Figure 8.1 Reasoning in the framework of the experimental model-organism approach to development. If developmental step B is changed or removed (indicated by BX), then developmental step C is absent or not properly formed (indicated by CX).
is justified and is exactly the basis of the Entwicklungsmechanik as founded by Wilhelm Roux (Roux 1894), which is the forerunner of modern molecular developmental biology (see Mocek 1998). If C develops normally despite the turning off or ablation of B, then there is no causal link between the two steps. This experimental analysis allows both positive and negative reasoning. One can conclude that C depends on, or is caused by B if the experiment indicates this, or C does not require B for a proper expression if the ablation of B does not affect C.
The experimental view of development is necessarily finalistic (even more than the descriptive approach) since the result of the normal developmental process is the differentiated adult or reproductive stage, and any severe disturbance of this process causes failure to reach the expected final product. In other words, according to this view development causes the adult stage.
Closely related to the direct experimental manipulation of developmental processes is the use of mutants or otherwise disturbed developmental patterns. In this case those naturally occurring aberrations of normal development are used to infer mechanism and causal relationships between developmental steps. The difference with respect to experimental manipulation is that the reason for the aberration remains largely unknown. The best-known example for this approach is the use of mutations in the fruit fly Drosophila melanogaster in understanding segmentation (Nusslein-Volhard and Wieschaus 1980).
How do these approaches relate to evolutionary change and evolutionary mechanisms? First of all one has to stress that there is no a-priori reason to generalise the outcome of these experiments. This is a fundamental problem of the model-organism centred approach. A single counterexample falsifies any generalisation. Furthermore, the experimental model-organism based approach is an indirect method for evolutionary inference. The experimental changes are just an analogy to evolutionary changes - developmental mechanisms are not evolutionary mechanisms. The new experimentally created phenotype is compared with naturally occurring differences between taxa and it is deduced that a similar developmental change has happened in the course of evolution. The problem is that the deduced evolutionary scenario might be correct or might be incorrect, since we simply do not know whether evolution carried out a corresponding experiment which led to evolutionary change. One has to discriminate between the functional proximate causes and the evolutionary ultimate causes or conditions for change (Mayr 1997, Sudhaus 2007). Experiments deal with proximate functional causes and not with ultimate evolutionary causes (Mayr 1997). Accordingly, there is an unbridgeable gap between experimental results and evolutionary changes.
The other aspect that has to be considered in the experimental framework is the fitness approach. According to this approach the performance of developmentally manipulated organisms or naturally occurring mutants is compared with that of normal embryos. This approach is often used to explain evolutionarily conservative characters as constraints caused by selective forces. An example is the investigation of Galis (1999) of the problem of why most mammals exhibit seven cervical vertebrae. Here malformed human and mouse embryos are studied showing that the occurrence of cervical ribs (i.e. a different number of true cervical vertebrae) is associated with a dramatic reduction in health and survival rate and fitness in general (Galis 1999). However, this view faces the problem that all living systems are functionally balanced and that mutations and experiments artificially disturb this functional balance. Accordingly, a reduced fitness may not be surprising and might not explain the evolutionary stability of characters in larger groups. Furthermore, the fact that there are mammals such as sloths and manatees that possess a different number of cervical vertebrae reveals that even in mammals this character has the freedom to evolve. Again, the proximate causes of functional disadvantages of onto-genetic change offer no direct explanation for the ultimate causes of evolutionary change. Obviously, the latter has a degree of freedom not realised by today's ontogenies.
In summary, experimental approaches cannot directly tell us about evolutionary change. They only show the possibility of what might have happened. They produce analogies to evolutionary processes but not direct evidence on them.
the descriptive comparative approach
In the descriptive comparative approach the developmental sequence ABCDE of one or more species is compared with the corresponding (homologous) development of other species which, for instance, exhibit an ABXDE sequence. This comparison allows the conclusion that C (or X) is not the necessary prerequisite of D and the following stages and that the third stage is independent of the other stages, i.e. not causally related to them (Figure 8.2). This is true in both developmental directions, namely the relationship to B and to D because C (or X) is not caused by B and C (or X) does not cause D. Furthermore, this conclusion is independent of the direction of the evolutionary transformation. It is
Was this article helpful?