HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

First published online December 19, 2008; doi:10.3732/ajb.0800126 American Journal of Botany 96: 366-381 (2009) © 2009 Botanical Society of America, Inc.

What's this?

This Article Abstract Full Text (PDF) All Versions of this Article: ajb.0800126v1 96/1/366    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (8) Citing Articles via Google Scholar Google Scholar Articles by Crepet, W. L. Articles by Niklas, K. J. Search for Related Content PubMed Articles by Crepet, W. L. Articles by Niklas, K. J. Agricola Articles by Crepet, W. L. Articles by Niklas, K. J. Social Bookmarking                            What's this?

Darwin’s second 'abominable mystery': Why are there so many angiosperm species?¹

Department of Plant Biology, Cornell University, Ithaca, New York 14853 USA

Received for publication 3 April 2008. Accepted for publication 13 October 2008.

ABSTRACT

The rapid diversification and ecological dominance of the flowering plants beg the question "Why are there so many angiosperm species and why are they so successful?" A number of equally plausible hypotheses have been advanced in response to this question, among which the most widely accepted highlights the mutually beneficial animal–plant relationships that are nowhere better developed nor more widespread than among angiosperm species and their biotic vectors for pollination and dispersal. Nevertheless, consensus acknowledges that there are many other attributes unique to or characteristic of the flowering plants. In addition, the remarkable coevolution of the angiosperms and pollination/dispersal animal agents could be an effect of the intrinsic adaptability of the flowering plants rather than a primary cause of their success, suggesting that the search for underlying causes should focus on an exploration of the genetic and epigenetic mechanisms that might facilitate adaptive evolution and speciation. Here, we explore angiosperm diversity promoting attributes in their general form and draw particular attention to those that, either individually or collectively, have been shown empirically to favor high speciation rates, low extinction rates, or broad ecological tolerances. Among these are the annual growth form, homeotic gene effects, asexual/sexual reproduction, a propensity for hybrid polyploidy, and apparent "resistance" to extinction. Our survey of the literature suggests that no single vegetative, reproductive, or ecological feature taken in isolation can account for the evolutionary success of the angiosperms. Rather, we believe that the answer to Darwin’s second "abominable mystery" lies in a confluence of features that collectively make the angiosperms unique among the land plants.

Key Words: agamospermy • angiosperm diversification • annual growth form • floral trait evolution • homeotic genes • polyploidy • species morphospace

Darwin’s "abominable mystery" has come to refer to a problem about timing and origin: When and from what ancient plant lineage did the angiosperms first evolve? At the time Darwin faced this problem comparatively little was known about the plant fossil record or the stratigraphy and taxonomic affiliations of those few angiosperm fossils to which Darwin had access. However, there is a second dimension to the "abominable mystery" and perhaps one that more precisely relates to Darwin’s famous phrase (Darwin, 1871, 1903): Why are the angiosperms so species-rich and ecologically successful? While we still do not understand angiosperm origins, we know more about the tempo and mode of flowering plant evolution today than Darwin did owing to the availability of compendia of standing paleospecies diversity and character state diversification as well as insights into angiosperm phylogeny gained from molecular biology (e.g., Dilcher, 1974, 2000; Dilcher et al., 1976; Niklas et al., 1980, 1983; Dilcher and Crane, 1984; Sun et al., 1998; Soltis et al., 2000; Chase, 2004; Friis et al., 2006). It is nevertheless fair to say that we are also still very much perplexed about Darwin’s second "abominable mystery."

To be sure, it is not easy to define precisely what is meant by "biologically successful," but by virtually every yardstick, the flowering plants can be said to lay claim to it. Nearly 90% of all extant land plant species are angiosperms (Fig. 1); they span every known plant body-plan and growth form, from annual nonwoody herbs and rhizomatous growth forms to woody perennial forbs, lianas, and trees; and they occupy every biome and habitat across the face of the planet, including intertidal communities otherwise dominated by the algae. This second "abominable mystery" has been the subject of intensive neontological investigation aimed principally at identifying and evaluating those unique aspects of angiosperms that might explain their success. The search for explanations has encompassed physiological, molecular, genetic, and epigenetic mechanisms and has generally used a reductionist approach to identify core underlying attributes among extant species. Likewise, tests for the hypotheses emerging from this approach have generally rested on the extent to which extant species diversity correlates with the presence or absence of the key functional traits purported to foster or impede species diversification within specific lineages (e.g., Burger, 1981; Stebbins, 1981; Tiffney, 1984, 1986; Eriksson and Bremer, 1992). A rich and growing literature dedicated to this double-pronged approach has made considerable progress in evaluating the extent to which angiosperm attributes affect the potential to speciate (e.g., Fenster et al., 2004). Nevertheless, the literature remains controversial for at least two reasons: (1) no single explanation appears to fit all angiosperm taxa, partly because the distribution of functional traits is convergent among angiosperm families that are otherwise very dissimilar (e.g., adaptations for pollination in grasses vs. orchids), and (2) correlation analyses using extant species numbers neglect the effects of extinction such that current species diversity sheds no light on past diversification trends.

View larger version (12K): [in this window] [in a new window]   Fig. 1. Percentage distribution of extant species assigned to different embryophyte plant groups. For convenience, some species groupings are polyphyletic and thus represent a grade-level of reproductive organization (e.g., gymnosperms). Data taken from primary literature estimates or species compendia (see Niklas, 1997).

Despite this growing body of paleontological knowledge and the development of powerful methodologies, virtually every hypothesis advanced to explain the success of the angiosperms remains controversial due to questions about the fundamental premises on which it rests and because each is vulnerable to distortions associated with an improper selection of the fossils used to calibrate phyletic or molecular models (Graur and Martin, 2004; Sanderson et al., 2004; Gandolfo et al., 2008; Nixon, 2008). For example, discrepancies between the observed first-appearance of fossil taxa and molecular clock-based models (e.g., Wikström et al., 2001) are often ascribed, directly or by implication, to a lag between the origin of new taxa and their deposition in basin sediments (e.g., Graur and Martin, 2004; Sanderson et al., 2004). However, studies of contemporary depositional basin environments provide no evidence for temporal lags of the magnitude sufficient to explain away discrepancies between predicted and observed first appearances in the fossil record (Burnham, 2008), which detracts from the credibility of some molecular clock based estimates of the timing of key angiosperm innovations.

In light of all these concerns, the philosophy adopted in this review of Darwin’s second "abominable mystery" is pluralistic in terms of drawing on both paleo- and neontological data, examining the extent to which the first appearances of functional trait innovations correlate with subsequent changes in standing species diversity, and discussing those attributes that appear to promote speciation or delay extinction. Examining patterns observed for both "the living and the dead" is essential because current species diversity reflects the effects of extinction as much as speciation and because "correlation" in the fossil record of plants or animals provides no evidence for "cause and effect." A pluralistic philosophy has additional merit in this context because there is little doubt that there are separate and possibly complementary reasons for angiosperm species diversification patterns (Raven, 1977; Stebbins, 1981; Crepet, 1984, 2008; Tiffney, 1984; Eriksson and Bremer, 1992; Tiffney and Mazer, 1995; Verdú, 2002; Fenster et al., 2004). Finally, space precludes a synoptic treatment of all attributes that might account for the success of the flowering plants, such as a propensity to form polyploids and hybrid species, many of which have been treated recently and comprehensively elsewhere (see Arnold, 1997; Rieseberg, 2001; Baack and Rieseberg, 2007; Rieseberg and Willis, 2007). For these reasons, we focus on a few features that individually or collectively appear to have contributed to delayed extinction, rapid speciation, or intense ecological niche partitioning.

