How quickly can populations evolve to combat disease spread?

A mathematical model sheds some light...
29 September 2023

Interview with 

Pratik Gupte, London School of Hygiene and Tropical Medicine


A woman wearing a facemask.


For more than two years during the Covid pandemic, the majority of us had to fight our natural instincts to socialise and instead stay away from those we love and like to mingle with to break the chain of disease transmission. Social behaviours evolve and become entrenched in populations because they confer great benefits: support, defence, food, growth. But if a disease emerges in a group, how quickly can evolved pro-social traits be shed to achieve nature's own equivalent of social distancing? Speaking with Chris Smith, Pratik Gupte is at the London School of Hygiene and Tropical Medicine…

Pratik - What happens to a population of individuals that have evolved rules for how to be social in the absence of a pathogen when an infectious disease then enters their population; how does that change their behaviour, especially in terms of being social when a disease first enters a population.

Chris - I suppose it's a balancing act, isn't it? Where you've got the benefits of being social, where there's learning from other individuals, there's safety in numbers, there's exploiting food, versus if an infection gets into that group, there's a price to pay and it's the balancing act of those two things?

Pratik - Yeah, absolutely. So what we've done is to say, well, each one of you can choose how you balance the costs and benefits of being social. And so what we found there is that you get a split. Some individuals will say, I want to be completely safe and I'll avoid everybody else. So that's very much like what you would think of as shielding during covid. But then there's other individuals that say there are serious benefits to being social, such as finding food. I will follow some of the other individuals that I can see, especially if they look like they've been successful in finding food themselves and risk it. I will take the costs. And the only reason that these two can exist sort of side by side is that the more social strategy pays a higher cost on average of infection, while the sort of safe strategy doesn't pay that cost. So on the whole they sort of balance out.

Chris - Before we talk more about the results, take us through how it actually works. How did you code this up? What does it do to model these scenarios?

Pratik - It's a computer simulation where every individual is just a little bit of computer code. Each individual is essentially just made up of a set of preferences. What's your preference for being near other individuals that have food? What's your preference for being near other individuals that don't have food and what's your preference for being near food itself? And this allows us to run lots of simulations sort of over and over again and then compare what happens in every instance and see whether there's any general patterns to be found.

Chris - Do you pre-specify those choices that each individual in your group that you're studying makes, or are those fluid, can they change?

Pratik - Well, we do pre-specify them in the sense that they can't change over an individual's lifetime. But because we take this sort of evolutionary approach, all the possible choices that exist in the population - and we allow every sort of combination that could exist to exist.

Chris - So that's like your gene pool effectively, you've got a range of possible behaviours in the population from which the selection may occur?

Pratik - Yeah, absolutely. So what we do then in the sort of evolutionary simulation, here, is to say, well, we've given the population all the possible options they could have. So we have initial phase of selection where that's whittled down to the preferences or let's say the genes, if you will, that are most suitable for moving and socialising when there's no infectious disease. And then once we introduce the infectious disease in terms of sort of randomly infecting a certain number of individuals and letting them spread the disease to others, we look to see how that whitling changes essentially so of the individuals that have made it thus far, which will continue to survive? And of course there's a mutational process here as well, which means that new combinations can always arise. So there's never a permanent loss of any of the genes or any of the preferences or any of the combinations as it were.

Chris - And so what emerges, and over what timeline, with the disease, when you introduce some sort of contagion into the group; what solution does the group settle on and how quickly?

Pratik - Yeah, so I would first start by saying that the group itself doesn't really settle on a single solution. And I think that's what's really interesting here. Every individual is making slightly different choices given the same information, even if they all at some point prefer avoiding their neighbours and they will avoid them to different extents. The two main things that we want to really focus on here is whether they avoid all their neighbours or whether they're willing to associate with some of them given some benefit. And that benefit is the social information as it were, the sort of information of who's been successful and where. In the main results of our paper, what we see is that the population essentially splits into two morphs - so two types of individuals - those that play it really safe, and avoid everyone else, and get infected much less. And then these others which are willing to take the risk of being infected given a certain sort of benefit, which is again, this useful social information and then pay the cost of being, on average, infected more frequently.

Chris - So bottom line then, what do you think this study adds and how does it change or influence our thinking?

Pratik - Yeah. So I would say that because this is a theoretical biology study that's really based on simulation models, what it does is that it adds another sort of toolkit to understanding potential scenarios for the future, especially long-term scenarios that can help with sort of not just management, but also understanding what the potential outcomes of letting an infectious disease spread to novel host populations might be. Which is not to say, of course this is exactly what will happen but rather it, it helps to set the, let's say, to set the tone for discussions around that sort of scenario. So again, to sort of come back to the avian influenza pandemicor panzootic, it's one of the things that perhaps people should consider when thinking about how to manage that, which is that there could be very long-term effects on how social these, the survivors of that, of this outbreak are likely to be and what that has, you know, what that means for sort of management of sort of populations of conservation concern. But again, I will emphasise that this is a theoretical study, which is really useful I think for setting the parameters within which one can interpret data that one collects from the real world in the future, or reinterpret data that have already been collected.


Add a comment