Research summary

Here I’ve put a summary of all the work I’ve done to date, and am doing now, along with the relevant published papers. You can either read about it all in one go here, or on each of the separate tabs.

Infectious disease transmission

I am interested in how within-host processes affect between-host disease transmission. This is summarised by the diagram below (adapted from Handel and Rohani 2015 Phil. Trans.). I’m working a bit on both (i) and (ii) in figure 1, and how they interact.

Fig 1. A figure I slightly adapted from the great review by Handel and Rohani 2015 in Phil. Trans.

How within-host processes affect host infectiousness

Recently, I’ve used the guppy-gyrodactylid host-parasite system to show that within host processes affect host infectiousness. One main result of this work is that if a host has transmitted successfully before, it transmits more worms on its second attempt (figure 2a). Intriguingly, worms that transmit from an experienced ‘donor’ seem to be less fit than those from inexperienced donors (figure 2b). I’m not sure why this happens, but I’ll hopefully work it out at some point. It could mean that while ‘superspreaders’ (the infected hosts in a population that transmit infection to disproportionately large numbers of uninfected hosts) do infect more hosts, these infections aren’t as severe as those from ‘normal spreaders’.

Fig 2. Panel (a) shows that donor guppies who had previously successfully transmitted an infection transmitted more parasites on their second attempt than donors on their first attempt. Panel (b) shows that the worm populations transmitted by these experienced donors had an equal probability of growing or shrinking (dashed line shows growth rate of 0), but those from inexperienced donors exclusively grew.

Another cool result from this experiment is that more resistant donors transmit more parasites – it’s as though the parasites are jumping ship from hosts that can effectively fight the infection (figure 3).

Fig 3. In these plots, ‘donor infection load’ is the number of parasites on the donor at the point of transmission, ‘donor infection integral’ is a measure of resistance, and is the area under the curve of infection load over time. This means that donors that have large infection loads for long periods (i.e. they are unable to limit parasite growth) have the largest infection integral. Panel (a) shows the raw data – donors that transmitted the largest number of worms are those that had heavy infection loads, but low infection integral (so they were able to effectively fight the infection). Panel (b) shows the data laid over the predictions from a model. This clearly shows that the hosts with the largest infection integrals (i.e. those that can’t limit the infection at all) don’t really transmit many worms, regardless of how heavy their infection load is.

These results (and others) are in an invited contribution to a Phil. Trans. special issue, to be published some time in 2017… Keep a look out for:

Stephenson J. F., Young K. A., Fox J., Jokela J., Cable J., Perkins S. Host heterogeneity affects both parasite transmission to and fitness on subsequent hosts. Phil. Trans. R. Soc. B.

How within-host processes affect the response infected hosts elicit from uninfected conspecifics

This is a bit of a mouthful, but basically I’m trying to work out how the healthy, uninfected individuals in a population may avoid their infected (and infectious) conspecifics. You’d think this would be a no-brainer, but actually, living in groups can be a good thing. Animals get better protection from predators, better foraging opportunities, better mating opportunities etc. etc. So leaving the group entirely because of the risk of becoming infected is a pretty bad option, probably almost as bad as actually being infected. This work is still in progress, but it looks as though guppies only avoid infected conspecifics that pose the biggest risk of transmitting the infection.

Natural enemy ecology

Direct effects of predators on host-parasite interactions

This seems like quite an obvious point – being sick makes you easier to catch, so parasites should make infected individuals more vulnerable to predation. Surprisingly, this hypothesis has been little tested outside systems in which the parasite benefits from the predation event, but could have important implications for the evolution of both host and parasite. In the guppy-gyro system, we used startle responses as a proxy for vulnerability to predation, and found that while infection had a limited influence on large fish, small fish covered significantly less ground when they were infected. Given that females are substantially larger than males in natural populations, we think this size effect is likely to result in sex-biased parasite-induced vulnerability to predation (Stephenson et al 2016, Ecology and Evolution).


Infected fish (dark points) moved less distance during the escape response than uninfected fish (light points), but the effect is greatest among small fish. Here the big points give the means±SE of two size classes, split at the median (denoted by the dashed line).

Indirect effects of predators on host-parasite interactions

Thinking about host-parasite systems in isolation ignores the fact that the host and the parasite are part of a community; as such, their interactions with other species are likely to affect the way they interact with each other. One obvious example is the impact of predators on their prey as parasite hosts. Do adaptations that prey species make to predation pressure change their interaction with parasites?

I used a large-scale, long-running study of parasite distribution in natural guppy populations to address this question. Below I’ve given the main results of two papers using these data.

  • Predator-driven evolutionary change in shoaling behaviour explains the higher prevalence of directly transmitted Gyrodactylus parasites in natural guppy populations that experience high levels of predation. Guppies also show sex-biased shoaling behaviour: females shoal more in high-predation populations, but males live more solitary lives. The sex difference in infection predicted by this sex difference in behaviour is also borne out by the data (Stephenson et al 2015, Ecology).
Where there are lots of predators, female guppies (white bars) shoal more, so the parasite can transmit between them more frequently. Where there are few predators, both male (black bars) and female guppies live relatively solitary, dispersed lives. This means the parasite can transmit less frequently, and that there’s less of a sex difference in transmission. Juveniles (grey bars) experience predation from adults across all populations, so they shoal everywhere – that’s why there’s no difference in prevalence across populations among juveniles.
  • Guppies from populations that experience these high levels of predation pressure generally have much shorter life histories (i.e. they mature faster, reproduce at a younger age etc.) than those from populations that experience low levels of predation pressure. We know from other systems and theory that organisms with longer life histories should invest more in defence against parasites because their bodies have to last longer. Correspondingly, we found that among guppies from populations experiencing low levels of predation pressure, infection with Gyrodactylus was not associated with decrease in body condition, whereas in high predation populations it was. As predicted, this suggests that low predation guppies invest more in defence against parasites. (Stephenson et al 2015, Biology Letters).

In populations of guppies that experience high levels of predation (lower course), being infected with Gyrodactylus is associated with a lower body condition (scaled mass index), but in populations that experience low levels of predation (upper course) it isn’t.