SPECIES AND SPECIATION : EVOLUTIONARY TEMPO AND MODE IN THE FOSSIL RECORD RECONSIDERED


ALAN TURNER


Hominid Palaeontology Research Group, Department of Human Anatomy and Ce# Biology, Universi(F of Liverpool, PO Box 147, Liverpool L69 3BX, England. HUGH PATERSON Department of Entomology, University of Queensland, St Lucia, Queensland 4067, Australia.


ABSTRACt The past twenty years has witnessed an unresolved palaeontological debate over the proposal that the fossil record demonstrates evolution by punctuated equilibria. We discuss the process of evolution in terms of a genetic view of species that considers anew the limits to the field for gene recombination and the origin of such limits at speciation. We argue that the resultant perspective offers a strongly-based resolution to the continuing argument about the evidence for evolutionary tempo and mode to be seen in the fossil record. ESPECE ET SPECIATION : RI~VISION DU RYTHME ET DES MODALITI]S D'I~VOLUTION DANS LES SI~RIES FOSSILES. R~UMI~ Les vingt derni6res ann6es ont montr6 que la notion d'6volution par 6quilibres intermittents, d6montrde par les archives fossiles, fait toujours l'objet d'un d6bat pal6ontologique.

Nous discutons ici du proeessus de l'6volution dans une perspective g6n6tique de l'esp6ce qui consid6re, sous un jour nouveau, les limites du champ de recombinaison des ggnes et l'originc de telles limites, au niveau de la sp6ciation. Nous soutenons que la perspective qui en r6sulte apporte une r6solution fortement 6tayEe au d6bat continuel sur l'6vidence du rythme et des modalit6s d'6volution, telle qu'on la per~oit darts le t6moignage fossile. KEY-WORDS : PUNCI'UATED EQUILIBRIA, SPECIES, SPECIATION. MOTS-CL ;ES : t~QUILIBRES INTERMITTENTS, ESPI~CE, SPt~CIATION. INTRODUCTION In an important but relatively neglected paper, Eldredge (1971) suggested that the dominance of a gradualist picture of species transformation tempo in palaeontological thought should be reassessed. In 1972, Eldredge and Gould made the much-quoted proposal that apparent gaps in the fossil record, with transitional forms absent, should not be dismissed as lacunae but should instead be seen as a true reflection of evolutionary tempo. They argued that the prevailing picture of phyletic gradualism, as they termed it, according to which evolution proceeds by slow cumulative changes with the record inevitably marred by an occasional break, has been adhered to largely through intellectual inertia. In place of that picture they proposed the idea of punctuated equilibrium, in which successive episodes of rapid change (punctuations) are followed by long periods of stable structure (equilibria). The notion that change should be episodic was suggested to the authors by the theory of allopatric speciation in small, peripherally isolated populations (Eldredge and Gould 1972 : 84). As they recognised (1972 : 91) " if we doubt phyletic gradualism, we should not seek to 'disprove' it 'in the rocks'. We should bring a new picture from elsewhere and see if it provides a more adequate interpretation of fossil evidence (emphasis added). They considered the nature of Manuscrit accept6 d6finitivement le 12.03.1991 SGeobios, 1991-'~ In~ 24, fasc. 6 ] /,~. 761-769 S 762 species, and pointed to the inconsistencies involved in efforts to integrate palaeontological species (including so-called chronospecies) with neontological biospecies. They went on to discuss the application of the allopatric theory of speciation to the fossil record ("because it is the aUopatric, rather than the sympatric, theory that is preferred by biologists" [Eldredge and Gould, 1972 : 94]), concluding that episodic change would be the expected result. In our view, however, they did not address the problem of the nature of species and the process of speciation from first principles, since the allopatric theory of speciation does not deal with the former and offers only a partial understanding of the latter. (Indeed, we argue that speciation can only be understood in the light of a clear understanding of the nature of species.) They thus left the way open for opponents of punctuated equilibrium to argue that evolutionary biology could equally well predict a gradual pattern of change, and to obscure the central issues in discussions about the precise meaning of fast and slow rates of evolutionary change. The result has been that many papers, based on the evidence of particular fossil data sets, have since been published with the aim of either supporting or (more commonly) of refuting Eldredge and Gould's ideas, although little evident agreement has emerged and no resolution seems to be in sight. In this paper we return to first principles and provide a logical argument about the nature of species in a way that gives a more complete model of the process of speciation. We thus aim to put forward a more precise statement of the picture to be seen in the fossil record and the reasons why it should be seen. We hope to show why the fossil record cannot be used to test the rival claims of evolutionary tempo and mode without such precise understanding of species and speciatiou, and why confusion will inevitably result from efforts lacking such understanding. We begin by offering a brief review of the effect of Eldredge and Gould's proposals upon the palaeontological community. PUNCTUATED EQUILIBRIUM AND THE PALAEONTOLOGICAL RESPONSE Many of the numerous and vocal opponents of punctuated equilibrium among the palaeontological community (for example, Brace 1981 ; Cronin et al. 1981 ; Gingerich 1984, 1985 ; Wolpoff 1984, 1986 ; Halstead 1984 ; Hecht & Hoffman 1986 ; Carroll 1988; Hoffman 1989) take a broadly