Evolutionary-developmental (evo-devo) dynamics of hominin brain size

Abstract

Brain size tripled in the human lineage over four million years, but why this occurred remains uncertain. Here, to study what caused this brain expansion, I mathematically model the evolutionary and developmental (evo-devo) dynamics of hominin brain size. The model recovers (1) the evolution of brain and body sizes of seven hominin species starting from brain and body sizes of the australopithecine scale, (2) the evolution of the hominin brain-body allometry and (3) major patterns of human development and evolution. I show that the brain expansion recovered is not caused by direct selection for brain size but by its genetic correlation with developmentally late preovulatory ovarian follicles. This correlation is generated over development if individuals experience a challenging ecology and seemingly cumulative culture, among other conditions. These findings show that the evolution of exceptionally adaptive traits may not be primarily caused by selection for them but by developmental constraints that divert selection.

Fig. 1. Evolution of brain and body sizes of seven hominin species solely by changing socio-genetic covariation.

Adult brain and body sizes of seven hominin species evolve in the model only by changing the EETB, the returns of learning and how the skills of cooperating partners interact. Squares are the observed adult brain and body sizes for the species at the top (data from refs. ^,,–). Dots are the evolved values in the model for a 40-year-old using the evo-devo dynamics approach. Pie charts give the EETB used in each scenario. The returns of learning are either strongly diminishing (power competence) for the left four scenarios or weakly diminishing (exponential competence) for the right three scenarios. Cooperation is either submultiplicative for the afarensis and right three scenarios or additive for the remaining scenarios. These EETBs and shapes of EEE were previously identified as evolving best-fitting adult brain and body sizes for the corresponding species, assuming evolutionary equilibrium. In principle, weakly diminishing returns of learning might arise from culture. I will show that varying EETBs and the shape of EEE only varies socio-genetic covariation L_z but not the direction of direct selection ∂w/∂z or where it is zero (it never is). I refer to the particular EETB and shape of EEE yielding the evolution of adult brain and body sizes of a given species as the species scenario. For the afarensis scenario, the ancestral genotypic traits are somewhatNaive2 (Supplementary equation (46)). For the remaining six scenarios, the ancestral genotypic traits are the final genotypic traits of the afarensis scenario started from the somewhatNaive2 genotypic traits. The final evolutionary time is 500 for all 7 scenarios. Pie charts reproduced with permission from ref. , Springer Nature Ltd.

Fig. 2. Brain–body allometries without and with evolution.

a, Brain size at 40 years of age versus body size at 40 years of age on a log–log scale, developed under the brain model from 10⁶ randomly sampled genotypes (that is, growth efforts, drawn from the normal distribution with mean 0 and standard deviation 4) using the parameter values of the sapiens scenario. Black dots are ‘non-failed’ organisms, whose body is not entirely composed of brain at 40 years of age, and are approximately 4% of 10⁶. Grey dots are ‘failed’ organisms having small bodies (<200 g) at 40 years of age entirely composed of brain tissue owing to tissue decay from birth, and are about 96% of 10⁶ (Supplementary Fig. 12). Coloured regions encompass extant and fossil primate species. b, Brain size at 40 years of age versus body size at 40 years of age over evolutionary time on a log–log scale for two trajectories. The bottom trajectory uses the parameter values of the afarensis scenario (Fig. 1) and somewhatNaive2 ancestral genotypic traits. The top trajectory uses the parameter values of the sapiens scenario (Fig. 1) and the evolved genotypic traits of the bottom trajectory as ancestral genotypic traits. A linear regression over the top trajectory yields a slope of 1.03 (red line). Adult values for 13 hominin species are shown in green squares. Brain and body size data for non-hominins are from ref. , excluding three fossil, outlier cercopithecines; brain and body size data for hominins are from refs. ^,,,,,– using only female data when possible. Fossil data may come from a single individual and body size estimates from fossils are subject to additional error. H., Homo; A., Australopithecus; P., Paranthropus.

