Math4Wisdom

M4W spaces

Investigations

Tools

Topics

Questions

Other Notes

Explore

Biology

Ants

Synthesizing the Born rule with reinforcement learning

Ants in EJ Applewhite

Gradients and transitions in the colony.

Auto & Allogrooming.

Temporal polyethism (Nurse, Midden, Foraging).

Possibility for a unified mind that makes decisions at a slower level (at very least this is “as if” for decisions which influence the colony as a whole)

⁠

https://link.springer.com/article/10.1007/s11721-024-00237-8⁠

A stochastic model of ant trail formation and maintenance in static and dynamic environments

Connecting to multiple pheromone models.

Single? Two with fast/slow decay? Multiple with modulator/tradeoff 3rd pheromone/mind?

Ant Questions

Ant Questions

Thoughts and questions by AK with responses by DAF in the notes!

Notes

Status

One issue is that there are so many species, with about 14,000 classified and perhaps half as many not yet classified. You and your advisor Deborah Gordon studied red harvester ants. But I suspect that focusing on a single species can give an incomplete and misleading view of what is going on in general.

⁠

https://pubmed.ncbi.nlm.nih.gov/25525497/

⁠

“Spatiotemporal resource distribution and foraging strategies of ants (Hymenoptera: Formicidae)” (M. Lanan 2014) — this is an example of review paper across

Deborah has worked with several other ant systems —

Turtle ants in the trees

⁠

in Mexico, and also

California local ant

⁠

ecologies. Further, she has always taken a broad

EcoEvolutionary perspective in reviews

⁠

, including ones we

have done together on ant genetics

⁠

“But I suspect that focusing on a single species can give an incomplete and misleading view of what is going on in general.”

Right. This is why with

broader formal frameworks

⁠

we are able to take species-specific (really experiment-specific unique biological data) and make sense of it within a broader theoretical framework (EcoEvoDevo).

Species have adapted to different situations but I suspect that in many cases it is simply that certain genes have been turned off. The full set of genes would tell the full story but in many species we see only part of the genes in action, all the more so if we are focusing on only what we can see, the ants outside of the nest, and ignoring what doesn't register with us, namely, the ants' sense of smell. So one question is whether ants across species have basically the same DNA and genes. Are ant species losing and gaining genes? Or are they simply turning off certain genes?

Absolutely. There are always complex genomic and epigenomic variants, within and across species (e.g. variation in coding/non-coding sequences, and structural variants, of Odorant receptors and other related proteins). This question of the variability and stability of different subtypes of odorant receptors, has always been major area of genomics research in ants.

⁠

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6211649/

⁠

The genomic architecture and molecular evolution of ant odorant receptors (2018)

Also it is far more complex than genes being turned on or off. Every cell is regulating a discrete/finite amount of RNA products, including non-coding, small RNA, etc. & there is RNA modification. This is why/how Systems Biology approaches, or multi-omics approaches, try to approach the interactions occurring within and across datasets.

We see ants acting by themselves. My impression is that they have one-track minds. They have a tendency to stick to what they are doing. An ant will forage, bring a seed to the nest, and then go out back to the very same spot and look for more seeds. It's not thinking or asking where it should go. It just repeats what it did before. It has the memory of that journey and it makes use of that memory. Similarly, I saw a video where ants were made to fight cockroaches. And I noticed that in the midst of this chaos, one ant was chasing one cockroach, with great focus.

“We see ants acting by themselves.” — isn’t much of the question, what a “self” is? E.g. Nestmate, Colony, etc. — why not Eukaryotic cell? What about a Diamond of all of them?

⁠

https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2019.02688/full

⁠

The Computational Boundary of a “Self”: Developmental Bioelectricity Drives Multicellularity and Scale-Free Cognition, Michael Levin

⁠

“It's not thinking or asking where it should go. It just repeats what it did before.”

This is not true that they simply repeat what they did before.

First, that wouldn’t explain the emergence of new behaviors within each nestmate life (e.g. as they proceed through the process of temporal polyethism, transitioning task from Nurse to Midden/Architecture to Foraging.

Second, they are continually responding to the changing environment (hence Active InferAnts and the OODA/action-perception cycle). They are not simply applying routines.

At a deeper level — I see ambiguity in this question, as to whether persistence and/or stability in task performance, is associated with (higher order) cognition.

e.g. there are rapidly-shifting things that think or don’t..... and there are things with persistence and high or

This recent discussion

https://www.youtube.com/watch?v=Xmu2QAKVuVU

⁠

A discussion between John Vervaeke, Gregg Henriques, Justin McSweeny, and Mike Levin.

Has a lot of relevant ideas here.

