Models of livestock disease
Another issue that's been causing environmental campaigners some concern recently is the badger cull, which is intended to restrict the spread of bovine tuberculosis. Rowland Kao is a Professor of Mathematical Population Biology at the University of Glasgow. He uses theoretical models to investigate the dynamics of infectious livestock diseases including bovine TB.
Ginny - So Rowland, why is tuberculosis important?
Rowland - Well, tuberculosis is an important human disease. The bovine part of it does infect humans. So, there's certainly a risk of zoonosis. Now, the risk is not very great, so one of the big reasons why we're worried about it is because our status as a country which has a high level of bovine TB means that we are seriously affected in terms of our trade within the EU.
Ginny - So, how do you go about studying something like this?
Rowland - Well, there's been all sorts of ways that people have been looking at the disease in the past 30 years. The key being, of course, epidemiological field investigations. You look at the situation and you see what is the most likely cause of infection. And the most controversial one is the randomised badger culling trial, conducted from about 1998 onwards for the next 8 years, where there was a large study which looked at comparing culling badgers as a way of controlling TB and not culling badgers.
Ginny - And what did they find?
Rowland - Well, they found that there's a definite link between the two of them. One thing we really know is that if you do something to the cattle population, it affects TB in badgers, if you do something to the badger population it affects TB in cattle. The problem is, we don't actually know the direction of that effect. So, it could make things better or it could make things worse.
Ginny - Okay, so how do you think TB spreads between animals?
Rowland - Well, TB is spread essentially through aerosolised particles from the lungs. There are a variety of mechanisms by which they're going to go from individual to individual. So for example, some people think the environment is involved. So, a cow might cough up TB bacteria then it might lay in the environment for a while and be picked up by another cow or it could be a badger doing the same thing or could be direct contact between animals. But the truth is that it's very difficult to do experiments to actually understand this conclusively and so we actually don't have that well quantified either.
Ginny - You're a professor of mathematical population biology. How can maths help you study something that seems to be to do with animals and bacteria?
Rowland - In this case, what we have is a whole series of partial pieces of evidence. So for example, we know something about the way cattle herds are linked together because we record the movements of the cows between them. We know something about the way the badgers are linked to the cattle because we know where the badgers are roughly speaking and we know where the cattle are. We know the timing of events. The stage we're looking at right now, we're looking at how the sequences of the bacteria that are taking on cattle and badgers are related to each other. And all these various bits of information need to be fit together in a way that makes sense and that's where the mathematics and the statistics come in because not one single piece of evidence actually gives us a conclusive picture.
Ginny - Could these kind of methods be applied to other diseases or do they only work for this kind of tuberculosis?
Rowland - Well, it certainly has been applied successfully for many viruses. So for example, in the pandemic flu outbreak we had a couple of years ago, whole genome sequencing. So, a very detailed look at the changes of the genetic structure of the virus in this case was used successfully to trace the diseases that pass from country to country. Now, we're in a very different situation here. Because the bacteria are much larger than the virus, it's much more expensive to do the work, and also, because the variation in the genetics is much less, there's a lot less information. But it's never the less incredibly valuable information. So, there's good precedent for using it. We just have to be a bit more clever about how we're using it. I'm not to say that those people who do the virus's are obviously very clever indeed. We have to be clever in a different way in order to use the same kind of data.
Ginny - So, what kind of differences are there and what do you have to do to actually make it work in this example?
Rowland - Well, there's two major things that are a problem. The first which is that the bacteria itself doesn't change that much from individual to individual. Now, our estimates are that, we get about 1 in every 4 bacteria might be changed over the course of the year. So, you get 0.25 changes in the number, in an actual mutation in the bacteria over the course of the year. Now, that's a very, very little change. What it means is we can't simply look at one individual or another, look at the bacteria in the two individuals and say, "Aha! This individual gives that one" and we don't have quite that much information. So, that's the first thing. The second one is that because badgers are almost certainly involved to some extent, we're missing half of the information. We have very little information form the badgers. Now, we're working with people at the HVLA to get more detailed information from the badgers, but nevertheless, there isn't anywhere near as much there available, say compared to a cattle which are tested every single year, and for which bacteria on a regular basis.
Ginny - So, what do we need to do? When will we actually know whether a badger cull would be effective?
Rowland - It's going to take a while. I mean, first of all, the current badger culls that are being done are not being done essentially as scientific trials. So, there's no way to gauge whether the effect of them is due to the culling itself. There's so many other things that's going to be going on at the same time. So, that alone isn't going to tell us anything. These data that we're collecting, I'm looking across much larger areas. It gives us in a sense a more comprehensive picture, but at the same time, it's very difficult still to identify any single cause. So, what we have is, the ability of that data, which is a much higher resolution ever before. So, we got a much better chance of picking up differences say, in an area with culling with without. But it would still undoubtedly be at least several years if, for no other reason than because TB is a very slow moving disease.
Ginny - So, still a lot more work to do.
Rowland - There is indeed.