Antagonistic coevolution between multiple quantitative traits: Matching dynamics can arise from difference interactions

Masato Yamamichi,Kelsey Lyberger,Swati Patel

doi:10.1002/1438-390x.12022

Masato Yamamichi, Kelsey Lyberger + Show 1 more

Open Access

https://doi.org/10.1002/1438-390x.12022

Copy DOI

Abstract

AbstractCoevolution is one of the major drivers of complex dynamics in population ecology. Historically, antagonistic coevolution in victim‐exploiter systems has been a topic of special interest, and involves traits with various genetic architectures (e.g., the number of genes involved) and effects on interactions. For example, exploiters may need to have traits that “match” those of victims for successful exploitation (i.e., a matching interaction), or traits that exceed those of victims (i.e., a difference interaction). Different models exist which are appropriate for different types of traits, including Mendelian (discrete) and quantitative (continuous) traits. For models with multiple Mendelian traits, recent studies have shown that antagonistic coevolutionary patterns that appear as matching interactions can arise due to multiple difference interactions with costs of having large trait values. Here we generalize their findings to quantitative traits and show, analogously, that the multidimensional difference interactions with costs sometimes behave qualitatively the same as matching interactions. While previous studies in quantitative genetics have used the dichotomy between matching and difference frameworks to explore coevolutionary dynamics, we suggest that exploring multidimensional trait space is important to examine the generality of results obtained from one‐dimensional traits.

Highlights

Theories developed from specific models are initially only as general as their underlying models, until they are shown to hold in other modeling frameworks
While previous studies in quantitative genetics have used the dichotomy between matching and difference frameworks to explore coevolutionary dynamics, we suggest that exploring multidimensional trait space is important to examine the generality of results obtained from one-dimensional traits
Victim-exploiter interactions are among the most fundamental type of ecological interactions. In addition to their importance in ecological communities, these interactions are widely recognized for playing an important role in ecological dynamics (e.g., extinction and predator-prey cycles: Cortez and Weitz (2014); Northfield and Ives (2013)) and evolutionary dynamics (e.g., sex, recombination, and epistasis: Otto and Lenormand (2002)), because each species needs to continue to adapt to new selection pressures by the other species

Summary

INTRODUCTION

Theories developed from specific models are initially only as general as their underlying models, until they are shown to hold in other modeling frameworks. Various coevolution models with differing genetic architectures (e.g., the number of genes involved) and interspecific interactions To account for genetic architecture, one approach is to consider coevolution with a few major genes that interact and produce a discrete phenotype Quantitative trait-based models (hereafter quantitative models) assume the genetic architecture consists of many genes of small effect and produce continuous phenotypes (e.g., Gavrilets, 1997; Saloniemi, 1993). As a proof of concept, we give an example model of theoretical quantitative genetics where a victim-exploiter coevolutionary dynamics arising from two pairs of difference traits can appear as a matching interaction in a single pair of traits

BACKGROUND

DISCUSSION