Geology 107: Paleobiology



Taxa at all levels are classified into evolutionary clades on the basis of derived characters. Shared ancestral characters are not used to define clades, but all clades have a list of them. Shared ancestral characters define in a clear way a concept that has been somewhat fuzzy: the Bauplan. Bauplan is a German word that means "blueprint" - no, not Blauplan! The Bauplan (plural Bauplane) is the set of plans from which a building is constructed. In general, any clade has its basic character set, the Bauplan, and superimposed on that as modifications or specializations is a set of rather more minor characters. So the Bauplan persists throughout the clade. It has come to be thought sometimes that because the Bauplan does not change, then somehow it cannot change. In some concepts, the Bauplan is hidden away from selection by the modifications that overlie it. The thought is rather like the concept of stasis in species: because the species does not change (a supposed fact), therefore it cannot change (unwarranted inference).

Bauplane can and do change. For example, the amphibian Bauplan must at some time have changed into the reptilian one, and so on. Again, as in speciation, one may postulate rather special circumstances in which stasis breaks down. Is the Bauplan modified only in some special process, or in some ordinary extrapolation of normal microevolutionary processes? Note that if the Bauplan is not ordinarily operated on by selection, changes in it cannot be accomplished by ordinary adaptation, and we must appeal to some unusual process.

I don't like any of this. I see stasis in species as maintained continually by stabilizing selection, and I see the Bauplan as similarly, continually, subject to selection. Clades persist because their constituent populations are well enough adapted to reproduce themselves, and the Bauplane of the constituent individuals are a part of that adaptive complex. We see dramatic selection against radical departures from normal Bauplane in many mutations in humans, domestic animals, and crops.

But we do have to face the question whether every character we see in organisms past and present is a good adaptation, molded and perfected by natural selection. Are there characters that are not adaptive?


NeoDarwinian theory accentuates the universal agency of natural selection, and it is tempting to see superbly close adaptation in every character of an organism. The more we study the living world, the more we find examples of the most incredibly intricate adaptations. But the German school of biology and paleontology has for some decades been liable to look for non-adaptative characters. For example, it was largely German scholars and their American colleagues who formulated the concepts of orthogenesis and the Law of Recapitulation, both of which called for, or allowed, large components in morphology and morphogenesis that are non-adaptive.


The British school of functional morphologists tended to stress the adaptive features of structures. The application of their style of thought to the fossil record was best articulated by Martin Rudwick, who proposed the paradigm method for evaluating the function of a structure in a fossil. One thinks of a possible function for the structure. To avoid the charge that one has simply made up a "Just So" story, one does a thought experiment, designing the very best, optimum, structure (the paradigm) which the organism could possibly have evolved to perform that function. If the actual structure corresponds closely to the paradigm, then one can be confident that it did perform that function. Rudwick's idea was specifically aimed at testing the German idea that many structures would be non-functional: as in orthogenetic sequences, for example.


Since the late 1960s, Adolf Seilacher, first at Tubingen University and now at Yale, has articulated concepts parallel to Rudwick's that have helped to clarify our thinking. His ideas continue the German tradition of stressing non-adaptive characters, but in a novel and convincing way. Seilacher proposes that any aspect of morphology has been molded by three distinct processes. One is adaptation, but adaptation has not been pushed by natural selection to perfection because two important constraints have prevented it.
  1. Historical Constraints. Any structure is the product of a long history of evolution from the first living organism, through continuous inheritance. The structure has arisen by small changes to a pre-existing state, and the final state is then the result of continuous tinkering. It's likely that its current form is not the optimum structure one would build, although it might be the best one could build given the initial configuration. For example, articulate brachiopods are anatomically inside two shells which are made of calcite, and secreted from a single mantle groove at the shell edges. Any evolutionary sequence within the group starts from this configuration, which is constraining. Bivalves, for example, have two grooves round the mantle edge, and this historical component of their morphology has allowed them to evolve siphons and other devices which never evolved in brachiopods. It would not have been impossible for brachiopods to have evolved siphons, but the historical constraint of a single mantle groove may have helped to prevent it.
  2. Constructional Constraints. Any organism that grows does so by adding tissue. Certain growth patterns are possible, but not all patterns can exist within the same animal. For example, brachiopods grow their shell by accretion at the edges. It's possible to modify the shell once it's laid down, but only with difficulty. It's not possible to reconstitute the original shell structure if the shell should be damaged. In contrast, ostracodes live inside a bivalved shell which they secrete around them, but it is secreted over all the mantle surface, rather than dominantly at the edges. It is not possible to modify the shell at all once it is laid down, and instead the ostracode molts it off and grows a new one. These contrasting styles of construction are associated with different solutions to similar adaptive problems. Some compromises between optimum adaptation and constructional style must be made for almost all morphological systems.
  3. Function, or Adaptation. Structures are subject to natural selection, and in order for the organism to survive to reproduce, it must function at some level of efficiency. The adaptive aspects of morphology that are not attributable to historical or constructional constraints compose the functional component.

