An Ode to Quantifying Infection Risk in Addition to Prevalence

When you’re studying parasites (or symbionts or pathogens), the prevalence of the parasite in the host population is one of the easiest response variables to measure. That’s not to say that it is easy; there are certainly a variety of methodological difficulties that crop up, and it can be expensive to run lots of blood tests if you’re looking at seroprevalence. But getting a prevalence estimate is certainly a lot easier than pinpointing when each host becomes infected (e.g., via mark-recapture methods) and/or calculating the actual risk of infection (i.e., the rate that susceptible hosts become infected = force of infection). For that reason, we often use prevalence as a response variable, and hope that we can infer things about parasite transmission based on those data. Sometimes, it works out great! For instance, in 1854, John Snow (the physician, not the Brother on the Wall) mapped the locations of Cholera cases in London. By pinpointing an area of high incidence on the map, he found a water pump that was probably an important source of infection in the epidemic. But do areas of high disease incidence or prevalence always occur in areas of high disease exposure?

Littorina littorea, the common periwinkle, is an abundant and widespread marine snail that hangs out in the intertidal zone (various levels of exposure to the air with the tides) and the subtidal zone (almost never exposed to air). Periwinkles are hosts for a few different trematode species, but for today, we’ll just focus on Cryptocotyle lingua, which infects snails, then fish, then shorebirds. Snails get infected when they consume trematode eggs from shorebird feces. ‘Loitering’ shorebirds are 6-20 times more likely to hang out in the high intertidal zone than the low intertidal zone, and as a result, the density of shorebird feces in the high intertidal zone is 70 times higher than in the low intertidal zone (Byers et al. 2015). Therefore, it is not surprising that when uninfected ‘sentinel’ snails were placed in field cages in the high and low intertidal zones, snails were four times more likely to become infected in the high intertidal zone (Byers et al. 2015). In fact, the probability that an uninfected snail would become infected in the low intertidal zone was effectively zero. That makes sense, because bird guano was almost never found in that zone.

So, when Byers et al. (2015) went out and sampled periwinkles in the high and low intertidal zones, they found way higher prevalences of infection in the high intertidal zone, where infection risk was high, right? WRONG! The prevalence of infection was much higher in the low intertidal zone, even though snails do not become infected there! How could that be?

First, let’s back up and talk about an important selection pressure in the low intertidal zone: predation. There are extreme size-dependent predation pressures in that zone that pretty much prevent small/young snails from living there. So, the only snails in the low intertidal zone are bigger/older snails. Big/old snails are much more likely to be infected by trematodes than small/young snails, because they have had longer to be exposed and become infected. But we know that the big snails aren’t becoming infected in the low intertidal zone, so where are they coming from? It may be that young snails hang out in the high intertidal zone, escaping predation but experiencing high infection risk, until they are big enough to safely live in the low intertidal zone. Once big enough, the snails migrate to that low zone, which provides better foraging opportunities, and the high density of big, infected snails results in high prevalences of infection (76% infection!) in an area that has effectively zero risk of infection. Isn’t that neat?!

So, as Byers et al. (2015) point out, “disease risk and prevalence patterns need not be tightly coupled in space.” I think that’s important to remember when we’re deciding what response variables we want to consider in ecological and epidemiological studies.



Byers, J.E., A.J. Malek, L.E. Quevillon, I. Altman, and C.L. Keogh. Opposing selective pressures decouple pattern and process of parasitic infection over small spatial scale. Oikos.

2 thoughts on “An Ode to Quantifying Infection Risk in Addition to Prevalence

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s