Fig 1.
Brain size-body size envelopes of the major vertebrate lineages, to illustrate both intra-lineage and inter-lineage variation.
The long-dashed outlines represent the 2 endothermic lineages (birds and mammals), the dotted outlines represent fishes, and the solid outlines the 2 ectothermic tetrapod lineages (amphibians and reptiles). Redrawn after [97].
Fig 2.
Categorization of the various hypotheses to explain evolutionary variation in relative brain size among vertebrates.
Two complementary clusters of hypotheses focus on either costs (expensive brain) or benefits (cognitive buffer). The brain size of a given species should reflect the balance between all relevant processes. In this review, we elaborate the developmental aspects of the expensive brain hypothesis. Key references are provided in superscripts: a: [124] (see also delayed benefits: [125]; b: [30]; 1: metabolic demands: [126,127]; 2: [45]; 3: [29, 30]; 4: [128,129]; 5: [55] (see also maturational constraints and brain malnutrition risks: [125]); 5: [83,88,108] (also: maternal energy); 7: [130,131]; 8: [132,133]; 9: [22]; 10: [134].
Fig 3.
The bootstrapping problem for developing brains.
In the absence of provisioning, immature animals developing their brains would likely face a long period of negative energy balance. Before the brain is fully grown and differentiated, it cannot provide adult-level cognitive benefits and the concomitant energy intake (green curve). The costs of growing, differentiating, and maintaining the brain (red curve) rise early and may even exceed adult values due to higher relative brain size of older immatures in mammals and costs of brain differentiation, before cognitive benefits, with their corresponding net energetic intake, stabilize at adult level. As a result, without parental provisioning the individual’s energy balance would be positive only after adulthood was reached.
Fig 4.
Schematic depiction of brain growth relative to body growth in different vertebrates as a function of parental provisioning.
The first phase (parental provisioning) shows the same, steep slope (virtually isometric: 1.0). After the end of parental provisioning, the slope becomes very low (ca 0.2) in mammals and altricial birds, whereas it become intermediate in precocial birds and ectotherms (ca 0.5) until adulthood is reached, and in most of the latter continues at the same relative rate after that due to indeterminate growth (sources are provided in the main text).
Fig 5.
Brain–body allometries among altricial and precocial birds.
Violin plots of slopes of the brain–body allometries of altricial (gray bars) and precocial (white bars) orders and families of birds. Data are taken from [104]. Orders or families were included when >5 species were available. Horizontal bars represent the median, red diamonds the mean, and boxes enclose the central 50% percentile range. The difference at the family level is significant (t = 2.60, DF = 26.46, p < 0.02). Sample sizes for altricial birds: 15 orders and 38 families; for precocial birds: 8 orders and 17 families.
Fig 6.
Brain size and the nature of cognition.
Natural selection is expected to optimize brain size, by finding the optimum balance between the cognitive abilities (in the broad sense) required to pay for the costs of maintaining the adult brain (A) and constructing it during development (B) on the one hand, and the brain-size-dependent cognitive abilities that are translated into adult performance (C) and immature performance (survival: D) on the other hand. The 4 sets of cognitive abilities no doubt show high overlap, but their nature remains poorly studied. For birds, set A would presumably contain abilities such as migratory habits, food storing, extractive foraging, and communal roosting; set B abilities like predation avoidance (especially of nest contents), efficient foraging, habitat and nest site selection, flexibility, coordination ability; set C many of the same abilities as A and B, but also avoidance of predation on adults, post-independence skill learning, optimal mate choice, and social skills; and set D also nest site selection, nest building, and predation-sensitive provisioning.