A WIDE SPECTRUM OF HYPOTHESES

Numerous hypotheses have been advanced to explain why there are so many angiosperm species compared to other terrestrial or aquatic plants and to explain why the angiosperms are so widespread ecologically (Raven, 1977; Regal, 1977; Doyle, 1978; Niklas et al., 1980, 1983; Burger, 1981; Stebbins, 1981; Crepet, 1984, 1996, 2008; Tiffney, 1984, 1986; Eriksson and Bremer, 1992; Stebbins, 1992; Waser, 1998; Grimaldi, 1999; Magallón et al., 1999; Verdú, 2002; Feild et al., 2004; Fenster et al., 2004; Grimaldi and Engel, 2005; Friis et al., 2006; Bomblies and Weigel, 2007). Some of these draw attention to the correlations between diverse angiosperm taxa and their animal pollinators, including vertebrates (Raven, 1977). And the suggestion that insect pollinators in particular had a significant role in angiosperm diversification actually extends to the very time when the phrase "abominable mystery" appeared in the famously quoted letter of 22 July 1879 from Charles Darwin to J. D. Hooker (in which Darwin mentions that "Saporta believes that there was an astonishingly rapid development of the high plants"(in context—angiosperms), "as soon [as] flower frequenting insects were developed" (Darwin, 1903 pp. 20, 21; see Saporta and Marion, 1881; for the complete letter, see Friedman, 2009, pp. 5–21, in this issue). Other hypotheses assess the potential significance of biotic seed dispersers (Tiffney, 1984, 1986); still others examine the possible synergistic effects among multiple factors (e.g., Raven, 1977; Regal, 1977; Crepet, 1984, 2008; Tiffney, 1984; Eriksson and Bremer, 1992), whereas other evolutionary biologists suggest that epigenetic versatility lies at the heart of angiosperm success. In addition, consideration has been given to functional traits and environments that minimize the probability of extinction, as well as to those that may promote speciation (Tiffney and Mazer, 1995; Niklas, 1997). More recently, hybrid necrosis presents another angiosperm characteristic that might promote the formation of new species and preserve the integrity of existing ones (Bomblies and Weigel, 2007). And traditionally held hypotheses, such as linking angiosperm success to various factors including pollinator specificity and efficiency following the model first proposed by V. Grant (1949), have been challenged (e.g., Midgley and Bond, 1991; Waser et al., 1996; Waser, 1998; Gorelick, 2001) in the face of apparent correlations between hyperdiverse taxa and specialized insect pollinators (e.g., orchids and euglossine bees; Dressler, 1968, 1982). Moreover, long-held views on pollination syndromes, fundamental to models proposing a relationship between species diversity and pollinator specificity, have come into question in a recent explosion of pollination biology literature (e.g., Herrera, 1996; Fenster et al., 2004; Raguso, 2008). Careful field studies have in certain instances revealed more relaxed pollinator specificity relative to floral syndromes challenging the "filtering" nature of such syndromes and their possible evolutionary significance (e.g., Waser et al., 1996; Waser, 1998) even while new "filters" are being discovered (Raguso, 2008). These investigations have focused attention on other angiosperm characteristics and relationships to explain the maintenance of divergent floral types. Such studies have unleashed publications dealing with a "specialization–generalization" debate that has flouted conventional thinking in pollination biology (Raguso, 2008). This turmoil, which reflects an increase in the number and quality of careful pollination studies in a field that was shy of such studies for years, remains very much unresolved because it is unclear if different "functional groups" (i.e., groups of unrelated pollinators with similar structural and thus functional characteristics) could have similar potentials for enhancing speciation in flowering plants (Fenster et al., 2004). Additional complications are presented by the recent discovery that at least some angiosperms have the ability to balance chemical attractants and repellants in ways that optimize pollinator visits, apparently, to maximize seed set (Kessler et al., 2008). This phenomenon, if widespread in angiosperms, might also contribute to their relative diversity.

Most hypotheses about angiosperm diversification fall into one of three camps: (1) those that have focused on vegetative attributes, which can confer ecological and thus evolutionary advantages, i.e., diverse organographic morphology and anatomy, rapid growth rates coupled with high hydraulic conductivity, and the capacity for extensive phenotypic plasticity; (2) those that draw attention on the importance of reproductive features, i.e., floral display and morphology, pollination syndromes employing nondestructive biotic as well as abiotic vectors, double fertilization, endosperm formation and rapid embryogenesis, and the benefits of broadcasting seeds by means of edible fruits; and (3) those that take a more pluralistic perspective by pointing out that the flowering plants manifest a constellation of structural or genetic and chemical attributes, which collectively confer vegetative and reproductive advantages as well as considerable plasticity.

As noted, while no single hypothesis advanced to account for angiosperm success has met with universal acceptance, it is fair to say that the most influential, intensively investigated, and widely accepted are those that have pointed to the strong and often very complex biological interconnections that have evolved between the flowering plants and animal species that facilitate pollination or seed/fruit dispersal. This focus is entirely justified given the large body of neontological studies that have documented numerous chemical, anatomical, morphological, and phenological plant adaptations that attract specific animal vectors for pollination and long-distance disseminule dispersal and a paucity of insect-pollinated extant gymnosperm species (e.g., Norstog et al., 1986). Likewise, paleontological studies indicate that plant–animal relationships predating the appearance of the flowering plants were, with few exceptions (generally inferred on the basis of pollen size and structure or evidence of phytophagy; see Taylor and Millay, 1979; Taylor and Scott, 1983; LaBandeira et al., 1994), either neutral or mutually antagonistic. Importantly, the ability of angiosperms to establish beneficial relationships with animals involves the capacity to evolutionarily modify many characters that are either absent or rare in each of the many other plant lineages. For example, many pollination syndromes involve adaptive chemical, anatomical, and morphological modifications of vegetative as well as reproductive organs.

Seen in this light, the remarkable coevolutionary history of the angiosperms and their animal compatriots can be interpreted as more of a manifestation of the intrinsic adaptability of the flowering plants than as a mechanistic explanation for the success of the angiosperms, suggesting that the focus of the search for underlying reasons should shift to an exploration of genetic or epigenetic mechanisms that might facilitate adaptive flexibility, which lies at the heart of angiosperm success. This agenda is particularly attractive in light of the suggestion that multicellular species may be largely unaffected by intense natural selection because of their propensity for gene duplication and divergence, pleiotrophy, and epistasis (Lynch, 2007). If true, rapid diversification within a clade is not a priori evidence for rapid adaptive diversification. It may simply reflect an evolutionary episode of genomic restructuring among related taxa that affords opportunities to occupy new niches (e.g., Verhoeven et al., 2008), some of which may involve coevolution among plant–animal interactions such as pollination syndromes. Whether these niches remain occupied is a consequence of natural selection and can only be judged in light of the fossil record of extinction patterns or after careful cladistic analyses of coevolutionary relationships (e.g., Pellmyr and Thompson, 1992; Brown et al., 1994; Pellmyr et al., 1996). Nonetheless, nonadaptive models provide no explanation for the high diversity of angiosperms relative to other groups because the phenomenology they predict should apply to all or at least most multicellular taxa.

PERCEPTION OR MISCONCEPTION?

However, before using new and more comprehensive data in our exploration of what makes the angiosperms so different and successful, it seems prudent to ask an important preliminary but often unarticulated question, viz.: "Is the diversification of angiosperms really significantly different from that of the pteridophytes or the gymnosperms when these grades of organization first appeared in Earth’s floras?" There was a time when the landscape was dominated by vascularized sporophytes and their free-living gametophytes and later by floras dominated by nonflowering seed plants. An evolutionary biologist viewing these ancient plant communities would probably have asked, "Why are there so many pteridophytes?" and more recently in geological time ask, "Why are there so many gymnosperm species?" Perhaps our perception of the aggressive rise and rapid ecological success of the angiosperms is erroneously colored by what the world looks like "here and now" rather than what happened in deep time when major evolutionary innovations like vascular tissues or seeds made their evolutionary debut.

To address this issue, we turn to a tabulation of the first and last occurrences of fossil plant species throughout the Phanerozoic (Niklas et al., 1980, 1983; see also Niklas, 1997), which permits a direct comparison of the tempo of origination rates of the pteridophytes, gymnosperms, and angiosperms. This tabulation is by no means complete; it is particularly deficient in terms of recent paleobotanical discoveries, but as discussed later, these discoveries have had more of an effect on understanding the significance of the affinities and nature of the fossils than they have had on the numbers of morphospecies they add to the record. This tabulation has one strength that justifies its use here, i.e., it represents a compendium of all vascular plant megafossils reported in the literature at the time the data were amassed. It therefore permits a direct, albeit admittedly imperfect comparison among the diversification patterns of pteridophytes, gymnosperms, and angiosperms. In terms of origination rates, it is important to note that this metric can be computed simply as the number of new species appearing in the fossil record per million years. However, it is far more instructive to compute origination rates as the number of new species appearing in the fossil record per million years divided by standing lineage diversity. By so doing, the probability of speciation increases as a function of the number of extant species existing at any particular time in a particular lineage. Diversification rates should be "normalized" with respect to standing diversity (see Niklas et al., 1980; Niklas, 1997).

With these limitations in mind, a comparison among the Phanerozoic origination rates and patterns of the three major vascular plant groups reveals that the early origination of the angiosperms is unexceptional in its tempo when compared to that of the pteridophytes or gymnosperms (Fig. 2A). The data indicate that each group radiates rapidly early in its evolutionary history and then subsequently declines in tempo. Likewise, the highest origination rates of the pteridophytes and the gymnosperms are comparable to those of the angiosperms. The only difference revealed by this comparison that can be judged to be numerically meaningful is the sustained tempo of speciation throughout the history of angiosperm evolution. That is, it appears that the angiosperms have maintained relatively high origination rates over longer periods of time, whereas the pattern of pteridophyte and the gymnosperm origination is characterized by an early burst, followed by a more or less steady decline. For example, our data indicate that angiosperm origination peaks in the mid-Cretaceous (Turonian) and in the Early Tertiary (Lower Eocene).

View larger version (24K): [in this window] [in a new window]   Fig. 2. Rates of (A) origination, (B) extinction, and (C) diversification for angiosperms, gymnosperms, and pteridophytes plotted against time. SD = standing species diversity. Data from Niklas et al. (1980, 1983) and Niklas (1997).

