Search

SUMMARYEvolvability is defined as the capacity of populations to generate adaptive variation that promotes evolution by natural selection1–4. Lineages in a fluctuating environment must rapidly adapt to different states of the environment to survive. One possible solution is for populations to evolve increased evolvability. However, in individual environmental states, there is no selection pressure on a population to make adaptive variation more accessible. Whether natural selection can drive organisms to become more evolvable is still unclear5–7. Here we show that evolution in fluctuating environments can lead to an increase in the mutational accessibility of adaptive variation. We find that genotypes from populations that evolve in environments fluctuating between two states have a larger proportion of mutants adapted to alternate environments when compared to genotypes in a static environment. These genotypes also encode more information about the nature of the fluctuating environments in their mutational neighborhood. Analysis of evolved lineages shows that the lineages survive by continually fixing adaptive mutations, and localizing on areas of the genotype space with increased access to alternate phenotypes. We find that this increase in adaptive mutants is contingent on the rate at which the environment fluctuates. To further understand this phenomenon, we explore evolutionary dynamics in a model with explicit genetic structure in fluctuating environments. Results from this simple model point toward a mechanism by which changing environments promote the localization of populations on edges in genotype space, where a diverse repertoire of variants is available through mutations. Populations then stochastically evolve along these boundaries and end up in highly evolvable locations between alternatively adapted phenotypes. Thus, we demonstrate that evolvability can and does evolve even in simple systems, and likely has been playing an important role in the adaptation of living systems.

Asexually reproducing populations of single cells evolve through mutation, natural selection, and genetic drift to enhance their reproductive fitness. The environment provides the contexts that allow and regulate their fitness dynamics. In this work, we used Avida - a digital evolution framework - to uncover the effect of mutation rates, maximum size of the population, and the relative abundance of resources, on evolutionary outcomes in asexually reproducing populations of digital organisms. We observed that over extended simulations, the population evolved predominantly to one of several discrete fitness classes, each with distinct sequence motifs and/or phenotypes. For a low mutation rate, the organisms acquired either of four fitness values through an enhancement in the rate of genomic replication. Evolution at a relatively higher mutation rate presented a more complex picture. While the highest fitness values at a high mutation rate were achieved through enhanced genome replication rates, a suboptimal one was achieved through organisms sharing information relevant to metabolic tasks with each other. The information sharing capacity was vital to fitness acquisition and frequency of the genotype associated with it increased with greater resource levels and maximum population size. In addition, populations optimizing their fitness through such means exhibited a greater degree of genotypic heterogeneity and metabolic activity than those that improved replication rates. Our results reveal a minimal set of conditions for the emergence of interdependence within evolving populations with significant implications for biological systems in appropriate environmental contexts.