Fig. 3. Evo-devo dynamics of hominin brain size.

a–d, Developmental dynamics over age (horizontal axis) and evolutionary dynamics over evolutionary time (differently coloured dots; bottom left label): evo-devo dynamics of brain size (a), follicle count (b), body size (c) and skill level (d). e–g, Evolutionary dynamics of brain size (green) and skill level (orange) (e), body size (f) and EQ (g non-dimensional) at 40 years of age. In a and c, the mean observed values in a cross-sectional modern human female sample are shown in black squares (data from Supplementary Table 2 in ref. , which fitted data from ref. ), the mean observed values in cross-sectional Pan troglodytes female samples are shown in grey triangles (body size data from Fig. 2 in ref. ; brain size data from Fig. 6 in ref. ), and the mean observed values in A. afarensis female samples are shown in pink stars (data from Table 1 in ref. ). One evolutionary time unit is the time from mutation to fixation. If gene fixation takes 500 generations and 1 generation for females is 23 years, then 300 evolutionary time steps are 3.4 million years. The age bin size is 0.1 year. Halving age bin size (0.05 yr) makes the evolutionary dynamics twice as slow, but the system converges to the same evolutionary equilibrium (Supplementary Fig. 6). I take adult phenotypes to be those at 40 years of age as phenotypes have typically plateaued by that age in the model. All plots are for the sapiens trajectory of Fig. 2b.

Fig. 4. The action of selection.

a–d, There is no direct selection for brain size (a), somatic tissue size (c), nor skill level (d). There is direct selection only for follicle count, and such selection decreases with age (b). e, The angle between the direction of evolution and direct selection, both of the geno–phenotype (that is, genotype and phenotype), is nearly 90 degrees over evolutionary time. f, Evolvability is small and decreases over evolutionary time. Evolvability equal to 0 here means no evolution despite selection (Supplementary Section 7 and equation (1) in ref. ). g, Population size increases over evolutionary time (plot of $\frac{1}{2}\mu {\overline{n}}^{*}{\eta }_0$, so the indicated multiplication yields population size). Mutation rate μ and parameter η₀ can take any value satisfying 0 < μ ≪ 1 and 0 < η₀ ≪ 1/(N_gN_a), where the number of genotypic traits is N_g = 3 and the number of age bins is N_a = 47 yr/0.1 yr = 470. If μ = 0.01 and η₀ = 1/(3 × 47 yr/0.1 yr), then a population size of 1,000 × 2/(μη₀) is 2.82 billion individuals (which is unrealistically large owing to the assumption of marginally small mutational variance to facilitate analysis). All plots are for the sapiens trajectory of Fig. 2b.

Fig. 5. Illustration of the fitness landscape in the brain model.

The fitness landscape w is a linear function (equation (5)) of the follicle count $x_{\mathrm{r}a}=g_{\mathrm{r},a-1}\left({\mathbf{x}}_{a-1},{\mathbf{y}}_{a-1},{\overline{x}}_{\mathrm{k},a-1}\right)$, which is a recurrence over age. The slope of the fitness landscape with respect to x_ra is positive and decreases with age a (Fig. 4b). Evaluating the recurrence at all possible genotypic trait values ${\mathbf{y}}_{\mathrm{j}}\in {\mathbb{R}}^{N_{\mathrm{g}}}$ for all ages j < a gives values $x_{\mathrm{r}a,\min }$ and $x_{\mathrm{r}a,\max }$ that depend on development g_rj for all ages j < a, the various parameters influencing it, and the developmentally initial conditions. The admissible follicle count ranges from $x_{\mathrm{r}a,\min }$ to $x_{\mathrm{r}a,\max }$. The admissible path on the landscape is given by the admissible follicle count. As there are no absolute mutational constraints, evolution converges to the peak of the admissible path (dot), where total genotypic selection vanishes, dw/dy = 0 (Extended Data Fig. 3).

