Brain evolution and development: adaptation, allometry and constraint

Abstract

Phenotypic traits are products of two processes: evolution and development. But how do these processes combine to produce integrated phenotypes? Comparative studies identify consistent patterns of covariation, or allometries, between brain and body size, and between brain components, indicating the presence of significant constraints limiting independent evolution of separate parts. These constraints are poorly understood, but in principle could be either developmental or functional. The developmental constraints hypothesis suggests that individual components (brain and body size, or individual brain components) tend to evolve together because natural selection operates on relatively simple developmental mechanisms that affect the growth of all parts in a concerted manner. The functional constraints hypothesis suggests that correlated change reflects the action of selection on distributed functional systems connecting the different sub-components, predicting more complex patterns of mosaic change at the level of the functional systems and more complex genetic and developmental mechanisms. These hypotheses are not mutually exclusive but make different predictions. We review recent genetic and neurodevelopmental evidence, concluding that functional rather than developmental constraints are the main cause of the observed patterns.

Keywords: adaptation; allometry; brain evolution; constraint; development.

Figure 1.

Origins of evolutionary constraints and covariance. Six scenarios that show how selection on one brain component (A) may cause coordinated changes throughout the system. The ancestral system is shown in the middle row; blue connections indicate developmental constraints (DC) and green connections indicate functional constraints (FC). Red outlines indicate the component(s) under primary selection; blue outlines indicate component(s) under secondary selection following changes in A. (i) Concerted brain evolution driven by DC: selection on A results in concerted expansion of all brain components. (ii) Concerted evolution with a small contribution of mosaicism: the evolution of new functions may be associated with an overall expansion of the system with a ‘top up’ for A driven by independent developmental mechanisms (top row). (iii) Mosaic evolution: a complete lack of constraint allows A to evolve independently. (iv) Mosaic evolution with FC: functional dependence between A and D means selection for A creates secondary selection for D to maintain the relationship between A and D (bottom row). If this functional relationship changes, A may be able to evolve without co-incident shifts in D (top row). (v) Mosaic evolution with system-wide functional dependence: selection on A will create secondary selection on the entire system (bottom row), patterns of covariance would appear identical to i and ii. If the functional connection changes between A and D, sub-networks A–C may evolve without co-incident shifts in A–D (top row). (vi) Mosaic evolution with partial DC and FC: If sub-networks A–C and B–D are developmentally linked internally, but functionally linked to other sub-networks, selection on A will result in a combination of secondary selection on D to maintain their functional relationship (lower row) and concerted expansion (of C and B) due to DC; the result is identical to i, ii and v. If the functional relationship changes between A and D, A may be able to respond without co-incident shifts in B–D but will still be accompanied by a ‘neutral’ change in C.

Figure 2.

Developmental routes to mosaic brain evolution. Selection can modify the relative size of individual brain components through three routes, as follows. (a) Modifying how the progenitor pool of cells that produce neurons is divided between regions by changing the boundaries of expression gradients of morphogens. A role for developmental patterning in creating variation in brain structure between species has been demonstrated in derived, cave dwelling populations of Atyanax mexicanus [88] and ecologically divergent cichlids in Lake Malawi [89]. (b) Prolonging the period of cell division in the progenitor pool of cells destined to form a specific component. Expansion of specific brain components has been linked to interspecific variation in region-specific duration of neurogenesis in Passerimorphae [–92], nocturnal Aotus monkeys [93] and Mammalia more generally [94]. (c) Accelerating the rate at which cells divide within a conserved developmental schedule. Variation in cell-cycle rate prior to the onset of neurogenesis is thought to contribute to interspecific differences in the relative size of the telencephalon in galliform birds [95].