similar position, one seemingly based on the view that the "modern synthesis" (Huxley 1942) of evolutionary theory remains substantially unchallenged so far as speciation is concerned. That view receives considerable support from among the ranks of evolutionary biologists (Stebbins & Ayala 1981 ; Barton & Charlesworth 1984; Carson & Templeton 1984 ; Butlin 1987 ; Maynard-Smith 1987 ; Coyne et al. 1988 ; Chandler & Gromko 1989) and was usefully summarised some years ago by Gould (1982) and more recently by Carroll (1988). In essence, opponents claim that punctuated equilibrium is not a new idea, since there have always been suggestions that variation in evolutionary rates may occur as proposed by Simpson (1944), and that much of the argument is actually semantic. In the view of such authors, the pattern of change over time in this or that species points to gradualism as the appropriate overall tempo, while punctuations are no more than errors in the record. We suspect that many arguments advanced by the gradualists stem directly from their "modern synthesis" notions of how evolution is assumed to occur, with new species expected to appear chiefly as a function of time over which within-species changes accumulate. A source of authority for many palaeontologists appears to be the idea formalised by Simpson (1961 : 165) and by Mayr (1969 : 35-36) that given fossil lineagrs can be divided somewhat arbitrarily into species at convenient points in the stratigraphy, if no other criteria for classification present themselves. This view had been strongly criticised by Eldredge and Gould (1972 : 90), but with such a guiding philosophy allied to the notion of change due to natural selection, one can easily see how the idea of constant gradual change within a species spilling over into new species from time to time would result. Various efforts to employ the chronostratigraphic approach to palaeontological taxonomy advocated by Mayr and Simpson have therefore been made, but perhaps the most extreme statement of this view was made by Campbell (1979 : 569) in his discussion of some of the problems of classification and nomenclature that confront physical anthropologists. As he put it then : "Since chronospecies are sequent, continuous and not discrete natural units, they cannot be distinguished by morphological characters alone. Somewhere, many times, an Australopithecus mother gave birth to a Homo child, and they were indistinguishable at the taxonomic level. Nothing is to be gained by following Tobias and creating intermediate taxonomic categories, for neither morphological nor behavioural boundaries exist in realfty, however carefully we define them. Simpson discussed this problem, and it is clear that the boundaries of sequent taxa, be they genera, species or subspecies, should be conveniently agreed time-lines, rather than diagnostic morphological features such as the famed cerebral rubicon of Keith. This means that both anatomy and dating are necessary to create the taxonomy of fossil lineages [original emphasis]. It follows that the development of a reliable chronology is one of the most important characteristics of the recent period of research, and as this paper will show, new dates give us new taxonomy [our emphasis]". 763 Such arguments about procedure may have some value for those interested in merely cataloguing the fossil record and putting items in convenient pigeon holes. But that record represents more than a source of obscurely-named specimens for incorporation in a museum or private collection. It is the record of the sequence of life on earth. It is certainly true that if one wishes to talk in terms of chronospecies then one must acknowledge that they are not discrete natural units. But then it is abundantly clear that one cannot be talking about species in any genetically-based sense of the term. Chronospecies are products of a division of the World erected to suit the stratigrapher, and while they may very-well have their part to play in the world of economic geology the best refinements to such taxonomies will not make them compatible with the real thing. If palaeontologists choose to seek refuge in the claim that they den with a different category of "species" from that employed by biologists studying modern organisms, then there can be little point of contact between the two fields. Most importantly of all, the palaeontologists would have no business becoming involved in the debate over the tempo and mode of evolution, because that debate is rooted in the question "what is the deployment of speciation in time?" Spedation can only have one meaning for that question to have any significance, and must be based on a single, coherent view of what makes a species. [For that reason we would look askance at efforts by those advocating phylogenetic species concepts (Bremer & Wanntorp 1979 ; Mischler & Brandon 1987; Mischler & Donoghue 1982 ; Donoghue 1985, 1987) ; to argue in favour of pluralism in defining taxa.] However, one interesting point does emerge from Campbell's discussion. His acceptance of the blurred nature of the morphological boundaries between taxa raises the idea of a decoupling between morphological change within species and that accompanying speciation. In our view, an unfortunate and continuing feature of the debate over punctuated equilibrium has been the failure to distinguish clearly between within-species changes and speciation events. We attribute part of the reason for this to El&edge and Gould's use of the term stasis to describe the periods between speciation events, and to the fact that they did not deal with the nature of species and the precise mechanism of