| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
2Department of Biological Sciences, Stanford University, Stanford, California 94305; and 3Department of Forest Resources, University of Minnesota, St. Paul, Minnesota 55108
Received for publication October 30, 1998. Accepted for publication February 18, 1999.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: comparative method • correlated evolution • independent contrasts • leaf life span • leaf size • photosynthesis • seed plants • specific leaf area
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Reich et al., 1999
). In general, species with short leaf life span exhibit high specific leaf area, mass-based nitrogen concentration, and carbon assimilation rates; in log-log analyses, these relationships are linear with r2 > 0.5 over a broad range of values. This suite of attributes is also broadly associated with high allocation to leaves, rapid growth rates, rapid attainment of reproductive age, and regeneration in high resource or disturbed environments (Bazzaz, 1979
; Reich, Walters and Ellsworth, 1992
; Cornellison, Castro-Diez, and Carnelli, 1998
; van der Werf et al., 1998
). Recently, it has been found that the scaling functions among leaf traits (slopes of log-log relationships) are similar across a range of contrasting biomes representing a diverse array of conifer and angiosperm taxa (Reich, Walters, and Ellsworth, 1997
). These relationships have been interpreted as evidence of coordinated leaf physiological strategies, reflecting adaptive and/or biomechanical constraints. For example, the inverse relationship between leaf life span and assimilation rate indicates that two trait combinations are not generally observed: high assimilation rates in long-lived leaves and low rates in short-lived leaves. The former combination would appear to be highly adaptive (high carbon gain sustained over long periods of time), but may be impossible as leaf longevity may require both thickness to resist damage, creating internal self-shading, which reduces assimilation, as well as investment in protective physical or chemical traits, which reduces resources invested in the photosynthetic apparatus. In contrast, a species with short-lived leaves and low assimilation rates would presumably be eliminated by selection due to low growth and fecundity. Optimality models, based on maximizing carbon gain and plant growth, also predict an inverse relationship between assimilation and leaf life span (Kikuzawa, 1991
; Ackerly, 1999b
). It is important to note, however, that up to threefold variation in physiological traits may be observed at a particular level of leaf life span, so correlations among these traits may be weak or nonapparent in samples that do not span a great range of values (cf. Diemer, Körner, and Prock, 1992
; Reich, 1993
).
In addition to these patterns in leaf physiological traits, considerable attention has been focused on the ecological significance of leaf size variation (Parkhurst and Loucks, 1972
; Givnish and Vermeij, 1976
; Chiariello, 1984
). Leaf size directly affects light interception, light penetration through the canopy, and leaf energy balance, and leaves are often smaller in species occupying habitats with high light, low nutrients, or low moisture availability (e.g., Hamann, 1979
; Körner et al., 1989
; Niinemets and Kalevi, 1994
; see references in Givnish, 1987
), conditions that are also associated with thick leaves, low nitrogen concentrations, and low photosynthetic rates (Reich, Walters, and Ellsworth, 1992
). In cool-temperate forests, it has also been suggested that the predominance of conifers is due to the association between evergreenness and small, steeply angled leaves (i.e., conifer needles), which enhance light penetration and low-temperature canopy photosynthesis (Sprugel, 1989
; but see Smith and Brewer, 1994
). Leaf size also varies considerably among species within habitats, partly in association with architectural traits and reproductive morphology (White, 1983
; Midgley and Bond, 1989
; Ackerly and Donoghue, 1998
). However, relationships among leaf size and the physiological traits listed above have not been systematically examined.
Adaptive interpretations of relationships such as these suggest that the traits have exhibited correlated evolutionary changes, as selection combined with biophysical constraints would maintain certain trait combinations as outlined above. However, the use of cross-species correlations in comparative ecology has come under intense scrutiny in recent years, particularly in relation to the question of adaptive correlations, due to the potential problem of similarity among related species and the statistical independence of species as data points (Harvey and Pagel, 1991
). Though it is not always explicitly stated, independence in this context is based on a null model of the independence of evolutionary changes (Felsenstein, 1985
). Thus we may pose the question in the following way: do the trait relationships observed among contemporary species reflect or provide evidence of statistically significant patterns of correlated evolutionary changes leading up to the present day? Alternatively, correlations among extant species may arise due to a small number of initial divergence events, in which biologically independent changes in two traits happen to be correlated, and these changes are passed on to descendent species. This will lead to correlations between traits among extant taxa despite the independence of evolutionary changes in the past. This question has been addressed in recent years through the development of a number of statistical techniques that incorporate the phylogenetic relationships among species in order to estimate the evolutionary patterns underlying present-day trait distributions. The most powerful of these is the method of independent contrasts, which calculates the standardized differences in trait values between sister taxa descended from each node of a phylogeny and then evaluates trait correlations or regressions between these contrasts (Felsenstein, 1985
; Garland, Harvey, and Ives, 1992
). This contrast correlation is quantitatively equivalent to the correlation between the evolutionary changes that have occurred along each branch of the phylogeny (Pagel, 1993
).
