Adaptation from interactions between metabolism and behaviour: self-sensitive behaviour in protocells

This thesis considers the relationship between adaptive behaviour and metabolism, using theoretical arguments supported by computational models to demonstrate mechanisms of adaptation that are uniquely available to systems based upon the metabolic organisation of self-production. It is argued how, by being sensitive to their metabolic viability, an organism can respond to the quality of its environment with respect to its metabolic well-being. This makes possible simple but powerful ‘self-sensitive’ adaptive behaviours such as “If I am healthy now, keep doing the same as I have been doing – otherwise do something else.” This strategy provides several adaptive benefits, including the ability to respond appropriately to phenomena never previously experienced by the organism nor by any of its ancestors; the ability to integrate different environmental influences to produce an appropriate response; and sensitivity to the organism’s present context and history of experience. Computational models are used to demonstrate these capabilities, as well as the possibility that self-sensitive adaptive behaviour can facilitate the adaptive evolution of populations of self-sensitive organisms through (i) processes similar to the Baldwin effect, (ii) increasing the likelihood of speciation events, and (iii) automatic behavioural adaptation to changes in the organism itself (such as genetic changes). In addition to these theoretical contributions, a computational model of self-sensitive behaviour is presented that recreates chemotaxis patterns observed in bacteria such as Azospirillum brasilense and Campylobacter jejuni. The models also suggest new explanations for previously unexplained asymmetric distributions of bacteria performing aerotaxis. More broadly, the work advocates further research into the relationship between behaviour and the metabolic organisation of self-production, an organisational property shared by all life. It also acts as an example of how abstract models that target theoretical concepts rather than natural phenomena can play a valuable role in the scientific endeavour.