speciation in their original treatment. Moroever, although Eldredge (1989) has now stressed the need to distinguish between within-species change and speciation we "also note that in a previous work (Eldredge, 1985 : 116) he appears to offer continued support for gradual change leading to speciation. Opponents of punctuated equilibrium have thus felt that a demonstration of gradual change over time in a given species is sufficient evidence to reject the punctuationist interpretation on its own terms. Perhaps the clearest example of this confusion may be seen in the debate over tempo and mode in the evolution of the Hominidae, as discussed in detail elsewhere (Turner 1985 ; 1987). Briefly, the issue has tended to centre on whether the various species referred to Homo show morphological stasis or change. The species taxon most commonly examined by both sides in the debate has been Homo erectus, first known in Africa c. 1.6 Myrs ago and last known in Eurasia perhaps 0.6 Myr ago. Over tiffs time a change in morphology may be seen, accompanied by an increase in brain size. Those in the gradualist camp have pointed to these changes as evidence in their favour (Cronin et aL 1981 ; Pope & Cronin 1984 ; Wolpoff 1984, 1986), while punetuationists have argued that the relative rate of such changes does not detract from their interpretation (Howells 1980; Delson 1981 ; Eldredge & Tattersall 1982). In our view, the gradualists are quite clearly in error, since within-species change does not address the issue of whether or not speciation occurs rapidly. (We are not concerned here with the separate issue of whether or not material referred to Homo erectus represents a single species : we are concerned with the argument between two camps who both broadly accept that it is, although there are strong grounds for considering it as one taxon [Turner & Chamberlain 1989].) Demonstrating that Homo erectus, or any other species for that matter, underwent morphometric changes over time offers no answer to questions about the speed at which it became that species or eventually gave rise to another. As we show in the next section, we believe that error to be based on a generally inadequate concept of the species itself, and we shall argue that once such a concept has been developed the problem and the debate tend to disappear. We believe that the debate just outlined points to a wider problem with the unqualified use of the fossil record as a testing ground for disputed ideas about evolutionary tempo. In a review of ideas on macroevolution, Levinton (1983 : 114) reiterates the point (Levinton & Simon 1980) that "paleontological data are used to recognize speciation by identifying changes in "species specific" characters. Thus, morphological evidence can readily be related to speciation since one is the source of the other's inference l" A similar criticism of the implied circularity in palaeontological interpretations of rapid phenotypic changes as speciation events has been made by Ayala (1983), who raises the further spectre of undetected sibling, or cryptic, species in the fossil record. In Ayala's view (1983 : 390), "when palaeontologists use evidence of rapid phenotypic change in favour of the punctuational model, they are committing a definitional fallacy". Such practical and procedural difficulties are of course well-known to the palaeontologist. Vrba (1980, 1984a, 1984b, 1985) has discussed the topic at length and pointed to a number 764 of options open to the worker faced with the problems. It is clear that speciations (punctuations) are likely to be recorded with least ambiguity in organisms that carry fossilisable species-specific characters of the kind that determine, or contribute to determining, the limits of the field for gene recombination (Lambert & Paterson 1982 ; Lambert et al. 1987 ; Geist 1988 ; Turner & Chamberlain 1989 ; Tattersall 1989), a point to which we shall return below. Models of the rate of speciation, in other words, should not be judged on the basis of just any group of fossil organisms, since not all are capable of providing unambiguous evidence for test purposes. At the very least, that implies the need for a careful selection of the test data employed if confusion is to be avoided and criticisms of the kind levelled by Levinton and Ayala are to be overcome. It is equally clear that the concept of species employed will have a major bearing on interpretation. We therefore argue that the first requirement in any argument about evolutionary tempo and mode is for a soundly based concept of species and of the process of speciation compatible with it. Thus armed, it becomes possible to examine the evidence of the fossil record in a rational manner, and to overcome the seemingly interminable arguments about whether one or other picture of evolutionary tempo is correct and whether any given lineage supports or refutes the notion of punctuated equilibrium. In 1977 Gould & Eldredge offered a response to then current criticisms of their proposal for punctuated equilibrium, pointing to a number of shortcomings in the arguments raised against it, but at that point they still did not address the problems of the nature of species and the process of speciation. Moreover, both have been prepared to modify the original arguments considerably, rejecting the primacy of allopatrie speciation (Eldredge and Cracraft, 1980; Gould 1980b) and, in the latter case, giving favourable mention to sympatric models involving the fixation of chromosomal variants as isolating mechanisms (Gould, 1980b: 123). Clearly, such new proposals, with their implications for episodic change, can be fitted into the punctuated