In recent years, a number of proposed trait relationships in plant functional ecology have been reevaluated in a phylogenetic context, and in several cases the phylogenetically structured correlations have been reported as considerably weaker and nonsignificant (e.g., seed size in relation to establishment conditions: Kelly and Purvis, 1993
; Kelly, 1995
; stomatal density in relation to life form: Kelly and Beerling, 1995
). These reexaminations have shown that many data sets may contain only a few appropriate contrasts from a phylogenetic perspective, greatly reducing the power of the analyses. The problem is particularly compounded in comparisons between a continuous and a categorical variable (e.g., seed size vs. gap/nongap regeneration), as there may be few contrasts for the categorical trait between related species. Further loss of power occurs when taxonomic schemes (e.g., Cronquist, 1981
) are used to construct a phylogeny, because of the lack of resolution.
In contrast to these case studies, meta-analyses of published results in both animal and plant ecology suggest that there is often little quantitative difference in the correlation coefficients or regression slopes resulting from cross-species and independent contrast analyses (Ricklefs and Starck, 1996
; Price, 1997
; Ackerly, 1999a
). These cases primarily involve associations between pairs of continuous traits (e.g., leaf size and branching density, body size and home range) combined with the use of fully resolved phylogenies, both of which maximize the number of available contrasts. Recently Ackerly and Donoghue (1998)
demonstrated quantitatively that substantial discrepancies between cross-species and independent contrast correlations only arise for traits that are very highly conserved from a phylogenetic perspective, i.e., when closely related species are also very similar in the functional and ecological traits of interest.
The objectives of this study were to reevaluate previously published analyses of leaf trait relationships for the functional traits described above, together with previously unpublished data for lamina area, based on the data set for 108 species presented in Reich, Walters, and Ellsworth (1997)
and Reich et al. (1999)
. A perusal of the data suggests that these relationships hold up across different phylogenetic groups and that most traits exhibit considerable convergence among the taxa in this study (Reich et al., 1999
). However, both leaf life span and lamina area exhibit consistent differences between conifers and angiosperms that might result in correlations across species that are not observed in independent contrasts. Here, we address the following questions: (1) Which of these traits exhibit the highest levels of evolutionary convergence (i.e., similarity among distantly related species)? (2) Is there a relationship across species between lamina area and leaf functional traits (i.e., specific leaf area, leaf life span, nitrogen concentration, or gas exchange rates)? (3) To what extent are trait correlations observed across species similar or different using phylogenetically independent contrasts, based on recent molecular phylogenetic analyses of seed plants? (4) What are the consequences of alternative seed plant phylogenies, and alternative approaches to resolving uncertainty in these phylogenies, for the conduct of comparative analyses?
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
. Lamina area was measured for the entire leaf for simple-leaved species, and for the leaflet in compound-leaved species. Analyses of trait correlations focused on the relationships between leaf area and the other seven traits, as these have not been previously reported and between leaf life span and the remaining six functional traits for comparison.
Phylogenetic trees
Data were collected for 108 species representing 78 genera and 45 families of seed plants, distributed over six distinct biomes. Four species were measured twice independently either within or across distinct biomes, resulting in 112 total records (see Reich et al., 1999
).
The construction of phylogenetic trees for broad ecological surveys such as this poses various difficulties. First, the species sampled in ecological surveys rarely coincide with those used in molecular phylogenetic analyses, necessitating extensive substitution of taxa based on taxonomic affiliation at the genus and family level. Secondly, uncertainty in the phylogeny at various levels (in part due to the substitution process just mentioned) may lead to numerous polytomies, which potentially reduce the number of contrasts and the power of statistical analyses. Finally, the necessity for combining information from various sources makes it impossible to obtain consistent branch lengths, in terms of the amount of evolutionary time represented along each branch. Recognizing these problems, we used the following procedure to generate a robust set of phylogenetic trees that combine high levels of resolution with extensive sensitivity testing to examine the consequences of phylogenetic uncertainty (cf. Donoghue and Ackerly, 1996
). This procedure is spelled out in detail to provide guidance for future comparative studies addressing broad samples of seed plants.
