CROWD

Page 1

CROWD Vol.1 from Vertex


Of crowds, behavior and how they form

7 The ant behavior program An incomplete summary of Dr. Gordon’s work on the collective behavior of ants

12

4 Article amore Science writing from around the web

8 Collective Behavior in Biology (India) Exploring a lab web-page and looking for interesting stuff

index

Matters of Order in Seemingly Disordered States

2


Hey, thanks for hopping by. Like all my personal projects, this one kick-started after a motivation-burst post reading Simone’s zine, ‘Antigone’. Vertex, my blog-cum-newsletter too was a post-motivation product, largely for keeping track of papers I’d been reading. I see this zine as being more long-form, exploring topics I know virtually nothing about and building on them (hopefully). Along with exploring design and creativity. Of course all of these are ideas. We’ll see how it turns out ;) This issue is centered around ‘CROWDS’ in biology.

vertex.substack.com

3


MATTERS OF ORDER IN SEEMINGLY DISORDERED STATES If I were to describe the city I live in one word, I’d say ‘Crowd’. Crowds are not only a feature, but a way of life here. The incessant energy often spills over to disorder. Human crowds are a result of a rising population, working hand in hand with urbanization, opportunity, and ultimately survival. Interestingly, so are animal crowds. I’d stretch and even call the a crowd an adaptation. In essence, crowds of a single species not only provide safety in numbers, but also make food and mates more easily available. Of course, through social structures. Crowds also evolved much before humans did. A bacterial colony, a school of fish, YOU (the organism) are a living collective--a crowd. And let’s not stop there. Inanimate substances form aggregates too. So what makes a biological crowd special? Illustration: A school of fish

4


Strikingly, there’s a lot common between biological and abiotic crowds. Both interact with their surroundings (the abiotic component I’m considering here is protein). Since the physical and biological worlds coincide where crowds are concerned, as with many things, scientists study them. With living systems, there always are a lot more factors to consider. The functioning (or goal) of a crowd, is perhaps most important. From philosophers to scientists, the attribution of a larger goal to a biological crowd is a common trend. All of our crowds form as a result of verbal communication, visual-auditory interest, social media and randomness. Not all crowds function the same way, and scientists believe they can be explained by certain non-mandatory but important principles.

Generating order: the unnecessary bare-bones

important

but

sometimes

A stimulus, without doubt, is the most important behaviour generator. Without it, there’s no response, and consequently no behaviour. However, it is hard for a single stimulus to sustain long-term behaviour that is typical of biological crowds. Neuroscience has exceptions, however sustaining behaviour from what I know is an elaborate process. Often, positive feedback loops are more effective in eliciting longer-term behaviour.

A termite is 350 times tinier than a human of average height. Its brain is decentralized into ganglia located throughout the body, and it meets more friends than I do on an average. Termites are also fantastic architects. Unlike the endeavours of their human counterparts, termites build mud-mounds (300 times larger sometimes!) intuitively. Well, the rules come in pheromone laden spit. You see, the walls of the mud-mounds are held together with spit. Worker ‘mites smell the pheromones and dump the mud they find at the site having the highest pheromone concentration. Positive feedback may not always be the most hygienic by human standards, but it gets the job done. Where positive feedback works by building the tempo and getting more work done, negative feedback gives similar results using the opposite route. Here, aversion to a particular stimulus drives action. The reducing space between two fish in a school, for instance, could be a negative feedback loop for maintaining fish school-structure. Given these factors, one cannot forget the existence random events in nature. Consider the fish-school example. Randomness could find its way at many levels here. If school formation begins with a single fish interacting with another, an event preventing interaction could be random. So could fish being passive to a stimulus and so on. This issue will look at some research from the field.

References: 1. Self-Organization in Relation to Several Similar Concepts: Are the Boundaries to Self Organization Indistinct? 2. The principles of collective animal behavior.

5


D

AN

ATA

EMEANOUR

VERTEX, Volume 1: CROWD

6


article amore

How Swarming Insects Act Like Fluids By studying a swarm of flying midges as though it were a fluid, physicists have learned how collective behaviors might stabilize a group against environmental disruptions. From Quanta magazine. Read here.

Animals Bow to Their Mechanical Overlords Robots are infiltrating insect, fish, and bird communities—and seizing control. From Nautilus magazine. Read here.

How Loners Are an Evolutionary Insurance Policy Organisms that don’t follow the herd may not be stragglers, but nature’s way of hedging its bets. From the Atlantic. Read here.

