Printer Friendly

Perspective: evolutionary patterns in the fossil record.

Some of the major questions in evolutionary paleobiology concern the origin and fate of biological innovations, the pace of evolution, clade replacements, patterns of origination and extinction, and other large-scale trends such as changes in organismal complexity, ecosystem structure, and biological diversity. The very nature of these questions, which involve not just single clades but the global biota, often demands that we document evolutionary patterns and search for underlying mechanisms with data spanning tens to hundreds of millions of years and cutting across phylogenetic lines. Although patterns within clades constitute a major area of paleobiology, they do not exhaust the rich evolutionary history that we find in the fossil record (Van Valen 1991; Gould 1995; Vermeij 1995). Genealogy is, of course, an essential part of many macroevolutionary analyses (Simpson 1953; Stanley 1979; Raup 1985; Schoch 1986; Fisher 1991; Smith 1994). It is sometimes central, as when we test for a pervasive tendency for evolutionary transitions to have a particular direction from ancestor to descendant (McShea 1994), and sometimes a relatively minor component of research, as when we proceed from the basic fact that the global biota whose evolution we study consists of multiple, independent clades. Whereas several recent works (e.g., Schoch 1986; Forey et al. 1992) have provided useful summaries of phylogenetic methods, a more important, and often more neglected, question concerns what to do with phylogenetic relationships once we have some estimate of them.

Armed with the potential power that phylogenetic analysis can bring to macroevolutionary studies, several works (e.g., Novacek and Norell 1982; Patterson and Smith 1987; Norell 1992; and Smith 1994, to name just a few) have broken with the statistical approach that has characterized much of evolutionary paleobiology. According to this approach, the noise inherent in the data of systematic biology can be surmounted with samples sufficiently large to allow a clear signal to emerge. One alternative (Smith 1994) proposes that problems such as the incompleteness of the fossil record can be solved by exclusively cladistic methods, even when the questions are not primarily genealogical. Among recent discussions of phylogenetic analysis in paleobiology, this attitude differs perhaps most from that of Fisher (1991). One of Fisher's main points is that, since a greater variety of information should generally help us understand evolutionary patterns and mechanisms more fully, we should strive to add data on genealogical relationships to other kinds of data. Phylogenetic analysis is seen to strengthen, rather than replace, other approaches.

In this essay I will argue that the broader approach, adding rather than substituting phylogenetic analysis, is more appropriate. Claims that phylogenetic analysis overcomes the problems inherent in other approaches have not been adequately supported theoretically or empirically. Traditional Linnaean taxa have been unduly dismissed by some workers on theoretical grounds without adequate consideration of their empirical properties. I will not present arguments based on the inaccuracy of cladograms. The reliability of cladograms and evolutionary trees is an important issue that must ultimately be taken into account (e.g., Felsenstein 1985; Archie 1989; Alroy 1994; Huelsenbeck 1994, 1995; Hillis 1995; Wagner 1995a), but, even if we knew the true evolutionary tree of some group, we would still not want to discard other (i.e., nongenealogical) information. I will emphasize the study of diversity patterns in paleobiology, but similar arguments could be made for other areas of research, such as origination and extinction trends, clade replacements, and other large-scale changes in ecosystems and biotic interactions (Vermeij 1987, 1995; Van Valen 1985; 1991; Gould 1995). This essay began as a review of Andrew B. Smith's (1994) recent book, Systematics and the Fossil Record. Because Smith touches on so much that is central to macro-evolution and paleobiology, I felt a broader scope was necessary. I thank the Evolution editors for kindly allowing me this latitude.


Taxonomic data (tabulations of originations, extinctions, and diversity of taxonomic units) are often readily available, and they have the potential to convey a wide variety of information, including morphology, evolutionary rates, and geographic distributions. Therefore, such data have played a central role in paleobiology (e.g., Lyell 1883; Simpson 1944, 1953; Muller 1955; Newell 1967; Valentine 1969, 1973; Van Valen 1973a, 1985; Raup 1976a,b, 1985; Stanley, 1979; Sepkoski 1981, among many others). To cite just a few lines of research, the pace of taxonomic origination and extinction has been used as a proxy for general evolutionary rates (Simpson 1944, 1953; Stanley 1979), patterns of taxonomic duration have contributed to hypotheses on species interactions in macroevolutionary dynamics (Van Valen 1973a), the discordance among temporal patterns of diversity at different taxonomic levels has enabled important inferences regarding changes in the ecological structure of marine communities over geologic time (Valentine 1969, 1973; Erwin et al. 1987), and similar discordances in environmental patterns of origination have given us insight into the ecological context of morphological innovations (e.g., Jablonski and Bottjer 1991). These areas of research usually assume implicitly that taxa, generally above the species level, provide a reasonable proxy for biological data that can be more difficult to observe or quantify directly.

All this taxon counting, as it is sometimes disparagingly called, raises important questions: (1) How good is the correlation between higher taxa and lineages in the history of origination, extinction, and diversity? (2) How can some of the shortcomings of the fossil record, for example its incompleteness, be taken into account in large-scale analyses of taxonomic data? (3) To what extent do taxonomic data provide reasonable proxies for other biologically interesting quantities, such as morphological divergence and ecological and functional innovation? Approaches to these questions have included sampling theory (e.g., Raup 1975, 1979; Marshall 1991), comparison between taxonomic data and morphological or ecological data (e.g., Van Valen 1971; Saundersand Swan 1984; Foote 1993), comparison of largely independent data compilations at different taxonomic levels (e.g., Sepkoski et al. 1981), comparison between well preserved and poorly preserved taxa (e.g., Sepkoski 1990, Jablonski and Bottjer 1991), phylogenetic analysis (e.g., Wagner 1995a,b,c), and mathematical modeling of evolutionary trees (e.g., Valentine and Walker 1986; Sepkoski and Kendrick 1993).

Patterns among Higher Taxa as Reflections of Lower-Level Patterns

Because tabulating taxonomic data at the species level in the fossil record is laborious and subject to relatively large sampling error (e.g., Raup 1979), it has been standard practice to compile large-scale databases, usually at a global scale and covering tens to hundreds of millions of years, at the genus level or higher. The theory beneath the assumption that higher-level taxonomic patterns can reflect lineage-level patterns is not very deep. The origination or extinction of a higher taxon ideally involves at least one lineage-level event, and, unless that taxon is polyphyletic, its presence within a time interval implies the presence of at least one lineage. This justification for using taxonomic data does not involve the assumption that higher taxa are "real" - by which a cladist generally means holophyletic - or that taxonomic ranks are assigned using uniform standards. [I follow Ashlock (1971) in referring to a holophyletic group as an ancestor and all its descendants, a paraphyletic group as an ancestor and some, but not all, of its descendants, and a monophyletic group as either holophyletic or paraphyletic.]

