Should you steal your (sick) co-worker’s food?

Do you have a co-worker, friend, or roommate? If you do, have you ever been in a position to “borrow” that person’s food? Did you? Ethical dilemmas aside, did you consider whether said person was sick?

Let’s say I have a co-worker named Gunter. If I take Gunter’s lunch, I will benefit from acquiring a “free lunch.” However, I might suffer costs if Gunter discovers who took his lunch. When I decide whether to purloin Gunter’s food, therefore, I am conducting a cost-benefit analysis. An additional factor I might include in this analysis could be whether Gunter is sick. If Gunter is sick, for example, I might risk acquiring his infection by eating his food. Such infection would likely more than negate the benefits I obtained from eating Gunter’s food. Gunter’s sickness would consequently dissuade me from “borrowing” his food. But, what if from past experience I know that Gunter gets incredibly angry when anyone borrows his food – except when he is sick. Naturally, he is more displeased about losing his lunch when he is sick than when he is healthy, but when he is sick he doesn’t have the energy to confront anyone about the loss. And now I have a conundrum. I am likely to suffer negative consequences from a healthy Gunter if I am discovered with his lunch. I may also acquire an infection from Gunter if I take his lunch when he is sick. But I am less likely to suffer negative consequences from Gunter at that time, even if I am discovered. So should I steal Gunter’s lunch, and when?

Unfortunately, no simple answer exists for this situation. The answer all depends on the cost:benefit ratio – the benefits I get from Gunter’s food relative to the costs I may suffer from stealing. If benefits are high enough, it would be optimal for me to steal all the time, regardless of Gunter’s state. If benefits are high, but costs due to Gunter’s behavioral response are high, it might only be optimal for me to steal when Gunter is sick.

If you think this example is a little contrived, well, you’re right. But not so much as you might think. For instance, Bouwman and Hawley (2010) discovered that male house finches preferentially fed near males who were infected with a bacterial parasite (mycoplasma), even though feeding sites enabled high transmission rates of the parasite. Bouwman and Hawley also noted that infected males were less aggressive than uninfected males. The authors therefore concluded that males who fed near an infected male benefited from encountering less aggressive males (meaning the former males were more likely to be able to feed without incident). This benefit might or might not outweigh the costs of infection for the population as a whole, but at least at the individual level, benefits appeared to override the consequences.

This house finch result has interesting implications for the evolution of host-parasite interactions. We often think of hosts evolving to maximize fitness, which means reducing parasite costs to an absolute minimum, but this does not have to come from reducing parasite infection to a minimum. If you engage in behaviors that lead to infection, but you get more benefits from those behaviors than the costs of infection, selection will favor that behavior, even though it increases infection.



Bouwman, K. M., and D. M. Hawley. 2010. Sickness behaviour acting as an evolutionary trap? Male house finches preferentially feed near diseased conspecifics. Biology Letters 6:462-465.

Does your behavior influence your immune system?

I previously wrote an article about how behavior can influence transmission. In that piece, I focused on behaviors that can increase transmission (in some cases, bolder individuals can be more likely to contract a particular parasite because their boldness may lead them to make risky decisions, such as fighting with infected individuals). However, to provide a fuller picture of host behavior in parasite transmission dynamics, I decided that I should now talk about how behaviors can influence parasite infection in a way that benefits the host.

I cannot speak with authority for disease ecologists as a group, but generally speaking I think we all realize that host behavior is very important for parasite transmission. Unfortunately, the effects of behavior can be very hard to quantify. Even in systems in which behavior and parasite transmission can be linked, demonstrating causality becomes incredibly difficult. For example, in my earlier post on boldness and parasite transmission, Dizney and Dearing (2013) could not determine whether infection caused the boldness, or whether boldness determined infection. It seemed likely that boldness was influencing infection given the biology of the system, but this inference was by no means proven beyond all doubt. Thankfully, such hindrances are not universal.

Daly and Johnson (2011), for instance, performed simple, yet elegant experiments with tadpoles. By comparing the number of infections anesthetized tadpoles and regular tadpoles received, the authors elucidated the effect of tadpole behavior on parasite infection (the authors had shown previously that anesthesia does not affect tadpole physiology or immunity). Anesthetized tadpoles had a much greater number of parasites (in this case trematodes) than the control tadpoles, demonstrating that tadpole behavior protected the tadpoles from infection. Expanding upon these results, another group determined that tadpole behavior also influenced the location of the parasite infections (Sears et al. 2013). Anesthetized tadpoles sustained the majority of their infections in the head, while untreated tadpoles suffered them in the tail. As one might imagine, parasites in the brain are a much more serious infection than in the tail, especially because these tadpoles lose their tails in the course of their development. Thus, tadpole behavior not only reduced the number of infections the tadpoles suffered, but also shunted the infections that they did receive to less critical areas.

