2017

Approaches to Macroevolution: 1. General Concepts and Origin of Variation

Approaches to macroevolution require integration of its two fundamental components, i.e. the origin and the sorting of variation, in a hierarchical framework.

Macroevolution occurs in multiple currencies that are only loosely correlated, notably taxonomic diversity, morphological disparity, and functional variety.

The origin of variation within this conceptual framework is increasingly understood in developmental terms, with the semi-hierarchical structure of gene regulatory networks (GRNs, used here in a broad sense incorporating not just the genetic circuitry per se but the factors controlling the timing and location of gene expression and repression), the non-linear relation between magnitude of genetic change and the phenotypic results, the evolutionary potential of co-opting existing GRNs, and developmental responsiveness to nongenetic signals (i.e. epigenetics and plasticity), all requiring modification of standard microevolutionary models, and rendering difficult any simple definition of evolutionary novelty.

The developmental factors underlying macroevolution create anisotropic probabilities—i.e., an uneven density distribution—of evolutionary change around any given phenotypic starting point, and the potential for coordinated changes among traits that can accommodate change via epigenetic mechanisms.

From this standpoint, “punctuated equilibrium” and “phyletic gradualism” simply represent two cells in a matrix of evolutionary models of phenotypic change, and the origin of trends and evolutionary novelty are not simply functions of ecological opportunity.

Over long timescales, contingency becomes especially important, and can be viewed in terms of macroevolutionary lags (the temporal separation between the origin of a trait or clade and subsequent diversification); such lags can arise by several mechanisms: as geological or phylogenetic artifacts, or when diversifications require synergistic interactions among traits, or between traits and external events.

The temporal and spatial patterns of the origins of evolutionary novelties are a challenge to macroevolutionary theory; individual events can be described retrospectively, but a general model relating development, genetics, and ecology is needed.

An accompanying paper (Jablonski in Evol Biol 2017) reviews diversity dynamics and the sorting of variation, with some general conclusions.

Whether these changes represent an overturning, an expansion, or a minor polishing of the neodarwinian theory of 50 years ago depends entirely on whose version of neodarwinism is used

The nonrandom distribution of phenotypes in morphospace, through time, and across environments, plays out at multiple hierarchical levels and so is amenable to integrated study in many systems, from the widest, most comprehensive focus to the narrowest and the most particular.

The analytical and modeling toolkits in both fields, and at their intersection, are gaining great power.

The next step is to hone a macroevolutionary theory that does not simply accommodate the observed patterns but is sufficiently mechanistic and multidimensional to gain predictive power

The complexity of the genotype-to-phenotype map undermines any direct correspondence between the differential probabilities of change at the genetic level and the nonrandom probabilities of change at the phenotypic level

On top of this potential for nonrandom mutation, another level of nonrandom variation is interposed between the genome and the organism by development, with its complex interactions involving multiple genetic pathways, cells, tissues, and attendant epigenetic effects 

Control Hierarchy

A starting point for a macroevolutionary view of variation is the now-commonplace observation that development is governed by semi-hierarchical networks of genes.

These gene regulatory networks (GRNs) are semi-hierarchical because, although much information on the time, place, and intensity of gene expression flows from high-level control genes through multiple intermediate steps to batteries of genes at the peripheries of these networks, GRNs contain feedbacks from lower levels within the network that can modulate expression of higher-level genes

we know little about which developmental differences tend to give rise to macroevolutionary ones.

transcription-factor binding sites evolve significantly more rapidly in mammals than in Drosophila, perhaps because of strong differences in effective population sizes, but whether these contrasts translate into systematic differences in phenotypic evolution has not been addressed.

Modularity

development, and therefore the evolution of development, is modular, i.e. organized into semi-independent regions, such that changes in gene expression in one module are more likely to affect GRNs, and ultimately the phenotype, of that module than of other modules.

“universal pleiotropy and epistasis” (the rule that every gene affects many traits and traits are determined by many genes of equal and small effect) is not as pervasive or chaotic as often assumed