First, the strict consensus trees from two distinct analyses of seed plant phylogeny (rbcL: Chase et al., 1993
; Rice, Donoghue, and Olmstead, 1997
; 18S: Soltis et al., 1997
) were pruned to show relationships among families represented in each analysis. Next, these trees were pruned to show only the families represented in the ecological data set in this study. Families that were present in the ecological data set but not represented in these phylogenetic analyses (four families for rbcL; 12 families for 18S) were then joined to affiliated taxa based on Chase et al. (1993)
for the smaller 18S analysis or, if not present in the Chase analysis, based on the classification presented by Bremer et al. (1998)
. The resulting rbcL-based tree had six polytomies for family-level relationships, while the 18S-based tree had four family-level polytomies (Table 1). Figure 1 shows the resulting trees for rbcL and 18S, and the correspondence between them. Species within each family were then attached to the tree, with genera within families and species within genera shown as polytomies. The four species with two independent measurements were represented as two adjacent branches at the tips of the trees. Both trees had six polytomies for generic relationships and six polytomies at the species level (Table 1).
|
|
), which resolves polytomies while preserving the remainder of the topology. This step provides a rapid approach to obtaining alternative, fully resolved phylogenies for sensitivity analyses. (2) Alternatively, "systematically resolved trees" were constructed based on extensive searches of the literature and consultation with experts for various groups to resolve the polytomies as listed in Table 1. Four polytomies remained unresolved for the rbcL tree and three of these also for 18S: three species of Vaccinium, three species of Protium, four families of the Violales (Chrysobalanaceae, Euphorbiaceae, Salicaceae, and Clusiaceae), and (rbcL only) relationships among Celastraceae, Violales, and a large clade including Fagaceae and Fabaceae. Polytomies of three and four taxa have three and 15 binary resolutions, respectively, so all possible resolutions of these polytomies produced 135 alternative phylogenies for the 18S-based tree and 405 alternatives for the rbcL-based tree, resulting in a final set of 540 systematically resolved trees for analysis.
Finally, a set of 250 fully random trees was created using MacClade's equiprobable trees algorithm, in order to evaluate the potential range of outcomes for the comparative analyses (cf. Maddison and Slatkin, 1991
; Losos and Adler, 1995
; Donoghue and Ackerly, 1996
).
Comparative methods
Comparisons of relative levels of evolutionary convergence were made using the Quantitative Convergence Index (QVI), introduced by Ackerly and Donoghue (1998)
for the analysis of continuous characters (convergence is abbreviated "V" to avoid confusion with the Consistency Index, CI; see Maddison and Maddison, 1992
). The QVI is based on linear parsimony algorithms for ordered discrete characters (see Swofford and Maddison, 1987
) and is equivalent to 1-Retention Index (Farris, 1989
); values range from 0, for traits in which closely related species are phenotypically most similar, to 1 for traits in which similar species are distantly related and closely related species are most dissimilar (see Ackerly, 1999a
). Trait relationships were evaluated based on correlation analyses, using phylogenetically independent contrasts (Felsenstein, 1985
) based on the various alternative phylogenies. In order to evaluate the robustness of the results, branch lengths were calculated with all lengths assumed to be equal, or following Grafen's (1989)
method, which assigns branch lengths based on the number of species in each clade such that the total branch lengths from the root to each terminal taxon are equal.
Comparative analyses were conducted with ACAP v2 (Ackerly and Donoghue, 1998
), which calculates QVI values and independent contrasts and their correlations over multiple fully resolved phylogenetic trees. Independent contrasts represent differences between the trait values of two sister taxa. As a result each contrast may be negative or positive depending on the direction chosen to calculate the difference between taxa, but the same direction must be used for all traits at each node. For graphical presentation, the contrasts in the trait plotted along the horizontal axis are all made positive by convention, and those for the trait on the vertical axis are positive or negative depending on whether the changes in the two traits covary positively or negatively, respectively (Garland, Harvey, and Ives, 1992
). In addition, due to the symmetry arising from the arbitrary direction of contrasts all correlation and regression analyses must be centered on the origin. In this paper, we focus on correlations rather than regressions to avoid a priori assignment of independent and dependent variables, but it is important to note that these correlations are centered on the origin, not on the means of the two variates as is customary.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). QVI values were higher within angiosperms considered separately, reinforcing this conclusion (Table 2; values were not calculated for conifers alone because they are paraphyletic in the rbcL phylogeny). QVI values were virtually identical over the 540 alternative phylogenies based on the 18S and rbcL trees (range of <0.05 between the lowest and highest values for all eight characters). Note that for this analysis, we have tested the null hypothesis that relationships among species are random with respect to their functional traits, leading to very high null expectations of convergence levels. In contrast Ackerly and Donoghue (1998)
randomized reconstructed changes over the branches of the tree, which leads to a null expectation of greater similarity among species, and much lower QVI values, as the phylogenetic structure of trait evolution is conserved. The difference between these two approaches to significance testing of evolutionary convergence is currently being further explored (D. Ackerly and D. Schwilk, unpublished data).