Some taxa are polyphyletic, and many are paraphyletic. Moreover, the reasons for assigning a given taxonomic rank are quite heterogeneous (see below). It is generally hoped that, if the database is large enough and the questions are properly framed, these heterogeneities will essentially cancel out. Nevertheless, taxonomic heterogeneity may obscure underlying patterns to an unknown extent. Partly for this reason, it has been suggested that we abandon traditional Linnaean taxa in favor of holophyletic taxa (e.g., Patterson and Smith 1987; Smith 1994). However, holophyletic taxa are likewise heterogeneous in the amount of diversity, morphological divergence, and distinctness they represent (Doyle and Donoghue 1993), even if they are sister-clades. Nevertheless, because holophyletic taxa are regarded by some as "real," many workers have supposed that they provide a better reflection of underlying lineage-level patterns.

This claim may prove to be true in the end, but it has scarcely been tested empirically. Depending on the question, heterogeneity and lack of "reality" among units of accounting need not hinder reasonable estimation of some quantity of interest. One would be very surprised, for example, if temporal changes in the size of the global human population were not matched by changes in the number of family units, even though families are extremely heterogeneous in inclusiveness and other qualities. The problem in paleobiology, of course, is that we do not know the underlying pattern of lineages that we are trying to estimate with higher taxa. Doyle and Donoghue (1993) give a hypothetical example in which patterns among lineages are not reflected by patterns in ranked taxa, even holophyletic taxa. Smith (1994) discusses similar cases. It is clear that we can contrive particular cases to support either position, but what about a broader approach? Sepkoski and Kendrick (1993), following Sepkoski (1989), modeled more general evolutionary trees to investigate the ability of arbitrarily delimited paraphyletic taxa and holophyletic taxa to detect overall diversity patterns and extinction events. Although the limited range of their simulations precludes complete generalization, Sepkoski and Kendrick found that paraphyletic and holophyletic taxa served as reasonable proxies for underlying patterns in lineages. Holophyletic taxa performed better with extremely good sampling, but with poorer sampling (perhaps more realistic), paraphyletic taxa were even more faithful to lineage patterns than were holophyletic taxa.

When it comes to predicting the effects of taxonomic practice on perceived evolutionary patterns, our intuition may often be misleading. Gould et al. (1977, 1987) found that higher taxa of marine animals originating early in the Paleozoic and mammalian taxa originating early in the Cenozoic have lower "centers of gravity" than taxa originating later, i.e., their taxonomic diversity is concentrated earlier in their history. It has been argued that, because many of the early originating taxa are paraphyletic, and therefore have their later descendants placed in holophyletic daughter taxa, this bottom-heaviness is an artifact (e.g., Smith 1994). There are even examples in which this has been demonstrated (e.g., early Paleozoic eocrinoids, a group of echinoderms from which ultimately descended blastoids, rhombiferans, and some other taxa) (Smith 1994). So, at first glance, this argument seems reasonable, until we stop to think that the descendants may come from anywhere in the history of the paraphyletic ancestral taxon. Thus, center of gravity may be biased upward or downward. Using simulation of evolutionary trees and empirical comparison of numerous holophyletic and paraphyletic mammalian families, Mark D. Uhen (pers. comm., 1995), in the broadest study of this kind to date, has found that paraphyletic taxa are not systematically more bottom-heavy. Further studies like Uhen's should allow more reliable insight into the effects of taxonomic practice on large-scale evolutionary patterns.

Empirically, the agreement between holophyletic and paraphyletic taxa (which is quite distinct from the agreement between higher taxa and species) remains an open question. Patterson and Smith (1987) found that certain peaks in the geologic history of extinction are apparent when all echinoderm and fish families are considered but not when only holophyletic families are included. In contrast, Wagner (1995c) found that the pattern of diversity among traditional taxa of early Paleozoic gastropods (an apparently more densely sampled group) broadly agrees with that based on strictly defined clades. In light of Sepkoski and Kendrick's (1993) conclusions regarding sampling, it is conceivable that Patterson and Smith's result reflects the relatively poor record of echinoderms and fishes (Sepkoski 1989, 1990), but this is by no means certain.

In exploring the effects of taxonomic practice on perceived evolutionary patterns, comparison among different taxonomic standards and different datasets is often enlightening. For example, Williams (1957) showed that, between the late nineteenth and mid twentieth centuries, apparent peaks of taxonomic evolution in brachiopods corresponded with the geologic period in which the leading worker of the day specialized. Therefore, Williams suggested, times of apparently high evolutionary activity should be regarded with caution. Sepkoski et al. (1981) demonstrated a correlation among diversity of traces of animal behavior, species diversity, genus diversity, family diversity, and species richness within communities for the Phanerozoic marine record. The last index of diversity is especially important, because it is measured at single localities, and therefore may not be strongly biased by the amount of globally available fossiliferous rock (though it may be sensitive to other sampling biases; see Raymond and Metz 1995). For trilobites, Foote (1993) found that diversity of genera named up to 1959 and species named after 1959 show broadly similar temporal patterns.

Given the paucity of empirical studies, we still have much to learn about the relationship between ranked higher taxa and species and the match between holophyletic and paraphyletic higher taxa. The hope (e.g., Smith 1994) that abandoning traditional taxa and allowing only holophyletic taxa will solve most problems concerning taxonomic data may prove to be justified in the end, but we are not now in an empirical position to have much confidence in this possibility.

Incomplete Sampling of Fossil Taxa

A biologic group is not preserved everywhere it existed, in space or in time; thus, observed geographic and temporal ranges generally underestimate true ranges. Because the relationship between apparent and observed ranges colors our perception of taxonomic patterns, it is crucial that we understand the effects of incomplete preservation. Incompleteness might not greatly hinder diversity studies if all taxa were equally affected at all times, but some organisms have instrinsically lower preservation potential because of their structure and habitat. Time-dependent biases, such as the increase in the amount of fossiliferous rock through the Phanerozoic, are difficult to factor out, although some noteworthy attempts have been made (e.g., Sepkoski 1976; Pease 1985, 1988a,b,c). Many time-dependent biases are subtle and hard to quantify, such as the possibility that the kinds of traits used in classifying living species may allow finer distinctions among their fossil representatives than can be made among the fossil representatives of extinct groups (e.g., Raup 1979). One way to deal with this and other aspects of the "Pull of the Recent" is to discard extant taxa (e.g., Raup 1991), but this introduces other problems. For example, extinct and living taxa represent different parts of the overall distribution of durations (Van Valen 1979).

