Evolution provides a creative fount of complex and subtle adaptations that often surprise the scientists who discover them. However, the creativity of evolution is not limited to the natural world: Artificial organisms evolving in computational environments have also elicited surprise and wonder from the researchers studying them. The process of evolution is an algorithmic process that transcends the substrate in which it occurs. Indeed, many researchers in the field of digital evolution can provide examples of how their evolving algorithms and organisms have creatively subverted their expectations or intentions, exposed unrecognized bugs in their code, produced unexpectedly adaptations, or engaged in behaviors and outcomes, uncannily convergent with ones found in nature. Such stories routinely reveal surprise and creativity by evolution in these digital worlds, but they rarely fit into the standard scientific narrative. Instead they are often treated as mere obstacles to be overcome, rather than results that warrant study in their own right. Bugs are fixed, experiments are refocused, and one-off surprises are collapsed into a single data point. The stories themselves are traded among researchers through oral tradition, but that mode of information transmission is inefficient and prone to error and outright loss. Moreover, the fact that these stories tend to be shared only among practitioners means that many natural scientists do not realize how interesting and lifelike digital organisms are and how natural their evolution can be. To our knowledge, no collection of such anecdotes has been published before. This article is the crowd-sourced product of researchers in the fields of artificial life and evolutionary computation who have provided first-hand accounts of such cases. It thus serves as a written, fact-checked collection of scientifically important and even entertaining stories. In doing so we also present here substantial evidence that the existence and importance of evolutionary surprises extends beyond the natural world, and may indeed be a universal property of all complex evolving systems.

Populations exposed to a high mutation rate harbor abundant deleterious genetic variation, leading to depressed mean fitness. This reduction in mean fitness presents an opportunity for selection to restore fitness through the evolution of mutational robustness. In extreme cases, selection for mutational robustness can lead to flat genotypes (with low fitness but high robustness) outcompeting fit genotypes (with high fitness but low robustness)—a phenomenon known as survival of the flattest . While this effect was previously explored using the digital evolution system Avida, a complete analysis of the local fitness landscapes of fit and flat genotypes has been lacking, leading to uncertainty about the genetic basis of the survival-of-the-flattest effect. Here, we repeated the survival-of-the-flattest study and analyzed the mutational neighborhoods of fit and flat genotypes. We found that the flat genotypes, compared to the fit genotypes, had a reduced likelihood of deleterious mutations as well as an increased likelihood of neutral and, surprisingly, of lethal mutations. This trend holds for mutants one to four substitutions away from the wild-type sequence. We also found that flat genotypes have, on average, no epistasis between mutations, while fit genotypes have, on average, positive epistasis. Our results demonstrate that the genetic causes of mutational robustness on complex fitness landscapes are multifaceted. While the traditional idea of the survival of the flattest emphasized the evolution of increased neutrality, others have argued for increased mutational sensitivity in response to strong mutational loads. Our results show that both increased neutrality and increased lethality can lead to the evolution of mutational robustness. Furthermore, strong negative epistasis is not required for mutational sensitivity to lead to mutational robustness. Overall, these results suggest that mutational robustness is achieved by minimizing heritable deleterious variation.

The role of historical contingency in the origin of life is one of the great unknowns in modern science. Only one example of life exists—one that proceeded from a single self-replicating organism (or a set of replicating hypercycles) to the vast complexity we see today in Earth's biosphere. We know that emergent life has the potential to evolve great increases in complexity, but it is unknown if evolvability is automatic given any self-replicating organism. At the same time, it is difficult to test such questions in biochemical systems. Laboratory studies with RNA replicators have had some success with exploring the capacities of simple self-replicators, but these experiments are still limited in both capabilities and scope. Here, we use the digital evolution system Avida to explore the interplay between emergent replicators (rare randomly assembled self-replicators) and evolvability. We find that we can classify fixed-length emergent replicators in Avida into two classes based on functional analysis. One class is more evolvable in the sense of optimizing the replicators' replication abilities. However, the other class is more evolvable in the sense of acquiring evolutionary innovations. We tie this tradeoff in evolvability to the structure of the respective classes' replication machinery, and speculate on the relevance of these results to biochemical replicators.