|
|
|
0.2 in all cases). In particular, in the association between leaf size and leaf life span, a negative correlation of -0.424 across species disappeared completely using independent contrasts (see Fig. 2). This occurred because the interspecific correlation reflects the differences in both traits between angiosperms and conifers (Fig. 3B), as reflected in their low QVI values. In the analysis of independent contrasts, these differences are revealed in the large contrast between the Pinaceae branch and the angiosperms, while there is no significant correlation between these traits within each of these groups (Fig. 3E). In contrast to SLA, leaf size was similar for angiosperms across the entire range of leaf life span, and leaves were larger than comparable conifers.
|
. The results were also very robust over the alternative phylogenies constructed for this analysis. For example, for leaf life span vs. specific leaf area, the mean correlations over the systematically resolved trees based on rbcL (N = 435) and 18S (N = 105) were -0.64 and -0.62, respectively, with a range of ±0.005 in each case (Table 4). Results were only slightly more variable in the two sets of 250 randomly resolved trees, with slightly lower means and a range of ±0.05. Similar patterns were observed for leaf life span vs. leaf size. These analyses were also repeated on an angiosperm phylogeny based on a combined analysis of rbcL sequences and morphology (Nandi, Chase, and Endress, 1998
), which appeared after this study had been completed, and the results were very similar to the rbcL-based analyses reported here (results not shown). In contrast, the average evolutionary correlations over random trees were virtually identical to the cross-species correlation (see Abouheif, 1998
). For leaf life span vs. SLA, the results over random trees barely overlapped the rbcL and 18S results, while for life span vs. leaf size the results were completely divergent, as expected due to the large differences between the cross-species and independent-contrast results in this case.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Reich et al., 1999
) are supported from a phylogenetic perspective. In addition, there were weak associations among leaf size and function, but several of these were primarily due to differences between conifers and angiosperms, not to relationships observed within each of these groups. The analysis of quantitative convergence levels also demonstrates significant phylogenetic structuring of these physiological traits, especially leaf size and leaf life span (Table 2). This illustrates that even in this phylogenetically diverse set of species there is a pattern of functional similarity among related species. This pattern is primarily driven by the overall similarities among conifers and angiosperms, relative to the divergence between these two groups; QVI values within angiosperms (Table 2) indicate considerable convergence in this subset of flowering plant species.
Various studies have documented interspecific leaf size variation both within and across habitats in association with light levels, nutrient and moisture availability, and temperature (Givnish, 1987
), and it was suggested that parallel variation in leaf ecophysiology would lead to associations among these traits. This hypothesis is partially supported by these data, in particular by the positive correlation among angiosperm species and among independent contrasts for the entire data set between leaf size and specific leaf area, i.e., smaller leaves tended to be thicker. Of all the traits, leaf size shows the greatest variation among communities in this study, with the smallest mean leaf size in the dry New Mexico site, and the largest mean size in Venezuelan rain forest (cf. Reich et al., 1999
). However, the species in this study were not sampled to compare mean trait differences between communities, and it is possible that random sampling of species from contrasting habitats would show stronger associations among leaf size and ecophysiological traits than those observed here. On the other hand, it is understandable that leaf size and function relationships, which reflect parallel responses to abiotic conditions, will be weaker than the functional relationships among specific leaf area, assimilation, nitrogen, and leaf life span, which reflect adaptive and/or physiological constraints on coordinated leaf function. Leaf size exhibits considerable variation among ecologically similar species and may be associated with other aspects of plant function. For example, in tropical rain forests variation in leaf size among pioneer and late-successional species is much greater than average differences between these groups (Ackerly, 1996
). Leaf size is strongly associated with canopy architecture and branching morphology, and to a lesser extent with reproductive morphology (White, 1983
; Midgley and Bond, 1989
; Ackerly and Donoghue, 1998
), and these traits may influence leaf size evolution independently of ecophysiological function. Despite the long-standing interest in leaf size, variation among species within habitats is still poorly explained (cf. Givnish, 1987
).
The results of the comparative analysis of leaf functional traits do provide strong additional support for broad patterns of convergence in leaf trait correlations. In addition, this study highlights several methodological points regarding comparative analyses of functional and ecological traits. As in many comparative ecological studies, the species in this data set are drawn from a broad phylogenetic spectrum. In this analysis, we constructed highly resolved "supertrees" for these species by substituting the species of interest onto appropriate locations of seed plant phylogenies based on rbcL and 18S sequences, pruning extraneous portions of the trees, and then resolving the numerous polytomies based on relevant published studies (see Table 1). This relatively time-consuming procedure left us with just four unresolved polytomies, but even this small number led to over 500 distinct phylogenies for examination. However, the results of both the convergence and correlation analyses were extremely similar over all of these trees, suggesting that reasonable results could be accomplished by picking any one resolution at random. Donoghue and Ackerly (1996)
also showed that the variations among the numerous alternative parsimonious trees of the seed plant rbcL phylogeny (Rice, Donoghue, and Olmstead, 1997
) have little influence on analyses of trait correlations.