An ant seems to have various kinds of memory. In the ant-on-ant interactions, the memory is short, on the order of 10 seconds, but with regard to its own tasks it is much longer. Or are they simply smelling their own distinct personal markings on trails? (Which could be considered a sort of written memory.)

Individual nestmates do not have like a personal odor that is left on the trail. Nestmates within a colony, essentially have the same repertoire of pheromones that they release. So there is a colony/species signature, not so much a nestmate-scale trail scent.

We did some dynamical equation modeling to explore the rich behaviors possible even with a single pheromone

https://www.sciencedirect.com/science/article/pii/S1389041723000207

⁠

— and in practice there are quite complex blends.

Either way, for sure the markings in the nest & outside, are a form of Stigmergy and extended memory:

http://pespmc1.vub.ac.be/Papers/Stigmergy-Springer.pdf

⁠

Have there been experiments done on the memory capabilities of single ants? Do they extend beyond an hour or so?

Yes this has been studied in many species/papers (of course still only a tiny fraction of possible settings, and usually with very incomplete methods (e.g. showing behavioral learning but not applying further neuroscientific methods to understand mechanistic causation)

There are memories that are certainly longer than days, of kinds ranging from chemosensory, to visual, to behavioral/associative, certainly longer than days

⁠

https://royalsocietypublishing.org/doi/10.1098/rsos.190778

⁠

Ants learn fast and do not forget: associative olfactory learning, memory and extinction in Formica fusca

⁠

https://www.pnas.org/doi/full/10.1073/pnas.1809433115

⁠

Desert ants possess distinct memories for food and nest odors

⁠

https://royalsocietypublishing.org/doi/10.1098/rsbl.2007.0224

⁠

Long-term memory of individual identity in ant queens

Also there is the stigmergic / architectural / niche memory, which is a true form of external/extended/embedded memory.

So suffice to say they have ample onboard and external memory across multiple scales.

Besides ant-on-ant interactions, there is the notion of ants as a liquid, like molasses, whereby ants can form living rafts and living bridges and living balls. Part of this can be explained by ant-on-ant interactions but I think also there must be some sense of the whole activity, and the state of the whole. These are cases where ants are peers and there is no queen.

Why does there need to

I suspect that the queen and her consort are much more important than Deborah Gordon makes them out to be.

I don’t believe Deborah thinks that the Queen is unimportant.

In Harvester ants, at the site we studied, colonies have a single long-lived queen. So obviously she is the mother of all other nestmates, and thus vital for intra- and inter-generational persistence.

However in terms of the minute-by-minute regulation of foraging activity — the queen does not play a direct role there.

A broader multiscale analysis of foraging behavior would reveal many interactions and feedbacks between e.g. Queen hormonal state,

Colony demographics

⁠

, seasonality, ecosystem context, etc.

I am curious, what happens when an ant colony loses its queen? Does it behave differently?

Great question.

This depends greatly on the species.

For a highly polygynous species like Argentine ants (e.g. millions of workers and hundreds-thousands of queens), probably the death of a single reproductive female has little impact. There are always new reproductive females arising.

For red harvester ants, where there is a single queen, when she dies (as I have seen in Lab colonies, but only inferred from the field), then the workers continue to forage / maintain the nest / raise the remaining brood. Sometimes, the workers can lay unfertilized eggs if they have any retained Ovarian activity (as unfertilized haploid eggs in Ants develop into males). However either way, the “stem cell” has died and the colony is essentially sterile and reproductively over, at that point.

There are some ants, where the worker nestmates retain some reproductive plasticity. So when the queen dies, the workers have a process of determining who rises to queen status:

https://en.wikipedia.org/wiki/Gamergate_(ant)

⁠

I have read or heard that there is communication going on through ants spreading their smells from one to another. In particular, smells are coming from the queen out by way of the nurses, and vice versa, to the queen. This suggests that the queen has a state of smell which may change and which can control the behavior of the nest. Furthermore, there are groups of ants inside the nest which do nothing. I suspect they are used as memory cells for states of smell, especially of the queen.

Yes. Through auto- & allo-grooming (e.g. preening themselves, and each other), there is a lot of chemical sharing among nestmates.

Largely the cuticular hydrocarbons are overlapping in composition between Queen and Nestmates (e.g. playing more basic roles in species identification, water loss protection, waxy coating, etc).

Also there are specific chemicals and blends/ratios that are reflective of the Queen:

https://www.sciencedirect.com/topics/agricultural-and-biological-sciences/queen-pheromones#:

⁠

so it is big area of research to understanding

“there are groups of ants inside the nest which do nothing” — at any given moment, there is a fraction of nestmates not doing some other active task. Those ants can play a variety of functions — chemical buffers/memories as you suggest, as well as providing extra capacity to the colony (e.g. able to mobilize/deploy if needed for a challenging task/battle or due to demographic loss).