Two important questions arise in interpreting absence of a taxon or lineage in time or space. (1) How confident can we be that a nonoccurrence reflects a true absence rather than a failure of preservation? (2) When should we infer the presence of a taxon or lineage that is not actually preserved? At least two probabilistic approaches have been used to address the first question. (1) The notion of taphonomic (preservational) control is that if we find remains of an organism with similar preservation potential as the organism in question, then we can probably take the absence of the latter at face value. This sort of reasoning has a long history; Bottjer and Jablonski (1988) present a recent application to the study of environmental patterns of taxonomic diversity. Although the approach is imperfect, since a taxon may be present and not preserved even if its control taxon is preserved, it seems clearly preferable to the alternative of assuming that what you see is what you get. (2) Some notion of quantified confidence limits on stratigraphic (temporal) first and last appearances goes back at least to Shaw (1964; see also Paul 1982; Strauss and Sadler 1989; Marshall 1991), although the intuitive concept has a longer history. The more fossil remains from which a lineage is known within its observed temporal range, the greater is its inferred preservation potential, and, therefore, the more confident we can be that its lack of preservation in environments similar to those in which it is otherwise found represents a true absence.

Most attention recently has focused on inferring the presence of a taxon or lineage not preserved at a particular time. Although it has been claimed (Smith 1994: p. viii) that cladistic analysis represents the most promising way to compensate for the effects of sampling and preservation, this claim has not been substantiated with a rigorous comparison among alternative methods. [An important, related issue concerns the effects of incompleteness and bias on the estimation of phylogenetic relationships. This is currently being studied in detail by John Huelsenbeck (pers. comm., 1995).]

The traditional range-through method, which can be thought of as a phylogenetic approach, is perhaps the simplest way to infer the presence of an unpreserved taxon. If a taxon is not polyphyletic, then its presence in intervening time intervals can be safely inferred from its presence at earlier and later times. Extending ranges beyond the observed first and last appearance is more problematic. If we are comfortable assuming that preservation potential does not vary systematically within the history of some taxon, then basic sampling theory allows an average range extension to be determined (e.g., Strauss and Sadler 1989; Marshall 1991; see also Raup 1989 and Marshall 1994b for approaches involving somewhat less stringent assumptions). Although the assumptions behind the range extension may not be met, it still seems preferable to attempt some correction than none at all, particularly in more poorly preserved taxa. Simulations and analytic approaches investigating the robustness of the range-extension method under various violations of the assumptions could be most illuminating (see Marshall 1991, 1994b). In addition to extending individual temporal ranges of taxa, it may be possible to estimate entire frequency distributions of true taxonomic durations based on the distributions of observed fossil ranges (Foote and Raup 1996).

An additional tool for extending taxonomic ranges is to incorporate estimates of phylogenetic relationships (Fig. 1; e.g., Fisher 1982, 1991; Novacek and Norell 1982; Norell 1992, 1993). These corrections, which provide minimal range extensions, are of at least three kinds. [The terminology, involving range extensions, ghost lineages, and ghost taxa is a bit complex (Norell 1992, 1993; Smith 1994: pp. 137-139), so I will avoid it; the concepts are clear enough without it.] (1) Suppose two taxa (A and B) having nonoverlapping temporal ranges are related as ancestor and descendant. Then there must have been at least one lineage during the intervening interval of nonpreservation. (2) Suppose two taxa are most closely related, and that the first appearance of one of them (B) is later than that of the other (A). Then, if one is not directly or indirectly ancestral to the other, there must have been at least one unobserved lineage ranging between the first appearances of A and B. (3) Assume that B and C are sister-taxa and that (B + C) is most closely related to A, whose first appearance is before those of both B and C. Again, if there are no direct or indirect ancestor-descendant relationships among A, B, and C, there must have been an unobserved lineage between the first appearance of A and the first appearance of B or C, whichever is older (see Fisher 1982: text-figure 1; Novacek and Norell 1982; Norell 1992, 1993). Clearly, the range corrections in (2) and (3) are sensitive to whether putative sister-taxa are in fact ancestor-descendant pairs. Treating sister-taxa as having an unobserved common ancestor, if they are in fact related as ancestor and descendant, tends to push originations further back in time than they should be, and may thus exaggerate the pace of a clade's early diversification. [This is not the place to discuss whether ancestor-descendant pairs can be identified with confidence in practice. Hennig (1966), Van Valen (1978), Fisher (1991, 1994), and Smith (1994: ch. 6), among others, present reasonable arguments and/or examples showing that they can, but this is still debated (e.g., Schoch 1986).]

Although one may be uneasy about the rarity with which ancestor-descendant relationships are postulated in some phylogenetic studies, there is perhaps a more fundamental issue. Suppose we are confident that relationships are well estimated, and that therefore many originations should be pushed back in time and many lineages interpolated according to an available cladogram. Should we then abandon probabilistic approaches in favor of phylogenetic approaches (Novacek and Norell 1982; Norell 1992; Smith 1994), or should we combine the two? A basic asymmetry in the phylogenetic approach suggests that, if our goal is to tabulate diversity, origination, and extinction, then we should retain the probabilistic approaches. Incomplete preservation truncates temporal ranges of taxa at both ends. Yet, under the phylogenetic approach, terminal lineages are not given forward range extensions. In randomly branching evolutionary trees with long-term speciation and extinction rates equal to each other (which must be the case for extinct groups), half of all lineages, on average, become extinct without issue. Even if long-term origination rate exceeds extinction rate, the proportion of lineages that terminate without descendants is, under simple branching models, not much less than one-half. [When long spans of geologic time have elapsed, this proportion can be shown to equal q/(p + q), where p is the origination rate per lineage per time unit, and q is the extinction rate (Kendall 1948: equation 52; Foote, unpubl. notes).] Thus, the problem of terminal lineages cannot be ignored. This problem is exacerbated by incompleteness; because many preserved lineages have descendants that are not preserved, we observe a large number of truly or apparently terminal lineages for which a forward range extension is desired.

Thus, the phylogenetic approach makes range corrections preferentially backward in time, even though sampling theory dictates that they be made (nearly) equally in both directions. The implications of this asymmetry have not been fully explored, but the effects in simple cases are apparent. For example, whether diversity declined gradually or abruptly during biotic crises is crucial to understanding the causes of major extinction events. However, incomplete preservation would make the simultaneous extinction of many lineages appear spread out in time (Signor and Lipps 1982). Forward range extension of terminal lineages yields a potential test of whether an apparent pattern of gradual lineage termination is consistent with an actual pattern of simultaneous loss of many lineages (Raup 1989; Springer 1990; Marshall 1994a). A purely phylogenetic approach would perforce take the ends of terminal lineages at face value, and would thereby forgo an opportunity to estimate the actual extinction pattern from the observed data.

