The Thin Line Between Parasites and Mutualists

If there is one thing that nature laughs at, it is any foolish human attempt to paint a situation as black and white, and the spectrum from parasitism to mutualism is no different.

Only fairly recently have scientists being grabbing hold of this idea that a parasite can become a mutualist and vice versa [1].

There are lots of possible examples: Humans adding nitrogen to soil could cause soil microbes that usually help a plant (by fixing nitrogen) to become parasitic since they still want their sugar even if the plant doesn’t need their help anymore.


coli is both a dreaded stomach flu and a normal occupant of the human body depending on location.

There are insects that steal nectar from the base of a plant, avoiding getting any pollen on them; could they have evolved from pollinators originally?While there seem to be examples that show that a strict mutualist or parasite could evolve into the other, there is no way of knowing for sure what happened or what will happen (unless you have a time machine hidden away somewhere).

Humans, however, are very good at spotting patterns, even when there isn’t actually a pattern there, so I have never been content to take such explanations for granted.

Since I personally don’t have a time machine stashed anywhere, I instead have created a simulation that can model mutualism, parasitism, and anything in between [2].

Symbulation, as it’s called, works using the concept of agent-based modeling.

Basically, I wrote code about how a generic host behaves and code about how a generic symbiont behaves.

(Side rant: symbiont is the proper term for a thing that lives with another thing and could be parasitic or mutualistic or something else.

Mutualist is the proper term for a thing that lives with another thing and they mutually benefit each other.

) Both the host and symbiont need to collect a certain number of resources to be able to reproduce and their offspring inherit their behavior with a chance of variation.

This very simple setup is all that is necessary for evolution to occur.

The behavior that I am interested in is how they behave towards each other.

The generic host and generic symbiont can be anywhere from fully cooperative (mutualists) and share all their resources to their mutual benefit all the way to fully antagonistic (host and parasite) where the host spends resources to defend themselves and the parasite takes as much as it can anyway.

The key is that it is a full spectrum and that I can create large populations of these hosts and symbionts that act independently.

In this way, I am able to model at the highest level how a population of host organisms and symbionts might evolve under different circumstances.

Now there are a lot of questions that could be asked with this software (and it’s open-source, so anyone is welcome to contribute), but science is all about focusing on one specific question at a time.

First, I asked if such a simple representation of a host and symbiont system was even capable of doing what most of us would expect of it.

The first thing that would be expected has to do with vertical transmission rate.

At a high-level, vertical transmission rate is the rate at which the symbiont is able to tag along for the ride when the host reproduces.

A high vertical transmission rate seems to be associated with mutualism, and that makes sense: if the symbiont is going to be reproduced when the host reproduces, then the symbiont should help the host reproduce.

Your mitochondria are a good example of a symbiont that relies 100% on vertical transmission now; the only way they are getting to the next generation is if the human cell they are in reproduces, making them very reliable mutualists.

A low vertical transmission rate seems to then naturally associate with parasitism and another form of transmission: horizontal transmission.

The cold virus that caused you to sneeze in the elevator isn’t likely to last until you next reproduce and so instead it needs to spread on its own and will take all the resources that it can from you to do just that.

As it turns out, I can now show that there is a causal link between the rate of vertical transmission and how mutualistic a symbiont population evolves to be:This is one of those graphs that at first seem really obvious and therefore boring.

However, I can’t stress enough that this is a hypothesis that has been really difficult to test without a simulation of this type and science is rife with hypotheses that seemed obviously right but turned out to be completely wrong.

However, this reassuring result laid the foundation for delving into some of those ‘seems obvious but might be wrong’ questions.

The paper explores a few others, including what happens when there are two tasks that need doing and the host can only do one of them.

But the one that turned out to be wrong is the most interesting, so that’s what I’ll focus on next.

How spatial structure changes the evolution of mutualismBear with me as we explore a brief tangent into the topic of cooperation and spatial structure.

It is well-known in the field of cooperation within a single species that spatial structure aids in cooperative strategies succeeding [5].

To understand the intuition behind that idea, think about a bacterium.

If that bacterium is prone to exerting effort to free resources that any surrounding cells can then use, it is very vulnerable to spatial structure.

If this is the open ocean where surrounding cells are unlikely to share its genes, then it is sharing resources and may get nothing back and not be able to reproduce.

However, if it is on a sticky surface and therefore all the cells surrounding it are nearly exact clones, any resources that it produces will go to those clones.

Since those clones share most of our friendly bacterium’s genes, they will also be exerting effort to free resources that our bacterium will use.

If a cooperator is surrounded by other cooperators, it is more likely to pass on its genes that if it is not surrounded by other cooperators.

Spatial structure that causes offspring to remain close to their parents promotes this cooperator-friendly environment.

Therefore, it is well-known that spatial structure makes it much more likely that cooperation will evolve as a successful strategy.

It turns out that that is only true when you are talking about a single species.

When spatial structure is added to host/symbiont systems, under certain vertical transmission rates, the symbionts become more parasitic!This result was so unexpected that when I first got the data, I and my colleagues were sure there was a bug somewhere.

Only after a lot of testing and the publication from another lab of a mathematical model predicting this result [3] was I satisfied that this is in fact correct.

This is a fairly jarring result, since decades of research into cooperation in a single species has pointed towards spatial structure enabling the evolution of cooperation.

The problem with that is that no species in the world lives in complete isolation.

There are bacteria, viruses, and fungi everywhere and they have a lot more influence on the fitness of us larger organisms than we like to think.

Through an agent-based simulation, I have shown that our understanding of cooperation gets a lot more complicated as soon as we factor in just one other species.

Imagine what we’ll realize when we factor in the hundreds that interact in all sorts of imagined and unimagined ways!References:[1] E.


Jones et al.

, Cheaters must prosper: Reconciling theoretical and empirical perspectives on cheating in mutualism.

(2015), Ecology Letters[2] A.


Vostinar and C.

Ofria, Spatial Structure Can Decrease Symbiotic Cooperation.

(2019), Artificial Life[3] E.

Akçay, Population structure reduces the benefits from partner choice in mutualism.

(2016), bioRxiv.. More details

Leave a Reply