Decephalisation in domestication, part 1

@tonyrebelo @jeremygilmore @botswanabugs @paradoxornithidae @matthewinabinett

As far as I know, all researchers in the field of brain size in mammals agree that domesticated species tend to be decephalised relative to their wild ancestors (Kruska 2005, https://karger.com/bbe/article-abstract/65/2/73/46634/On-the-Evolutionary-Significance-of?redirectedFrom=fulltext).
 
By ‘decephalised’ I mean that, relative to body mass, the brain is less massive in the domestic form than in its ancestral, wild form.

(Please note the difference between encephalisation quotient (EQ) and intelligence quotient (IQ). The former refers to brain size, and can be measured for any animal, including humans; the latter refers to intelligence, and is specific to humans. In this Post, I do not refer to IQ.)
 
However, the more I think about the decephalisation (reduced EQ) of domestic mammals, the more questions it raises.
 
It makes sense that artificial selection during the process of domestication should generally have the effect of reducing the braininess of a given species of mammal. This is mainly because domestic species must be tractable, and the intractability of wild species is partly owing to their intelligence.

In effect, part of what the domestication process achieves is to increase the difference in intelligence between Homo sapiens and the form in question, so that we, the human domesticators, can boost our advantage in intelligence in the relationship.
 
However, there are at least four doubts/questions associated with the well-accepted generalisation that domestication involves decephalisation.
 
Firstly, the facts are actually rather poorly established.

The wild ancestor has long been extinct in the case of domestic horse (Equus caballus), European bost (Bos taurus), Indian bost (Bos indicus), river buffalo (Bubalus bubalis), dromedary (Camelus dromedarius), perhaps domestic dog (Canis familiaris), and, at least in part, ass (Equus asinus).

The ancestors have survived in the case of water buffalo (Bubalus carabanensis), and bactrian camel (Camelus bactrianus, https://www.inaturalist.org/taxa/524409-Camelus-bactrianus), but as far as I know the encephalisation quotient has never been measured in these species.

I am unaware of figures and statistics in the case of domestic rat and domestic mouse, despite the availability of both wild (ancestral) and laboratory populations. The same may apply to the guinea pig (Cavia porcellus, https://www.inaturalist.org/taxa/119405-Cavia-porcellus) and hamsters (https://en.wikipedia.org/wiki/Hamster).

The domestic pig (Sus scrofa) is rather mixed-up taxonomically, with possible several species involved both ancestrally and in domestication.

In the case of both domestic sheep (Ovis aries) and domestic goat (Capra hircus), not one but several species seem to have contributed to domestication.

In the case of the domestic dog, various breeds may differ greatly in encephalisation quotient, if intelligence of behaviour is any guide (think of border collie vs livestock landraces bred for meat to be eaten by Homo sapiens).

The encephalisation quotient of the domestic cat (Felis catus) has been measured. However, I have yet to see a clear comparison with its wild ancestor, which remains available for study. My impression is that there is little difference in encephalisation between ancestral Felis lybica (https://www.inaturalist.org/taxa/371414-Felis-lybica) and its domestic descendant, despite the fact that the domestic form is different enough to warrant a different species-name.

Indeed, the more one looks at this whole picture the more complicated and uncertain it seems to become.
 
Secondly, in the case of rodents the measurement of encephalisation quotient has long been unsatisfactory.

It is obvious to any scientist who knows Rattus norvegicus (https://www.inaturalist.org/taxa/44576-Rattus-norvegicus) well that this is an intelligent species, showing social complexity, play behaviour in adulthood, and versatility in response to different situations.

However, by most allometric methods the laboratory rat and its ‘wild’ ancestor (which is actually profoundly commensal today, rather than wild) have EQ values much less than the 1.0 characterising the average mammal.

This makes little sense. As long as we use methods that fail to correlate the obvious intelligence of rats with their EQ values, it must remain in doubt how domestication has affected their encephalisation. Mus musculus (https://www.inaturalist.org/taxa/44705-Mus-musculus) seems even more poorly documented in this way.

A similar problem occurs in the case of the domestic pig. This species has puzzlingly small values for EQ relative to other cloven-hoofed mammals, even in the case of the wild species of suids which have been studied. Why is it that even the fully wild common warthog (Phacochoerus africanus) has EQ of only about 0.6, whereas antelopes and deer of similar body size tend to have EQ values of about 1.0?

All suids seem to be ‘naturally decephalised’. So, how sure can we be of the decephalisation-by-domestication in the domestic pig, even if the latter were clearly just one species (which it is not because domestication of pigs seems to have proceeded from different wild ancestors in Europe vs China, notwithstanding the inevitable mixing)?
 
Thirdly, to what degree has the decephalisation-by-domestication been the result of the boosting of body size in domestication?

The domestic horse is a good example of this problem.

The ancestral, wild species (which may have been more than one, for all we know) has long been extinct. Therefore, we have no way of knowing precisely how encephalised it was.

However, even if we assume that the ancestral encephalisation was similar to that of the plains zebra (EQ about 1.2, or slightly more encephalised than the average mammal), the trouble is that there is such a wide range of body sizes in the domestic horse that it is hard to come up with a representative EQ value for the species as a whole.

It seems that carthorses, which weigh about 1 tonne, have far smaller values for EQ than ponies, owing to differences in body mass. The actual values involved have yet to be clearly documented, but seem to range from >1 (ponies) to 0.6 (largest-bodied breeds).

If one focuses on ponies, of similar body size to the hypothetical wild ancestor, is it really true that decephalisation has occurred? How can we accept that the domestic horse is decephalised relative to the average mammal (EQ < 1 ) when ponies in particular act in such an intelligent way that it is obvious that they are more intelligent than the average mammal?
 
Fourthly, no study has yet documented any reversion towards ancestral, wild EQ in feral populations of any domesticated mammal. Such data as are available indicate that the feral populations remain decephalised regardless of how long the feral condition exists, with its at least partial reversion to ‘natural’ selection for fitness.

Why is the domestically reduced brain size so permanent, given the great individual variation in brain size that has been documented even within fully domestic populations? How can we reconcile the fact that, even in domestication, brain size varies so much relative to body size, with the supposed fact that such variation fails to result in an evolutionary reversion towards increased average brain size after several generations of feral life?
 
To exemplify these problems, please bear in mind the following w.r.t. the domestic dog.
 
a) What is the EQ of the breed we are considering?

The facts are suprisingly obscure. I think the answer is usually ‘about 1’. This would mean that the domestic dog has an encephalisation about average for a mammal, something that seems to make fair sense.

However, how much does EQ differ between e.g. border collie and e.g. pit-bull?
 
b) How can we really know what the ancestral, wild EQ was? Do we really know which species the domestic dog has been bred from?

The prevailing belief is that the wild ancestor was a ‘wolf’. However, there are various problems with this, including the fact that there are various species/subspecies of wolves, and all surviving forms of wolf have been subject to some interbreeding with the domestic dog.

On the face of it, wolves do seem to have far greater values for EQ than the domestic dog, i.e. about 1.8 compared to about 1.0. However, the more one thinks about how we actually know this, the more dubious the facts of decephalisation become.
 
c) As in the case of the domestic horse, how can we know the EQ of the domestic dog without fully taking body size into consideration?

No doubt, EQ would vary considerably were we to measure it in the chihuahua (https://en.wikipedia.org/wiki/Chihuahua_(dog)) vs the St Bernard (https://en.wikipedia.org/wiki/St.Bernard(dog)). Which is the ‘true’ value? And what if the real wild ancestor was not some large form of wolf, but instead a jackal-size species, similar in appearance to the black-and-tan kelpie (https://en.wikipedia.org/wiki/Australian_Kelpie)?
 
d) Imagine that a large founder-population of a given breed were set free in a post-apocalyptic world, where it would be subject to ‘natural’ selection for many generations. How long – if at all – would it take the population to revert to the ancestral EQ? For that matter, is the EQ of the dingo (https://en.wikipedia.org/wiki/Dingo) greater than that of fully domestic breeds?
 
e) Consider the border collie (https://en.wikipedia.org/wiki/Border_Collie), with its conspicuously intelligent behaviour. Is it possible that this breed is still as decephalised as any other breed of domestic dog, but simply represents a reorganisation of the brain, favouring certain functions?

Another way of asking this question:
If one did a thought-experiment in which one feral post-apocalyptic population was founded purely in the border collie, and another was founded purely in the pit-bull, would there be any difference in EQ, and fitness, between the two populations after, say, 100 generations?
 
Dear readers, do you see how something that is passed off, in the most recent literature (e.g. Kruska 2005), as a ‘simple fact’, is actually so complicated that it makes the mind spin?

Posted on June 19, 2022 12:26 AM by milewski