It is difficult to evaluate the strengths and weaknesses of probabilistic and phylogenetic approaches in any particular case, since we do not know the true underlying pattern of diversity, origination, and extinction. Simulation studies in progress (Christine M. Janis and J.J. Sepkoski, Jr., pers. comm., 1995), are exploring various models of diversity and preservation. These should help suggest under which combinations of taxonomic rates, paleontological completeness, and tree topology the benefits of the probabilistic and phylogenetic approaches emerge most clearly.

Taxa as Proxies for Nontaxonomic Biological Data

Many insights into the history of life have come from the use of taxa as proxies for biological information. For example, the morphological differences among taxa of higher rank are generally thought to be greater than those among taxa of lower rank. Although this perception has not been thoroughly tested, there is reason to believe that it is more than a mere hope. For example, morphological data on Paleozoic blastozoans, trilobites and crinoids, the groups with which I am most familiar, suggest the unsurprising pattern that differences among species within the same genus are smaller than differences among genera within the same order, which are smaller than differences among orders within the same class. There are great theoretical and practical difficulties with compiling morphological data for all living organisms, but the apparent concordance between morphological divergence and taxonomic rank has enabled some seminal research into the history of life. For example, Valentine (1969) noted that the early appearance of taxa of high rank has important implications for the evolution of novelties and the long-term evolution of global ecosystems, given that taxonomic rank seems to convey some information about morphological and ecological divergence.

Before proceeding to the practical use of taxonomic rank, I should make a few general comments on the matter. (1) The factors that go into assigning taxonomic rank, from morphological divergence to taxonomic convenience, are quite varied (e.g., Smith and Patterson 1988; Smith 1988, 1994; Mayr and Ashlock 1991). However, we must make an important distinction regarding heterogeneity of ranks within, versus among, major biologic groups. We know that a family of clams and one of mammals may not mean the same thing objectively, and it is hard to imagine how we would compare taxa of a given rank across such disparate groups. Cherry et al. (1982) present an approach for tetrapods. Van Valen (1973b) has made perhaps the broadest attempt, in which the comparison is not among the properties of organisms directly, but rather the properties of classifications. It may be that the difficulty of comparing the same rank among disparate groups implies equally large problems with assuming a meaningful difference among ranks within the same group, but this does not follow logically. Empirically, in fact, there is some evidence that patterns of morphological diversification vary predictably among taxa of different rank within the same major biologic group (see Other Evolutionary Patterns, below). Holman (1989) has made a similar suggestion regarding taxonomic durations and origination and extinction patterns (see discussion below).

(2) Of course, the problems of ranking persist whether taxa are holophyletic, paraphyletic, or polyphyletic (Van Valen 1978, 1989; Mayr and Ashlock, 1991; Doyle and Donoghue 1993). In one scheme, rank is based mainly on inclusiveness (Smith 1994: p. 98). However, inclusiveness alone essentially requires that fossil and Recent taxa be ranked separately, one of several cladistic procedures that can lead to serious practical difficulties (e.g., Mayr 1974; Van Valen 1978, 1989; Mayr and Ashlock 1991). It may be that ranks never have been strictly comparable, but does this mean that an additional, forced disparity between fossil and Recent taxa should not concern us? Systematists have spent decades trying to achieve some sort of taxonomic standardization in many groups. The result may not be perfect, but the effort has yielded positive steps toward some sort of internal consistency. To abandon this goal, and to suggest that, since rank is arbitrary, it does not really matter how arbitrary, seems like a definite step backward.

(3) The history of classification of major biologic groups is very complex. One should therefore be cautious in evaluating why ranks have been assigned in any particular case. For example, Smith's (1994: p. 91) statement, that the living classes of echinoderms are given the class rank for no reason other than that they have achieved high taxonomic diversity, suggests that no other factors, such as the morphological distinctions among sea cucumbers, echinoids, starfish, brittle stars, and sea lilies, have played a role. However, these and other factors should be explicitly evaluated before being so casually dismissed.

It is helpful to focus on two principal problems in the practical use of taxonomic rank. (1) Given that taxa are defined heterogeneously and somewhat arbitrarily, how can one have any confidence in the patterns we infer from taxonomic data? (2) To what extent can large-scale temporal patterns reflect simple constraints inherent in the branching geometry of evolutionary trees?

(1) It is unduly negative to suppose that, since ranked taxa may or may not represent what we think they do (e.g., with respect to morphological divergence), that we should assume they do not (Smith 1988; 1994). It seems again that an empirical approach is much more promising. Surely, we cannot always assume that orders or classes occupy distinctive adaptive zones, for example, but this can be tested. Van Valen (1971) has done so for mammals. Studies like Van Valen's allow some confidence that, at the appropriate scale, taxa of higher rank can serve as reasonable proxies for ecological data (Bambach 1985). Likewise, a tabulation of taxa of higher rank need not provide a good proxy for overall morphological diversity, but it may. Figure 2 compares estimates of morphological diversity, based on the dispersion of species in multivariate morphological space, to the number of traditional higher taxa (including holophyletic, paraphyletic, and even some polyphyletic groups) for three large clades of Paleozoic invertebrates. The temporal patterns of morphological diversity and higher taxonomic diversity are not perfectly concordant, but, at least for crinoids and blastozoans, one would not have been too far off the mark if one had used the number of higher taxa as a proxy for morphological diversity.

If the few cited examples prove to be typical, then the heterogeneity of taxa does not ipso facto preclude their ability to reflect biologically interesting patterns (cf. Smith 1988). If morphological data are available, one should use them. However, until such data are gathered for most fossil groups (an enormous task that is not close to completion), it is shortsighted to ignore the morphological information that is implicit, albeit imperfectly, in taxonomic data.

(2) Many striking patterns in the fossil record may reflect the geometry of evolutionary trees. Raup and Gould (1974) showed that trends, character correlations, and other features could result from randomly branching trees. Similarly, Foote (1991, 1993) pointed out that the concentration of morphological diversity late in a clade's history, observed in a number of groups, is to be expected if origination and extinction are effectively random with respect to morphology. Raup (1983) suggested that, because contemporaneous pairs of species have a high probability of sharing a most recent common ancestor distantly removed in time, the early appearance of higher-ranked taxa might be seen as a constraint of branching geometry. This topological constraint does not explain away all interesting patterns that higher taxa may exhibit, however. If phyla prove to reflect greater degrees of morphologial divergence than classes, then, as Raup emphasized, the pattern of initially rapid morphological evolution implied by the early origin of phyla is not explained by the branching constraint. Moreover, Valentine (1969, 1990) showed that, not only are the origins of phyla concentrated before those of classes, which are concentrated before those of orders, and so on, but also that the peaks in diversity are offset systematically, phyla before classes, and so on (see also Simpson 1953: chapter VII). Valentine and Walker (1986) have demonstrated one way to obtain staggered diversity peaks with an ecologically grounded diversification model.

