Determining the relationships between genes, traits, and the environment is a broad goal of biology. Within this broad goal, understanding the interplay of ecological and evolutionary dynamics is a key goal of modern ecology. Species possess traits that mediate ecological dynamics; when the traits rapidly evolve, population dynamics may be altered. These traits are underlain by dozens, hundreds, or thousands of genes that guide development. Refining the mapping from genes to traits, and from traits to the environment, is being facilitated by recent technological advances. We would like to have sets of expectations for this variety of mappings. For example, do we expect systematic eco-evolutionary differences when traits are underlain by gene networks of different sizes or topologies?
I tested hypotheses relating gene networks underlying an ecologically-critical trait to a variety of ecological scenarios using agent-based models. Agents possess an ecologically-important trait encoded by a Boolean gene regulatory network of 16, 32, 64, 128, or 256 genes, initiated in either a random or a scale-free topology. The model includes sexual reproduction, mutation, and recombination. Ecological scenarios include a single, sudden environmental change; a constantly-fluctuating environment; two competing species in a single patch; and a two-species, three-patch metacommunity.
Results/Conclusions
Smaller, scale-free networks confer higher trait heritability than larger, random networks. Keeping with prior research that relates heritability to population recovery, I find that populations in which the gene network is small and scale-free (thus, higher heritability) recover faster than populations with larger and more random gene networks. Smaller, scale-free networks also permit longer population persistence in rapidly-changing environments, but larger network populations persist equally well when the rate of environmental change is slow.
A trade-off between adaptation speed and accuracy, driven by the sizes of the underlying gene networks and the rate of environmental change, is recovered when two species compete. Smaller networks confer an adaptive speed advantage, but given slow environmental change, the greater adaptive accuracy afforded by larger networks leads to competitive superiority. Lastly, when two species compete in a metacommunity, differences in dispersal rate can override adaptive advantages afforded by specific gene networks.
These results provide a set of theoretical expectations when considering genomic information in the light of ecological dynamics. For example, traits that map to environmental variables with high spatio-temporal heterogeneity should be encoded by networks as small as possible, yet large enough to permit high adaptive accuracy relative to other species.