For example, significant changes in angiosperm floral trait shifts may have initiated concomitant changes in fruit types that resulted in changes in dispersal mechanisms. One of these changes appears to have occurred in the early Tertiary when seed and fruit size conspicuously increased to modern-day standards (Tiffney, 1984; Wing and Boucher, 1998). Two explanations have been advanced to account for this significant increase in propagule size. One emphasizes the adaptive radiation of mammals and birds (Tiffney, 1984), while the other focuses on climatic changes that permitted the spread of multistratal closed-canopy forests (Eriksson et al., 2000). The diversification and radiation of frugivores during the Early Tertiary may have paved the way for the evolution of larger seeds/fruits with concomitantly larger endosperm reserves for embryo development, germination, and seedling establishment as adaptations to low light conditions. This, in turn, would have paved the way for angiosperm species adapted to open habitats (which are generally characterized as producing small seeds/fruits with high dispersability) to evolve into closed-canopy habitats (for an extensive presentation of this hypothesis, see Tiffney, 1984).

By the same token, major climatic changes in the early Tertiary may have initiated changes in plant community structure in which seedling recruitment was limited by light. In theory, this could have favored species capable of producing seeds with large endosperm reserves that in turn permitted recruitment under low light conditions. It would have also favored species with fleshy fruits that attract larger frugivores capable of dispersing propagules greater distances.

Importantly, these two explanations are not mutually exclusive. The adaptive radiation of frugivores capable of larger dispersal ranges would have been favored by vegetation dominated by plants producing larger edible disseminules. In turn, these species would have been well adapted to closed-canopy communities perhaps made possible by climatic changes. The reciprocity between climatic and biotic changes during the early Tertiary would in turn have resulted in runaway selection favoring an increase in the size-ranges of frugivores and edible angiosperm disseminules (see Tiffney, 1984), even if there was no net gain in relative fitness. Recent phylogenetic analyses of phyletic shifts in fleshy vs. nonfleshy fruit types among 50 phylogenetically independent angiosperm lineages indicate that fleshy fruit types have multiple independent evolutionary origins, which are correlated with shifts between shaded and nonshaded habitats (Bolmgren and Eriksson, 2005). However, these analyses do not support the hypothesis that climatic changes favoring the spread of closed-canopy communities catalyzed the early Tertiary shift from smaller to larger angiosperm disseminules because it is equally plausible that runaway selection for larger angiosperm and frugivore species adapted to open or semiopen habitats permitted species to progressively tolerate more shaded conditions.

More important from the perspective of why some angiosperm lineages are species-rich is the report that extant species numbers are higher among woody lineages that have undergone a shift from nonfleshy to fleshy disseminules, whereas comparable shifts in nonwoody lineages have had the opposite effect (Bolmgren and Eriksson, 2005). The quantitative differences in the levels of species-richness among these lineages are insufficient to adduce fleshy disseminules as innovations supporting the success of the flowering plants. Rather, these differences are more in keeping with the suggestion that increased biotic dispersal ranges may reduce extinction rates and thus increase standing diversity (Tiffney and Mazer, 1995). Indeed, any hypothesis to explain the diversification of the flowering plants must address factors that increase origination rates and reduce extinction rates.

Turning to the same data set used to calculate origination rates and calculating extinction rates as the number of species not continuing from one geological period to the next younger per million years per species standing diversity, we see that the angiosperms tend to have significantly lower extinction rates compared to their pteridophyte and gymnosperm temporal counterparts (Fig. 2B).

We freely admit that the data set used to estimate origination and extinction rates is not quite as robust as it could be in light of recent discoveries of new species in the fossil record, particularly of angiosperms. Nonetheless, the nature of the data set (i.e., the number of morphospecies appearances and disappearances over time) is not going to be radically affected by these new discoveries because, with respect to angiosperms, they are qualitatively different. New discoveries, particularly of flowers have radically changed our view of angiosperm tempo by providing data on the timing of precisely identified taxa and their reproductive biology. It is not the number of these fossil discoveries, but their nature that has had such a dramatic impact on our understanding of angiosperm fossil history (Friis et al., 2006; Crepet et al., 2004, 2008). These discoveries represent a rather minor number of new taxa that even if omitted from the morphospecies data set would not affect the pattern derived from that data set dramatically due to the scale of those numbers.

If we limit ourselves to qualitative aspects of the patterns shown in Fig. 2, we believe that two statements are justified. First, the rate at which new angiosperm species enter the fossil record is consistently and continuously higher than that of any other group, and, second, the rate at which these species die out is consistently lower. We re-emphasize, therefore, that any robust hypothesis must focus on those biological features that permit species to resist extinction as well as those that foster speciation.

A NULL HYPOTHESIS

Arguably, any hypothesis attempting to explain the rapid taxonomic and ecological diversification of any clade requires a null hypothesis that permits us to focus on the tempo of the appearance of evolutionarily novel characters in the fossil record and the extent to which this tempo correlates with changes in standing species diversity. The logic of this kind of null hypothesis rests on three propositions, which we fully acknowledge are problematic without access to extremely robust data bases: (1) most species are (and have been) identified on the basis of their morphological-anatomical phenotypic character states as opposed to their molecular-genomic biology or their reproductive biology (i.e., the majority of extant species and all fossil species are morphospecies), (2) the ability to recognize taxonomic distinctions among morphospecies increases in proportion to the number of their phenotypic attributes, and (3), as the number of character states unique to a clade changes over time, so will the number of theoretically possible morphospecies. To illustrate this numerically, note that every clade is taxonomically defined by (and identifiable on the basis of) some number N of unique characters. If we make the simplifying assumption that each unique character has only two character states, it follows that the number of unique character state combinations (the number of potential morphospecies) in any clade is given by the formula 2^N. Thus, in the case of one clade distinguishable from all others based on five unique characters, the maximum number of theoretical morphospecies is 2⁵ = 32, whereas for another clade defined by 10 unique characters, the number is 2¹⁰ = 1024. Clearly, the number of theoretical morphospecies will increase rapidly as the number of character states per unique character in each clade increases. For example, if each character has three states, the two hypothetical clades have 3⁵ = 243 and 3¹⁰ > 59 000 potential morphospecies.

As noted, this approach has distractions and it must be used cautiously for a number of reasons. For example, the number of theoretically possible morphospecies provides only an upper limit on the predicted number of real species. Some character state combinations may be physically impossible or highly improbable (e.g., zygomorphic corollas with fused spirally arranged tepals are possible but unlikely for geometric and developmental reasons). All biologically possible combinations may not appear even in an ancient clade due to genomic and epigenetic constraints, whereas in an evolutionarily young clade, there may be insufficient time for some or even many phenotypes to evolve. However, in theory, we might expect the number of morphospecies to (1) increase rapidly early in the evolutionary history of a clade (as the numbers of characters and character states unique to a clade increases or as unique character combinations appear), (2) subsequently plateau (as developmental and genetic potential become canalized), and (3) at some point late in the history of the clade, dwindle (as unique character permutations are eliminated as a consequence of extinction events). If this prediction holds true, the correlation between the number of unique character state combinations (i.e., theoretical morphospecies) and the number of real morphospecies should be stronger early in the history of the clade than at any other time.

Our null hypothesis says nothing about the kind of characters or the number of their states that should be used to construct a morphospace. Obviously, this will vary depending on the group being examined. It will also depend on research predilections and the kinds of questions being asked. However, when dealing with the flowering plants, we are immediately drawn to floral characters and character states, particularly in light of what appears to be a correlation between the appearance of new floral characters and the standing diversity of extant species in angiosperm families coappearing in the fossil record between the Aptian and the Upper Paleocene (Fig. 3).

View larger version (22K): [in this window] [in a new window]   Fig. 3. First appearances of key floral characters and character states (number of floral character states indicated in rear panel; see Table 1) and standing extant species diversity in angiosperm families co-occurring in the fossil record (number of species indicated in front panel). Bold lines highlight geological periods for which data are available; thin lines and question marks indicate predicted numbers of floral character states based on standing species numbers. Adapted from Crepet (2008).

View larger version (112K): [in this window] [in a new window]   Fig. 4. Oldest fossils representing certain derived floral types compared with examples of modern counterparts: (A) Brush-type flower, Protomimosoidea Crepet and Taylor (Mimosaceae), from the Paleocene-Eocene of Tennessee, North America. Scale bar = 5 mm. (B) Calliandra sp. (Mimosaceae) inflorescence (example of an extant "brush flower" inflorescence, courtesy of M. A. Luckow, Cornell University and D. Kearns, University of Texas). (C) Inflorescence of Cladrastis kentuckea (Fabaceae) illustrating zygomorphic flowers with wing, standard, and keel petals (courtesy of K. C. Nixon, Cornell University). (D) Fossil papilioniod legume flower (Barnebyanthus Crepet and Herendeen, Fabaceae) from Paleocene-Eocene deposits of Tennessee, USA, showing typical fabaceous floral organization and expanding ovary. Scale bar = 1 cm. (E) Asterid I flower from Middle Eocene deposits in Tennessee, USA, with funnelform corolla. Scale bar = 5 mm. (F) Ipomoea pes-caprae (Convolulaceae), modern taxon with flowers having funnelform corollas (courtesy of K. C. Nixon, Cornell University).

View larger version (15K): [in this window] [in a new window]   Fig. 5. Cumulative number of angiosperm species first appearances and cumulative number of new floral character states plotted against Cretaceous-Tertiary geological stages. Data taken from Niklas et al. (1980, 1983) and Crepet (2008).