It is important to emphasize that Raup's analytic approach to divergence times does not concern patterns among nested taxa of different rank. Given the heterogeneity of taxonomic practice, these are probably not tractable analytically (Raup 1983). In an attempt to step beyond the domain of Raup's approach, Smith (1994: p. 100) tries to explain away Valentine's empirical results, arguing that these are mere artifacts of the greater inclusiveness of higher taxa. To investigate the plausibility of this argument, I generated evolutionary trees according to simple branching models designed to yield various patterns of taxonomic diversification (see Raup 1985). I then agglomerated lineages into nested taxa at three levels, using monophyly and inclusiveness as the sole criteria of ranking (i.e., "phyla" have greater diversity than "classes," which are larger than "orders"). A few simulations yield the Valentinian pattern of sequential diversity peaks, but this is far from the most common outcome. So it is possible that a most intriguing pattern in the history of life may reflect topological constraints of trees formed under relatively uninteresting biological rules, but nobody has yet shown that it is terribly likely. Nor has it been shown that branching constraints can account for Holman's (1989) finding that taxonomic durations and rates of origination and extinction vary systematically depending on the rank of the taxon within which they are analyzed.


Some recent paleontological works (e.g., Fisher 1991; Smith 1994) have reemphasized the ways in which genealogical relationships can help us understand large-scale evolutionary patterns. However, the value of genealogical data does not imply that we have much to gain by viewing phylogenetic analysis as an alternative, rather than a complement, to other approaches (see Fisher 1991). To illustrate this point, I will discuss one area with which I am familiar, namely the study of morphological diversity; others could equally well be cited.

The concern that taxonomic data may not always provide a good proxy for morphological diversity led me to extend previous work (e.g., Raup 1967; Saunders and Swan 1984) by documenting evolutionary patterns in multivariate morphological space for a number of fossil clades (e.g., Foote 1993). The general approach has been to contrast certain idealized evolutionary models, and to compare their expectations with observed patterns. Though there is no one-to-one correspondence between large-scale patterns and smaller-scale mechanisms, certain patterns in morphospace occupation narrow down the plausible range of underlying mechanisms. For example, if morphological transitions are large early in the history of a group and small later, then, in a diversifying clade, variance in morphology among species will tend to increase rapidly at first, then more slowly. This expectation is easily demonstrated analytically and by simulation (Foote, unpubl. manuscript). Just such an initial acceleration in the differences among species is observed in blastozoan echinoderms (Foote 1992), consistent with the initially higher rates of morphological evolution advocated by some workers (e.g., Sprinkle 1973), but questioned by others (e.g., Smith 1988). A similar pattern has been found in other groups (e.g., bryozoa; Anstey and Pachut 1992).

It should go without saying that different approaches are appropriate for different questions, and that more than one approach can usefully be applied to a particular question. My criticism (1992) of quantifying differences among species based exclusively on derived characters was meant to imply only that this approach is not appropriate for what I was addressing, namely net differences among species, which are reflected in both primitive and derived characters. If one is interested in the rate at which novelties arise (e.g., Derstler 1982), then the focus on derived characters is quite appropriate. Smith (1994), among others, has criticized the use of phenetic distance among species because it "takes no account of character convergence in the tree, mistaking homoplasy for homology" (p. 169), and therefore "underestimate[s] the amount of morphological change that has occurred between end members in comparison to cladistic methods" (p. 170). The case he uses to illustrate this point is a concocted one; it is not clear how representative it is.

More importantly, however, the distance method does not "mistake" homoplasy for homology because it is not intended to tease the two apart; it is not a phylogenetic method (see Gould 1991; Foote 1995). The distance method simply says that if character state 0 goes to 1 and back to 0, there has been no net change, even though there have been two units of total change. The obvious analogy is with taxonomic turnover and diversity. A clade can achieve high diversity through rapid turnover (e.g., genera of Cambrian trilobites) or slow turnover (e.g., families of insects; Labandeira and Sepkoski 1993). Turnover (analogous to total character change) and diversity (analogous to net change) reflect different aspects of the kinetics and dynamics of taxonomic evolution. Diversity metrics do not "mistake" low turnover for high turnover, because they concern a completely different aspect of taxonomic data than do measures of turnover.

For a given problem, either net or total change may be more interesting, but neither has logical pre-eminence in general. In fact, in the case of Paleozoic crinoids, the discordance between the two aspects of character change is itself revealing (Foote 1994, 1995). For over 200 million years, there was abundant total change (estimated indirectly by taxonomic turnover) and limited net change (estimated more directly with morphological data). This suggests some limits to crinoid form that could not be or were not surpassed (Ausich 1988; see Fisher 1982 for an example of the bearing of genealogical information on arguments for constraint). If we had insisted that only total change or net change were relevant, we might have overlooked an interesting evolutionary pattern.

One might be tempted to dismiss the insight of many paleontologists (summarized by Valentine 1980) that the pattern of morphological diversification of the major clades of life differs substantially from that of their subclades, since this view was not based on extensive genealogical analysis, and since heterogeneity of taxonomic practice supposedly precludes any such comparison among ranks in the first place. However, the discordance among taxonomic levels advocated by Valentine and others has been supported by morphological analysis. For the few groups investigated so far, clades of the highest rank filled their design space relatively rapidly, while nested subclades proliferated more gradually in morphospace (e.g., Foote 1993, 1995). If this difference among taxonomic levels holds more generally (and we have a long way to go before we know whether it does), it suggests that rank may not be as arbitrary as sometimes claimed. More importantly, it shows that we would overlook biologically significant patterns in the fossil record if we dismissed all evolutionary insight prior to the development of cladistics.