If these examples seem a tad different from any human disease (who has heard of someone being able to control whether a cold causes headaches or congestion), house finches may provide a more relevant example. Recent work with house finches has demonstrated a number of interesting correlations among behavior, infection status, and the immune system. For example, in a study testing house finch preference to associate with a sick or healthy bird, house finch males chose to associate with a healthy male more than a sick one, and males that were closer to the sick bird invested more in immune defense than finches that chose to be farther away (Zylberberg et al. 2013). In another house finch study, similarly, male birds that invested more in surveillance behaviors (which are thought to reduce parasite exposure because the finch would not be blindly entering a novel area or eating a novel food) invested less in immune defense. Furthermore, as male avoidance of novelty decreased, investment in immunity increased (Zylberberg et al. 2014). Thus, house finch males appear to balance investment in anti-parasite immunity and behavior – as investment in behavioral defenses increase, immunity declines, and vice versa. These results may be more applicable to humans in that while we cannot control where a parasite infects us (as far as I know), we can perform behaviors akin to the house finches: We can control whether we choose to go to work when we are sick, or how much we wash our hands, or whether we eat food off the floor (e.g., “the 10 second rule”). It remains unknown whether such behaviors, or lack thereof, correlate with the immune system in humans as they do in house finches. If they do, the next question becomes “why?” Why should some individuals invest in behavior and others the immune system? Do individuals differ at birth or during childhood in some way that set them on a behavior or immune system course? Are these differences selectively favored? Perhaps a population with mixed defense strategies survives better through time than one in which all individuals use the same parasite defense strategy? Perhaps things to think about the next time you are debating whether the 10 or 30 second rule applies.


Daly, E. W., and P. T. J. Johnson. 2011. Beyond immunity: quantifying the effects of host anti-parasite behavior on parasite transmission. Oecologia 165:1043-1050.

Dizney, L., and M. D. Dearing. 2013. The role of behavioural heterogeneity on infection patterns: implications for pathogen transmission. Anim Behav 86.

Sears, B. F., P. W. Snyder, and J. R. Rohr. 2013. Infection deflection: hosts control parasite location with behaviour to improve tolerance. Proceedings of the Royal Society B-Biological Sciences 280:5.

Zylberberg, M., K. C. Klasing, and T. P. Hahn. 2013. House finches (Carpodacus mexicanus) balance investment in behavioral and immunological defenses against pathogens. Integrative and Comparative Biology 53:E400.

Zylberberg, M., K. C. Klasing, and T. P. Hahn. 2014. In house finches, Haemorhous mexicanus, risk takers invest more in innate immune function. Animal Behaviour 89:115-122.

The Coevolution Continues

A while back, I made a post about parasite-host coevolution.

Here’s a cartoon, because my posts are too often filled with text.


Anyway, to continue on the topic. In the last article about coevolution, I covered the idea of host and parasite coevolution leading to continual and reciprocal evolutionary changes – the hosts and parasites are constantly running, but they stay in the same place relative to each other. The (excellent) cartoon makes sense now, right?

In the article about coevolution, I also touched on the issue of multiple parasites infecting multiple hosts, and how this can complicate host-parasite coevolution. Here, I’m going to further complicate matters by talking a little about environmental heterogeneity and other species that interact with host-parasite associations.

Harrison et al. (2013) dealt with the issue of environmental heterogeneity. They conducted an evolutionary experiment with bacteria and their viral parasites, varying the bacterial habitat over time. They found that varying the habitat prevents coevolutionary dynamics between the bacteria and their viral parasite.

Conversely, Friman and Buckling (2013) experimented with bacteria, their viral parasites, and protist predators of the bacteria. These authors found that the bacteria tend to coevolve to either the viruses or the predator, not both. The presence of the bacterial predator also constrained the coevolution between the bacterial and virus.

