The Polysemy of Genes: Analogical and Agent-Based Models Investigation of Phenotypic Plasticity

The aim of this thesis is to investigate phenotypic plasticity andits evolution from two different, yet complementary angles: whilethe first part is theoretical and philosophical, the second is basedon a computational approach. In the first part, I try to develop aconceptual framework which is alternative, but not necessarily inconflict with the so-called “gene’s eye view”, that considers genes asthe fundamental units of biological evolution, continuously competingto ensure survival, in the form of copies, in the next generation.To expand this narrow perspective, I start my analysis by isolatingthe main structural aspects shared by phenotypic plasticity, that is,the ability of a given genotype to produce alternative phenotypesin response to environmental conditions, and polysemy, that is, theability of a word or expression to generate alternative meanings inresponse to contextual factors. Based on this linguistic analogy, Iintroduce a broader conceptual model which could combine the centralityof genes and the outstanding importance of their interactionwith the environment and with other regions of the genome, alsoincluding the pivotal role of developmental processes in the equation.Besides its limitations, this alternative view may serve as anupdated heuristic tool to better understand and communicate biologicalevolution, or even to suggest some new hypotheses. Thesecond part of the thesis involves a specific gene that is known toenable/disable this “biological polysemy”. In this section, I introduceagent-based modelling, that is, programming the behaviourof individual computational entities and letting them interact toproduce emerging patterns spontaneously. In some contexts, thismodelling technique may be suitable for investigating evolutionaryprocesses in a much faster and controllable way than laboratory orfield research. Using the modelling environment NetLogo, I implementan agent-based model that simulates the competition betweenplastic and non-plastic strains of the nematode worm Pristionchuspacificus, which is an exemplary organism in the field of evolutionarydevelopmental biology. The global behaviour of the system is monitoredin terms of frequency of fixation and mean time to fixationof two alleles of the developmental switch gene eud-1, one enablingmouth-form dimorphism and predation (+), the other leading to theconstitutive expression of a single morph (e, which stands for eud-1 ). The model includes benefits in fecundity for each phenotype,costs in fecundity due to starvation or to the production and maintenanceof plasticity (for plastic strains only), different degrees ofenvironmentally-induced variation, and two alternative food sources(i.e., bacteria and nematodes’ larvae), whose availability may varyover time at regular intervals. Using this model, I simulate severalexperimental scenarios with the aim of identifying relevant emergingpatterns resulting from the interplay between the key variables thatwere introduced into the system.