Dogs Have the Most Neurons, Though Not the Largest Brain: Trade-Off between Body Mass and Number of Neurons in the Cerebral Cortex of Large Carnivoran Species

Abstract

Carnivorans are a diverse group of mammals that includes carnivorous, omnivorous and herbivorous, domesticated and wild species, with a large range of brain sizes. Carnivory is one of several factors expected to be cognitively demanding for carnivorans due to a requirement to outsmart larger prey. On the other hand, large carnivoran species have high hunting costs and unreliable feeding patterns, which, given the high metabolic cost of brain neurons, might put them at risk of metabolic constraints regarding how many brain neurons they can afford, especially in the cerebral cortex. For a given cortical size, do carnivoran species have more cortical neurons than the herbivorous species they prey upon? We find they do not; carnivorans (cat, mongoose, dog, hyena, lion) share with non-primates, including artiodactyls (the typical prey of large carnivorans), roughly the same relationship between cortical mass and number of neurons, which suggests that carnivorans are subject to the same evolutionary scaling rules as other non-primate clades. However, there are a few important exceptions. Carnivorans stand out in that the usual relationship between larger body, larger cortical mass and larger number of cortical neurons only applies to small and medium-sized species, and not beyond dogs: we find that the golden retriever dog has more cortical neurons than the striped hyena, African lion and even brown bear, even though the latter species have up to three times larger cortices than dogs. Remarkably, the brown bear cerebral cortex, the largest examined, only has as many neurons as the ten times smaller cat cerebral cortex, although it does have the expected ten times as many non-neuronal cells in the cerebral cortex compared to the cat. We also find that raccoons have dog-like numbers of neurons in their cat-sized brain, which makes them comparable to primates in neuronal density. Comparison of domestic and wild species suggests that the neuronal composition of carnivoran brains is not affected by domestication. Instead, large carnivorans appear to be particularly vulnerable to metabolic constraints that impose a trade-off between body size and number of cortical neurons.

You Do Not Mess with the Glia

“…Importantly, this cell quantification technique has been found by three independent groups to yield results that are comparable to those obtained with unbiased stereology, but are much faster to obtain and far less prone to user error and undersampling [66][67][68]. The consistency of the approach and technique across studies allowed us to collect data that could be compared systematically across structures in individual brains; across individuals of the same species; across species within a clade; across mammalian clades ( Figure 4); and even across mammals, birds, and non-avian reptiles [53][54][55][56][57][58][59][60][61][62][63][64][69][70][71]. While published results have so far been limited to numbers of neuronal and non-neuronal cells, one advantage of the isotropic fractionator is that, because all tissue heterogeneities in cell distribution are literally dissolved, only very small samples are required for counting, which allows for storage of the remaining suspension at −20 • C for later studies employing new markers or morphological criteria [65].…”

“…Over the last 12 years, we and our collaborators have generated a wealth of data on the numbers of neuronal and non-neuronal cells that compose brain structures in over 50 species of mammals [53][54][55][56][57][58][59][60][61][62][63][64]. Our systematic approach to determine the cellular composition of brain structures in a manner that was readily comparable across species, using reproducible dissection criteria, employed a quantitative technique that we developed-the isotropic fractionator [65].…”

“…While that initially appeared to be the case for the cerebral cortex alone, at a time when only a handful of species were analyzed [50], systematic examination of species representing all mammalian clades benefitting from the ease of the isotropic fractionator method showed that neuronal densities do not universally decrease with increasing brain size [53][54][55][56][57][58][59][60][61][62][63][64]-which would have been the requirement for universally increasing GNRs with increasing brain size. Despite the ease with which authors once could claim that the human brain had 10 times more glia than neurons [42,99,100], there are at best similar numbers of neuronal and glial cells in the human brain as a whole [55].…”

Longevity and sexual maturity vary across species with number of cortical neurons, and humans are no exception

“…Data on maximal longevity (years), time to reach sexual maturity from birth (days) in males and females, body mass (g) and metabolic rate (W) for mammalian and bird species were obtained from the carefully curated AnAge database compiled by de Magalhães and Costa (). Species belonging to the following 17 orders or superorders (“clades”) were analyzed, chosen as the scaling relationship between brain mass and number of cerebral cortical neurons that apply to at least five species in the respective clades are known (Dos Santos et al, ; Herculano‐Houzel et al, ; Jardim‐Messeder et al, ; Olkowicz et al, ): Afrosoricida ( n = 7 in the AnAge database), Artiodactyla ( n = 172), Carnivora ( n = 205), Cetacea ( n = 45), Dasyuromorphia ( n = 38), Didelphimorphia ( n = 20), Diprodontia ( n = 72), Erinaceomorpha ( n = 10), Hyracoidea ( n = 3), Lagomorpha ( n = 20), Macroscelidea ( n = 9), Passeriformes ( n = 417), Primata ( n = 174), Proboscidea ( n = 2), Psittaciformes ( n = 169), Rodentia ( n = 357) and Soricomorpha ( n = 33), in a total of 1,753 species (1,167 mammals and 586 birds). We thus did not examine the order Chiroptera, which is known to include extraordinarily long‐lived species for their small body size, simply because data on numbers of neurons in the brain of these animals are not yet available.…”

“…Maximal longevity, age at female sexual maturity, and longevity postmaturity across the ensemble of bird and mammalian species (including humans) are best predicted by the number of neurons in the cerebral cortex than by numbers of neurons in other brain structures or in the whole brain. Each circle represents values for one species (black, birds; red, primates; pink, artiodactyls; blue, afrotherians; brown, marsupials; ochre, carnivorans; dark green, glires; kaki, eulipotyphlans) from the AnAge database (de Magalhães & Costa, ) or obtained with the isotropic fractionator (Dos Santos et al, ; Herculano‐Houzel et al, ; Jardim‐Messeder et al, ; Olkowicz et al, ), for the entire cerebral cortex (a–c), cerebellum (d–f), rest of brain (brainstem, diencephalon [including the hypothalamus], striatum; g–i), and whole brain (j–l). Values of r 2 indicated in each graph apply to the power function fitted to all bird and mammalian species together (listed in Table ).…”