Of more practical relevance, results of this study were also fairly similar when the families, genera, and species were pasted onto the rbcL and 18S trees and the resulting polytomies were resolved at random in MacClade. This is a very rapid procedure which provides fully bifurcating trees, maximizing the power of comparative analyses. The results of this study suggest that alternative random resolutions of uncertain portions of the phylogeny (cf. Losos and Adler, 1995
) are superior to the treatment of each polytomy as a single contrast (Pagel, 1992
), which results in a rapid loss of degrees of freedom and power. To facilitate the construction of phylogenies for comparative analyses, a simple utility has been written that prunes and pastes branches in order to partially automate this process (D. Ackerly and B. Bennett, unpublished data). Although much remains to be learned about seed plant phylogeny, the current state of knowledge coupled with these approaches to phylogenetic uncertainty provide ample basis to conduct phylogenetically structured analyses of ecological data. The rationalization that phylogenies are still too crude for these analyses is not justified.
The results of this study also illustrate that in many cases the results of phylogenetically structured correlation analyses are quite similar to traditional cross-species correlations (Fig. 2; cf. Ackerly, 1999a
). However, it is very important to pay attention to those cases in which results of the two approaches are quantitatively different from each other, especially when the difference is not due to a loss of power as discussed in the Introduction. In this study we found an example of such divergence in the relationship between leaf life span and leaf size (Fig. 3B, E). The negative relationship between these two traits results entirely from the divergence between the angiosperms and the conifers, and there is no relationship observed within either group nor in the analysis of independent contrasts for the entire set of species. How should this difference in the results be interpreted? One view is that evolutionary changes in trait values along particular phylogenetic lineages may occur for many reasons or even at random, and the fact that both leaf size and leaf life span exhibit a shift along the branch leading to the angiosperms does not provide statistical evidence of a biological association between the two traits. It represents a single event, so there is no support for a repeated pattern of correlated evolutionary change. From this perspective, the correlation observed at the species level might be considered spurious and biologically meaningless, and the results of cross-species analyses positively misleading. From another view, the changes in both of these traits may in fact be related to the shifts in physiology and life history accompanying the evolution of angiosperms (cf. Bond, 1989
). Larger laminas and shorter leaf life span of angiosperms may be two functional traits associated with increased hydraulic conductance, enhanced photosynthetic rates, and rapid growth, relative to conifers as a whole. And, as Sprugel (1989)
argued, the combination of small leaves and evergreenness in conifers may in fact be related to their success in cold environments. In this sense, the divergence event between these groups does represent an instance of biologically significant coordinated evolutionary change, which at the very least has influenced the subsequent success of each group. Furthermore, the maintenance of the differences between the groups may reflect continued stabilizing selection within each group, rather than "phylogenetic inertia" due to lack of genetic variation or directional selection (Hansen, 1997
). Each of these traits exhibits high levels of convergence within angiosperms and conifers, suggesting rapid evolution and minimal constraints for response to selective pressures. While this may be true, it is still worth noting that there is no relationship between leaf size and leaf life span within conifers or angiosperms, supporting the notion that there is no intrinsic biological linkage between evolutionary changes in these two traits. The choice between the two alternative views above cannot be made simply based on the quantitative results themselves. Ultimately, the biological and evolutionary interpretation of these correlative patterns depends on both functional understanding of the traits, as well as a philosophical decision about the significance and interpretation of unique historical events (Donoghue, 1989
).
One final aspect of this analysis that deserves attention is the problem of species sampling in relation to the phylogeny. The species in this study were drawn from a very broad phylogenetic range, with individual species serving as representatives of lineages with hundreds or thousands of taxa. On the one hand, this sampling means that the patterns observed may be safely inferred to apply across all seed plants; they are not idiosyncrasies of one family or one biological community (cf. Reich, Walters, and Ellsworth, 1997
). On the other hand, this sparse sampling from a phylogenetic perspective seriously draws into question the meaning of the trait reconstructions used to calculate convergence indices and independent contrasts. For example, Acomastylis (= Geum) rosii and Prunus serotina are the only two Rosaceae in this study and appear as sister taxa in the phylogeny, and they also exhibit relatively similar functional traits (deciduous leaves with moderately high SLA and assimilation rates). Based only on this sample, these values are interpreted as phylogenetically conserved in these two taxa and representative of the Rosaceae, contributing to lower convergence values. However, there is no question that each of these species has numerous closer relatives with contrasting functional traits (e.g., Prunus serotina is more closely affiliated with evergreen chaparral species, such Prunus ilicifolia and Adenostema fasciculata; see Morgan, Soltis, and Robertson, 1994
), so the similarity of the two species in this study is clearly misleading as a representation of Rosaceae as a whole. Analyses of independent contrasts are robust to random species sampling, but may be biased if the associations in question are stronger within the communities sampled in a particular study, relative to the entire evolutionary distribution of the traits (D. Ackerly, unpublished data). Such biases are probably minimized in these data as they have been drawn from six widely contrasting habitats. However, reconstructions of the sequence and timing of evolutionary changes, especially in these relatively labile physiological traits, require greater study of closely related species to complement the broad comparative approach employed here.
|
FOOTNOTES |
|---|
4
Author for correspondence (dackerly{at}stanford.edu
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Ackerly, D. D. 1996 Canopy structure and dynamics: integration of growth processes in tropical pioneer trees. In S. S. Mulkey, R. L. Chazdon, and A. P. Smith [eds.], Tropical forest plant ecophysiology, 619–658. Chapman and Hall, London.
