Search

Organisms are constantly faced with the challenge of adapting to new circumstances. In this thesis, I argue that the ability to adapt to new circumstances, “evolvability”, is deeply ingrained in the genetic, developmental, morphological, and physiological architecture of organisms. Using a blend of conceptual research, theoretical modelling, and multidisciplinary studies, I demonstrate how organismal architecture can evolve so that organisms can cope better and better with future environmental challenges. As a first step, I systematically classify the many factors contributing to evolvability. Then I use a simulation approach to show how evolvability-enhancing structures can readily evolve in gene-regulatory networks. This happens via the evolution of "mutational transformers" - structural elements that convert random mutations at the genetic level into adaptation-enhancing mutations at the phenotypic level. In another thesis chapter, I demonstrate that even if selection acts only sporadically, complex adaptations can evolve and persist over long time periods. In other words, complex adaptations do not require constant selection pressure. In an interdisciplinary contribution, I apply biological insights regarding the properties of an evolvability-enhancing mutation structure to the design of algorithms used in Artificial Intelligence. The result is the “Facilitated Mutation” method which enhances the performance of the algorithms in various respects, highlighting the potential for leveraging biological principles in computational sciences. Finally, I embed my research findings in a philosophical context. I emphasise the importance of organismal architecture in retaining evolutionary memories and suggest future research directions to further enhance our understanding of evolvability.

Numerous cases of evolutionary trait loss and regain have been reported over the years. Here, we argue that such reverse evolution can also become apparent when trait expression is plastic in response to the environment. We tested this idea for the loss and regain of fat synthesis in parasitic wasps. We first show experimentally that the wasp Leptopilina heterotoma switches lipogenesis on in a fat-poor environment, and completely off in a fat-rich environment. Plasticity suggests that this species did not regain fat synthesis, but that it can be switched off in some environmental settings. We then compared DNA sequence variation and protein domains of several more distantly related parasitoid species thought to have lost lipogenesis, and found no evidence for non-functionality of key lipogenesis genes. This suggests that other parasitoids may also show plasticity of fat synthesis. Last, we used individual-based simulations to show that a switch for plastic expression can remain functional in the genome for thousands of generations, even if it is only used sporadically. The evolution of plasticity could thus also explain other examples of apparent reverse evolution.