equilibrium picture of evolutionary tempo, although the causal connection between chromosomal rearrangements and speciation is open to very serious question (Paterson, 1981) and speciation in sympatry has little theoretical or empirical support (Mayr, 1963; Paterson, 1981, 1985). But it is equally clear that such incorporation points to a degree of frailty in the theoretical underpinning of the original argument for punctuated equilibrium so far as the concept of the species and the model of speciation that results are concerned. Eldredge (1989) has now moved to rectify some of those problems, and it is clear that his view of species and the process of speciation now falls substantially in line with that discussed here in the next section. THE CONCEPT OF SPECIES Paterson (1985) has recently stressed that "any view of species must be cast in genetic terms if it is to be useful in understanding the process of evolution", and we believe that this must apply equally to fossil organi.~ms. We recom~im the special practical problems that face the palaeontologist forced to deal with taxonomic species, but argue that the underlying concept of species and the process of speciation must be that employed by the neontologist. This is not a novel approach since, as we have outlined above~ the gradualist viewpoint stems from an attempt to extrapolate the slow pace of change implied by classic population biology to events over geological time. The original proposal for punctuated equilibria was itself firmly tied to the theory of allopatric speciation. We argue, however, that previous concepts of species used as a basis for these extrapolations have been inadequate. Indeed, as we pointed out, such dubious concepts as chronospecies have even been invoked, a taxonomic category with little evident relationship to any biological reality. To date, the widely accepted view among biologists has involved what has been termed the Biological Species Concept (= "Isolation Concept" of species), as expressed by Mayr (1963) when he said that "species are groups of interbreeding populations that are reproductively isolated from other groups". Reproductive isolation in this view is considered to result from a series of ad hoc characters, known as "isolating mechanisms", and the species themselves are seen as adaptive devices, as authors like Futuyma (1979) have been eager to stress. This latter concept was also exemplified in the writings of Dobzhansky (1976 : 104), who argued that species represent "the way in which the living world has deployed itself to master a progressively greater range of environments and ways of living". This statement echoes an earlier one of Mayr's, in which he argued that "speciation is thus an adaptive process towards the most efficient utilization of the environment" (Jepson, Mayr & Simpson, 194 : 284), thereby implying the existence of group adaptations that have evolved in some way "for the good of the species". Both Mayr and Dobzhansky appear to have had problems with various aspects of their Isolation Concept model. Chief among these was the difficulty posed by speciation in allopatry, as expressed by Mayr (1963 : 548) when he noted that isolating mechanisms "are ad hoc mechanisms. It is therefore somewhat difficult to comprehend how isolating mechanisms can evolve in isolated populations". As Lambert et aL (1987) have pointed out, Mayr himself has consistently preferred to stress the negative, isolating aspect of his definition, arguing that "species are more unequivocally defined by 765 their relatioa to non-couspecific populations ("isolation") than by the relation of conspecific individuals to each other" (Mayr 1963 : 20). Recent efforts to play down the illogicalities inherent in the arguments for isolating mechanisms and their origins in allopatry (Raubenheimer and Crowe 1987 ; Donoghue 1987 ; Coyne et al. 1988) really cannot escape the tangle of semantic confusion that has resulted, as Masters and Spencer (1989) have shown. In fact we argue that any proposed character arising in allopatry can clearly not be an isolating mechanism, since it has not evolved in any sense to protect the "integrity of the species". It must be considered an effect in the sense employed by Williams (1966), and we would argue that any "reproductive isolation" that results, if one wishes to speak in such negative terms, is better understood as an incidental effect of characters formerly with other functions. We may also note that while the original proposal for punctuated equilibria (Eldredge & Gould 1972) sought support from the concept of speciation in allopatry, it took no account of the difficulties posed by the requirement for isolating mechanisms to arise in such circumstances. In contrast to the Isolation Concept, we believe that a more satisfactory notion of the nature of species is afforded by the fundamentally different Recognition Concept proposed by Paterson (1978, 1981, 1982, 1985, 1986 ; Paterson & Macnamara 1984), as Eldredge (1989) has recently underlined. This concept has the desirable property of clearly delineating "the species as a field for geue recombination", as Carson (1957) expressed it. The species is thus defined as "that most inclusive population of individual, biparental organisms which share a common fertillxntion system" (Paterson 1985). The concept stems from the premise that ferfili:,ation will not occur fortuitously and that the form of the fertilization system will inevitably specify the gene recombination field. Each system consists of a number of components that are adapted to perform a contributory function in bringing about fertilization in the organism's normal