———. 1999a Phylogeny and the comparative method in plant functional ecology. In M. Press and J. D. Scholes [eds.], Advances in physiological plant ecology, in press. Blackwell Scientific, Oxford.
———. 1999b Self-shading, carbon gain and leaf dynamics: a test of alternative optimality models. Oecologia 119: 300–310.[CrossRef][ISI]
———, and M. J. Donoghue. 1998 Leaf size, sapling allometry, and Corner's rules: a phylogenetic study of correlated evolution in maples (Acer). American Naturalist 152: 767–791.[CrossRef][ISI]
Bazzaz, F. A. 1979 The physiological ecology of plant succession. Annual Review of Ecology and Systematics 10: 351–371.
Bond, W. J. 1989 The tortoise and the hare: ecology of angiosperm dominance and gymnosperm persistence. Biological Journal of the Linnean Society 36: 227–249.[ISI]
Bremer, K. 1994 Asteraceae: cladistics and classification. Timber Press, Portland, OR.
———. 1996 Major clades and grades of the Asteraceae. In D. J. N. Hind and H. J. Beentje [eds.], Compositae: systematics. Proceedings of the International Compositae Conference, Kew, 1994, vol. 1, 1–7. Royal Botanic Gardens, Kew.
———, et al. 1998 An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 12: 438–445.
Chase, M., et al. 1993 Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 528–580.[CrossRef][ISI]
Chiariello, N. 1984 Leaf energy balance in the wet lowland tropics. In E. Medina, H. A. Mooney, and C. Vázquez-Yanes [eds.], Physiological ecology of plants of the wet tropics, 85–98. Dr. W. Junk, The Hague.
Cornellison, J. H. C., P. Castro-Diez, and A. L. Carnelli. 1998 Variation in relative growth rate among woody species. In H. Lambers, H. Poorter, and M. M. I. Van Vuuren [eds.], Inherent variation in plant growth: physiological mechanisms and ecological consequences, 363–392. Backhuys Publishers, Leiden, The Netherlands.
Cronquist, A. 1981 An integrated system of classification of flowering plants. Columbia University Press, New York, NY.
Diemer, M., C. Körner, and S. Prock. 1992 Leaf life spans in wild perennial herbaceous plants: a survey and attempts at a functional interpretation. Oecologia 89: 10–16.[CrossRef][ISI]
Donoghue, M. J. 1989 Phylogenies and the analysis of evolutionary sequences, with examples from the seed plants. Evolution 43: 1137–1156.[CrossRef][ISI]
———, and D. D. Ackerly. 1996 Phylogenetic uncertainties and sensitivity analyses in comparative biology. Philosophical Transactions of the Royal Society of London, Series B 351: 1241–1249.[CrossRef]
Doyle, J. J. 1995 DNA data and legume phylogeny: a progress report. In M. D. Crisp and J. J. Doyle [eds.], Advances in legume systematics, 11–30. Whitstable Litho, Ltd., Whitstable, Kent.
Farris, J. S. 1989 The retention index and the rescaled consistency index. Cladistics 5: 417–419.[ISI]
Felsenstein, J. 1985 Phylogenies and the comparative method. American Naturalist 125: 1–15.
Garland, T., Jr., P. H. Harvey, and A. R. Ives. 1992 Procedures for the analysis of comparative data using phylogenetically independent contrasts. Systematic Biology 41: 18–32.[CrossRef][ISI]
Givnish, T. J. 1987 Comparative studies of leaf form: assessing the relative roles of selective pressures and phylogenetic constraints. New Phytologist 106: S131–S160.
———, and G. J. Vermeij. 1976 Sizes and shapes of liane leaves. American Naturalist 110: 743–778.[CrossRef][ISI]
Grafen, A. 1989 The phylogenetic regression. Philosophical Transactions of the Royal Society of London, Series B 326: 119–157.[ISI][Medline]
Hamann, O. 1979 On climatic conditions, vegetation types, and leaf size in the Galápagos Islands. Biotropica 11: 101–122.[CrossRef][ISI]
Hansen, T. F. 1997 Stabilizing selection and the comparative analysis of adaptation. Evolution 51: 1341–1351.[CrossRef][ISI]
Harvey, P. H., and M. Pagel. 1991 The comparative method in evolutionary biology. Oxford University Press, Oxford.