THE INSECT–PLANT CONNECTION

Taken at face value, the trend shown in Fig. 5 suggests that the angiosperm floral morphospace increased as new floral characters (and their character states) appear in the fossil record and that this morphospace was comparatively rapidly occupied by real species until it was "saturated." This scenario would gain credibility if it were possible to show that individual families manifested the same general pattern throughout their evolutionary history. Unfortunately, it is currently not possible to rigorously examine the extent to which the first appearances of floral characters in individual families correlate with the species diversification patterns within these families.

However, even if this was possible, every correlation, regardless of how strong, is readily contested as an explanation on the logical grounds that association provides no proof of causality. True, much of scientific theory and empirical investigation rests on the evidentiary basis of statistical correlation coupled with reasonable mechanistic explanation. However, the difference between correlation and causation is particularly important in the context of paleontology, which permits no direct experimentation and therefore no direct test of proposed mechanism, although experimentation and tests for proposed mechanisms can be accomplished indirectly with considerable success. This caveat is specifically important when considering early angiosperm evolution, which has been undoubtedly profoundly influenced by many factors, not the least of which are plant–animal synergistic coevolution. Certainly, plant–animal interactions have been invoked many times by many workers to explain a panoply of plant adaptations as well as the rapid early diversification in angiosperm history (Raven, 1977; Burger, 1981; Crepet, 1984; Eriksson and Bremer, 1992; Grimaldi, 1999; Grimaldi and Engel, 2005). Many of these adaptationist scenarios are not only plausible, but, in a limited number of cases, they have been empirically substantiated by neobotanical enquiries, which show that a change in one plant character state (e.g., white to red corollas or the acquisition of floral nectar spurs) can result in a shift in reproductive biology (e.g., moth or bee to hummingbird pollinators) that in theory can lead to a rapid genetic isolation of populations and character displacement (for an in-depth review, see Fulton and Hodges, 1999; Fenster et al., 2004; and Thomson and Wilson, 2008). Indeed, biotic pollination may not be a necessary condition for angiosperm diversification. For example, using quantitative trait linkage (QTL) analyses and reciprocal transplant experiments of two wild barley populations, Verhoeven et al. (2008) have shown that the target of habitat-specific natural selection contributing to population-level divergence can be flowering time. Thus, phenological shifts even among abiotically pollinated species may contribute significantly to angiosperm diversification.

Floral trait recognition and pollinator consistency have been extensively studied, i.e., insect pollinators tend to exhaust one floral morph for resources before moving on to other floral morphs (see Grant, 1949, 1950; Kugler, 1955; Faegri and van der Pijl, 1966). Indeed, A. R. Wallace (1889) may have been the first to suggest that sympatric plant species pollinated by "flower constant" pollinators will profit from having different floral recognition traits. Consequently, we would expect statistically significant correlations among the first appearances of critical floral traits and the diversification of the flowering plants and their insect pollinators. As noted, the first appearances of floral traits and the diversification of flowering plant species are significantly correlated (see Fig. 5). Likewise, the first appearances of key floral traits and insect families in the fossil record (data from Dmitriev and Ponomarenko, 2002) are significantly correlated (r² = 0.844, P < 0.0001; N = 11) as are angiosperm species number and insect family number (r² = 0.746, P < 0.0001; N = 16) (Fig. 6). Statistical analysis also reveals that over 96% of the variance in the cumulative number of angiosperm species throughout the Cretaceous and Cenozoic is described by changes in the number of floral traits and the number of insect families. Of central importance here is the fact that floral morphological characters and derived pollinators co-occur at times when highly diverse families appear in the fossil record (Fig. 6).

View larger version (15K): [in this window] [in a new window]   Fig. 6. Cumulative number of angiosperm species first appearances and cumulative number of new insect families plotted against Cretaceous-Tertiary geological stages. Data taken from Niklas et al. (1980, 1983) and Dmitriev and Ponomarenko (2002).

A NEO-GOLDSCHMIDTIAN PERSPECTIVE

The occurrence of rapid changes in critical reproductive character states is well documented (Garcia-Bellido, 1983; Hilu, 1983; Gottlieb, 1984). For example, Galen (1996) showed that populations of the alpine wildflower Polemonium viscosum can rapidly adapt to abrupt changes in pollinator assemblages. Her data indicate that the broadly flared flowers of the bumblebee pollinated P. viscosum could have evolved from narrower ones in a single generation because corolla flare increased by 12% from populations pollinated by a wide assemblage of insect pollinators to those pollinated only by bumblebees. Although this shift in floral phenotype is not the result of mutation, it demonstrates that mutations have the potential to result in rapid changes in gene flow within plant populations. By way of another example, species within the family Asteraceae are distinguished in part by whether their inflorescences contain radially symmetrical "disk" flowers, bilaterally symmetrical "ray" flowers, or both. Yet, by performing artificial crosses between two species of Haplopappus that have rayed and rayless florets (H. aureus and H. venetus subspecies venetus, respectively), Jackson and Dimas (1981) discovered that the presence or absence of ray flowers is controlled by a single gene, which can mutate to effect phenotypic differences reflected by the two species. Along similar lines, Singh and Jha (1978) examined X-ray induced mutants of soybean (Glycine max) that result in phenotypes bearing flowers with two or more carpels rather than one carpel, which characterizes the family Fabaceae.

Perhaps the best known examples of single gene mutations with significant floral phenotypic effects are those altering homeotic genes (i.e., genes that contain the genetic information required to direct development along a particular morphogenetic pathway), which have the ability to shift the developmental fate of cells, tissues, or entire organs. In the majority of cases, mutations of homeotic loci change the type (rather than the number) of organs produced, which suggests that the developmental patterns affected by these mutations involve genes that regulate organ identity and not those that regulate organ number.

The most extensively studied floral homeotic mutations occur in the mouse-ear cress, Arabidopsis, and in the snapdragon, Antirrhinum. Like many angiosperms with "perfect" flowers, these plants have four whorls of floral organs of which the outermost develop into sepals and the innermost develop into carpels. Mutations of AP3 and PI genes of Arabidopsis or the DEF gene of Antirrhinum cause petals to be replaced by sepals and stamens by carpels (Koornneef et al., 1983; Bowman et al., 1989; Sommer et al., 1990), which results in "imperfect" flowers incapable of self-fertilization. Mutations in the AG gene of Arabidopsis and the PLENI gene of the snapdragon convert stamens into petals and carpels into sepals (Carpenter and Coen, 1990), which are also incapable of self-fertilization. Because homeotic mutations such as these have the potential to establish reproductive barriers, they can serve as a genomic vehicle for character displacement, genetic divergence, and in theory the eventual appearance of new species.

The foregoing examples and those documented by Hilu (1983), Gottlieb (1984), and many other showing how single-gene mutations can produce major phenotypic shifts can evoke a neo-Goldschmidtian hypothesis for the rapid speciation of the angiosperms. However, enthusiasm for this hypothesis must be tempered in light of the low probability that any allele of this kind will become fixed in a population. Specifically, Kimura (1962) showed that a new mutant arising as a single copy in a diploid population of size N has a probability of fixation P given by the formula P = (1 – e^–2Nes/N) / (1 – e^–4Ne•s), where Ne is the effective population size and s is the selective advantage of the allele. Note that this formula reduces to P = (1 – e^–2s) / (1 – e^{–4 N•s}) when N = Ne, and becomes P = 1 / 2N when s = 0.0 (i.e., a neutral mutation). Therefore, in the case of a hypothetical mutant with a selective advantage of s = 0.01 appearing in a population of 1000 individuals, we see that P = (1 – e^–0.002) / (1 – e^–4) = 0.00199 or roughly 0.2%. Likewise, if the mutant is selectively neutral, we find that P = 0.05%. This hypothetical case illustrates that Goldschmidtian "hopeful monsters" created even by advantageous or neutral allelic changes (which arguably represent best-case scenarios) have exceedingly low probabilities of becoming fixed in a population. Indeed, if our hypothetical mutant is even slightly deleterious (e.g., 0.001), we see that P = 0.004% when N = 1000.

SMALL POPULATIONS OF ANNUAL GROWTH FORMS

Even though the fixation of single-gene mutations has a low probability, a number of theories have been proposed that predict the rapid spread of novel alleles that can quickly become embedded in adaptive gene complexes (e.g., genetic drift, shifting balance, and crash-founder mathematical models). Interestingly, all of these theories require two conditions: (1) recurrent mutations of the same alleles and (2) very small, isolated populations. We can say very little about the natural mutation rates of homeotic plant genes other than to point out that they are likely to be faster in annual rather than perennial species simply as a consequence of the disparity between the life spans, generation times, and growth rates of these two life forms (Fig. 7). It is also reasonable to suggest that annual species are more likely to exist in smaller and more isolated populations than perennial woody counterparts. If both of these conjectures are generally true, it is reasonable to suggest that the mutation and fixation rates of alleles are probably faster among annual as opposed to perennial species such that speciation attributable to Goldschmidtian hopeful monsters might be faster among annual as opposed to perennial species.