Note that the two constraints were built into Rudwick's arguments, in which he specified that the paradigm was the best structure that the organism in question could have built. But the power of Seilacher's analysis is that it explicitly separates out the constraints that prevent a structure from being optimal from the functional component would otherwise drive it toward optimal morphology.

Seilacher's approach has come to be known as Konstruktionsmorphologie, or Constructional Morphology. Later modifications and discussions have not altered its basic structure. Thus Raup added the components of chance and phenotypic response to the list of agents that can prevent a structure from being optimal in morphology, and Hickman separated out further subcategories of constructional and historical constraints. Cowen suggested that constraints might not be as important as we might think. Structures might be closer to optimal than they look because they are built to be cost-effective rather than perfect. For example, people might like to own a Mercedes but often buy a Volkswagen instead. If you like, you can think of cost-benefit considerations as a separate, and very important, constraint on morphology.

Altogether, then, it's clear that we cannot expect perfection in structures that have evolved in real organisms under natural selection. The question then becomes, how can we tell which aspects of a structure are adaptive, and which are molded more by the various constraints that act on them? The techniques involved are included in the exercise of functional analysis of fossils, and they have two major aspects.

1. Comparison with Living Organisms. We can make observations and experiments on living animals which allow us to identify adaptations. For example, the peppered moth of England suffers predation that depends on its capacity to find a background that will give it camopuflage against the birds that hunt it by sight. Dark and light morphs are differentially selected, depending whether the trees they use as perches are dark or light. In this case, color is an important component of adaptation.

In assessing the morphology of fossils, then, an immediate task is to find homologous or analogous morphology that has been carefully studied among living organisms. Direct comparison can then be made against the detailed morphology of the fossil.

2. Engineering or Thought Analogues. Sometimes there is no available homologue or analogue, and so we make comparisons with structures that are man-made. Thus the flight of pterosaurs is being studied by building full-size replicas, and the plates on the back of Stegosaurus were studied by building models. In these cases the analogues can be made as exactly applicable to the fossil as ingenuity permits, whereas the availability of homologous or analogous living organisms is not predictable, and the comparisons may not be directly relevant. But the non-biological comparisons are often difficult because the constructional materials do not match exactly, and there are no comparable historical constraints on man-made objects.

Any functional analysis of a fossil can be tested only by the goodness of fit to the predictions of a hypothesis, and goodness of fit is not a numerical, objective parameter. It is certainly best if one can design a series of tests, all of which must be successfully met. Thus the lens of a trilobite eye must be interpreted in a way that makes optical sense, the lenses of the eye must then work together to make up an operational organ, and the eyes must operate in a biological reasonable way in the animal as a whole. Any reconstruction of trilobite vision therefore must survive tests at multiple levels, and we may feel confident about a hypothesis that survives such testing.

Such analyses, applied to individual species, or evolutionary sequences of species, may tell is a great deal about the way they operated in their environments, and therefore about the selective agents that brought about the adaptive features noted on the organisms. In the end, they add more information that we can use to reconstruct evolutionary mechanisms, so that we can deal with the paleobiology of the organism in its environment, instead of lists of characters that its fossils share. Return to 107 main menu