7


The ant behavior program VERTEX, Volume 1: CROWD

8


T

wo rod like antennae tipped with bulbous sensillae pop like prongs on the ant’s head. These are the instruments an ant uses to navigate the world.

studying them is difficult, given the few tools available. Besides there are over 12,000 ant species worldwide. Many, insular, adapted to

Now ants don’t see very well. They have lenses split into a million rooms called ommatidia. Millions of tiny split lenses, it turns out, don’t focus an image very well. So ants default to their trusty antennae to sniff their world instead.

WHAT IT TAKES DESERT FOOD-TRAIL

Professor Deborah Gordon has been studying ant behavior for over two decades. Her most recent project includes determining the ant behavior code among others. To work out the possible ant-code, her lab studies desert ants near the Mexican border.

Now ants aren’t particularly a model system. Which is to say that not too much is known about their lives, and

FOR

A

Desert ants can cover an area about 2000 times their length. For comparison, that is equivalent to a human travelling slightly short of the east-west expanse of Indian territory! A chosen few scour the expanse for food.

Dr. Gordon hadn’t stepped into the field with the intention of studying ant-behavior necessarily. The techniques we use today weren’t available in those times, she says in one interview. Instead she sought for principles that pushed crowds of cells to organize during development. The closest visual-approximation were social animals—ants. How are desert harvester ants similar to cells during development, one may ask. For one, there is absolutely no hierarchy. The ant-queen exists for the sole purpose of reproduction. The burden of labour however, falls upon the sterile, female worker ants that go about their duties in the absence of a rulebook. A desert harvester ant colony is a kingdom run by the masses, without central control.

the variation in ant colonies over time. How new colonies form and how they grow. Such extensive work has lead the group to understand a lot about how hierarchy in ants is arranged for instance, and how roles in ants are allocated. The only interactions the group has discovered so far, are olfactory ones between the worker-harvesters.

environments as diverse as rainforests to deserts. And given their range, also showing distinct behaviour sometimes. Nevertheless, Dr. Gordon and her team have spent years studying

The search for fat-heavy seeds is an oasis in a desert search. The fat is metabolised into water, and water is scarce, ill distributed, and a matter of survival. The arid squeezes water out through dehydration every second. Foraging food is a balance between sustenance and death. Meaning it’s irrational to drain water from a thousand gatherers than it is to send ten to get the job done. Worst comes to worst, the ten may die of dehydration, but the colony is saved. If the gatherers return however-a sign of temporary abundance, a touch from the incoming comrades is all it takes for the next slew to take over.

9


Dr. Gordon believes such collective behaviour may have an underlying algorithm, a code. One that is related closely to the environment the ants live in.

The team in collaboration with Dr. Leonard and Dr. Pagliara had sights for a new model.

The environment dictates the availability of resources, the amount of energy taken in versus given out by means of food, and the number of times an incoming ant touches the one in the entrance chamber. The only player that could be subjected to manipulation in the wild was the contact from incoming ants.

GIST: Ant-interactions and the nest ants’ responses form an open loop. This is to say that the responses of the nest-ants don’t influence the number of interactions that take place. The much larger closed loop however, is the incoming foragers interact, the nestants respond and subsequently become outgoing foragers and on the occasion that they find food, return to the nest as incoming foragers again. Food availability feeds into this system, allowing greater foraging ability, greater chances of return, and consequently greater numbers of interactions.

The team resorted to an ingenious methodology. By trapping unsuspecting incoming ants into boxes, the team managed to reduce the rate of contact from incoming ants. As if in response, fewer ants from the entrance chamber left it. When the incoming harvesters were set free in contrast, the entrance chamber ants seemed to emerge from the nest after a short lag. The process is stochastic or random, meaning the number of interactions, the number of incoming ants and the number of outgoing ants are not the same every single time. Of course all this is attributed to the randomness in environment itself. However, the trend is the same. Number of outgoing ants always increase after interactions with incoming ants. It struck Dr. Prabhakar, who worked in collaboration, that the internet—yes, the world wide web that we use today works on a similar algorithm. In the TCP-IP protocol, new packets of information are sent only when the previous packet is received. The data from the ants was fitted into this model, and indeed the model was a close fit. This model considers only the inflow and outflow of information. If I may, it may be a little reductionist.

The presence of holes chamber formed as a experiment interesting. interactions and its behavior be possible?

in the nesting result of the Could greater influence on

To understand the response of the nest-ants to the incoming foragers, the team collected interaction data. Now, the nest-ants, are present in the entrance chamber located slightly below ground-level. The upper mud layer was dug up and covered with a transparent sheet instead. Interactions were measured and colour-coded based on the type of ants involved.