View larger version (11K): [in this window] [in a new window]   Fig. 7. Annual growth rate (dry mass production per plant per year, G) plotted as a function of body mass (total dry mass per plant, MT) of nonwoody and woody plants. Solid lines are regression curves, which have the same slope but differ in elevations, indicating that nonwoody plants have on average faster growth rates for their body mass compared to woody plants. Data taken from numerous sources (see Niklas, 1994).

View larger version (11K): [in this window] [in a new window]   Fig. 8. Size (stem height; original units cm) frequency distribution of numbers of genera characterized as dominated by annual herbs and woody perennials. Taken from Niklas (1994).

View this table: [in this window] [in a new window]   Table 2. Net rates of chromosome evolution and speciation as a function of growth form (taken from Levin and Wilson, 1976).

View this table: [in this window] [in a new window]   Table 3. Distribution of the numbers of incompatibility and sterility barriers in 72 plant groups (genera, sections, or subtribes) sorted into life forms on the basis of taxogenetic analyses (taken from Grant, 1971).

The role of polyploidy in plant evolution and speciation has been reviewed and discussed many times (most recently by Arnold, 1997; Rieseberg, 2001; Baack and Rieseberg, 2007; Rieseberg and Willis, 2007), particularly in terms of the distribution of amphiploidy among genera containing species with different life forms (e.g., Müntzing, 1936; Stebbins, 1938; Grant, 1956, 1971). Three very broad generalizations emerge from these surveys: (1) polyploidy is far more common among plants, particularly pteridophytes, mosses, and angiosperms, than animals (Fig. 9); (2) among the angiosperms, it typically takes the form of amphiploidy (resulting most commonly from meiotic nonreduction followed by the union of unreduced gametes), which (3) provides a genetic system for perpetuating adaptive hybrid genotypes by means of sexual reproduction (for a detailed summary, see Grant, 1971). In addition, there is ample evidence to suggest that polyploidy is more frequent among species with long-lived herbaceous life forms coupled with some means of vegetative propagation than among species with other life forms (see Müntzing, 1936; Stebbins, 1938; Grant, 1956). It is noteworthy that polyploidy is absent in some ancient plant lineages, such as the cycads, which argues against the proposition that the frequency of polyploidy in a lineage is merely of symptom of lineage age (Fig. 9).

View larger version (16K): [in this window] [in a new window]   Fig. 9. Haploid chromosome number frequency distribution for (A) dicot, (B) moss, and (C) cycad species (data taken from Darlington and Ammal, 1945; Darlington and Wylie, 1955). Supernumerary or B chromosomes not counted; odd numbers are arithmetic means (e.g., 2N = 7 is N = 3.5).

Embryo formation without benefit of fertilization (i.e., agamospermy, which includes diplospory, apospory, and adventitious embryony) is prevalent among land plants but no where more so than among angiosperms, ferns, and mosses, which are the three largest species-rich embryophyte groups. This phenomenon also permits the proliferation of hybrid genotypes asexually. Darlington (1939) referred to this strategy as a means by which genotypes can "escape from sterility," as for example among hybrid triploids or pentaploids. However, it cannot escape attention that agamospermy also occurs among species that are sexually fertile (e.g., Citrus hybrids) and thus provides an escape strategy for the 50% cost of sexual reproduction (i.e., genotypes that are well adapted to their environment convey only 50% of their genes to their offspring; see Williams, 1975; Maynard Smith, 1978; Niklas, 1997). Specifically, sexually fertile agamospermous organisms can produce genetically different progeny, which favors the colonization of new habitats and adaptive evolution by means of genetic recombination, but they can also reproduce asexually to produce disseminules that are as adapted to local environmental conditions as they are. These disseminules have all the advantages of being encapsulated in seeds and fruits, which may have specialized biotic dispersal agents. To the best of our knowledge, this "agamospermous syndrome" occurs nowhere in the plant kingdom other than among the flowering plants.

ASEXUAL ADVANTAGE

Agamospermy is only one of many methods that permit plants to reproduce asexually (Table 4), which under many circumstances affords a clonal advantage (Tiffney and Niklas, 1985). New clones can be continuously generated by the same or different members of a population. These can provide a short-term method for the coexistence of genetically different individuals produced by a population of sexually reproductive genotypes. Such clones may be adapted to different seasonal variations of the environment, or they may be adapted to different biotic or abiotic "subniches" in a stable but heterogeneous environment. Likewise, the ability to reproduce asexually permits a sexually sterile organism to escape death and to capitalize on subsequent genomic changes that may confer fertility. Finally, asexual reproduction may permit Goldschmidtian-like "hopeful monster" mutants to survive and spread in a population.

View this table: [in this window] [in a new window]   Table 4. Developmental mechanisms of plant growth (horizontal axis) and growth stimuli (i.e., traumatic vs. programmed reiteration) that can result in different modes of asexual reproduction. Adapted from Tiffney and Niklas (1985).

The advent of the angiosperms and their subsequent ecological success undoubtedly reflect the synergy among many functional traits that allowed them to "escape" from the developmental constraints experienced by their gymnosperm progenitors. Although the seed habit contributed to the survival of gymnosperm lineages during the Permo–Triassic extinction event, the presence of secondary growth in those that did pass through this gauntlet apparently restricted clonal growth, particularly by means of the rhizome or stolon fragmentation. Among extant gymnosperms, the closest approximations to clonal growth are seen in the root sprouting of conifers, the basal suckering of some cycads and cycadeoids, and the liana growth habit of a few Gnetum species (Tiffney and Niklas, 1985). In contrast, many species in the most species-rich flowering plant families are capable of vegetative cloning, particularly by means of rhizome or stolon fragmentation. For example, among the monocots, which contain some of the largest angiosperm families (e.g., grasses, which have over 500 genera and 8000 species), over 70% of the 65 modern-day families are rhizomatous, while most of the remaining 30% appear to be derived from rhizomatous ancestors.

Finally, it is potentially significant that mathematical models predict that the community dynamics resulting from the convergence of sexual and asexual reproductive mechanisms is particularly efficacious for the survival of a species in environments experiencing episodic or random physical disturbance (see Sebens and Thorne, 1985).

REINVENTING THEMSELVES

The pattern of angiosperm diversification preserved in the fossil record is not diagnostic of a monotonic process, one in which species numbers increase steadily at the same pace. Rather, it is characterized by episodes of comparatively rapid species diversification followed by periods of quiescence (see Fig. 2). The intrinsic limitations of the fossil record should never be ignored, particularly the poor fidelity of its signal in understudied or poorly documented geological periods (Burnham, 2008). Nevertheless, this pattern in angiosperm evolution can be interpreted as evidence for saltational evolution or punctuated equilibrium resulting from the appearance of new functional traits, which opened the door to new niches or different modes of more rapid speciation. The acquisition of the flower and the subsequent elaboration of phenotypic innovations that facilitated animal-assisted pollination have been asserted historically as the causative agents underwriting the early burst in angiosperm diversification during the Cretaceous (e.g., Saporta and Marion, 1881; Dilcher, 2000). Likewise, the evolutionary acquisition, exploitation, and "rediscovery" of the annual growth form, which is entirely absent among extant gymnosperms, in tandem with continued coevolution with animal pollinators and dispersal vectors may have been important ingredients in the subsequent evolutionary success of the flowering plants during the Cenozoic (Crepet, 2008).

Seen in this light, hallmarks of the angiosperms include their capacity to "reinvent themselves" by mixing old vegetative and reproductive traits into different combinations in addition to their ability to innovate new traits throughout their history (Crepet, 2004, 2008; Friis et al., 2006). These characteristics surface in neobotanical studies that have drawn particular attention to the phenomenon called phenotypic plasticity, which is nowhere better revealed than through studies of angiosperms (e.g., Sultan, 1992, 1995; DeWitt and Scheiner, 2004). Because phenotypic plasticity confers potential adaptive diversity to individual genotypes, it is likely to increase both the ecological distribution of a taxon and its pattern of diversification. Provided that each genotype within an individual species is sufficiently plastic to be broadly tolerant of environmental diversity, the ecological range of the species is expected in increase in proportion to the response breadth of its individual genotypes (Lewontin, 1957; Gross, 1984; Bazzaz and Sultan, 1987). As the geographic range of a species increases, the potential for ecotypic divergence and genetic isolation of subpopulations is also expected to increase (Levins, 1968; Van Tienderen, 1990), which can, in theory, foster the appearance of protospecies, particularly among plants with an annual life form. This hypothesis requires extensive research. However, a few studies indicate that more plastic traits can evolve more rapidly than less plastic ones (e.g., Andersson, 1989)

Curiously, phenotypic plasticity may contribute to survival as well as rapid speciation. Because many aspects of adaptive differentiation may be obviated in taxa manifesting functionally appropriate phenotypes in response to key environmental pressures, phenotypic plasticity can shield genetic diversity from the effects of natural selection and it can enhance the long-term survival of taxa by means of species selection (see Schlichting, 2004). For example, using protein electrophoresis for 25 isozyme loci, Novak et al. (1991) report little genetic divergence among 60 populations of the widespread introduced species Bromus tectorum, which is particularly surprising given that this species is self-pollinating. Naturally, individual studies shed little light on the relative frequency of evolutionary phenomena, which requires analysis of many case studies to determine generalizable patterns and their underlying mechanisms.

