Moderate Environmental Variation Promotes the Evolution of Robust Solutions
📝 Abstract
Previous evolutionary studies demonstrated how evaluating evolving agents in variable environmental conditions enable them to develop solutions that are robust to environmental variation. We demonstrate how the robustness of the agents can be further improved by exposing them also to environmental variations throughout generations. These two types of environmental variations play partially distinct roles as demonstrated by the fact that agents evolved in environments that do not vary throughout generations display lower performance than agents evolved in varying environments independently from the amount of environmental variation experienced during evaluation. Moreover, our results demonstrate that performance increases when the amount of variations introduced during agents evaluation and the rate at which the environment varies throughout generations are moderate. This is explained by the fact that the probability to retain genetic variations, including non-neutral variations that alter the behavior of the agents, increases when the environment varies throughout generations but also when new environmental conditions persist over time long enough to enable genetic accommodation.
💡 Analysis
Previous evolutionary studies demonstrated how evaluating evolving agents in variable environmental conditions enable them to develop solutions that are robust to environmental variation. We demonstrate how the robustness of the agents can be further improved by exposing them also to environmental variations throughout generations. These two types of environmental variations play partially distinct roles as demonstrated by the fact that agents evolved in environments that do not vary throughout generations display lower performance than agents evolved in varying environments independently from the amount of environmental variation experienced during evaluation. Moreover, our results demonstrate that performance increases when the amount of variations introduced during agents evaluation and the rate at which the environment varies throughout generations are moderate. This is explained by the fact that the probability to retain genetic variations, including non-neutral variations that alter the behavior of the agents, increases when the environment varies throughout generations but also when new environmental conditions persist over time long enough to enable genetic accommodation.
📄 Content
1 Moderate Environmental Variation Promotes the Evolution of Robust Solutions
Nicola Milano*, Jônata Tyska Carvalho*#, Stefano Nolfi*
*Institute of Cognitive Sciences and Technologies National Research Council (CNR), Roma, Italia,
#Center for Computational Sciences (C3), Federal University of Rio Grande (FURG), Av. Italia, km 8, Rio Grande, Brasil jonata.carvalho@istc.cnr.it
Abstract Previous evolutionary studies demonstrated how evaluating evolving agents in variable environmental conditions enable them to develop solutions that are robust to environmental variation. We demonstrate how the robustness of the agents can be further improved by exposing them also to environmental variations throughout generations. These two types of environmental variations play partially distinct roles as demonstrated by the fact that agents evolved in environments that do not vary throughout generations display lower performance than agents evolved in varying environments independently from the amount of environmental variation experienced during evaluation. Moreover, our results demonstrate that performance increases when the amount of variations introduced during agents’ evaluation and the rate at which the environment varies throughout generations are moderate. This is explained by the fact that the probability to retain genetic variations, including non-neutral variations that alter the behavior of the agents, increases when the environment varies throughout generations but also when new environmental conditions persist over time long enough to enable genetic accommodation.
Keywords—environmental variations; evolvability; stability; artificial evolution.
Introduction
The last two decades have seen an increasing recognition of the role of environmental variations in
evolution.
The interaction between environmental conditions and the expression of genetic variation influences the
evolutionary dynamics. Genes influencing a trait in one environment may not be important in a different
one (Viera et al., 2000). Mutations often have environment-dependent effects (Kawecki, 1994;
Szafraniec, Borts and Korona, 2001). The environmental conditions influence the genetic interactions
among traits, i.e., the correlation between the genetic influences on a trait and the genetic influences of
another trait, which are known to influence the evolutionary dynamics (Sgro and Hoffmann, 2004). For
instance, the genetic correlations among certain traits can be positive in an environment and negative in
another one. Consequently environmental variations influence evolutionary trajectories in populations
(Sgro and Hoffmann, 2004).
2
Moreover, as stressed by West-Eberhard (2003), phenotypic variation arises not only as a result of genetic
variations but also as a result of environmental variations. “Environmentally induced phenotypic changes
can give rise to adaptive evolution as readily as mutational induced changes; both are equally subject to
genetic accommodation.” (West-Eberhard, 2003, pp.498).
In this paper we analyze the impact of environmental variation on the evolution of neuro-controlled
agents situated in an external environment. More specifically we analyze whether agents evolved in
environmental conditions that vary over generations outperform agents evolved in non-varying
environments. The obtained results demonstrated that indeed agents evolved in varying environments
outperform agents evolved in environments that do not vary throughout generations independently from
the amount of environmental variation experienced by evolving individuals.
This study is related to the area of optimization in dynamic environment (Jin and Branke, 2005; Cruz,
Gonzalez and Pelta, 2011) that focus on how evolving solutions can cope with optimization problems that
are dynamic and change over time stochastically and/or periodically (see also Kashtan, Noor and Alon,
2007; O’Donnell at al., 2014; Janssen et al., 2016). However, the objective of these studies is to evolve
flexible solutions, i.e. agents capable of adapting to new environmental conditions during a certain
number of generations, rather than robust solutions, i.e. agents capable of operating effectively in new
environmental conditions immediately without the need to further evolve. Consequently, also
methodological issues differ. For example, the inclusion of mechanisms that preserve population diversity
is clearly important for the evolution of flexible agents capable of adapting to the new environmental
conditions in few generations but is not necessarily important for the evolution of robust agents.
Moreover, the formalization of a method for measuring the speed with which agents adapt to the new
environmental conditions is crucial for studying the evolution of flexible agents but is not relevant for the
study of robust agents.
Previous research demonstrated how exposing evolving candidate solutions to (deli
This content is AI-processed based on ArXiv data.