Rates of Evolution

K.S.W Campbell, M.F. Day

• Publisher: Routledge
• Year: 2019
• ISBN:
  9781000063691
• Edition: 1
• Topic: Biological Sciences
• Subtopic: Evolution
• Index: Biological Sciences

Originally published in 1987 Rates of Evolution is an edited collection drawn from a symposium convened to bring together palaeontologists, geneticists, molecular biologists and developmental biologists to examine some aspects of the problem of evolutionary rates. The book asks questions surrounding the study of evolution, such as did large morphological changes really occur rapidly at various times in the geological past, or is the fossil record too imperfect to be of value in assessing rates of morphological change? What is the measure of 'rapid' change? Is stasis at any taxonomic level established? Is it possible to relate genomic and morphological change? What is the role of regulatory and executive genes in controlling evolutionary change? Does the transfer of genetic material between different taxa provide the possibility of increasing evolutionary rates? Featuring contributions from leading researchers, this book will interest anthropologists, palaeontology and scientists of evolution and genetics.

1

Major features of the fossil record and their implications for evolutionary rate studies

DAVID M. RAUP

ABSTRACT

The fossil record contains a vast array of data on the distribution of extinct species in space and time. Sampling problems are often severe, however, so that great care must be exercised when using palaeontological data for the analysis of evolutionary rates. But, because the fossil record makes rate data available over geological time spans, information from fossils is indispensable to the complete analysis of rates.

Cladogenetic rates of evolution involve branching and termination of lineages and can be approached through the use of a simple branching model of the birth-death type. This model often provides accurate predictions of the behaviour of the cladogenetic process with generalised, real-world data. In many cases, however, analysis with higher resolution data shows cladogenesis to be much more episodic (punctuated) than would be predicted from the model. The branching model thus provides a time-averaged picture of evolution as well as a null hypothesis with which significant departures from statistical expectations can be uncovered.

It has been suggested from molecular-clock data that rates of genomic change are higher in the early stages of an adaptive radiation, although this view has been challenged. To the extent that clock rates do change predictably and systematically, the cause may lie in the possibility that genomic change is more a function of the frequency of lineage branchings (speciation events) than of total elapsed time. This proposition should be tested with biological groups where genomic divergence data are optimal and where well-dated fossil records are available.

INTRODUCTION

The fossil record contains information on about 250 000 plant and animal species that lived in the geologic past (Raup 1976a). Although recognisable fossils are found in rocks as old as about 3.5 × 10⁹ years (Ga), the vast majority are in rocks of Phanerozoic age (600 million years (Ma) and younger). In spite of the large size of the fossil sample, it represents but a tiny fragment of past life. Mean durations of species vary from one biological group to another but average in the range of 1-10 million years (Valentine 1970, Van Valen 1973, Raup 1978). Thus, the biological world has turned over many times during the Phanerozoic so that the number of species that have lived in the past should be some large multiple of the number living today. Estimates of the total number of species that have lived in the past range as high as 50 × 10⁹ (Simpson 1952). It is for this reason that the 250 000 species known as fossils constitute an extremely small sample of past life.

The fossil record provides a wealth of information on evolutionary rates. Within the limits of the resolution of geological dating, we can document two basically different rates: cladogenetic and morphological. The first has to do with the branching pattern which constitutes the evolutionary tree. Critical cladogenetic elements include rates of lineage branching (speciation) and rates of lineage termination (extinction) and these combine to produce change in taxonomic diversity (richness). Morphological rates refer to change in the phenotype over geological time and this has been of particular interest in the past decade because it bears on the problem of punctuated equilibrium. Is morphological change distributed over time in the evolution of a lineage, or is it concentrated at the branch points in the evolutionary tree?

Cladogenetic and morphological rates of evolution can be studied at various scales. If one defines an evolutionary lineage to be the temporal sequence of populations of a single species, then branching events are true speciation events which serve to split the gene pool into reproductively isolated species. The termination of a lineage is the extinction of a single biological species. The morphological change that may take place during the life of the lineage is caused by population-level mechanisms such as change in allelic frequencies by directional selection or random drift. Although this kind of change (called anagenesis or phyletic transformation) may lead to a new species, the term ‘speciation’ is here reserved for true lineage branching.

The lineage can also be defined at higher taxonomic levels. Thus, we may define a lineage as all species of a genus or family. If so, the taxon is initiated by a branching event (speciation) but we are not concerned with the details of branching within the genus or family. The taxon becomes extinct when all its species are extinct. In this sort of higher level lineage, morphological change may be dominated by processes such as species selection, wherein trends in morphological change are determined by differential survival of species rather than by changes at the population level (Eldredge & Gould 1972, Stanley 1975).

Although the fossil record provides the only means for observing evolutionary change over long spans of time, recent developments in molecular genetics make it possible to infer rates of change over geological time through the analysis of genetic distance data for pairs of living species. I refer to the use of the molecular-clock logic to reconstruct cladogenetic patterns. Given a few calibration points from the geological record, it is possible to say something about variation in rates of genomic change from one biological group to another or through time. It is thus of special interest to compare evolutionary rates observed in the fossil record with those inferred from the genetics of living organisms.