Because I have been interested in the large-scale unfolding of morphological diversity, I have focused on net morphological change. Nevertheless, I agree with Fisher (1991), Bodenbender (1994), and Wagner (1995b) that more detailed analysis of pathways of change, when genealogical data are available, can be most illuminating. A good example of the union of genealogical and phenetic data is found in the work of B. E. Bodenbender (1994; pers. comm., 1995). After extensive phylogenetic study involving careful character analysis, he found that the Permian survivors among blastoid echinoderms were drawn from many branches of the evolutionary tree. This strengthened the conclusion, based solely on the distributions of species in a landmark-based morphospace, that large-scale patterns of survival appeared not to be strongly selective with respect to morphology (Foote 1991, 1993). Similarly, Wagner (1995d) combined some existing phylogenetic schemes with character data from phenetic analysis of blastozoan echinoderms (Foote 1992) to estimate the number of morphological transitions per node on the cladogram. Consistent with the inference drawn from a purely phenetic approach (Foote 1992), Wagner found that mean step size decreased through the Paleozoic. Wills et al. (1994) also found that phenetic and cladistic analyses of morphological data supported the same conclusion, that the morphological disparity of Cambrian and Recent arthropods is comparable.


The utility of incorporating genealogical information into paleobiological analyses, neither novel nor radical in itself, is clear enough. Unfortunately, reliable genealogies are not yet available for most fossil groups. The work of Van Valen (1971), Sepkoski and Kendrick (1993), Bodenbender (1994), Wagner (1995b,c,d), and Wills et al. (1994), among others, suggests that this lack of genealogical data need not stop us in our tracks. When evolutionary trees become available, there will be much to gain by linking them with other data (Fisher 1991). But we will sacrifice too much information and understanding if we discard everything else we know while we produce estimates of evolutionary trees.

The desire to focus exclusively on genealogical data (e.g., Smith 1994) seems to reflect partly a misunderstanding of what noncladistic (not anticladistic!) data represent, and partly a presumption that if a phylogeny is necessary for some questions, it must be sufficient for any question we might wish to ask. But the history of paleobiology and evolutionary biology shows that this is not the case. I have illustrated this point with the study of taxonomic and morphological diversity. However, there are many other important areas of research needed to round out the broader macroevolutionary picture, such as ecology, geography, energy flow, and climate (Van Valen 1976, 1991; Bambach 1993; Vermeij 1995). Much seminal research and many entire areas of inquiry, whose main focus is not genealogical, would be disallowed under a solely phylogenetic research program. These include aspects of theoretical and constructional morphology (e.g., Raup 1967), the study of rates and trends (e.g., Simpson 1953; Stanley 1979; McShea 1994; Valentine et al. 1994), evolutionary paleoecology, particularly long-term changes in the composition and structure of the biosphere (e.g., Sepkoski 1981; Vermeij 1987, 1995; Bambach 1985, 1993), and, in general, the study of biotal evolution (Van Valen 1991). Yet these areas of research have been crucial in our understanding of the history of life (see Gould 1995 and Vermeij 1995 for recent summaries). And, as illustrated by the example of morphological diversification, the conclusions of this "forbidden" research have been supported by a number of different approaches, including phylogenetic analysis. To abolish this understanding, built on decades of careful, multidisciplinary work, merely because it does not rest solely on cladistic analysis, would be, as Howard Margolis said in another context (1987: p. 223), "like swatting a fly that had landed on a priceless painting."


I thank B. E. Bodenbender, D. Jablonski, C. M. Janis, D. J. Miller, J. J. Sepkoski Jr., M. D. Uhen, G. J. Vermeij, and P. J. Wagner for comments and discussion. I am grateful to G. J. Vermeij for liberally allowing me to extend a review of A. B. Smith's book into a broader essay. Supported by the National Science Foundation (DEB-9207577 and DEB-9496348).


Alroy, J. 1994. Four permutation tests for the presence of phylogenetic structure. Systematic Biology 43:430-437.

Anstey, R. L., and J. F. Pachut. 1992. Cladogenesis and speciation in early bryozoans. Geological Society of America Abstracts with Programs 24:A139.

Archie, J. W. 1989. A randomization test for phylogenetic information in systematic data. Systematic Zoology 38:253-269.

Ashlock, P. D. 1971. Monophyly and associated terms. Systematic Zoology 20:63-69.

Ausich, W. I. 1988. Evolutionary convergence and parallelism in crinoid calyx design. Journal of Paleontology 62:906-916.

Bambach, R. K. 1985. Classes and adaptive variety: The ecology of diversification in marine faunas through the Phanerozoic. Pp. 191-243 in J. W. Valentine, ed. Phanerozoic diversity patterns: Profiles in macroevolution. Princeton University Press, Princeton, NJ.

-----. 1993. Seafood through time: Changes in biomass, energetics, and productivity in the marine ecosystem. Paleobiology 19:372-397.

Bodenbender, B. E. 1994. Stratocladistic reconstruction of blastoid evolutionary history. Geological Society of America Abstracts with Programs 26:A427.

Bottjer, D. J., and D. Jablonski. 1988. Paleoenvironmental patterns in the evolution of post-Paleozoic benthic marine invertebrates. Palaios 3:540-560.

Cherry, L. M., S. M. Case, J. G. Kunkel, J. S. Wyles, and A. C. Wilson. 1982. Body shape metrics and organismal evolution. Evolution 36:914-933.

Derstler, K. L. 1982. Estimating the rate of morphological change in fossil groups. Proceedings, Third North American Paleontological Convention 1:131-136.

Doyle, J. A., and M. J. Donoghue. 1993. Phylogenies and angiosperm diversification. Paleobiology 19:141-167.

Erwin, D. H., J. W. Valentine, and J. J. Sepkoski Jr. 1987. A comparative study of diversification events: The early Paleozoic versus the Mesozoic. Evolution 41:177-186.

Felsenstein, J. 1985. Confidence limits on phylogenies: An approach using the bootstrap. Evolution 39:783-791.

Fisher, D. C. 1982. Phylogenetic and macroevolutionary patterns within the Xiphosurida. Proceedings, Third North American Paleontological Convention 1:175-180.

-----. 1991. Phylogenetic analysis and its application in evolutionary paleobiology. Pp. 103-122 in N. L. Gilinsky and P. W. Signor, eds. Analytical paleobiology (Short Courses in Paleontology, number 4). The Paleontological Society, Knoxville, TN.

-----. 1994. Stratocladistics: Morphological and temporal patterns and their relation to phylogenetic process. Pp. 133-171 in L. Grande and O. Rieppel, eds. Interpreting the hierarchy of nature: From systematic patterns to evolutionary process theories. Academic Press, San Diego, CA.

Foote, M. 1991. Morphological and taxonomic diversity in a clade's history: the blastoid record and stochastic simulations. Contributions from the Museum of Paleontology, University of Michigan 28:101-140.

-----. 1992. Paleozoic record of morphological diversity in blastozoan echinoderms. Proceedings of the National Academy of Sciences, U.S.A. 89:7325-7329.