Jansen, R. K., and K. Kim. 1996 Implications of chloroplast DNA data for the classification and phylogeny of the Asteraceae. In D. J. N. Hind and H. J. Beentje [eds.], Compositae: systematics. Proceedings of the International Compositae Conference, Kew, 1994, vol. 1, 317–339. Royal Botanic Gardens, Kew.
Kato, H., R. Terauchi, F. H. Utech, and S. Kawano. 1995 Molecular systematics of the Trilliaceae sensu lato as inferred from rbcL sequence data. Molecular Phylogenetics and Evolution 4: 184–193.[CrossRef][ISI][Medline]
Kelly, C. K. 1995 Seed size in tropical trees: a comparative study of factors affecting seed size in Peruvian angiosperms. Oecologia 102: 377–388.[CrossRef][ISI]
———, and D. J. Beerling. 1995 Plant life form, stomatal density and taxonomic relatedness: a reanalysis of Salisbury (1927). Functional Ecology 9: 422–431.[CrossRef][ISI]
———, and A. Purvis. 1993 Seed size and establishment conditions in tropical trees: on the use of taxonomic relatedness in determining ecological patterns. Oecologia 94: 356–360.[CrossRef][ISI]
Kikuzawa, K. 1991 A cost-benefit analysis of leaf habit and leaf longevity of trees and their geographical pattern. American Naturalist 138: 1250–1263.[CrossRef][ISI]
Körner, C., M. Neumayer, S. P. Menendez-Riedl, and A. Smeets-Scheel. 1989 Functional morphology of mountain plants. Flora 182: 353–383.[ISI]
Kron, K. A. 1996 Phylogenetic relationships of Empetraceae, Epacridaceae, Ericaceae, Monotropaceae, and Pyrolaceae: evidence from nuclear ribosomal 18S sequenced data. Annals of Botany 77: 293–303.
Lavin, M. 1995 Tribe Roninieae and allies: model groups for assessing Early Tertiary northern latitude diversification of tropical legumes. In M. D. Crisp and J. J. Doyle [eds.], Advances in legume systematics, 141–160. Whitstable Litho Ltd., Whitstable, Kent.
Losos, J. B., and F. R. Adler. 1995 Stumped by trees? A generalized null model for patterns of organismal diversity. American Naturalist 145: 329–342.[CrossRef][ISI]
Maddison, W. P., and D. R. Maddison. 1992 MacClade: analysis of phylogeny and character evolution, version 3. Sinauer, Sunderland, MA.
———, and M. Slatkin. 1991 Null models for the number of evolutionary steps in a character on a phylogenetic tree. Evolution 45: 1184–1197.[CrossRef][ISI]
Manos, P. S., and K. P. Steele. 1997 Phylogenetic analyses of higher Hamamelidae based on plastid sequence data. American Journal of Botany 84: 1407–1419.[Abstract]
Midgley, J., and W. Bond. 1989 Leaf size and inflorescence size may be allometrically related traits. Oecologia 78: 427–429.[CrossRef][ISI]
Morgan, D. R., D. E. Soltis, and K. R. Robertson. 1994 Systematic and evolutionary implications of rbcL sequence variation in Rosaceae. American Journal of Botany 81: 890–903.[CrossRef][ISI]
Nandi, O. I., M. W. Chase, and P. K. Endress. 1998 A combined cladistic-analysis of angioserpms using rbcL and non-molecular data sets. Annals of the Missouri Botanical Garden 85: 137–212.
Niinemets, Ü,. and K. Kalevi. 1994 Leaf weight per area and leaf size of 85 Estonian woody species in relation to shade tolerance and light availability. Forest Ecology and Management 70: 1–10.[CrossRef][ISI]
Olmstead, R. G., B. Bremer, K. M. Scott, and J. D. Palmer. 1993 A parsimony analysis of the Asteridae sensu lato based on rbcL sequences. Annals of the Missouri Botanical Garden 80: 700–722.[CrossRef][ISI]
Pagel, M. D. 1992 A method for the analysis of comparative data. Journal of Theoretical Biology 156: 431–442.[CrossRef]
———. 1993 Seeking the evolutionary regression coefficient: an analysis of what comparative methods measure. Journal of Theoretical Biology 164: 191–205.[CrossRef][ISI][Medline]
Parkhurst, D., and O. Loucks. 1972 Optimal leaf size in relation to environment. Journal of Ecology 60: 505–537.[CrossRef]
Price, T. 1997 Correlated evolution and independent contrasts. Philosophical Transactions of the Royal Society of London, Series B 352: 519–529.[CrossRef][ISI][Medline]
Reich, P. B. 1993 Reconciling apparent discrepancies among studies relating life span, structure and function of leaves in contrasting plant life forms and climates: ‘the blind men and the elephant retold’. Functional Ecology 7: 721–725.[CrossRef][ISI]
———, D. S. Ellsworth, M. B. Walters, J. M. Vose, C. Gresham, J. C. Volin, and W. D. Bowman. 1999 Generality of leaf trait relationships: a test across six biomes. Ecology 80: 1955–1969.[CrossRef][ISI]
———, M. B. Walters, and D. S. Ellsworth. 1992 Leaf life-span in relation to leaf, plant, and stand characteristics among diverse ecosystems. Ecological Monographs 62: 365–392.[CrossRef]
———, ———, and ———. 1997 From tropics to tundra: global convergence in plant functioning. Proceedings of the National Academy of Sciences, USA 94: 13730–13734.
