Multivariate Phenotypic Divergence Due to the Fixation of Beneficial Mutations in Experimentally Evolved Lineages of a Filamentous Fungus

Sijmen E Schoustra,Rees Kassen,David Punzalan,Rola Dali,Howard D Rundle,William J Etges

doi:10.1371/journal.pone.0050305

Sijmen E Schoustra, Rees Kassen + Show 4 more

Open Access

https://doi.org/10.1371/journal.pone.0050305

Copy DOI

Abstract

The potential for evolutionary change is limited by the availability of genetic variation. Mutations are the ultimate source of new alleles, yet there have been few experimental investigations of the role of novel mutations in multivariate phenotypic evolution. Here, we evaluated the degree of multivariate phenotypic divergence observed in a long-term evolution experiment whereby replicate lineages of the filamentous fungus Aspergillus nidulans were derived from a single genotype and allowed to fix novel (beneficial) mutations while maintained at two different population sizes. We asked three fundamental questions regarding phenotypic divergence following approximately 800 generations of adaptation: (1) whether divergence was limited by mutational supply, (2) whether divergence proceeded in relatively many (few) multivariate directions, and (3) to what degree phenotypic divergence scaled with changes in fitness (i.e. adaptation). We found no evidence that mutational supply limited phenotypic divergence. Divergence also occurred in all possible phenotypic directions, implying that pleiotropy was either weak or sufficiently variable among new mutations so as not to constrain the direction of multivariate evolution. The degree of total phenotypic divergence from the common ancestor was positively correlated with the extent of adaptation. These results are discussed in the context of the evolution of complex phenotypes through the input of adaptive mutations.

Highlights

The study of adaptive evolution has, in recent years, proceeded along two largely independent lines
As a first step towards a more unified view of the impacts of adaptive evolution on phenotypes, we present the results of a multivariate phenotypic analysis of the response to selection during replicated experimental evolution of microbial populations
Analyses of rank can reveal whether phenotypic evolution was restricted in its trajectory, at the extreme occurring in only a single trait combination, or whether it occurred in multiple independent phenotypic directions

Summary

Introduction

The study of adaptive evolution has, in recent years, proceeded along two largely independent lines. Insight into patterns of multivariate trait evolution may be gained by characterizing the dominant axes of such covariance matrices via eigenanalysis [13]. This produces a set of orthogonal dimensions of variation, whose directions are described by their eigenvectors and the extent of variation along each are described by the corresponding eigenvalues. The distribution among the latter can be informative regarding the effective dimensionality of the matrix (i.e. its ‘rank’). Analyses of rank can reveal whether phenotypic evolution was restricted in its trajectory (i.e. a D-matrix of low rank), at the extreme occurring in only a single trait combination (i.e. dmax, the first or leading eigenvector of D), or whether it occurred in multiple independent phenotypic directions (i.e. independent trait combinations, reflected in a Dmatrix of high rank)

Methods

Results

Conclusion