Animal grouping behaviors have been widely studied due to their implications for understanding social intelligence, collective cognition, and potential applications in engineering, artificial intelligence, and robotics. An important biological aspect of these studies is discerning which selection pressures favor the evolution of grouping behavior. In the past decade, researchers have begun using evolutionary computation to study the evolutionary effects of these selection pressures in predator-prey models. The selfish herd hypothesis states that concentrated groups arise because prey selfishly attempt to place their conspecifics between themselves and the predator, thus causing an endless cycle of movement toward the center of the group. Using an evolutionary model of a predator-prey system, we show that how predators attack is critical to the evolution of the selfish herd. Following this discovery, we show that density-dependent predation provides an abstraction of Hamilton's original formulation of domains of danger . Finally, we verify that density-dependent predation provides a sufficient selective advantage for prey to evolve the selfish herd in response to predation by coevolving predators. Thus, our work corroborates Hamilton's selfish herd hypothesis in a digital evolutionary model, refines the assumptions of the selfish herd hypothesis, and generalizes the domain of danger concept to density-dependent predation.

We study complex networks in which the nodes are tagged with different colors depending on their function (colored graphs), using information theory applied to the distribution of motifs in such networks. We find that colored motifs can be viewed as the building blocks of the networks (much more than the uncolored structural motifs can be) and that the relative frequency with which these motifs appear in the network can be used to define its information content. This information is defined in such a way that a network with random coloration (but keeping the relative number of nodes with different colors the same) has zero color information content. Thus, colored motif information captures the exceptionality of coloring in the motifs that is maintained via selection. We study the motif information content of the C. elegans brain as well as the evolution of colored motif information in networks that reflect the interaction between instructions in genomes of digital life organisms. While we find that colored motif information appears to capture essential functionality in the C. elegans brain (where the color assignment of nodes is straightforward), it is not obvious whether the colored motif information content always increases during evolution, as would be expected from a measure that captures network complexity. For a single choice of color assignment of instructions in the digital life form Avida, we find rather that colored motif information content increases or decreases during evolution, depending on how the genomes are organized, and therefore could be an interesting tool to dissect genomic rearrangements.

We present a system that can evolve the morphology and the controller of virtual walking and block-throwing creatures (catapults) using a genetic algorithm. The system is based on Sims' work, implemented as a flexible platform with an off-the-shelf dynamics engine. Experiments aimed at evolving Sims-type walkers resulted in the emergence of various realistic gaits while using fairly simple objective functions. Due to the flexibility of the system, drastically different morphologies and functions evolved with only minor modifications to the system and objective function. For example, various throwing techniques evolved when selecting for catapults that propel a block as far as possible. Among the strategies and morphologies evolved, we find the drop-kick strategy, as well as the systematic invention of the principle behind the wheel, when allowing mutations to the projectile.

We evolved multiple clones of populations of digital organisms to study the effects of chance, history, and adaptation in evolution. We show that clones adapted to a specific environment can adapt to new environments quickly and efficiently, although their history remains a significant factor in their fitness. Adaptation is most significant (and the effects of history less so) if the old and new environments are dissimilar. For more similar environments, adaptation is slower while history is more prominent. For both similar and dissimilar transfer environments, populations quickly lose the ability to perform computations (the analogue of beneficial chemical reactions) that are no longer rewarded in the new environment. Populations that developed few computational “genes” in their original environment were unable to acquire them in the new environment.

One of the central questions in evolutionary biology concerns the dynamics of adaptation and diversification. This issue can be addressed experimentally if replicate populations adapting to identical environments can be investigated in detail. We have studied 501 such replicas using digital organisms adapting to at least two fundamentally different functional niches (survival strategies) present in the same environment: one in which fast replication is the way to live, and another where exploitation of the environment's complexity leads to complex organisms with longer life spans and smaller replication rates. While these two modes of survival are closely analogous to those expected to emerge in so-called r and K selection scenarios respectively, the bifurcation of evolutionary histories according to these functional niches occurs in identical environments, under identical selective pressures. We find that the branching occurs early, and leads to drastic phenotypic differences (in fitness, sequence length, and gestation time) that are permanent and irreversible. This study confirms an earlier experimental effort using microorganisms, in that diversification can be understood at least in part in terms of bifurcations on saddle points leading to peak shifts, as in the picture drawn by Sewall Wright.

We study the evolution of robustness in digital organisms adapting to a high mutation rate. As genomes adjust to the harsh mutational environment, the mean effect of single mutations decreases, up until the point where a sizable fraction (up to 30% in many cases) of the mutations are neutral. We correlate the changes in robustness along the line of descent to changes in directional epistasis, and find that increased robustness is achieved by moving from antagonistic epistasis between mutations towards codes where mutations are, on average, independent. We interpret this recoding as a breakup of linkage between vital sections of the genome, up to the point where instructions are maximally independent of each other. While such a recoding often requires sacrificing some replication speed, it is the best strategy for withstanding high rates of mutation.