Rice, K. A., M. J. Donoghue, and R. G. Olmstead. 1997 Analyzing large data sets: rbcL 500 revisited. Systematic Biology 46: 554–563.[CrossRef][ISI][Medline]
Ricklefs, R. E., and J. M. Starck. 1996 Applications of phylogenetically independent contrasts: a mixed progress report. Oikos 77: 167–172.[CrossRef][ISI]
Smith, D. E., and A. S. Klein. 1994 Phylogenetic inferences on the relationship of North American and European Picea species based on nuclear ribosomal 18S sequences and the internal transcribed spacer 1 region. Molecular Phylogenetics and Evolution 3: 17–26.[CrossRef][Medline]
Smith, R. L., and K. J. Sytsma. 1990 Evolution of Populus nigra (sect. Aigeiros): introgressive hybridization and the chloroplast contribution of Populus alba (sect. Populus). American Journal of Botany 77: 1176–1187.[CrossRef][ISI]
Smith, W. K., and C. A. Brewer. 1994 The adaptive importance of shoot and crown architecture in conifer trees. American Naturalist 143: 528–532.[CrossRef][ISI]
Soltis, D. E., et al. 1997 Angiosperm phylogeny inferred from 18S ribosomal DNA sequences. Annals of the Missouri Botanical Garden 84: 1–49.
Sprugel, D. G. 1989 The relationship of evergreenness, crown architecture and leaf size. American Naturalist 133: 465–479.[CrossRef][ISI]
Swensen, S. M. 1996 The evolution of actinorhizal symbioses: evidence for multiple origins of the symbiotic association. American Journal of Botany 83: 1503–1512.[CrossRef][ISI]
Swofford, D. L., and W. P. Maddison. 1987 Reconstructing ancestral states under Wagner parsimony. Mathematical Biosciences 87: 199–229.[CrossRef][ISI]
Tsumura, Y., K. Yoshimura, N. Tomaru, and K. Ohba. 1995 Molecular phylogeny of conifers using RFLP analysis of PCR-amplified specific chloroplast genes. Theoretical and Applied Genetics 91: 1222–1236.[ISI]
van der Werf, A., et al. 1998 The importance of relative growth rate and associated traits for competition between species during vegetation succession. In H. Lambers, H. Poorter, and M. M. I. Van Vuuren [eds.], Inherent variation in plant growth. Physiological mechanisms and ecological consequences, 489–502. Backhuys Publishers, Leiden, The Netherlands.
White, P. S. 1983 Corner's rules in eastern deciduous trees: allometry and its implications for the adaptive architecture of trees. Bulletin of the Torrey Botanical Club 110: 203–212.[CrossRef][ISI]
This article has been cited by other articles:
|
|
 |
|
  D. L. ROYER NUTRIENT TURNOVER RATES IN ANCIENT TERRESTRIAL ECOSYSTEMS Palaios, July 1, 2008; 23(7): 421 - 423. [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Knight and J. M. Beaulieu Genome Size Scaling through Phenotype Space Ann. Bot., April 1, 2008; 101(6): 759 - 766. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Royer, L. Sack, P. Wilf, C. H. Lusk, G. J. Jordan, U. Niinemets, I. J. Wright, M. Westoby, B. Cariglino, P. D. Coley, et al. Fossil leaf economics quantified: calibration, Eocene case study, and implications Paleobiology, December 1, 2007; 33(4): 574 - 589. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. G. Abrahamson Leaf traits and leaf life spans of two xeric-adapted palmettos Am. J. Botany, August 1, 2007; 94(8): 1297 - 1308. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Niinemets, A. Portsmuth, D. Tena, M. Tobias, S. Matesanz, and F. Valladares Do we Underestimate the Importance of Leaf Size in Plant Economics? Disproportional Scaling of Support Costs Within the Spectrum of Leaf Physiognomy Ann. Bot., August 1, 2007; 100(2): 283 - 303. [Abstract] [Full Text] [PDF]
|
|
100