-----. 1993. Discordance and concordance between morphological and taxonomic diversity. Paleobiology 19:185-204.

-----. 1994. Morphological disparity in Ordovician-Devonian crinoids and the early saturation of morphological space. Paleobiology 20:320-344.

-----. 1995. Morphological diversification of Paleozoic crinoids. Paleobiology 21:273-299.

Foote, M., and D. M. Raup. 1996. Fossil preservation and the stratigraphic ranges of taxa. Paleobiology 22. In press.

Forey, P. L., C. J. Humphries, I. J. Kitching, R. W. Scotland, D. J. Siebert, and D. M. Williams. 1992. Cladistics: A practical course in systematics. Clarendon Press, Oxford.

Fortey, R. A. 1990. Ontogeny, hypostome attachment and trilobite classification. Palaeontology 33:529-576.

Gould, S. J. 1991. The disparity of the Burgess Shale arthropod fauna and the limits of cladistic analysis: Why we must strive to quantify morphospace. Paleobiology 17:411-423.

-----. 1995. A task for Paleobiology at the threshold of majority. Paleobiology 21:1-14.

Gould, S. J., D. M. Raup, J. J. Sepkoski Jr., T. J. M. Schopf, and D. S. Simberloff. 1977. The shape of evolution: A comparison of real and random clades. Paleobiology 3:23-40.

Gould, S. J., N. L. Gilinsky, and R. Z. German. 1987. Asymmetry of lineages and the direction of evolutionary time. Science 236: 1437-1441.

Hennig, W. 1966. Phylogenetic systematics. University of Illinois Press, Urbana.

Hillis, D. M. 1995. Approaches for assessing phylogenetic accuracy. Systematic Biology 44:3-16.

Holman, E. W. 1989. Some evolutionary correlates of higher taxa. Paleobiology 15:357-363.

Huelsenbeck, J. P. 1994. Comparing the stratigraphic record to estimates of phylogeny. Paleobiology 20:470-483.

-----. 1995. Performance of phylogenetic methods in simulation. Systematic Biology 44:17-48.

Jablonski, D., and D. J. Bottjer. 1991. Environmental patterns in the origins of higher taxa: The post-Paleozoic fossil record. Science 252:1831-1833.

Kendall, D. G. 1948. On the generalized "birth-and-death" process. Annals of Mathematical Statistics 19:1-15.

Labandeira, C. C., and J. J. Sepkoski Jr. 1993. Insect diversity in the fossil record. Science 261:310-315.

Lyell, C. 1833 [1991]. Principles of geology, vol. III. John Murray, London [reprinted, University of Chicago Press, Chicago].

Margolis, H. 1987. Patterns, thinking, and cognition: A theory of judgment. University of Chicago Press, Chicago.

Marshall, C. R. 1991. Estimation of taxonomic ranges from the fossil record. Pp. 19-38 in N. L. Gilinsky and P. W. Signor, eds. Analytical paleobiology (Short courses in paleontology, number 4). The Paleontological Society, Knoxville, TN.

----. 1994a. Using the ammonite fossil record to predict the position of the K/T boundary Ir anomaly on Seymour Island, Antarctica. Geological Society of America Abstracts with Programs 26:A394.

-----. 1994b. Confidence intervals on stratigraphic ranges: partial relaxation of the assumption of randomly distributed fossil horizons. Paleobiology 20:459-469.

Mayr, E. 1974. Cladistic analysis or cladistic classification? Zeitschrift fur Zoologische Systematik und Evolutionsforschung 12: 94-128.

Mayr, E., and P. D. Ashlock. 1991. Principles of systematic zoology, 2d ed. McGraw-Hill, New York.

McShea, D. W. 1994. Mechanisms of large-scale evolutionary trends. Evolution 48:1747-1763.

Moore, R. C., and C. Teichert [eds.]. 1978. Treatise on invertebrate paleontology, Part T, Echinodermata 2. The Geological Society of America and the University of Kansas, Boulder, CO and Lawrence, KS.

Muller, A. H. 1955. Der Grossablauf der stammesgeschichtlichen Entwicklung. Fischer, Jena, Germany.

Newell, N. D. 1967. Revolutions in the history of life. Geological Society of America Special Paper 89:63-91.

Norell, M. A. 1992. Taxic origin and temporal diversity: The effect of phylogeny. Pp. 89-118 in M. J. Novacek and Q. D. Wheeler, eds. Extinction and phylogeny. Columbia University Press, New York.

----. 1993. Tree-based approaches to understanding history: Comments on ranks, rules, and the quality of the fossil record. American Journal of Science 293A:407-417.

Novacek, M. J., and M. A. Norell. 1982. Fossils, phylogeny, and taxonomic rates of evolution. Systematic Zoology 31:366-375.

Patterson, C., and A. B. Smith. 1987. Is the periodicity of mass extinctions a taxonomic artefact? Nature 330:248-251.

Paul, C. R. C. 1982. The adequacy of the fossil record. Pp. 75-117 in K. A. Joysey and A. E. Friday, eds. Problems of phylogenetic reconstruction. Academic Press, London.

Pease, C. M. 1985. Biases in the durations and diversities of fossil taxa. Paleobiology 11:272-292.

----. 1988a. Biases in the total extinction rates of fossil taxa. Journal of Theoretical Biology 130:1-7.

-----. 1988b. Biases in the per-taxon origination and extinction rates of fossil taxa. Journal of Theoretical Biology 130:9-30.

-----. 1988c. Biases in the survivorship curves of fossil taxa. Journal of Theoretical Biology. 130:31-48.

Raup, D. M. 1967. Geometric analysis of shell coiling: Coiling in ammonoids. Journal of Paleontology 41:43-65.

-----. 1975. Taxonomic diversity estimation using rarefaction. Paleobiology 1:333-342.

-----. 1976a. Species diversity in the Phanerozoic: A tabulation. Paleobiology 2:279-288.

-----. 1976b. Species diversity in the Phanerozoic: An interpretation. Paleobiology 2:289-297.

-----. 1979. Biases in the fossil record of species and genera. Bulletin of the Carnegie Museum of Natural History 13:85-91.

-----. 1983. On the early origins of major biologic groups. Paleobiology 9:107-115.

-----. 1985. Mathematical models of cladogenesis. Paleobiology 11:42-52.

-----. 1989. The case for extraterrestrial causes of extinction. Philosophical Transactions of the Royal Society of London, Series B 325:421-435.

-----. 1991. A kill curve for Phanerozoic marine species. Paleobiology 17:37-48.