Fig. 6. The action of constraint on brain expansion.

a–d, Mechanistic socio-genetic cross-covariance matrix between brain size and follicle count at evolutionary time τ = 1 (a), τ = 10 (b), τ = 100 (c) and τ = 500 (d) for the sapiens trajectory of Fig. 2b. For instance, in b, the highlighted box gives the socio-genetic covariance between brain size at 20 years of age and follicle count at each of the ages at the top horizontal axis. Thus, at evolutionary time τ = 10, socio-genetic covariation between brain size at 20 years of age and follicle count at 6 years of age is negative (bottom bar legend) but between brain size at 20 years of age and follicle count at 30 years of age is positive. The positive socio-genetic covariation between brain size and follicle count (for example, yellow areas in b and c) causes brain expansion. Bar legends have different limits so that patterns are visible (bar legend limits are {−l, l}, where $l=\max \left(\mid L_{x_{\mathrm{b}a},x_{\mathrm{r}j}}\mid \right)$ over a and j for each τ).

Extended Data Fig. 1. Causal diagram of the brain model analysed under the evo-devo dynamics framework.

The evo-devo dynamics framework clarifies how to separate the direct and total effects of traits on fitness in the model. Variables have age-specific values. The phenotype comprises brain size, follicle count, somatic tissue size, and skill level, all constructed by a developmental process. Each arrow indicates the direct effect of a variable on another one. The total effect of a variable on another one is that across all the arrows directly or indirectly connecting the former to the latter. A mutant’s genotypic traits at a given age directly affect brain size, follicle count, somatic tissue size, and skill level at the immediately subsequent age (with the slope quantifying developmental bias from genotype). A mutant’s phenotypic traits at a given age affect themselves at the immediately subsequent age (quantifying developmental bias from the same phenotypic trait), thus the direct feedback loop from phenotypic traits to themselves. A mutant’s phenotypic traits at a given age also directly affect each other at the next age (quantifying developmental bias from immediately previous phenotypes). A mutant’s follicle count is the only trait directly affecting fitness (direct selection on follicle count). The social partner’s skill level at a given age directly affects its own development at an immediately subsequent age (quantifying developmental bias from the same phenotypic trait), thus the direct feedback loop. The social partner’s skill level at a given age also directly affects all the mutant’s phenotypic traits at the next age (quantifying indirect genetic effects from the phenotype). The genotype is assumed to be developmentally independent (that is, controls y are open-loop), which means that there is no arrow towards the genotype. This diagram is a simplification of that considered by the evo-devo dynamics framework, so the brain model can be extended and the framework can still be used to analyse it.

Extended Data Fig. 2. Evo-devo dynamics of brain size under afarensis scenario.

a-d, Developmental dynamics over age (horizontal axis) and evolutionary dynamics over evolutionary time (differently coloured dots; bottom left label): Evo-devo dynamics of a, brain size; b, follicle count; c, body size; and d, skill level. Evolutionary dynamics of (e, green) brain size, (e, orange) skill level, (f) body size, and (g) encephalisation quotient (EQ) at 40 years of age. a,c, The mean observed values in a modern human female sample are shown in black squares (data from ref. who fitted data from ref. ). One evolutionary time unit is the time from mutation to fixation.

Extended Data Fig. 3. The action of total selection.

a-d, Total selection on brain size, follicle count, somatic tissue size, and skill level at each age over evolutionary time. Total selection for skill level over life persists at evolutionary equilibrium (red dots in d). e-g, Total selection on effort for brain growth, follicle production, and somatic growth at each age over evolutionary time. Total selection for genotypic traits nearly vanishes at evolutionary equilibrium (red dots in e-g), indicating that a path peak on the fitness landscape is reached. All plots are for the sapiens trajectory of Fig. 2b.

Extended Data Fig. 4. The action of constraint on body, skill, and follicle count expansion.

Mechanistic socio-genetic cross-covariance matrix between: a-d, body size and follicle count, e-h, skill level and follicle count, and i-l, follicle count and itself. All plots are for the sapiens trajectory of Fig. 2b.