Allometry, evolution and development of neocortex size in mammals

Abstract

Variation in neocortex size is one of the defining features of mammalian brain evolution. The paramount assumption has been that neocortex size indicates a monotonic allometric relationship with brain size. This assumption holds the concomitant neurodevelopmental assumption that the ontogenetic trajectory of neocortex size is so stable across species that it restrains changes in the direction of evolution. Here we test this fundamental assumption. Whereas previous research has focused exclusively on changes in mean size among groups (i.e., intercept), we additionally investigate changes in covariation (i.e., slope) and strength of allometric integration (i.e., residual variation). We further increase data resolution by investigating 350 species representing 11 mammalian orders.

Results identify nine shifts in covariation between neocortex and brainstem in different mammalian groups, indicate that these shifts occur independently of shifts in size, and demonstrate that the strength of allometric integration across different neocortical regions in primates is inversely related to the neurodevelopmental gradient such that later developing regions underwent more evolutionary change.

Although our results confirm that variation in brain organization is structured along a neurodevelopmental gradient, our results suggest two additional principles of size reorganization in brain evolution: (1) repatterning of growth allocation among brain regions may occur independently of size and (2) later developing regions indicate faster evolution, not necessarily directional evolution toward larger size. We conclude that the evolution of neocortex size in mammals is far more variable than previously assumed, in turn suggesting a higher degree of evolutionary flexibility in neurodevelopmental patterning than commonly suggested.

Introduction

One of the defining features of mammalian evolution is the extraordinary variation in overall brain size. Explaining the patterns and processes that characterize this variation has been a major driving force of brain evolutionary studies. Whereas the study of pattern is concerned with how variation is ordered across species, the study of process investigates the mechanisms that generate and maintain this order.

One of the central questions in the study of the pattern of mammalian brain evolution is whether certain brain regions can be identified that vary in size more than others (Passingham, 1975), and whether such highly variable regions explain variation in overall brain size (Smaers and Soligo, 2013, Smaers and Vanier, 2019). The neocortex has received particular attention in this regard (Barton and Harvey, 2000). Mammals have evolved a unique set of neocortical features, including six lamina, radial patterning of axonal connections across the lamina, and a clear differentiation between cortical gray matter and axonal white matter (Striedter, 2005). These features are thought to have allowed the so-called scaling up of the mammalian brain. Indeed, when considering the amount of change that has occurred among gross-anatomical brain regions, the neocortex does vary in size more than other regions (Finlay and Darlington, 1995).

However, fundamental questions as to which processes underpin the observed patterns of neocortical evolution remain. The main discussion here mirrors a longstanding discussion in evolutionary biology about the relative influence of selective pressure and biological constraints on evolutionary change. Whereas those emphasizing adaptation in response to directional selective pressures propose that traits are primarily determined by natural selection (Beldade et al., 2002a, Beldade et al., 2002b; Frankino et al., 2005, Frankino et al., 2007), those that emphasize the importance of biological constraints propose that the type and amount of trait change that can be accomplished over evolutionary time is largely determined by genetic and developmental mechanisms (Alberch, 1982; Brakefield, 2006; Jernvall, 2002; Maynard Smith et al., 1985; Schluter, 1996; Stern, 2000). In the context of brain evolution these assumptions translate into suggestions that variation in brain organization either changes flexibly according to behavioral selective pressures (Barton and Harvey, 2000; termed the “mosaic” hypothesis of brain evolution), or that variation in brain organization is restricted to what is possible given a conserved neurogenetic schedule (Yopak et al., 2010 Finlay and Darlington, 1995; termed the “concerted” hypothesis of brain evolution).

Here, we aim to contribute to this discussion by better describing patterns of neocortex size evolution and exploring whether these patterns may have resulted from comparative shifts in the neurodevelopmental schedule. In doing so, we increase both the analytical and empirical resolution compared to previous research. Whereas previous research has focused exclusively on variation in relative neocortex size (differences in intercept) (Barton and Harvey, 2000) or the interpretation of monotonic allometric differences among brain regions (Clancy et al., 2001), we extend our analysis to investigate shifts in covariation (differences in slope) and strength of allometric integration (differences in residual variation). We further increase comparative data resolution compared to previous research by investigating 350 species representing 11 mammalian orders.