ANTS AS NEURONS When it comes to neurons, excitability is dependent on the interactions with neighbouring neurons. A neuron spikes or sends a stimulus only when it crosses a 10


“ Food availability feeds into this system, allowing greater foraging ability, greater chances of return, and consequently greater interactions.

certain threshold. Continuous stimulation by adjacent neurons slowly raises the membrane potential until its pushed over this threshold. And then it spikes. The team believes the nest ants show similar behavior. Stimulation from incoming ants should be consistent and continuous to push the ant from its resting state. Except an ant is not a neuron after all. And so, instead of going back to its ‘resting state’ in the entrance chamber, in the absence of an ‘excitatory touch’, it does deeper down the nest.

REFERENCES: •

Deborah’s talk: The Ecology of collective behavior at BPPB

In the seminar, Deborah presents a third model that factors in humidity. However, its still in the thinking stages. She also gives a glimpse of another ant population she’s been studying that show starkly different behavior, given the circumstances. •

Talks at iBiology

The two talks here are relatively older, but are very telling about the research, especially in the context discussed here. •

Interview with Quanta Magazine

This interview takes us through the course of Dr. Gordon’s journey, I believe. Its during the same time as the iBiology videos roughly.

Summary at ‘Collective behavior: How animals work together’ with Knowable magazine

This is a more recent offering showcasing other types of collective behavior.

“Hey mate! Stop being a dog”

11


12 12


True to its name, the lab uses theoretical approaches, and it does so at various macroscopic levels. Individual, flock and ecosystem levels. A recent study dealing with schooling of karimeen, a fish found along the western coast, fits snugly into the collective behavior region described previously. Two words, however, caught my attention—’Game Theory’.

GAME THEORY

(a result of searching for papers and chickening out from actually posting this on Twitter for alternate view-points) Game theory was developed originally to understand economics and solve allied problems, but I’ll take it to be a series of mind-bending thought experiments. Originally, game theory considered two party games. Two parties with different strategies compete to win a given round. Given multiple rounds, outcomes may or may not change. From the outside, game theory seems to be built on probability, internal abilities or strategies, sets of parameters that may or may not change and a large amount of mental gymnastics. Probability, internal abilities, sets of parameters that may vary, and interactions occur at another place-biology. Population dynamics and species-species interactions that ultimately shape evolution are ultimately games over generations and timescales. Internal abilities and strategy for instance, are analogous to alleles. The population of a

species, being a collection of individuals with different alleles and thereby, different characteristics (strategies), cause the game to exist at multiple levels. Intra-species survival of the fittest optimises for the strategies that lead to the greatest reproductive success. Within a population, reproductive success corresponds to transmission of the genes across generations, ensuring the young are healthy before they face the world, among other things. Social interactions play a large part in it. So, it’s safe to say there are various games at different levels and winning at the larger gameevolution, involves winning a lot of tinier games. I think game theory delves into a tiny bit of cognition and how we perceive punishment and reward. I’d after all like to think of myself as a thinking-feeling person rather than a body programmed to evolutionary success, no? As for me, there’s no way my choices are rational, let alone optimised for success. Which is to say, I like to believe I have free-will. That must mean, all my interactions, are in whatever proportion influenced by my perception of situations. I found a paper that, I think, wanted to say free-will in essence is a magnification problem. That if you zoom in, you could find order, or principles that govern it. Or maybe a matter of how you see it.

In philosophy, there exists a term called reductionism. Which means reducing a phenomenon to its absolute bare-bones. In many cases it’s an oversimplification. Hard sciences like physics and chemistry operate at this level. Biology

TEELAB

T

he first name that popped post a quick Google search—’collective behavior research in India’ was the Teelab website. It stands for Theoretical Ecology and Evolution Laboratory.

https://teelabiisc.wordpress.com/

13


doesn’t. Due to the interaction of various scales in biology, everything is more complicated, than it looks. Reductionism doesn’t fit the bill when understanding free-will and where it works. Or maybe looking for free-will is like trying to find answers to the wrong question. In any case, game theory gives determined results for a given problem. So, you could write the rule-book for a computer to do the mind-bending part. You could change parameters. Look for similarities, possibilities. And then map it to interactions you see in the wild, perhaps. At least, that’s what I imagine game theory in biology to look like.

REFERENCES 1.

Evolutionary game theory primer: Find it here.

2.

Game Theory, Evolutionary Stable Strategies and the Evolution of Biological Interactions: Find it here.

3.

RANDOMNESS, GAME THEORY AND FREE WILL : Find it here.

More labs? Honestly, Teelab is the only lab that said outright they did collective behaviour research. At least, it was the only one that popped instantly on Google. But, like any kind of research, I’m not sure if collective behaviour can be a different field altogether. Mostly, because there different levels.

N D

Biofilm research, self-organising systems, cooperation between populations, cancer spread research, developmental patterning, all could exist as tiny islands. Each of these are, at heart, interactions between groups of ‘single-units’.

14


Curated and edited by: Sanjeevni Sign up for the newsletter: vertex.substack.com 15


Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.