Do the Large Cetacean Brains Deny the "Human Advantage" in the Cerebral Cortex?

                                                                                      Orcas

I have agreed with Suzana Herculano-Houzel's argument that the (very rough) estimate of 16 billion neurons in the human cerebral cortex--out of the (very rough) estimate of 86 billion neurons in the whole human brain--gives us a "human advantage" that explains why we are the dominant animal on the Earth.  No other animal, she predicts, has more cortical neurons than the 16 billion in the human cerebral cortex.

But what about those animals that have brains larger than the human brain--elephants, whales, and dolphins?  Herculano-Houzel and her colleagues have reported that the African elephant brain is about three times larger than the human brain, and the elephant cerebral cortex is about twice the weight of the human cerebral cortex (Herculano-Houzel, et al. 2014).  Using the isotropic fractionator method for counting neurons, they counted an estimated 257 billion neurons in the elephant brain, which is three times more than the average human brain.  They also found that 97.5% of the neurons in the elephant brain (251 billion) are in the cerebellum.  So the elephant cerebral cortex has only about 5.6 billion neurons, which is only about one-third of the neurons in the human cerebral cortex.  This confirmed Herculano-Houzel's prediction that animals with brain larger than human brains would not have as many cortical neurons as in the human cerebral cortex.

As I have indicated in previous posts, these claims come with at least two kinds of reservations.  First, all of these numbers are only very rough estimates, because there is such a great range of variation in brains.  Second, the estimates here are based on a very small sample size--only the right half of one adult male elephant brain!  Counting the neurons in other elephant brains will result in widely varying numbers.  And there's always the problem of relying on "convenience samples": whether one has access to even one fresh elephant brain is happenstance.

But now if you look at Wikipedia's "List of Animals by Neuron Count," you will see (at the end of the article) the highest estimated numbers of cortical neurons.  And while humans rank near the top--with Herculano-Houzel's estimate of 16 billion--you will also see the Orca (Orcinus orca), or killer whale, has 43.1 billion cortical neurons (Ridgway et al. 2019), about two and a half times more than in the human brain.  You will also see estimates of 37.2 billion cortical neurons for the long-finned pilot whale (Mortensen et al. 2014), 12.8 billion for the common Minke whale (Eriksen and Pakkenberg 2007), and 11.8 billion for the short-finned pilot whale (Avelino-de-Souza 2018).  So here we have two cetacean species with many more cortical neurons in their brains than in the human brain, and two cetacean species whose number of cortical neurons are very close to the number in the human brain.

These data seem to falsify Herculano-Houzel's prediction in 2016 that the number of cortical neurons in the cetacean cerebral cortex would always be much lower than the 16 billion neurons in the human cerebral cortex (Herculano-Houzel 2016, 103-107).  But if you look more carefully at the various reports of neuron counts for cetacean brains, it's not so clear that her prediction has failed.

Eriksen and Pakkenberg (2007) reported using the stereological technique in counting 12.8 billion neurons in the cerebral cortex of the Northern common minke whale based on averaging the counts for five brains.  But Kamilla Avelino-de-Souza, Nina Patzke, Karl Karlsson, Paul R. Manger, and Suzanna Herculano-Houzel (2025) have recently reported using the isotropic fractionator technique in counting 3.2 billion neurons in the Northern minke whale based on a count for one adult male brain.

Why was the stereological estimate so much higher than the isotropic fractionator estimate?  Herculano-Houzel and her colleagues argue that the stereological technique suffers from undersampling.  Eriksen and Pakkenberg (2007, 87-88) report that they sampled only 12-13 sections out of over 3000 sections of the Northern minke whale cerebral cortex, counting only up to 215 neurons per brain.  Sampling from so few and minute sites creates undersampling because the neuronal densities in different areas of the brain are highly variable.  By contrast, the isotropic fractionator method eliminates undersampling by collecting samples from the brain structure of interest only after it has been homogenized into a soup in which the neuronal cell nuclei are evenly distributed (Avelino-de-Souza et al. 2025, 17; Herculano-Houzel 2016, 25-26, 28-34, 104-105).

Previously, in 2018, in her dissertation at the Federal University of Rio de Janeiro, Avelino-de-Souza had reported using the isotropic fractionator technique to calculate that the Minke whale's cerebral cortex had 3.1 billion neurons--virtually the same number as in her recent report.  But in her dissertation, she also reported 11.8 billion neurons in the cerebral cortex of the short-finned pilot whale.  That's the number given in the Wikipedia list.  And it contradicts what Herculano-Houzel had predicted in 2016--that the number of cortical neurons in the pilot whale would be around 3 billion cortical neurons (Herculano-Houzel 2016, 104).  

I asked Herculano-Houzel about this, and she responded (in an email message):

In my recollection, the pilot whale [in Avelino-de-Souza's study] was a VERY old specimen--old in the sense that it had been sitting in fixative for a very, very long time, contrary to the minke whale we reported, which was freshly caught.  The nuclei still had detectable DNA, but were very autofluorescent.  Those counts were therefore an estimate based on total numbers of nuclei, but we cannot be really certain of what fraction were neurons.  We have a better specimen of a beluga whale now that we can work on and get a proper estimate for an odontocete.

I told her that I will be happy to see the numbers for that beluga whale.

Herculano-Houzel still has to explain the 43.1 billion count for the cortical neurons of the Orca (Ridgway et al. 2019) and the 37.2 billion count for the cortical neurons of the long-finned pilot whale (Mortensen et al. 2014).  Presumably, she will argue that these calculations suffer from the undersampling problem of the stereological technique.  But if some study using the isotropic fractionator technique were to confirm these counts for these cetaceans, that would falsify her predictions.

REFERENCES

Avelino de Souza, Kamilla. 2018. "Analysis of the Cellular Composition of the Cetacean Brain."  Ph.D. Dissertation. Program in Morphological Sciences. Federal University of Rio de Janeiro.

Avelino-de-Souza, Kamilla, Nina Patzke, Karl Karlsson, Paul R. Manger, and Suzana Herculano-Houzel. 2025. "Cellular Composition of the Brain of a Northern Minke Whale." Journal of Comparative Neurology 533:e70089.

Eriksen, Nina, and Bente Pakkenberg. 2007. "Total Neocortical Cell Number in the Mysticete Brain." The Anatomical Record 290:83-95.

Herculano-Houzel, Suzana, Kamilla Avelino-de-Souza, Leber Neves, Jairo Porfirio, Debora Messeder, Larissa Mattos Feijo, Jose, Maldonado, and Paul R. Manger. 2014. "The Elephant Brain in Numbers." Frontiers in Neuroanatomy 8: Article 46.

Herculano-Houzel, Suzana. 2016. The Human Advantage: A New Understanding of How Our Brain Became Remarkable. Cambridge: MIT Press.

Mortensen, Heidi S., Bente Pakkenberg, Maria Dam, Rune Dietz, Christian Sonne, Bjami Mikkelsen, and Nina Eriksen. 2014. Frontiers in Neuroanatomy 8: article 132.

Ridgway, Sam H., Robert H. Brownson, Kaitlin R. Van Alstyne, and Robert A. Hauser. 2019. "Higher Neuron Densities in the Cerebral Cortex and Larger Cerebellums May Limit Dive Times of Delphinids Compared to Deep-Diving Toothed Whales." PLoS ONE 14 (12): e0226206.