habitat. In the case of motile organisms, a subset of characters of the system serves the function of bringing mating partners together. These characters, which involve signalling between mating partners or their cells, have been termed the Specific-Mate Recognition System or SMRS (Paterson 1978, 1981). The response of one mating partnerto a signal from the other is regarded as an act of recognition, although this implies no act of judgement or choice on the part of the responding partner. Given the fundamentally hnportant nature of reproduction, we assume that .all the signal and response steps in a fertilization system are likely to have evolved to serve particular roles, and that they are adaptations sensu stn'cto. This concept, in contrast to the Isolation Concept, is not a relational one. In other words, under the Recognition Concept, species are defined in terms of criteria meaningful to the members of the species, and not in terms of their relationship with non-eonspecific populations. Armed with this concept of the species it becomes clear how we may understand speciation. The coadapted nature of the signal and response steps will impose stability via stabiliTiug selection while the organism occupies its normal habitat (Paterson 1985, 1986), a point to which we shall return below. In view of the imposed stability that operates on the SMRS, change is only likely to occur if a small section of the original population becomes restricted to a new habitat in which different selective pressures operate on the adaptive characters of the fertilization system to produce a retnning of the signal and response reaction chain. If conditions should change, and selection now operate to produce a new set of component characters, then organisms equipped with that new set may effectively, though incidentally, be "reproductively isolated" from those maintaining the old set. Speciation cannot be said to occur inevitably in such circumstances, since the new habitat may either not produce sufficient change to bring about destabilization of the fertilization system or it may be so different that the isolated group becomes extinct. But we can at least speak in terms of a necessary condition. Quite clearly, the resultant pattern of speciation will be episodic, and the fossil record is indeed likely to show a sequence of punctuations and equilibria. The punctuations are likely to coincide with periods of environmental change sufficient to provoke range fragmentation into small, allopatric communities where the adaptive characters of the fertilization system can experience different selective pressures. If speciations are environmentally driven, then they are also likely to coincide across a range of organisms, as recoguised by Vrba (1985), as part of an overall pattern of speciation, extinction and dispersion. At first sight this view of speciation may be thought to differ not at all from that originally proposed by Eldredge & Gould (1972), who also argued for speciatien in allopatry. But we stress that it differs fundamentally from the theory of allopatric speciation proposed by Mayr (1963) under his Isolation Concept of species, the speciation model that underlay Eldredge & Gould's original argument, as Eldredge (1989) has acknowledged. In contrast to the Isolation Concept, the Recognition Concept invokes no necessary pleiotropic modification of isolating mechanisms, and sees spedation as a simple result of adaptive changes to the fertilization system in allopatry (Paterson 1986). Because the Recognition Concept relates speciation to environmentally-induced changes in the SMRS of each species, it carries obvious predictions about the correlations of spedation events with major environmental swings (Paterson 1986). 766 DISCUSSION When considered within the framework of the Recognition Concept, speciation is clearly seen to be an event that occurs in precise circumstances. Most importantly, it can no longer be seen as the result of accumulated change over time, and processes of within-species change cannot be argued to spill over into speciation. Species arc not arbitrary divisions of the morphologic~ pattern seen over time in a lineage but the effect of processes that we can understand and which have a clear genetic basis. We should expect to detect speciafions with greatest reliability in the fossil record when the fossils in question represent biparental organisms that bear characters of importance in the fertilization system of the species, such as the gerfi'talia of insects, horn cores of the Bovidae, antlers of the Cervidae or cranial ornaments in dinosaurs. We would draw that point to the attention of those who seek to test ideas about evolutionary tempo and mode using the fossil record, as well as those who consider all statements of evolutionary rate based on the fossil record as inevitably circular. By way of further emphasis, we repeat the point made earlier : argtunents about the rate of speciafion should not, indeed cannot, be judged on the basis of just any group of fossil organisms. It would, for example, be dangerous to select a lineage of fossil rodents whose modern representatives are known to comprise a number of morphologically cryptic species and to use the lineage in support of an argument for evolutionary tempo and mode without taking that factor into account. The detailed study of European PlioPleistocene arvicolid rodents by Chaline and Laurin (1986) could perhaps be criticised on that basis. On the other hand, an understanding of the importance or otherwise of a non-visual SMRS in the modern species might lead one to predict the likely level of morphological change over time in