Raup, D. M., and S. J. Gould. 1974. Stochastic simulation and evolution of morphology - Towards a nomothetic paleontology. Systematic Zoology 23:305-322.

Raymond, A., and C. Metz. 1995. Laurussian land-plant diversity during the Silurian and Devonian: mass extinction, sampling bias, or both? Paleobiology 21:74-91.

Romano, M., W. T. Chang, W. T. Dean, G. D. Edgecombe, R. A. Fortey, D. J. Holloway, P. D. Lane, A. W. Owen, R. M. Owens, A. R. Palmer, A. W. A. Rushton, J. H. Shergold, D. J. Siveter, and M. A. Whyte. 1993. Arthropoda (Trilobita). Pp. 279-296 in M. J. Benton, ed. The fossil record 2. Chapman and Hall, London.

Saunders, W. B., and A. R. H. Swan. 1984. Morphology and morphologic diversity of mid-Carboniferous (Namurian) ammonoids in time and space. Paleobiology 10:195-228.

Schoch, R. M. 1986. Phylogeny reconstruction in paleontology. Van Nostrand Reinhold, New York.

Sepkoski, J. J., Jr. 1976. Species diversity in the Phanerozoic: Species-area effects. Paleobiology 2:298-303.

-----. 1981. A factor analytic description of the Phanerozoic marine fossil record. Paleobiology 7:36-53.

-----. 1989. Periodicity in extinction and the problem of catastrophism in the history of life. Journal of the Geological Society, London 146:7-19.

-----. 1990. The taxonomic structure of periodic extinction. Geological Society of America Special Paper 247:33-44.

-----. 1992. A compendium of fossil marine animal families, 2d ed. Milwaukee Public Museum Contributions in Biology and Geology 83:1-156.

Sepkoski, J. J., Jr., and D. C. Kendrick. 1993. Numerical experiments with model monophyletic and paraphyletic taxa. Paleobiology 19:168-184.

Sepkoski, J. J., Jr., R. K. Bambach, D. M. Raup, and J. W. Valentine. 1981. Phanerozoic marine diversity and the fossil record. Nature 293:435-437.

Shaw, A. B. 1964. Time in stratigraphy. McGraw-Hill, New York.

Signor, P. W., III, and J. H. Lipps. 1982. Sampling bias, gradual extinction patterns, and catastrophes in the fossil record. Geological Society of America Special Paper 190:291-296.

Simms, M. J., A. S. Gale, P. Gilliland, E. P. F. Rose, and G. D. Sevastopulo. 1993. Echinodermata. Pp. 491-528 in M. J. Benton, ed. The fossil record 2. Chapman and Hall, London.

Simpson, G. G. 1944. Tempo and mode in evolution. Columbia University Press, New York.

-----. 1953. The major features of evolution. Columbia University Press, New York.

Smith, A. B. 1988. Patterns of diversification and extinction in early Palaeozoic echinoderms. Palaeontology 31:799-828.

----. 1994. Systematics and the fossil record: Documenting evolutionary patterns. Blackwell Scientific, Oxford.

Smith, A. B., and C. Patterson. 1988. The influence of taxonomic method on the perception of patterns of evolution. Evolutionary Biology 23:127-216.

Springer, M. S. 1990. The effect of random range truncations on patterns of evolution in the fossil record. Paleobiology 16:512-520.

Sprinkle, J. 1973. Morphology and evolution of blastozoan echinoderms. Special Publication, Museum of Comparative Zoology, Harvard University, Cambridge, MA.

Stanley, S. M. 1979. Macroevolution: Pattern and process. W. H. Freeman, San Francisco, CA.

Strauss, D., and P. M. Sadler. 1989. Classical confidence intervals and Bayesian probability estimates for ends of local taxon ranges. Mathematical Geology 21:411-427.

Valentine, J. W. 1969. Patterns of taxonomic and ecological structure of the shelf benthos during Phanerozoic time. Palaeontology 12:684-709.

----. 1973. Evolutionary paleoecology of the marine biosphere. Prentice-Hall, Englewood Cliffs, NJ.

-----. 1980. Determinants of diversity in higher taxonomic categories. Paleobiology 6:444-450.

-----. 1990. The fossil record: A sampler of life's diversity. Philosophical Transactions of the Royal Society of London, Series B 330:261-268.

Valentine, J. W., and T. D. Walker. 1986. Diversity trends within a model taxonomic hierarchy. Physica 22D:31-42.

Valentine, J. W., A. G. Collins, and C. P. Meyer. 1994. Morphological complexity increase in metazoans. Paleobiology 20:131-142.

Van Valen, L. 1971. Adaptive zones and the orders of mammals. Evolution 25:420-428.

-----. 1973a. A new evolutionary law. Evolutionary Theory 1: 1-30.

-----. 1973b. Are categories in different phyla comparable? Tax-on 22:333-373.

-----. 1976. Energy and evolution. Evolutionary Theory 1:179-229.

-----. 1978. Why not to be a cladist. Evolutionary Theory 3: 285-299.

-----. 1979. Taxonomic survivorship curves. Evolutionary Theory 4:129-142.

----. 1985. A theory of origination and extinction. Evolutionary Theory 7:133-142.

-----. 1989. The poverty of cladism. Evolutionary Theory 9: 109-110.

-----. 1991. Biotal evolution: A manifesto. Evolutionary Theory 10:1-13.

Vermeij, G. J. 1987. Evolution and escalation: An ecological history of life. Princeton University Press, Princeton, NJ.

----. 1995. Economics, volcanoes, and Phanerozoic revolutions. Paleobiology 21:125-152.

Wagner, P. J. 1995a. Stratigraphic tests of cladistic hypotheses. Paleobiology 21:153-178.

-----. 1995b. Testing evolutionary constraint hypotheses: with early Paleozoic gastropods. Paleobiology 21:248-272.

----. 1995c. Diversity patterns among early Paleozoic gastropods: contrasting taxonomic and phylogenetic descriptions. Paleobiology 21:410-439.

-----. 1995d. Systematics and the fossil record. Palaios 10:383-388.

Williams, A. 1957. Evolutionary rates of brachiopods. Geological Magazine 94:201-211.

Wills, M. A., D. E. G. Briggs, and R. A. Fortey. 1994. Disparity as an evolutionary index: A comparison between Cambrian and Recent arthropods. Paleobiology 20:93-130.
COPYRIGHT 1996 Society for the Study of Evolution
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Foote, Mike
Date:Feb 1, 1996
Previous Article:The Major Transitions of Evolution.
Next Article:Conducting phylogenetic comparative studies when the phylogeny is not known.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters