Ecologists are obsessed with variation, in any form, the more bizarre, the better. We really love it! But why?
The textbook explanation is that variation among individuals, if heritable, work as a template for selection and thus drives evolution. Without variation, little can change. Evolutionary important variation relates to genetic traits that make the organism better adapted to its environment, a better competitor, more disease resistant, or relates to traits that make him/her more attractive to the other sex, thereby increasing the likelihood of siring offspring.
And additional explanation, and sometimes equally important, is that it is fun with variation: an animal may be short or long, have a peculiar nostril shape, vary in the curvature of antlers, or have striking plumage colors. Simply, humans like variation, and the diversity in itself therefore drives curiosity-driven researchers.
This said, when it comes to disease in animals most researchers tend to neglect variation. Disease is commonly treated as a constant; the animal is either infected with parasite X or is not. However, in reality what the researcher denotes as parasite X may actually be a plethora of different pathogen genotypes, all seemingly dressed in the same costume (the phenotype), or sometimes even consist of cryptic species. This is dangerous, as things that look the same in the microscope (or in a conserved gene used for molecular screening) may have fundamental differences in traits that are relevant for infection processes, such as pathogenicity, transmission and virulence. Simply, we may run the risk of not seeing patterns that are there, or jump to the wrong conclusion based on simplified assumptions.
Further, surprisingly often wildlife diseases are treated at the level of the population (especially abundant in veterinary medicine), and not at the level of the individual animal. For instance, prevalence, the proportion of individuals carrying a particular disease at a given time, is much more frequently used than estimates of incidence, which relates to the risk of acquiring infection. In the former you can adhere to a ‘hit and run’ sampling approach, in the latter you need to monitor individuals across time and take repeated samples.
For a long time, actually since 2002, we have studied influenza A virus in a migratory population of Mallards in SE Sweden. We also started at the level of population, describing temporal variation in influenza A virus prevalence in the duck population, and describing differences in prevalence among ages and sexes. And yes, we treated the virus as pathogen X, not at the level of subtype (which there are many of in flu). But with time we have moved to assessing what is happening at the individual level, and how differences among individuals in susceptibility drive disease dynamics, and how disease histories and immunity patterns in turn drive evolution in the virus.
These efforts are starting to pay, and in a paper published this week (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0061201) we address the issue of individual variation among Mallards in influenza A virus infection risk. The question we asked is how individuals with the same background, in a shared environment with similar exposure to influenza, differ in disease histories and immune responses.
In our monitoring program we use a large duck trap to catch wild ducks. By providing grain we give the birds an incentive to visit the trap, and as additional attraction we have a compartment with lure ducks, that are supposed to get the wild ducks to enter. In this study, we used the lure ducks as a natural infection experiment. Ten immunologically naïve, juvenile Mallards from a farm were placed in the trap and were then followed throughout an autumn season, and then for the next spring, summer and autumn. Fecal samples were collected daily and blood samples approximately every second week. A lot of samples, and collected with a precision that allowed us to give very detailed infection histories for each individual.
In turned out that our study ducks varied tremendously in disease patterns, despite being of the same age, raised in the same farm, sharing the same little experimental enclosure and being exposed to the same environmental variation. All ducks became infected with flu within the first five days of being placed in the trap, but the number of infection days varied tremendously. And so did the number of retrieved virus subtypes, thus different individuals were infected with varying number of virus variants, in this case equal to different infection events.
Furthermore, we got really nice long-term patterns. After the initial primary infections early on in the first autumn, and a number of secondary infections later the same autumn, we recorded only a single infection day the next spring and summer. It wasn’t until the second autumn, when migration of wild ducks started in earnest again, that new infections were seen in the lure ducks. And in this case, no infection was of a subtype the individual had experienced the year before, suggesting very strong and long-lasting homosubtypic immunity.
Individuals also varied profoundly in their immune responses. We measured the humoral immune response, manifested as anti-influenza-antibodies (raised against the conserved nucleoprotein of the virus), across time. Have a look at the figure below; it really shows variation both on a temporal scale, but also at the individual scale, both in patterns and in height of response.
So what does it tell us? To start with, there is a large difference between individuals in resistance/susceptibility to influenza A virus infection in Mallards. This difference is not only manifested in different infection histories, but also as very variable immune responses. Second, these differences are very likely determined by genetic differences, meaning that there are heritable differences, and thus traits that could be selected for by natural selection. Not all ducks are equal – and this important for our ability to model disease dynamics in this system. Is it really the mean that is important for assessing the transmission probabilities along migration? Perhaps it is the outliers that are driving the processes?
This study is a first step to adress individual variation, and there are already a couple of follow-up publications in the peer-review tube, so we will have opportunities to get back to this topic.
That’s all for now. Live long and prosper – and don’t treat disease simply as a property of the population.