In the film, "Empire of the Ants", they explained how for the honeypot ants, the colony is started by one queen along with other queens that join her. They all lay eggs and the eggs becomes nurses for the queens. I was struck that ultimately some of the nurses kill some of the queens. I suspect that the queens that were killed lost out in the smell contest, by which I mean, they were ultimately not able to produce the smell that is needed for dominance, at which point the nurses turned on them.

Yes. The exact details there are species-specific.

Broadly though, there are always dynamic-oppositional processes in Biology. E.g. a Fat cell is both making fat and breaking it down (both enzyme chains are present, with an overall Anabolic balance when nutrients are high and Catabolic balance when nutrients are low).

So the colony is generally “over-producing” eggs, many of which are eaten by Nurses (this is called a “trophic egg” because it is used for food). There could be various reasons and patterns for why this is.

Probably they made it very narrative and dramatic in Empire of the Ants.

My impression from that was that the ant colony is ruled by an insular royal court made up of queens and nurses, much like an imperial Chinese court with an emperor but also wives, concubines, and notably, eunuchs. These insular rules live in a semiotic world of smells that they receive from the outside world and produce, especially the queen. It is like Jere's AutoGnome, where there is a chain of modules, each more removed from the external reality. This distance from reality is very important for purity of thought and the associated autonomy of will. That is why ants that go outside of the nest are not allowed back into the depths of the nest. The ants that have gone out into the world have had experiences that would disrupt the purity of semiotic thinking that goes on in the depths. So they are kept out.

Very interesting about the inner semiosis.

The entire consort of workers are female/diploid. There are species where the worker nestmates have no ovaries at all (e.g. “neuter”) and ones that have stunted-like development of female reproductive anatomy.

Many of the Morphological and Genomic investigations into ants, compare (Male/Female) and then (Queen/Nestmate) within females, and then (Nursing/Foraging) within nestmates.

The question then is what could the state of smell of a queen control? One possibility is that when ants close up the openings to the colony, in a state of high humidity, with a chance of flash flood, then the question is, who decides when to open the entrances again? Is it simply a matter of humidity going down? But who is the one to make the decision and open the entrance? Or is there an overall state of mind in the colony?

In red harvester ants, the queen is 6 feet down. So probably over the hours-days timescale, she is deeply in feedback with hormonal, nutritional, and other decentralized physiological processes

https://pubmed.ncbi.nlm.nih.gov/32305342/

⁠

About the Rainstorm “to plug or not to plug” decision, it is totally not known how that process relates to on-nestmate cognition (e.g. humidity and temperature detection), among-nestmate cognition (e.g. interactions in the nest lobby), and possibly deeper inputs from the nest.

I agree that half of the story can be explained by ant-on-ant interactions. But I think, at least for consciousness, there needs to be another half of the story. And given the importance of the queen to the colony, and likewise the nurses of the queen and of the brood, I suspect that they play a key role in certain actions or reactions of the ant colony. Certainly, they send out the ants that fly to create new colonies.

Yes.

One metaphor is that “

physiology loads the bow

⁠

” and interactions are the proximate cause that “pulls the trigger”.

This is still a purely mechanistic account of the distal and proximal causation:

⁠

https://en.wikipedia.org/wiki/Tinbergen%27s_four_questions

⁠

https://link.springer.com/article/10.1007/s00265-018-2606-3

⁠

Synthesis of Tinbergen’s four questions and the future of sociogenomics (Kapheim 2018)

I suspect that if we look at the wide variety of species, then we can figure out what is possible of ants in general, and that will inform how the ant colony functions as a mind. Then that model can serve as a guide to understanding how, in particular species, that general model gets implemented. But if we just focus on particular species, then given the paucity of knowledge, we may likely miss the relevant concepts.

Yes!

This is basically one of my core long-term research agenda — looking

⁠

https://docs.google.com/presentation/d/1iPIBjEAghkRjnD6qZw0y1ssRcvwo382uYt00gJPy9Ww/edit?usp=sharing

⁠

— This was a meme slide I made with a friend. Now I can see that it is like a Stigmergic 3 cycle.

⁠

Here was my postdoc plan from 2020. I am still in this game long-term, to look across facets of the biological system (genomes to ecologies) and unify the analysis within the Variational / Renormalization Group / Multiscale Active Inference scaffold

https://www.mdpi.com/1099-4300/25/7/964

⁠

https://zenodo.org/records/10400745

⁠

A snapshot and pipeline for tissue-specific gene expression meta-analysis in honey bees (Apis mellifera)

There are no rows in this table

⁠

AK has read half of Deborah Gordon's book on ant interactions, and have also watched various videos, including "Empire of the Desert Ants" (2011) about the honeypot ant in Arizona. The movie was made with the help of Bert Hölldobler of Arizona State University, a collaborator of E.O.Wilson.

