evolutionary mechanisms

Figure 8.2 Reasoning in the framework of the descriptive comparative approach to development.

Figure 8.2 Reasoning in the framework of the descriptive comparative approach to development.

true in any instance independent of whether C or X is the plesiomorphic developmental character.

The comparative approach falsifies generalisations that two or more developmental steps are necessarily causally connected. At the same time this approach reveals that the developmental steps that differ are 'evolutionary developmental steps', i.e. developmental steps that have had the freedom to evolve independently of other developmental steps. This does not automatically imply an evolutionary direction or polarisation of the change but deals with the general possibility of change. Only a phylogenetic analysis allows assessment of the direction of change. Moreover, from comparison alone one cannot show that there was a causal nexus between developmental steps in the first place. If we assume that, based on a phylogenetic analysis, we 'know' that evolution went from C to X, then it can be concluded that the causal relationship between C and D has been broken. However, this can be concluded only if experiments have shown that C does indeed cause D in the normal development of that particular species or group of species (see above). It means that at some stage in evolution C has become an evolutionary developmental step which gained the potential for independent evolutionary alteration. In evolutionary terms, C is no longer the cause of D etc. and is not caused by B. If we concluded from experiments or based on comparisons that C is the necessary cause of D, this is falsified because the evolutionary transformation of C to X under conservation of the other stages reflects the independence of the subsequent stages from the occurrence of C.

The literature is full of examples for this phenomenon at all levels from genes to morphogenesis (e.g. Sewertzoff 1931, Remane 1952, Dohle 1989, Raff 1999, Richardson et al. 2001, Scholtz 2005). Hence, these examples will not be repeated here.

This result provides a different perspective on development. According to this view, development is not finalistic since every developmental step including the adult stage can be altered or lost in the course of evolution (see Scholtz 2005, and for arguments against the current general 'adultocentric' view see Minelli 2003). Accordingly, in evolutionary terms development does not necessarily cause the final stage. Furthermore, only negative causal conclusions ('falsifications') can be drawn, i.e. it can be concluded that C is independent of B and D or that C is not caused by B and does not cause D. In no case is it possible to make a positive inference that B depends on A even if this is the case in all observed examples.

The comparative approach can show that a developmental step is an independent evolutionary step. In other words, the developmental step is free to evolve independently of other steps of developmental sequence of which it is a part. The comparative approach shows, so to speak, experiments carried out by evolution (Dohle and Scholtz 1988, Scholtz and Dohle 1996). The absence of a necessary causal relation can be directly shown if the outcome of a developmental sequence is different.

In summary, comparison is a direct means to study changes in evolution because it deals directly with the differences and causes in developmental sequences.

the experimental comparative approach

As stated above, the descriptive comparative approach allows the conclusion that a developmental step is not the necessary condition or cause for subsequent developmental steps. If this is the case, an experimental comparative analysis does not necessarily add further support to this conclusion. An example for this is the role of the pair rule gene even-skipped (eve) in hexapods. From experiments in Drosophila it has been deduced that eve is a necessary prerequisite for proper segmentation because it regulates the segment polarity gene engrailed (en) which in turn establishes segmental boundaries (Fujioka et al. 1995). Comparative analyses have shown that in the grasshopper Schistocerca americana, eve is not expressed as a pair rule gene, but proper segmental en stripes are formed nevertheless (Patel et al. 1992). Accordingly, this result falsified the generalisation drawn from the Drosophila condition that en expression and thus segmentation is necessarily caused by eve. Liu and Kaufman (2005) tested the functional role of eve with RNAi experiments in the bug Oncopeltus fasciatus. This study revealed that indeed eve does not play a role as a pair rule gene in Oncopeltus, where nevertheless, as in Schistocerca, normal en stripes are generated (Liu and Kaufman 2005).

Comparative experiments add useful information if they reveal that in corresponding sequences ABCDE in two different species, in one case B is the cause of C and C is the cause of D, and in the other case B is not the cause of C and C is not the cause of D. This indicates that despite similar sequences of developmental stages, the underlying causal chains have been altered during evolution. C is an independent developmental step (Figure 8.3). Again, together with a phylogenetic analysis this allows the stepwise analysis of changes, including those that are not visible with a descriptive comparative approach alone

Observation species sequence

2 A^B^C^D^E Manipulation species 1: if BX then CX then DX species 2: if BX then C, if CX then D Inference species 1: B is the necessary cause ofC etc.

species 2: B is not the cause of C, C is not the cause of D, C has the potential to evolve independently Proximate causes set the starting point for ultimate causes

Figure 8.3 Reasoning in the framework of the experimental comparative approach to development. Experimentally manipulated steps and the resulting alterations of subsequent steps are indicated by an x.

A^B^C^D^E B does not cause C C does not cause D

C is an independent step with the possibility for evolutionary change

Figure 8.4 Putative evolutionary sequence showing the stepwise independence of developmental step C. In species 1 and 2 there is the same sequence of developmental steps. In species 1 there are causal relationships between all developmental steps in the sequence. In species 2 the developmental step C is not caused by the preceding developmental step B and does not cause the next step D. It can be concluded that C has the potential to evolve independently of B and D. In species 3 the developmental step C is replaced by X. The comparative analysis reveals that C is an independent evolutionary developmental step.

Was this article helpful?

0 0

Post a comment