the fossils, even if it did not help one to establish a taxonomy. As Paterson (1986) has stressed, the likely effect of stabilising selection under the Recognition Concept provides a robust explanation for the stasis that the proponents of the punctuated equilibrium claim. Indeed, that explanation provides a direct answer to the comment on punctuated equilibrium made by MaynardSmith (1981) when he said in discussing the Chicago Macrocvolution Conference "as a geneticist I was left with the impression that the "sudden" appearance of a new species callcd for no new explanations, but that stasis may do". Moreover, it affords a more coherent understanding of the equilibrium phases than that originally put forward by Eldredge & Gould (1972 : 112). They proposed an important but rather nebulous role for homeostasis, which included the interesting suggestion that the coherence of a species existing in nearly-indcpendcnt local populations might result from the species origin "as a peripherally isolated population that acquired its own powerful homeostatic system". They rightly stressed the reinforcing effect of such a homeostatic mechzni.~m, but were unable to offer a concrete model for its action and could only suggest that "the answer probably lies in a view of species and individuals as homeostatic systems". Although they later modified that view (see below), once again the door was left open for those who simply chose to disagree with them, although, to our knowledge, their suggestion has been largely ignored rather than challenged. The effect of stabilizing selection on the signal and response chain of the SMRS may now be seen as the key point in the argument for stasis. So far as predictions that follow from the Recognition Concept are concerned, we may note that evolutionary stasis is well demonstrated in the fossil record by the records of Neogene Coleoptera recorded by Coope (1979). His results point to movements by populations seeking to remain within normal habitats in the face of quite massive fluctuations in environmental conditions, thereby avoiding range fragmentation and the possible effects of altered selection pressures upon the SMRS of the individual species. The point to stress in relation to Coleoptera is that we do not simply select them for mention because they suite our argument and ignore others that do not. Taxonomic division in Coleoptera is based in large part on genital morphology, highly visible to the systematist and obviously an important part of the fertilization system of the organisms. Coleoptera therefore meet the requirements specified in section II of this paper for the palaeontological recognition of speciations with least ambiguity: they carry fossilisable species-specific characters of the kind that determine, or contribute to determining, the limits for gene recombination. Interestingly, Eldredge & Gould (1974) went on to suggest that stasis might be based on habitat-tracking, and Eldredge (1985, 1989) has specifically cited Coope's work in evidence of that phenomenon, although none of these arguments seems to have been based on an appreciation of the absence of change in SMRS related characters as the basic reason for stability. Among mammals, the cyclical appearance of cold- and warmth- tolerant forms moxang in response to environmental changes is well demonstrated in the faunas that characterise the Eurasian and North American glacial and interglacial stages of the Plio-Pleistocene (Kurten 1968 , Kurten & Anderson 1980 ; Savage & Russell 1983 ; Azzaroli 1983 ; Azzaroli et al. 1988), and particularly evident in the case of a restricted area such as the British Isles (Stuart 1982 ; Sutcliffe & Kowalski 1976 ; Turner 1984). Indeed, it is now evident that the entire Cenozoic has been characterised by considerable dispersions and cyclical movements of manunalian populations. We would predict that many if not all of 767 the above examples reflected response to environmental change and an effort on behalf of organisms to remain within normal habitats. In the case of plants, the distribution of habitats determines the distribution of orgzni.~ms, whether considered as individuals or communities. We note here in passing that such evidence of movement in response to conditions is of considerable relevance to the Red Queen Hypothesis proposed by Van Valen (1973), which sees natural selection as a force attempting to maintain the status quo in the face of environmental decay. If environmentaly provoked movement is the more likely outcome then the Red Queen is no longer the force she was once considered to be, except perhaps in restricted circumstances where movement is impossible (Paterson 1982 : 56). Since the normal habitat is of fundamental importance to the maintenance of stability, it is clear that specialist species, tied to particular environments, are more likely to speciate or become extinct over a given time period than generalists. This has been confirmed for sister taxa of African bovids (Vrba 1984a, 1984b), and increasingly appears to be true for a range of organisms (Vrba 1985). The more general prediction of a correllation between evolutionary events and environmental change across a range of organisms, if speciation and extinctions are environmentally driven, has also found some support. Studies presented in Vrba et al. (1985) and Partridge et aL (1986) have produced preliminary evidence from Africa and the southern oceans that evolutionary pulses in diverse lineages of the aquatic and terrestrial biota do indeed correlate with past climatic changes. It now remains to be seen whether such correlations can be established on a wider scale, but since