CONCLUSIONS

Darwin’s second "abominable mystery" is likely to remain unresolved as long as a single explanation is sought for the rapid diversification and subsequent ecological success of the angiosperms. Comparisons among the diversification patterns of angiosperms, gymnosperms, and pteridophytes reveal that all three groups experienced comparably high diversification rates early in their respective evolutionary histories (Fig. 2). The only unique feature of the angiosperm pattern revealed by these comparisons is a record of episodic and high diversification rates. In contrast, the diversification patterns of the pteridophytes and gymnosperms are characterized by high initial rates followed by a more or less steady decline in the appearance of new species. The angiosperm pattern is all the more remarkable considering that the "pteridophytes" and the "gymnosperms" each designate a grade of reproductive organization (rather than a clade) that reappears at different times in the early history of the embryophytes. If each of the lineages within these two grades of reproductive organization manifested an initial diversification "burst," we would expect the overall pattern of each grade to look much more like the diversification pattern of the angiosperms. In our view, this dissimilarity tells us that the evolutionary history of the angiosperms is truly unique not because of the features angiosperms possessed when they "burst onto the evolutionary scene" but because they sustained their initial tempo of speciation and, at times, exceeded it considerably.

This impression is reinforced by insights gained from population and evolutionary theory and from neobotanical investigations, which in tandem revealed a number of factors that can sustain and increase the tempo and mode of speciation. These insights indicated to us that the success of the angiosperms is Vielzeitigkeit (multifaceted) and not the product of any one functional trait or syndrome of traits. Indeed, we believe that the available evidence suggests that the angiosperms have an unparalleled capacity for evolutionarily "reinventing themselves" and that each reinvention has allowed them to reiterate their pattern of species diversification and ecological success.

It is clear that plant–animal interactions were critical to the success of the earliest flowering plants in light of a reciprocal driving mechanism for angiosperm and animal diversifications. However, if this were the single evolutionary innovation of the flowering plants, we would expect to see a monotonically decreasing pattern in the rate of angiosperm diversification throughout the Cretaceous–Tertiary. In contrast, the available (albeit limited and therefore suspect) paleobotanical data indicates a pattern of saltational evolution and species diversification, which is more in keeping with repeated bursts of phenotypic and behavioral innovations resulting from the evolutionary introduction of novel functional traits throughout the history of the flowering plants. Whether these bursts indicate adaptive evolution sensu stricto remains problematic in light of the suggestion that multicellular organisms are generally little affected by intense natural selection (Lynch, 2007). In this paper, we have selected a few among the many traits for special consideration only because neobotanical studies have provided sufficient evidence that each can foster adaptive phenotypic divergence within populations. Some among these traits provide the raw materials for phenotypic innovation (e.g., single-gene mutations capable of producing striking morphological changes in a single generation); others provide an avenue for fixing these traits in plant populations or escaping extinction (e.g., the annual growth habit, substantial phenotypic plasticity, and asexual/sexual reproductive syndromes).

Many other traits, such as the propensity for polyploidy and hybridization are also likely to have been important to the evolution of the angiosperms because they provide opportunities for duplicate genes to functionally diverge. Moreover, we cannot yet even estimate the significance of a newly discovered autoimmune mechanism that might stimulate speciation by acting as a gene flow barrier (hybrid necrosis; Bomblies and Weigel, 2007), and we have to allow for the possibility that new genomically based discoveries may some day illuminate the underlying cause or causes of the evolutionary plasticity of the angiosperms. However, if our perspective is judged valid in general, it requires a continued pluralistic approach to the "problem of speciation" in general and to the predilection of the angiosperms toward this process, which is particularly well expressed compared to all other terrestrial plant lineages. What are the intrinsic features of organisms that provide the materials for speciation, and what are the external environmental factors that permit protospecies to survive and mature into full-fledged new species? These questions are likely to be answered differently in their details depending on the individual biological features of the taxa being examined. But they are also likely to share features in common across a spectrum of lineages sharing a last common ancestor, such as those comprising the embryophytes. When viewed in this manner, Darwin’s second "abominable mystery" could be better cast as the question: Why are there so few species of other embryophyte lineages compared to the angiosperms?

FOOTNOTES

¹ The authors thank Drs. B. H. Tiffney (University of California, Santa Barbara) and K. Pigg (Arizona State University) for valuable suggestions to improve an early draft of this paper, which we dedicate to Verne Grant whose pioneering work in the field of angiosperm speciation and biology is as relevant today as it was in the last century.

² Author for correspondence (e-mail: wlc1{at}cornell.edu)

LITERATURE CITED

Andersson, S. 1989. Phenotypic plasticity in Crepis tectorum (Asteraceae). Plant Systematics and Evolution 168: 19–38.[CrossRef][Web of Science]

Arnold, M. L. 1997. Natural hybridization and evolution. Oxford University Press, Oxford, UK.

Baack, E. J., AND L. H. Rieseberg. 2007. A genomic view of introgression and hybrid speciation. Current Opinion in Genetics & Development 6: 513–518.

Bazzaz, F. A., AND S. E. Sultan. 1987. Ecological variation and the maintenance of plant diversity. In K. M. Urbanska [ed.], Differentiation patterns in higher plants, 69–93. Academic Press, London, UK.

Bell, G. 1982. The masterpiece of nature: The evolution and genetics of sexuality. Croom Helm, London, UK.

Bolmgren, K., AND O. Eriksson. 2005. Fleshy fruits—Origins, niche shifts, and diversification. Oikos 109: 255–272.[CrossRef][Web of Science]

Bomblies, K., AND D. Weigel. 2007. Hybrid necrosis: Autoimmunity as a potential gene flow barrier in plant species. Nature Reviews Genetics 8: 382–393.[CrossRef][Web of Science][Medline]

Bowman, J. L., D. R. Smyth, AND E. M. Meyerowitz. 1989. Genes directing flower development in Arabidopsis. Plant Cell 1: 37–52.[Abstract/Free Full Text]

Brown, J. M., O. Pellmyr, J. N. Thompson, AND R. G. Harrison. 1994. Mitochondrial DNA phylogeny of the Prodoxidae (Lepidoptera: Incuvarioidae) indicates rapid ecological diversification of yucca moths. Annals of the Entomological Society of America 87: 795–802.[Web of Science]

Burger, W. C. 1981. Why are there so many kinds of flowering plants? Bioscience 31: 572, 577–581.

Burnham, R. J. 2008. Hide and go seek: What does presence mean in the fossil record? Annals of the Missouri Botanical Garden 95: 51–71.[CrossRef][Web of Science]

Carpenter, R., AND E. S. Coen. 1990. Floral homeotic mutations proposed by transposon-mutagenesis in Antirrhinum majus. Genes & Development 4: 1483–1493.[Abstract/Free Full Text]

Chase, M. W. 2004. Monocot relationships: An overview. American Journal of Botany 91: 1645–1655.[Abstract/Free Full Text]

Crepet, W. L. 1984. Advanced (constant) insect pollination mechanisms: Patterns of evolution and implications vis-à-vis angiosperm diversity. Annals of the Missouri Botanical Garden 71: 607–630.[CrossRef][Web of Science]

Crepet, W. L. 1996. Timing in the evolution of derived floral characters: Upper Cretaceous (Turonian) taxa with tricolpate and tricolpate-derived pollen. Review of Palaeobotany and Palynology 90: 339–360.[CrossRef][Web of Science]

Crepet, W. L., K. C. Nixon, AND M. A. Gandolfo. 2004. Fossil evidence and phylogeny: The age of major angiosperm clades based on mesofossil and macrofossil evidence from Cretaceous deposits. American Journal of Botany 91: 1666–1682.[Abstract/Free Full Text]

Crepet, W. L. 2008. The fossil record of angiosperms: Requiem or renaissance? Annals of the Missouri Botanical Garden 95: 3–33.[CrossRef][Web of Science]

Darlington, C. D. 1939. The evolution of genetic systems. Cambridge University Press, Cambridge, UK.

Darlington, C. D. 1956. Natural populations and the breakdown of classical genetics. Proceedings of the Royal Society of London, B. Biological Sciences 145: 350–364.[Free Full Text]

Darlington, C. D., AND E. K. J. Ammal. 1945. Chromosome atlas of cultivated plants. G. Allen & Unwin, London, UK.

Darlington, C. D., AND A. P. Wylie. 1955. Chromosome atlas of flowering plants. G. Allen & Unwin, London, UK.

Darwin, C. 1871. On the origin of species by means of natural selection. Appleton, New York, New York, USA.

Darwin, F. [ed.]. 1903. More letters of Charles Darwin, a record of his work in hitherto unpublished letters, vol. 2. John Murray, London, UK.

Dewitt, T. J., and S. M. Scheiner [eds.]. 2004. Phenotypic plasticity: Function and conceptual approaches. Oxford University Press, Oxford, UK.

Dilcher, D. L. 1974. Approaches to the identification of angiosperm leaf remains. Botanical Review (Lancaster) 40: 1–157.