DATA BASE AND SAMPLING PROBLEMS

The fossilisation and subsequent discovery of an ancient species is a most unlikely event. This follows inevitably from the fact that such a small proportion of past species are known as fossils. It follows that truly unusual circumstances are required for fossilisation, and this means that our sample of past life may be distinctly non-random. As a general rule, dead organisms are consumed by scavenger activity (including bacterial action) shortly after death so that an organism that dies in a biologically active environment has little chance of being preserved. The probability of preservation is greatly enhanced if the dead organism is moved to a biologically inactive environment and this general process is probably responsible for most fossils. A shallow-water marine animal may be thrown up on a beach by a storm or covered in situ by a sudden influx of sediment, or a terrestrial animal may be buried suddenly by volcanic or flood debris. These are examples of the large variety of mechanisms that can isolate plants and animals from biologically active environments and thus favour preservation. The study of this process constitutes the important subdiscipline of palaeobiology called taphonomy (see Kidwell & Jablonski 1983, for a recent treatment).

Taphonomic studies have demonstrated that the fossil record is a record of rare accidents, many of which are catastrophic. It is thus not surprising that the record is strongly biased in favour of certain kinds of organisms and certain environments. As a general rule, environments where there is a net accumulation of sediment contain a disproportionate number of preserved plants and animals. This means that preservation of marine organisms is more common than preservation of terrestrial species. Within the terrestrial realm, species living near lakes and rivers are more likely to be preserved than upland forms, and so on. One result of taphonomic differentials such as this is that aquatic (and especially marine) forms provide much the best sample for any analysis of evolutionary rates. Hard-shelled marine animals are thus far more appropriate for rate studies than, for example, insects, birds, or land plants.

Although the fossil record is systematically biased as described above, there are occasional spectacular exceptions. These are the so-called Lagerstätten and include unusual accidents such as the preservation of a complex mammalian fauna in the La Brea tarpits of Pleistocene age in Southern California. Such Lagerstätten provide invaluable windows to the past although they have a strangely disruptive effect on synoptic studies of evolutionary rates. It can be noted, for example, that cladogenetic activity seems to be greatly elevated at Lagerstätten but this is only an artifact of the unusual preservation. Because Lagerstätten often provide the only record of certain species and higher taxa, these taxa will appear to originate and become extinct at a single point in time. Branching rates, extinction rates and total taxonomic diversity increase markedly at these points in the geological record, but this is only an artifact of the absence of preservation of the same taxa above and below this point. It is thus important to remove data from Lagerstätten before analysis of rates. To the extent that Lagerstätten represent extreme cases in a continuum of varying preservation, this variation adds noise to palaeontological data which can never be removed completely.

Another kind of bias in the fossil record is that which has been called The Pull of the Recent’ (Raup 1979). It is generally true that the probability of occurrence and discovery of fossils increases as one moves toward the Recent (present day). Younger rocks are more widely exposed simply because they are closer to the ‘top of the pile’ and because they have had less time to be destroyed by metamorphism or removed by erosion. For this reason, the number of fossils found per million years of rock increases toward the Recent, with or without a true increase in biological diversity. Thus, rate of cladogenetic activity gives the appearance of increasing toward the Recent. There are several ways of coping with this time-dependent bias, including normalising for volume or area of exposed sedimentary rock (Raup 1976b). But the problem remains a serious one and it has given rise to considerable debate over basic questions such as whether total taxonomic diversity has increased through time (Bambach 1977, Sepkoski et al. 1981).

A related aspect of the ‘Pull of the Recent’ stems from the fact that modern organisms are far better known (sampled) than fossil organisms. A higher percentage of living species has been found and described than their extinct counterparts and this leads to a systematic bias in certain kinds of evolutionary rate analysis. This can be explained with reference to the observed time ranges of taxonomic groups. Ordinarily, the range of a taxon is simply the time interval between its first and last occurrences in the fossil record. Suppose we have a taxon which has a low probability of preservation and has been found, by chance, at only one horizon, such as in the Cambrian. Suppose further that the organism is not living today. The range of the taxon will thus be recorded as Cambrian. This is acceptable as long as it is understood that the taxon may have had a longer, unrecorded range, in line with the general observation that ranges in the fossil record are often truncated (shortened) by lack of preservation. Alternatively, suppose the taxon is still living today. Because of better sampling in the modern world, we are unlikely to miss its existence today. If the taxon is living today and is found as a fossil only in the Cambrian, the apparent range will be Cambrian to Recent. This general principle has the insidious effect of...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Original Title Page
• Original Copyright Page
• Preface
• Contents
• List of tables
• List of contributors
• 1 Major features of the fossil record and their implications for evolutionary rate studies
• 2 The timing of major evolutionary innovations from the origin of life to the origins of the Metaphyta and Metazoa: the geological evidence
• 3 Rates and modes of evolution in the Mollusca
• 4 Rates of evolution among Palaeozoic echinoderms
• 5 The initial radiation and rise to dominance of the angiosperms
• 6 Selection or constraint?: a proposal on the mechanism for stasis
• 7 Developmental pathways and evolutionary rates
• 8 Population biology and evolutionary change
• 9 Comparative rates of molecular, chromosomal and morphological evolution in some Australian vertebrates
• 10 Evolution of gene structure in relation to function
• 11 Population genetics, evolutionary rates and Neo-Darwinism
• 12 Genetic systems and evolutionary rates
• 13 The origin, nature and significance of genetic variation in prokaryotes and eukaryotes
• 14 Old and new theories of evolution
• 15 From genome to phenotype
• 16 The new gene and its evolution
• Index