the likely causes of the African pattem involve major climatic changes related to development of the Antarctic ice sheet (Denton 1985), it seems highly likely that a more global pattern will emerge. In conclusion, we stress that the fact that the proposal for a punctuated pattern of change was made by interpreting the fossil record stands as a major insight that we in no way wish to diminish. Rather, we argue that the practicalities of resolving the debate over punctuation versus gradualism now require a fresh approach based on a return to the question of what species are and how speciation occurs. The Recognition Concept of species provides a strong body of theory for the pattern of evolutionary tempo likely to be observed over long time spans, and helps to resolve polarised arguments based on particular fossil data sets. Such a solution does not remove creativity from the discipline of palaeontology, nor does it condemn it to the role of inductive data gathering so rightly castigated by Gould (1980a, p. 97). With a clearer understanding of the basis of evolutionary change we shall be in a better position to use the information about the development of life uniquely provided by the fossil record. Acknowledgements - We thank Bernard Wood and Andrew Chamberlain for critical reading of earlier versions of the manuscript, and Niles Eldredge and Elisabeth Vrba for constructive refereeing. REFERENCES AYALA FJ. 1983 - Microevolution and Macroevolution. In D.S. BENDALL (ed.) : Evolution From Molecules to Men. Cambridge University Press : 387- 402. AZZAROLI A. 1983 - Quaternary mammals and the "end Villafranchian" dispersal event - a turning point in the history of Eurasia. Palaeogeogr., PalaeoclimatoL, PalaeoecoL, 44 : 117-139. AT.7AROLI A., DE GIUH C., FICCARELLI G. & TORRE D. 1988 - Late Pliocene to early Mid-Pleistocene mammals in Eurasia : faunal succession and dispersal events. Palaeogeogr., Palaeoclimatol., Palaeoecol., 66 : 77-100. BARTON N.H. & CHARLESWORTH B. 1984 - Genetic revolutions, founder effects, and speciation. Ann. Rev. EcoL Syst., 15 : 133-164. BRACE C.L. 1981 - Tales of the phylogenetic woods: the evolution and significance of evolutionary trees. Am. J. Phys. Anthrop., 56 : 411-429. BIU3MER K. & WANNTORP H.E. 1979 - Geographic populations or biological species in phylogeny reconstruction ? Syst. Zool., 28 : 220-224. BUTLIN R.K. 1987 - Species, speciation and reinforcement. Amer. Nat., 130 : 461-464. CAMPBELL B.G. 1979 - Some problems in hominid classification and nomenclature. In C. JOLLY (ed.) : Early Hominids of Africa, St Martin's Press, New York : 567-581. CARROLL R.L. 1988 - Vertebrate Paleontology and Evolution. Freeman, New York. CARSON H.L. 1957 - The species as a field for gene recombination, bt E. MAYR (ed.) : The Species Problem. Amer. Assn. Advt. Sci., 50 : 23-38. CARSON H.L. & TEMPLE'TON A.R. 1984 - Genetic revolutions in relation to speciation phenomena : the founding of new populations. Ann. Rev. EcoL Syst., 15 : 97-131. CHALINE J. & LAURIN B. 1986 - Phyletic gradualism in a European Plio-Pleistocene Mimomys lineage (Arvicolidae, Rodentia). PaleobioL, 12 : 203-216. CrLANDLER C.R. & GROMKO M.H. 1989 - On the relationship between species concepts and speciation processes. Syst. Zool., 37 : 190-200. COOPE G.R. 1979. - Late Cenozoic fossil coleoptera : evolution, biogeography and ecology. Ann. Rev. EcoL Syst., 10 : 247-267. COYNE J.A., ORR Hak. & FtYrt.~MA D.J. 1988 - Do we need another new species concept ? Syst. ZooL, 37 : 190-200. 768 CRONIN J.E., BOAZ N.T., NTRINGER C.B. & RAK Y. 1981 - Tempo and mode in hominid evolution. Nature, 292 : 113-122. DELSON E. 1981 - Paleoanthropology : Pliocene and Pleistocene human evolution. Paleobiolo~, 7(3) : 298-305. DENTON G.H. 1985 - Did the Antarctic ice sheet influence Late Cainozoic climate and evolution in the southern hemisphere ? Afr. J. Sci., 81 : 224-229. DOBZHANSKY T. 1976 - Orgnni.~mic and molecular aspects of species formation, In A.I. AYALA (ed.) : Molecular Evolution Sinauer Associates, Sunderland Mass. : 95-105. DONOGHUE M.J. 1985 - A critique of the Biological Species Concept and recommendations for a phylogenetic alternative. Bryologist, 88 : 172-181. DONOGHUE M.J. 1987 - South African perspectives on species : an evaluation of the recognition concept. Cladistics, 2 : 285-294. ELDREDGE N. 1971 - The aUopatric model and phylogeny in Paleozoic invertebrates. Evolution, 25 : 156-167. ELDREDGE N. 1985 Time Frames. Princeton University Press, Princeton. ELDREDGE N. 1989 - Macroevolutionary Dynamics. McGraw-Hill, New York. ELDREDGE N. and CRACRAIq" J. 1980. - Phylogenetic Patterns and the Evolutionary Process. Columbia University Press, New York. ELDREDGE N. and GOULD SJ. 1972 - Punctuated equilibria : an alternative to phyletic gradualism. In T.J.M. SCHOPF (ed.) : Models in Paleobiology. Freeman, San Francisco : 82-115. ELDREDGE N. and GOULD S.J. 1974 - Reply to Hecht. Evol. Biol., 7 : 303-308. ELDREDGF. N. and TATI'ERSALL I. 1982 - The Myths of Human Evolution. Columbia University Press, New York : 187p. FUTUYMA D.J. 1979. - Evolutionary Biology. Sinauer Associates, Sunderland. Mass. GEIST V. 1988 - On the evolution of antler form in Quaternary deer. In C. WEMMER (ed.) : Biology and Management of the Cervidae. Smithsonian Instit. : 235-255. GINGEmCII P.D. 1984 - Punctuated equilibria: where is the evidence ? Syst. ZooL, 33 : 335-338. GINGERICII P.D. 1985 - Species in the fossil record : concepts, trends and transitions. Paleobiology, 11 : 27-41. GOULD S.I. 1980a - The promise of paleobiology as a nomothetic evolutionary discipline. Paleobiology, 6 : 96-118. GOULD S.J. 1980b - Is a new and general theory of