Dilcher, D. L. 2000. Toward a new synthesis: Major evolutionary trends in the angiosperm fossil record. Proceedings of the National Academy of Sciences, USA 97: 7030–7036.[Abstract/Free Full Text]

Dilcher, D. L., AND P. R. Crane. 1984. Archaeanthus: An early angiosperm from the Cenomanian of the western interior of North America. Annals of the Missouri Botanical Garden 71: 351–383.[CrossRef][Web of Science]

Dilcher, D. L., W. L. Crepet, C. B. Beeker, AND H. C. Reynolds. 1976. Reproductive and vegetative morphology of a Cretaceous angiosperm. Science 191: 854–856.[Abstract/Free Full Text]

Dmitriev, V. Yu., AND A. G. Ponomarenko. 2002. In A. P. Rasnitsyn, and D. L. J. Quicke [eds.], History of insects, 325–331. Kluwer, Dordrecht, Netherlands.

Doyle, J. A. 1978. The origin of angiosperms. Annual Review of Ecology and Systematics 9: 365–392.[CrossRef][Web of Science]

Dressler, R. L. 1968. Pollination by euglossine bees. Evolution 22: 202–210.[CrossRef][Web of Science]

Dressler, R. L. 1982. Biology of the orchid bees (Euglossini). Annual Review of Ecology and Systematics 13: 373–394.[CrossRef][Web of Science]

Ehrendorfer, F. 1964. Cytologie, Taxonomie und Evolution bei Samenpflanzen. Vistas in Botany 4: 99–186.

Eriksson, O., AND B. Bremer. 1992. Pollination systems, dispersal modes, life forms and diversification rates in angiosperm families. Evolution 46: 258–266.[CrossRef][Web of Science]

Eriksson, O., E. M. Friis, AND P. Löfgren. 2000. Seed size, fruit size, and dispersal systems in angiosperms from the early Cretaceous to the late Tertiary. American Naturalist 156: 47–58.[CrossRef][Medline]

Faegri, K., AND L. van der Pijl. 1966. The principles of pollination biology. Pergamon, New York, New York, USA.

Feild, T. S., N. C. Arens, J. A. Doyle, T. E. Dawson, AND M. J. Donoghue. 2004. Dark and disturbed: A new image of early angiosperm ecology. Paleobiology 30: 82–107.[Abstract/Free Full Text]

Fenster, C. B., W. S. Armbruster, P. Wilson, M. R. Dudash, AND J. D. Thompson. 2004. Pollination syndromes and floral specialization. Annual Review of Ecology, Evolution, and Systematics 35: 375–403.[CrossRef][Web of Science]

Friedman, W. E. 2009. The meaning of Darwin’s "abominable mystery." American Journal of Botany 96: 5–21.[Abstract/Free Full Text]

Friis, E. M., K. R. Pedersen, AND P. R. Crane. 2006. Cretaceous angiosperm flowers: Innovation and evolution in plant reproduction. Palaeogeography, Palaeoclimatology, Palaeoecology 232: 251–293.[CrossRef]

Fulton, M., AND S. A. Hodges. 1999. Floral isolation between Aquilegia formosa and Aquilegia pubescens. Proceedings of the Royal Society of London, B, Biological Sciences 266: 2246–2252

Galen, C. 1996. Rates of floral evolution: Adaptation to bumblebee pollination in an alpine wildflower Polemonium viscosum. Evolution 50: 120–125.[CrossRef][Web of Science]

Gandolfo, M. A., K. C. Nixon, AND W. L. Crepet. 2008. Selection of fossils for calibration of molecular dating models. Annals of the Missouri Botanical Garden 95: 34–42.[CrossRef][Web of Science]

Garcia-Bellido, A. 1983. Comparative anatomy of cuticular patterns in the genus Drosophila. In B. C. Godwin, N. Holder, and C. C. Wylie [eds.], Development and evolution, 227–255. Cambridge University Press, Cambridge, UK.

Gleason, H. A. [ed.] 1968. The new Britton and Brown illustrated flora of the northeastern United States and adjacent Canada. Hafner, New York, New York, USA.

Gorelick, R. 2001. Did insect pollination cause increased seed plant diversity? Biological Journal of the Linnean Society 74: 407–427.[CrossRef][Web of Science]

Gottlieb, L. D. 1984. Genetic and morphological evolution in plants. American Naturalist 123: 681–709.[CrossRef][Web of Science]

Grant, V. 1949. Pollination systems as isolating mechanisms in flowering plants. Evolution 3: 82–97.[CrossRef][Web of Science][Medline]

Grant, V. 1950. The flower constancy of bees. Botanical Review 16: 379–398.[CrossRef]

Grant, V. 1956. The influence of breeding habit on the outcome of natural hybridization in plants. American Naturalist 90: 319–322.[CrossRef][Web of Science]

Grant, V. 1971. Plant speciation. Columbia University Press, New York, New York, USA.

Grimaldi, D. A. 1999. The co-radiations of pollinating insects and angiosperms in the Cretaceous. Annals of the Missouri Botanical Garden 86: 373–406.[CrossRef][Web of Science]

Grimaldi, D. A., AND M. S. Engel. 2005. Evolution of the insects. Cambridge University Press, Cambridge, UK.

Graur, D., AND W. Martin. 2004. Reading the entrails of chickens: Molecular timescales of evolution and the illusion of precision. Trends in Genetics 20: 80–86.[CrossRef][Web of Science][Medline]

Gross, L. J. 1984. On the phenotypic plasticity of leaf photosynthetic capacity. In S. Levin, and T. Hallam [eds.], Mathematical ecology: Proceedings of the autumn course, Trieste, Italy, 1982, 2–14. Springer-Verlag, Berlin, Germany.

Herrera, C. M. 1996. Floral traits and plant adaptations to insect pollinators: A devil’s advocate approach. In D. G. Lloyd, and S. C. H. Barrett [eds.], Floral biology, 65–87. Chapman and Hall, New York, New York, USA.

Hilu, K. W. 1983. The role of single-gene mutations in the evolution of flowering plants. In M. K. Hecht, B. Wallace, and G. T. Prance [eds.], Evolutionary biology, vol. 16, 97–128. Plenum, New York, New York, USA.

Jackson, R. C., AND C. T. Dimas. 1981. Experimental evidence for systematic placement of the Haplopappus phyllocephalus complex. Systematic Botany 6: 8–14.[CrossRef][Web of Science]

Kessler, D., K. Gase, AND I. T. Baldwin. 2008. Field experiments with transformed plants reveal the sense of floral scents. Science 321: 1200–1202.[Abstract/Free Full Text]

Kimura, M. 1962. On the probability of fixation of mutant genes in populations. Genetics 47: 713–719.[Free Full Text]

Koornneef, M., J. Van Eden, C. J. Hanharth, P. Stam, F. J. Braaksma, AND W. J. Feenstra. 1983. Linkage map of Arabidopsis thaliana. Journal of Heredity 74: 265–272.[Abstract/Free Full Text]

Kugler, H. 1955. Einführung in die Blütenökologie. Gustav Fischer, Stuttgart, Germany.

Labandeira, C. C., D. L. Dilcher, D. R. Davis, AND D. L. Wagner. 1994. Ninety seven million years of angiosperm insect association: Paleobiological insights into the meaning of co-evolution. Proceedings of the National Academy of Sciences, USA 91: 12278–12282.[Abstract/Free Full Text]

Levin, D. A., AND A. C. Wilson. 1976. Rates of evolution in seed plants: Net increase in diversity of chromosome numbers and species numbers though time. Proceedings of the National Academy of Sciences, USA 73: 2086–2090.[Abstract/Free Full Text]

Levins, R. 1968. Evolution in changing environments. Princeton University Press, Princeton, New Jersey, USA.

Lewontin, R. C. 1957. The adaptation of populations to varying environments. Cold Spring Harbor Symposia on Quantitative Biology 22: 395–408.[Abstract/Free Full Text]

Lynch, M. 2007. The origins of genome architecture. Sinauer, Sunderland, Massachusetts, USA.

Magallón, S., P. R. Crane, AND P. S. Herendeen. 1999. Phylogenetic pattern, diversity and diversification of eudicots. Annals of the Missouri Botanical Garden 86: 297–372.[CrossRef][Web of Science]

Maynard Smith, J. 1978. The evolution of sex. Cambridge University Press, Cambridge, UK.

Midgley, J. J., AND W. J. Bond. 1991. How important is biotic pollination and dispersal to the success of the angiosperms? Philosophical Transactions of the Royal Society of London, B, Biological Sciences 333: 209–215.[CrossRef]

Müntzing, A. 1936. The evolutionary significance of autopolyploidy. Hereditas 21: 263–378.[Web of Science]

Niklas, K. J. 1994. Plant allometry. University of Chicago Press, Chicago, Illinois, USA.

Niklas, K. J. 1997. The evolutionary biology of plants. University of Chicago Press, Chicago, Illinois, USA.

Niklas, K. J., B. H. Tiffney, AND A. H. Knoll. 1980. Apparent changes in the diversity of fossil plants: A preliminary assessment. Evolutionary Biology 12: 1–89.