It seems that there is much that is not known about ants and ant colonies. Even so, I am starting to formulate some ideas about the consciousness of an ant colony.

⁠

Loading…⁠

⁠

KU: Ants in Pop Culture has been a focus. Ant Colonies as metaphors and models.

Interconnecting Codas as tunnels⁠

How useful is anthropomorphism? Are warring factions of ants manifesting Consciousness in the sense of a 3rd mind, per the WW cogsys?

Perhaps humans in equilibrium tend towards Conscious + Unconscious only? Minds 1 & 2? The everyday mundane is about reinforcing current habits, teaching maths with just a calculator and/or with no questioning of authority, staying in well-worn ruts. Why rock the boat?

When there’s no earnest prayer to overcome, only past momentum and a resolute shoulder-to-the-wheel commitment to stay the course, to double down, is 3rd mind present?

Escape from Freedom: When the steering wheel doesn’t steer; When voting makes no difference.

⁠

https://royalsocietypublishing.org/doi/10.1098/rsbl.2015.0695⁠

Nest architecture shapes the collective behaviour of harvester ants

⁠

3rd mind for

Ants⁠

1st — Streams of nestmates

2nd — Colony and streams

3rd — Internest

⁠

https://www.aliusresearch.org/uploads/9/1/6/0/91600416/friston_-_of_woodlice_and_men.pdf⁠

⁠

friston_-_of_woodlice_and_men.pdf

756.2 kB

⁠

Friston’s response to question about selfness/consciousness in

Ants⁠

Consciousness is a process. A process of inference.

Variational Free Energy = Bayesian Evidence Maximization = Surprise Minimization = ...

See

Active Inference Textbook⁠

equations 2.5 & 2.6 for more information.

Minimal selfhood:

Counterfactual breadth

Temporal depth

Generative models with predictions about consequences of action.

Generative models distinguishing Self and Other.

Generative models distinguishing Others of different kinds/grades of similarities and differences.

⁠

Biology⁠

Mechanistic modeling of alarm signaling in seed-harvester ants

https://www.aimspress.com/article/doi/10.3934/mbe.2024244⁠

(2024)

⁠

http://www.bruno-latour.fr/sites/default/files/166-BODY-CINI-pdf.pdf⁠

When I had the chance to meet Deborah M. Gordon at Stanford, I realized several things at once:

first, that ANT was aptly named after all!

Two, that the long kidnapping of ants to play a role in the fight between organicist versus market-based models of society could finally come to an end.

And third, that alternative ligatures could be invented empirically for composing the anthill and thereby escape the appeal to any superorganism.

⁠

@Daniel Ari Friedman⁠

in my yard, I have this mound, which keeps growing. Is it an ant colony? I have never seen any ants around it. But the dirt seems dug up. Andrius.

The movement of large pieces of grass seems too big for Ants. This looks more like a bird, or a small ground mammal. That being said, there may be some European ants with incredible construction capacities.

⁠

@Daniel Ari Friedman⁠

Thank you! The grass and plants were always there. The mound just grows taller, which means that everything is being lifted up. But by what? Andrius.

⁠

@Andrius Kulikauskas⁠

we will have to wait & see!

⁠

@Daniel Ari Friedman⁠

It has been about three or four years that it has been growing taller.

Loading…⁠

⁠

Here is a close up shot which shows that its a pile of fine dirt, not soil, underneath.

⁠

Loading…⁠

⁠

Andrius: In other ant news, today 2024.04.17 I noticed this ant colony in my rose garden. It is about 35-40 cm across. In the afternoon, ants were pouring out and I think working on the nest. Maybe they had been wintering underground and then are building this nest because it is spring time. I had never seen it before. There are similar looking ant nests in the forests that can be 150 cm across and 80 cm high. In the evening, when I made this photo, there were fewer ants around. I haven’t mown my lawn for twelve years, since I got my house, so my yard is very natural.

⁠

Loading…⁠

⁠

KU: picture of ant colony not far from Berlin shared by

Struppi⁠

on Telegram:

⁠

KU: I realize the current Coda Topic is Topics | Biology | Ants, whereas here I’m XREFing on: insect colonies : discovered @ home : amateur photography : Google Earth (not Math4Wisdom Coda topics per se, just out there amidst the chatter boxen).

KU: Tunnel Bee Habitat (known as Miner Bees in the local vernacular):

⁠