Zoologist 36: 61-69
Galloway G. L. R. L. Malmberg R. A. Price 1998 Phylogenetic utility of the nuclear gene arginine decarboxylase: an example from Brassicaceae. Molecular Biology Evolution 15: 1312-1320
Grant V. 1979 Character coherence in natural hybrid populations in plants. Botanical Gazette 140: 443-448[CrossRef]
Jaccard P. 1908 Nouvelles recherches sur la distribution florale. Bulletin Societe Vaudoise des Sciences Naturelle 44: 223-270
Klingenberg C. P. S. D. Zaklan 2000 Morphological integration between developmental compartments in the Drosophila wing. Evolution 54: 1273-1285[CrossRef][ISI][Medline]
Manly F. J. 1986 Randomization and regression methods for testing for associations with geographical, environmental and biological distances between populations. Research in Population Ecology 28: 201-218
Mezey J. G. J. M. Cheverud G. P. Wagner 2000 Is the genotype-phenotype map modular? A statistical approach using mouse quantitative trait loci data. Genetics 156: 305-311
Nicotra A. B. R. L. Chazdon C. D. Schlichting 1997 Patterns of genotypic variation and phenotypic plasticity of light response in two tropical Piper species. American Journal of Botany 84: 1542-1552[Abstract]
Olson E. C. R. L. Miller 1958 Morphological integration, 1st ed. University of Chicago Press, Chicago, Illinois, USA
Phillips P. C. S. J. Arnold 1999 Hierarchical comparison of genetic variance-covariance matrices. I. Using the Flury hierarchy. Evolution 53: 1506-1515[CrossRef][ISI]
Pigliucci M. 1992 Spatio-temporal variation of phenotypic plasticity and integration in natural populations of Anchusa italica Retz. Acta Oecologica 13: 255-267
Pigliucci M. C. Paoletti S. Fineschi M. E. Malvolti 1991 Phenotypic integration in chestnut (Castanea sativa Mill.): leaves versus fruits. Botanical Gazette 152: 514-521[CrossRef]
Pigliucci M. C. D. Schlichting C. S. Jones K. Schwenk 1996 Developmental reaction norms: the interactions among allometry, ontogeny and plasticity. Plant Species Biology 11: 69-85
Riska B. 1985 Group size factors and geographic variation of morphometric correlation. Evolution 39: 792-803[CrossRef][ISI]
Roff D. 2000 The evolution of the G matrix: selection or drift?. Heredity 84: 135-142
Rohlf F. J. 1998 NTSYS pc. Exeter Software, Setauket, New York, USA
Schlichting C. D. 1989a Phenotypic integration and environmental change. BioScience 39: 460-464[CrossRef][ISI]
Schlichting C. D. 1989b Phenotypic plasticity in Phlox. II. Plasticity of character correlations. Oecologia 78: 496-501[CrossRef][ISI]
Schlichting C. D. M. Pigliucci 1998 Phenotypic evolution: a reaction norm perspective, 1st ed. Sinauer Associates, Sunderland, Massachusetts, USA
Shubin N. H. D. Wake 1996 Phylogeny, variation, and morphological integration. American Zoologist 36: 51-60
Smouse P. E. J. C. Long R. R. Sokal 1986 Multiple regression and correlation extensions of the Mantel test of matrix correspondence. Systematic Zoology 35: 627-632[CrossRef]
Song K. P. Lu K. Tang T. C. Osborn 1995 Rapid genome change in polyploids of Brassica and its implications for polyploid evolution. Proceedings of the National Academy of Sciences, USA 92: 7719-7723
Song K. T. C. Osborn P. H. Williams 1990 Brassica taxonomy based on nuclear restriction fragment length polymorphisms (RFLPs). Theoretical and Applied Genetics 79: 497-506[ISI]
Stearns S. G. de-Jong B. Newman 1991 The effects of phenotypic plasticity on genetic correlations. Trends in Ecology and Evolution 6: 122-126
Steppan S. J. 1997 Phylogenetic analysis of phenotypic covariance structure. I. Contrasting results from matrix correlation and common principal component analyses. Evolution 51: 571-586[CrossRef][ISI]
Stirling G. D. Roff D. Fairbairn 1999 Four characters in a trade-off: dissecting their phenotypic and genetic relations. Oecologia 120: 492-498[CrossRef][ISI]
Sultan S. E. 1995 Phenotypic plasticity and plant adaptation. Acta Botanica Neederlandica 44: 363-383
Terentjev P. V. 1931 Biometrische Untersuchungen iiber die morphologischen Merkmale von Rana ridibunda Pall. (Amphibia, Salientia). Biometrika 23: 23-51
Thompson D. W. 1917 On growth and form. Cambridge University Press, Cambridge, UK
Wagner G. 2001 The character concept in evolutionary biology, 622. Academic Press, San Diego, California, USA
Wagner G. P. 1995 Adaptation and the modular design of organisms. In F. Moran, A. Moreno, J. J. Merelo, and P. Chacon [eds.], Advances in artificial life, 317–328. Springer, Berlin, Germany
Wagner G. P. K. Schwenk 2000 Evolutionarily stable configurations: functional integration and the evolution of phenotypic stability. Evolutionary Biology 31: 155-217
Waldmann P. S. Andersson 2000 Comparison of genetic (co)variance matrices within and between Scabiosa canescens and S-columbaria. Journal of Evolutionary Biology 13: 826-835[CrossRef][ISI]
Westoby M. M. R. Leishman J. M. Lord 1995 On misinterpreting the ‘phylogenetic correction’. Journal of Ecology 83: 531-534[CrossRef]
Williams P. H. 1989 Wisconsin fast plants manual. Carolina Biological Supply Company, Burlington, North Carolina, USA
Zelditch M. L. F. L. Bookstein B. L. Lundrigan 1992 Ontogeny of integrated skull growth in the cotton rat Sigmodon fulviventer. Evolution 46: 1164-1180[CrossRef][ISI]
This article has been cited by other articles:
|
|
 |
|
  C. J. Murren and M. Pigliucci Morphological responses to simulated wind in the genus Brassica (Brassicaceae): allopolyploids and their parental species Am. J. Botany, May 1, 2005; 92(5): 810 - 818. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Schranz and T. C. Osborn De novo variation in life-history traits and responses to growth conditions of resynthesized polyploid Brassica napus (Brassicaceae) Am. J. Botany, February 1, 2004; 91(2): 174 - 183. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Preston and D. D. Ackerly Hydraulic architecture and the evolution of shoot allometry in contrasting climates Am. J. Botany, October 1, 2003; 90(10): 1502 - 1512. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, P = 0.05 *, P = 0.01 **, P = 0.001 ***, and P < 0.001 ****. Module A is composed of life-history characters. Module B is composed of architectural characters and Module C is composed of leaf characters