evolution emerging ? Paleobtology, 6 : 119-130. GOULD S.J. 1982 - Darwinism and the expansion of evolutionary theory. Science, 216, p. 380-387. GouLD S.J. & ELDREDGE N. 2977 - Punctuated equilibria : the tempo and mode of evolution reconsidered. Paleobiology, 3 : 115-151. HALSTEAD B. 1984 - Neo-Darwinism rules. New Scient&t, 102 : 1-40. HECHT M.K. & HOFFMAN A. 1986 - Why not neoDarwinism ? A critique of paleobiological challenges. Oxford Survey EvoL BioL, 3 : 1-47. HOFFMAN A. 1989 - Arguments on Evolution. Oxford University Press, London. HOWELLS W.W. 1980 - Homo erectus : who, when and where : a survey. Yearbook of Phys. Anthrop., 23 : 1-23. HUXLEY J. 1942 - Evolution, the Modern Synthesis. Allen and Unwin, London. JEPSON G.L., MAYR E. & SIMPSON G.G. (Eds) 1949 - Genetics, Paleontology and Evolution. Princeton University Press. KURTEN B. 1968 - Pleistocene Mammals of Europe. Weidenfeld and Nicolson, London. KURTEN B. & ANDERSON E. 1980 - Pleistocene Mammals of North America. Columbia University Press, New York. LAMBERT D.M., MlCrlAUX B. & WHITE C.S. 1987 - Are species self-deflnlng ? Syst. ZooL, 36 : 196-205. LAMBERT D.1V[. & PATERSON E.H. 1982 - Morphological resemblance and its relationship to genetic distance measures. Evol. Theory., 5 : 291-300. LEVIN~FON J.S. 1983 - Stasis in progress: the empirical basis of macroevolution. Ann. Rev. EcoL Syst., 14 : 103-137. LEVINTON J.S. & SIMON C.M. 1980 - A critique of the punctuated equilibria model and implications for the detection of speciation in the fossil record. Syst. Zool., 29 : 130-142. MASTERS J.C. & SPENCER H.G. 1989 - Why we need a new genetic species concept. Syst. ZooL, 38 : 270- 279. MAYNARD-SMITH J. 1981 - Macroevolution. Nature, 289 : 13-14. MAYNARD-SMITH J. 1987 - Darwinism stays unpunctured. Nature, 330 : 516 MAYR E. 1963 Animal Species and Evolution. Harvard University Press, Cambridge, Mass. MAYR E. 1969 - Principles of Systematic Zoology. McGraw-Hill, New York. MISCHLER B.D & BRANDON R.N. 1987 - IndMduality, pluralism, and the phylogenetic species concept. Biol. Philos., 2 : 397-414. MISCHLER B.D. & DONOGHUE M.J. 1982 - Species concepts : a case for pluralism. Syst. Zool., 31 : 491-503. PARTRIDGE T.C., VRBA E.S., BURCKLE L.H. & DENTON G.H. (eds) 1986 - Palaeoclimate and Evolution II. S. Aft'. Z Sci., 82 : 62-96. PATERSON H-F.H. 1978 - More evidence agMnst speciation by reinforcement. S. Afr. J. Sci, 74 : 369-371. PATERSON H.E.H. 1981 - The continuing search for the unknown and unknowable : a critique of centempoary ideas on speciation. S. Afr. J. Sci., 77 : 113-119. 769 PATERSON H.E.H. 1982 - Perspective on speciation by reinforcement. S. Afr. J. Sci., 78 : 53-57. PATERSON H.EM. 1985 - The recognition concept of species. In E.S. VII_I3A (ed.) : Species and Speciation. Transvaal Museum Monograph, 4 : 21-29. PATERSON H.E.H. 1986 - Environment and species. S. Afr. Z Sci., 82 : 62-65. PATERSON H.E.H. & MACNAMARA M. 1984 - The recognition concept of species. S. Afr. Z Sci., 80 : 312-318. POPE G.G. & CRONIN J.E. 1984 The Asian Hominidae. Jnl. Human Evol., 13 : 377-396. RAUBENHEIMER D. &/; CROWE T.M. 1987 - The Recognition Species Concept: is it really an alternative? S. Afr. J. Sci., 83 : 530-534. SAVAGE D.E. & RUSSELL D.E. 1983 - Mammalian Paleofannas of the World. Addison-Wesley Publishing Co., London. SIMPSON G.G. 1944 - Tempo and Mode in Evolution. Columbia University Press, New York. SIMPSON G.G. 1961 - Principles of Animal Taxonomy. Columbia University Press, New York. STEBBINS G.L. & AYALA F.J. 1981 - Is a new evolutionary synthesis necessary ? Scio~ce, 213 : 967-971. STUART AJ. 1982 - Pleistocene Vertebrates in the British Isles. Longman, London. SUTCLIFFE A.J. & KOWALSKI K. 1976 - Pleistocene rodents of the British Isles. Bull. Brit. Mus. (Nat. [list.) (Geology), 27, (2) : 31-147. TA'ITERSALL I. 1989 - The roles of ecological and behavioural observation in species recognition among primates. Human Evol., 4 : 117-124. TURNER A. 1984 - Human migration into high latitudes as part of a large-mnmrnal community. In R. FOLEY (ed.) : Hominid Evolution and Community Ecology. Academic Press : 193-217. TURNER A. 1985 - The recognition concept of species in palaeontology, with special consideration of some issues in hominid evolution. In E.S. VRBA (ed.) : Species and Speeiation. Transvaal Museum Monograph, 4 : 153-158. TURI, ZER A. 1987 - Species, speciation and human evolution. Human EvoL, 1 : 419-430. TURNER A. & CHAMBERLAIN A. 1989 - Speciation, morphological change and the status of African Homo erectus. Z Human EvoL, 18 : 115-130. VAN VALEN L. 1973 - A new evolutionary law. EvoL 77~eory, 1 : 1-30. VRBA E.S. 1980 - Evolution, species and fossils: how does life evolve ? S. Aft. Y. Sci., 76 : 61-84. VRBA E.S. 1984a - Evolutionary pattern and process in the sister group Alc~laphini-Aepycerotini (Mammalia : Bovidae). In N. ELDREDGE & S.M. STANLEY (cds) : Living Fossils. Springer-Verlag : 62-79. VRBA E.S. 1984b - Patterns in the fossil record and the evolutionary process. In M.W. Ho & P.S. SAUNDERS (eds) : Beyond Nco-DarMni~m, Academic Press, London : 115-142. VRBA E.S. 1985 Environment and evolution: alternative causes of the temporal distribution of evolutionary events. S. Afr. Z Sci., 81 : 229-236. VRBA E.S., BURCKLE L.H., DEI',rrON G.H. & PARTRIDGE T.C. (eds) 1985 -Palaeoclimate and evolution I. S. Afr. Z Sci., 81 : 224-275. WILLIAMS G.C. 1966 Adaptation and natural Selection. Princeton University Press, Princeton. WOLPOFF M.H. 1984 - Evolution in Homo erectus : the question of stasis. Palr 10 : 389-406. WOLPOFF M.H. 1986 - Stasis in the interpretation of evolution in Homo erectus: a reply to Rightmire. PaleobioL, 12 : 325-328.