So where does this leave us? It seems that when we consider only single host-single parasite systems, we may find the treadmill coevolutionary pattern seen in the cartoon above, but the likelihood of this situation breaks down when we consider other factors like habitat heterogeneity and multiple natural enemies (i.e., predators and parasites). However, one could reasonably argue that the discussed experiments with bacteria are not relevant for all organisms, and systems in which parasites and hosts do engage in treadmill (aka arms race) coevolution could exist. Whether such coevolution does or does not occur, though, will likely be system specific.

Friman, V. P. and A. Buckling (2013). Effects of predation on real-time host-parasite coevolutionary dynamics. Ecology Letters 16(1): 39-46.

Harrison, E., A. L. Laine, M. Hietala and M. A. Brockhurst (2013). Rapidly fluctuating environments constrain coevolutionary arms races by impeding selective sweeps. Proceedings of the Royal Society B-Biological Sciences 280(1764).

When Sharing is Caring: A Parasite’s Perspecitve

We’ve all heard of Influenza A, or Flu. It’s a viral parasite that can and has caused massive pandemics in humans. However, Influenza A does not infect only humans. It also parasitizes birds, particularly waterfowl, and other mammals. Many parasites have this diverse infection ability, i.e., many parasites infect multiple host species.

For such parasites, questions regarding their transmission dynamics become more interesting (and complicated). Instead of considering the interaction between only one parasite and one host species, one has to consider how different host species contribute to the parasite’s population. One host species may be the parasite’s main host, adding more parasites to the parasite population than any other host species. On the opposite side of the spectrum, the parasite may have no main host; each host species may contribute equally to parasite numbers in the community at large. In all likelihood, most parasites probably fall somewhere in the middle of this range; the parasites do not have a single main host, but the host species also contribute unequally to the parasite population.

In the cases where a parasite has a few (or one) main hosts, those hosts are called reservoirs. By producing most of the parasite population, reservoirs maintain that population. Reservoirs are thus important research areas for parasite population control, and therefore wildlife and human health. For example, many parasites that infect humans have non-human animal reservoir hosts. These parasites, ones that spillover into humans from other species, are called enzootic parasites. If one can determine their specific wild reservoirs, then one can eliminate parasite transmission to humans in a number of ways, including: limiting human-reservoir contact and preventing successful transmission among the reservoir species members (which itself can be accomplished through different mechanisms such as: vaccination, culling, habitat modification). However, such tools and goals are not simple, even when managing only a few host species.

More than the practical and financial considerations in appreciably affecting even a few host species, the different types of reservoir hosts complicate management efforts. Not all reservoir hosts are so for the same reasons. Streicker et al. (2013) argue three main mechanisms create reservoir host species: 1) The host species can be super abundant and have high population density; 2) The host species can be very susceptible to infection; 3) The host species can be highly infectious, producing numerous new parasites when infected. In the first mechanism, the host maintains the parasite simply by being so numerous. While a small percentage of the host population is infected, the host population is so large that it produces many parasites and maintains the parasite population. In the second mechanism, because hosts are so susceptible, a very high proportion of the host population is infected and can maintain the parasites at relatively lower host population sizes. In the third mechanism, hosts are highly infectious, so even though few hosts are infected, a single infected host produces many new parasites and has a high chance of infecting another host. These three situations are not mutually exclusive, but Streicker et al. argue that which one mostly applies to a particular reservoir species should affect management practices regarding that species. For example, the authors analyzed the effect of targeted removal of infected hosts and random removal of hosts from populations where a host species was either only super abundant, only super susceptible, or only super infectious. For super abundant hosts, random removal of hosts did not efficiently decrease the host’s contribution to parasite population because there was a small chance of removing an infected host. Targeted removal of infected hosts resulted in a moderate reduction, however. For super susceptible hosts, where a large proportion of the host population is infected, both random and targeted removal of hosts efficiently reduced the host’s contribution to parasite numbers. Because so much of the host population is infected, the chance that a random removal gets an infected host is not much less than targeted removal of infected hosts. For super infectious hosts, random removal finds infected hosts in proportion to their frequency in the population, and since each infected host contributes a large parasite number, each removal of an infected host will noticeably decrease the host’s contribution to the parasite population. Targeted removal of infected infectious hosts will decrease the host contribution to parasite population much more than random, although. This occurs because when there are few infected individuals, and each produces a large number of parasites, specific removal of the infected hosts will remove a large proportion of the parasite population pool. These results indicate that – for removal of hosts as a management tool – the different mechanisms creating reservoir hosts affect the outcomes of management actions.


Streicker, D.G., A. Fenton, and A.B. Pedersen, