Is Brain Connectivity the Same among All Mammals?

[.Scott]

TL;DR Summary

Using diffusion MRI, they reconstructed the brain connectomes of 123 mammalian species. Network analysis revealed that both connectivity and the wiring cost are conserved across mammals.

This paper was published in Nature: Mammal Brain Connectivity
And was described by the University in this press release: https://www.aftau.org/press-release---brain-connectivity---july-20-2020

What interests my is the editorial spin:
From the press release:

The intriguing results, contradicting widespread conjectures, revealed that brain connectivity levels are equal in all mammals, including humans.

From the Abstract:

Over 100 years ago, Ramon y Cajal hypothesized that two forces played a role in the evolution of mammalian brain connectivity: minimizing wiring costs and maximizing conductivity speed.

Cajal and "widespread conjecture" notwithstanding, I have never had any expectation that the basic component technology in the human brain was any better than any other mammal. In fact, I would not be surprised if a few component enhancements are found in a much smaller species - for example mice.

Here is my take:
It's all about evolution and circuitry. It's hard to evolve any system one small change at a time. So step one would be to evolve it in an evolvable direction - tending towards modularity. Then later, small changes can have narrow results. If a small change is made in one place, it will not be at the cost of destroying other parts of the system.

With a nervous system, this become problematic. It needs to be modular, but any change is likely to create a system-wide compatibility problem.
For example, once I have established a communications protocol - say RS-232 at 1Mbaud - I would have to make matching changes to the transmitter and receiver to improve it - two exactly complementary changes in a single mutation - very unlikely. So, if the system is to be evolvable, it needs to have fairly tolerant components. Ex, a transmitter and receiver that can operate at a range of current levels and baud rates.

But that is just a communication link. What about the encoding of basic notions (thoughts) and the subsequent processing of that data? There will be some system-wide standards (such as how notions are encoded) that will be effectively set in stone for the entire evolutionary future of an organism. As the data processing becomes more elaborate, more and more subsystems will depend on these standards and the only avenue for changing the standard will be extinction and replacement. In other words: starting over.

How would you get "big improvements" when the evolutionary process is such a burden? Answer: Lot's of experimentation: lots of time; big populations; small gestation periods; lot's of competition during that gestation period. Mice have it over people, paws down.

Then how would you evolve a person - in 66 million years, populations in the thousands; gestation period of little more than a decade? You start with something like a mouse that has evolved a good basic nervous system - one that has evolved to be adaptable through evolution. Then you create variations on that theme: language areas that works off existing audio and socialization circuits; analytical areas that work off the language. Stuff like that.

 

[Mr Green T]

The study does lead me to conclude that due, both connectivity and the wiring cost are conserved across mammals, the evolution of these is optimized currently. What improvements do you propose?

...
For example, once I have established a communications protocol - say RS-232 at 1Mbaud - I would have to make matching changes to the transmitter and receiver to improve it - ...

Why not improve the protocol?

 

[.Scott]

The study does lead me to conclude that due, both connectivity and the wiring cost are conserved across mammals, the evolution of these is optimized currently. What improvements do you propose?

Why not improve the protocol?

I was anthropomorphizing evolution. I am not attempting to participate in that design process myself.
That apparent connectivity optimization occurs within the context of a mammal neural "technology".

If we think of this in terms of VLSI technology ("computer chips"), it's as though VLSI with 5nm design rules was developed 66 million years ago and the same technology is still being used today.
But I suggest that it's likely more than that. Even the basic circuitry architecture has likely remained unchanged. So if the PCI bus was in use 66 million years ago, we'd still be using that basic standard now - because allowing only one change at a time, it would take a lot more evolution to change something that has become the foundation of so many other "features". Changing something that basic would be literally maddening.

 

[BillTre]

I looked at the abstract and the
Perhaps I am not understanding the use of "connections" as they are using it and/or how the diffusion MRI relates to their connection (or connectome) concept.

Here is one problem I have. To me it looks like diffusion MRI reveals (non-invasive) what the same thingas a white matter histological stain, but in a nice 3D manner.
Wikipedia:

Diffusion imaging is a MRI method that produces in vivo magnetic resonance images of biological tissues sensitized with the local characteristics of molecular diffusion, generally water (but other moieties can also be investigated using MR spectroscopic approaches).[9] MRI can be made sensitive to the motion of molecules. Regular MRI acquisition utilizes the behavior of protons in water to generate contrast between clinically relevant features of a particular subject. The versatile nature of MRI is due to this capability of producing contrast related to the structure of tissues at the microscopic level.

Another problem is when I think if neural connectivity, I think of synapses between neurons. I can accept fibers that can be traced from one region to another as probably making connections between some cells in the two regions, but that can be selective.
Do these diffusion MRIs (answer probably in the paper which I have not seen) have such good 3D resolution that specific fibers can be continuously traced from one region to the other? I can understand using fiber tracing techniques like tracing tracts by using lots of different colors to distinguish different finbers over long distances. This makes following individual fibers easily. I'd be impressed if they could do this.
They say diffusion MRI:

detects the white matter in the brain, enabled the researchers to reconstruct the neural network: the neurons and their axons (nerve fibers) through which information is transferred, and the synapses (junctions) where they meet.

Because of my background in neural development, I think if connectivity in the manner of making specific connections between specific cells.
however, the authors seem to be thinking about regional interconnections and how easily (how many different step are required) for a region to be connected with all other regions.
This seems like a degree of separations argument to me. How many steps does it take to connect to every other region. Thieir connections seem to be based on anatomy rather electro-physiological data (more directly reflecting brain functioning).

With respect to neural evolution, I would agree it is modular to design, but I would also argue that signicant changes have happened through evolution.
The vertebrate nervous system is influenced by the periphery it is connected to. As new body features evolve, new brain parts evolve to deal with them. Similarly new control and analytical systems are evolved to handle new demands.
These changes however, don't usually replace what came before. The usually just lay a new control system on top of older ones. This can be seen in the descending fiber paths from the brain down the spinal cord, to control motor outputs for different limbs, limb functioning, and behavioral control.

There are some basic repeated parts like: neurons/synapses (and cell physiological properties), local neural tissue structures (small local operating neural regions and processing units among local populations of neurons).
There is also a complex anatomical structure, inherited from a vertebrate ancestor, with all vertebrate nervous systems are built upon.
This provides a common starting point for different vertebrate nervous systems to elaborate their different species specific structures, using their different developmental programs.

The structures of vetebrate nervous systems are therefor both limited in some ways, opened ended in others, while different modules are reused in different developmental situations.