Niklas, K. J., B. H. Tiffney, AND A. H. Knoll. 1983. Patterns in vascular land plant diversification. Nature 303: 614–316.[CrossRef][Web of Science]

Nixon, K. C. 2008. Paleobotany, evidence, and molecular dating: an example from the Nymphaeales. Annals of the Missouri Botanical Garden 95: 43–50.[CrossRef][Web of Science]

Norstog, K., D. W. Stevenson, AND K. J. Niklas. 1986. The role of beetles in the pollination of Zamia furfuraceae L. f. (Zamiaceae). Biotropica 18: 300–306.[CrossRef][Web of Science]

Novak, S. J., R. N. Mack, AND D. E. Soltis. 1991. Genetic variation in Bromus tectorum: Population differentiation in its North American range. American Journal of Botany 78: 1150–1161.[CrossRef][Web of Science]

Pellmyr, O., AND J. N. Thompson. 1992. Multiple occurrences of mutualism in the yucca moth lineage. Proceedings of the National Academy of Sciences, USA 89: 2927–2929.[Abstract/Free Full Text]

Pellmyr, O., J. N. Thompson, J. M. Brown, AND R. G. Harrison. 1996. Evolution of pollination and mutualism in the yucca moth lineage. American Naturalist 148: 827–847.[CrossRef][Web of Science]

Raguso, R. A. 2008. Start making scents: The challenge of integrating chemistry into pollination biology. Entomologia Experimentalis et Applicata 128: 196–207.[CrossRef][Web of Science]

Raven, P. H. 1977. A suggestion concerning the Cretaceous rise to dominance of the angiosperms. Evolution 31: 451–452.[CrossRef][Web of Science]

Regal, P. J. 1977. Ecology and evolution of flowering plant dominance. Science 196: 622–629.[Abstract/Free Full Text]

Rieseberg, L. H. 2001. Polyploid evolution: Keeping the peace at genomic reunions. Current Biology 11: R925–R928.[CrossRef][Web of Science][Medline]

Rieseberg, L. H., AND J. H. Willis. 2007. Plant speciation. Science 317: 910–914.[Abstract/Free Full Text]

Sanderson, M. J., J. L. Thorne, N. Wickström, AND K. Bremer. 2004. Molecular evidence on plant divergence times. American Journal of Botany 91: 1656–1665.[Abstract/Free Full Text]

de Saporta, G., AND A.-F. Marion. 1881. L’évolution du règne végétal. Les Cryptogams. Germer Bailliére et Cie., Paris, France.

Schlichting, C. D. 2004. The role of phenotypic plasticity in diversification. In T. J. DeWitt, and S. M. Scheiner [eds.], Phenotypic plasticity: Functional and conceptual approaches, 191–200. Oxford University Press, Oxford, UK.

Sebens, K. P., AND B. L. Thorne. 1985. Coexistence of clones, clonal diversity, and the effects of disturbance. In J. B. C. Jackson, L. W. Buss, and R. E. Cook [eds.], Population biology and evolution of clonal organisms, 357–398. Yale University Press, New Haven, Connecticut, USA.

Singh, B. B., AND A. N. Jha. 1978. Abnormal differentiation of floral parts in a mutant strain of soybean. Journal of Heredity 69: 143–144.[Free Full Text]

Smith, S. A., AND M. J. Donoghue. 2008. Rates of molecular evolution are linked to life history in flowering plants. Science 322: 86–89.[Abstract/Free Full Text]

Soltis, D. E., P. S. Soltis, M. W. Chase, M. E. Mort, D. C. Albach, M. Zanis, V. Savolainen et al.. 2000. Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnean Society 133: 381–461.[CrossRef][Web of Science]

Sommer, H., J.-P. Beltrán, P. Huijser, H. Pape, W.-E. Lönnig, H. Saedler, AND Z. Schwarz-Sommer. 1990. Deficiens, a homeotic gene involved in the control of flower morphogenesis in Antirrhinum majus: The protein shows homology to transcription factors. EMBO Journal 9: 605–613.[Web of Science][Medline]

Stebbins, G. L. 1938. Cytological characteristics associated with the different growth habits of dicotyledons. American Journal of Botany 25: 189–198.[CrossRef][Web of Science]

Stebbins, G. L. 1958. The inviability, weakness, and sterility of interspecific hybrids. Advances in Genetics 9: 147–215.[Web of Science][Medline]

Stebbins, G. L. 1981. Why are there so many species of flowering plants? Bioscience 31: 573–577.[CrossRef][Web of Science]

Stebbins, G. L. 1992. Comparative aspects of plant morphogenesis: A cellular, molecular, and evolutionary approach. American Journal of Botany 79: 589–598.[CrossRef][Web of Science]

Sultan, S. E. 1992. What has survived of Darwin’s theory. Evolutionary Trends in Plants 6: 61–71.[Web of Science]

Sultan, S. E. 1995. Phenotypic plasticity and plant adaptation. Acta Botanica Neerlandica 44: 363–383.[Web of Science]

Sun, G., D. L. Dilcher, S. Zheng, AND Z. Zhou. 1998. In search of the first flower: A Jurassic angiosperm, Archaefructus, from northeast China. Science 282: 1692–1695.[Abstract/Free Full Text]

Taylor, T. N., AND M. A. Millay. 1979. Pollination biology and reproduction in early seed plants. Review of Palaeobotany and Palynology 27: 329–355.[CrossRef][Web of Science]

Taylor, T. N., AND A. C. Scott. 1983. Interactions of plants and animals during the Carboniferous. Bioscience 33: 488–493.[CrossRef][Web of Science]

Thomson, J. D., AND P. Wilson. 2008. Explaining evolutionary shifts between bee and hummingbird pollination: Convergence, divergence, and directionality. International Journal of Plant Sciences 169: 23–38.[CrossRef][Web of Science]

Tiffney, B. H. 1984. Seed size, dispersal syndromes, and the rise of the angiosperms: Evidence and hypothesis. Annals of the Missouri Botanical Garden 71: 551–576.[CrossRef][Web of Science]

Tiffney, B. H. 1986. Fruit and seed dispersal and the evolution of the Hamamelidae. Annals of the Missouri Botanical Garden 73: 394–416.[CrossRef][Web of Science]

Tiffney, B. H., AND S. J. Mazer. 1995. Angiosperm growth habit, dispersal and diversification reconsidered. Evolutionary Ecology 9: 93–117.[CrossRef][Web of Science]

Tiffney, B. H., AND K. J. Niklas. 1985. Clonal growth in land plants: A paleobotanical perspective. In J. B. C. Jackson, L. W. Buss, and R. E. Cook [eds.], Population biology and evolution of clonal organisms, 35–66. Yale University Press, New Haven, Connecticut, USA.

Tischler, G. 1955. Der Grad der Polyploidie bei den Angiospermen in verschiedenen Grossarealen. Cytologia 20: 101–118.

Van Tienderen, P. H. 1990. Morphological variation in Plantago lanceolata: Limits of plasticity. Evolutionary Trends in Plants 4: 35–43.[Web of Science]

Verdú, M. 2002. Age at maturity and diversification in woody angiosperms. Evolution 56: 1352–1361.[Web of Science][Medline]

Verhoeven, J. K. F., H. Poorter, E. Nevo, AND A. Biere. 2008. Habitat natural selection at a flowering time QTL is a main driver of local adaptation in two wild barley populations. Molecular Ecology 17: 3416–3424.[Medline]

Wallace, A. R. 1889. Darwinism: An exposition of the theory of natural selection. Macmillan, London, UK.

Waser, M. 1998. Pollination, angiosperm speciation, and the nature of species boundaries. Oikos 82: 198–201.[CrossRef][Web of Science]

Waser, N. M., L. Chittka, M. V. Price, N. M. Williams, AND J. Ollerton. 1996. Generalization in pollinations systems and why it matters. Ecology 77: 1043–1060.[CrossRef]

Wikström, N., V. Savolainen, AND M. W. Chase. 2001. Evolution of the angiosperms: Calibrating the family tree. Proceedings of the Royal Society, B, Biological Sciences 268: 2211–2220.[Abstract/Free Full Text]

Williams, G. C. 1975. Sex and evolution. Princeton University Press, Princeton, New Jersey, USA.

Wing, S., AND L. D. Boucher. 1998. Ecological aspects of the Cretaceous flowering plant radiation. Annual Review of Earth and Planetary Sciences 26: 379–421.[CrossRef][Web of Science]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				R. A. Stockey, S. W. Graham, and P. R. Crane Introduction to the Darwin special issue: The abominable mystery1 Am. J. Botany, January 1, 2009; 96(1): 3 - 4. [Full Text] [PDF]

				G. W. Rothwell, W. L. Crepet, and R. A. Stockey Is the anthophyte hypothesis alive and well? New evidence from the reproductive structures of Bennettitales Am. J. Botany, January 1, 2009; 96(1): 296 - 322. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: ajb.0800126v1 96/1/366    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (8) Citing Articles via Google Scholar Google Scholar Articles by Crepet, W. L. Articles by Niklas, K. J. Search for Related Content PubMed Articles by Crepet, W. L. Articles by Niklas, K. J. Agricola Articles by Crepet, W. L. Articles by Niklas, K. J. Social Bookmarking                            What's this?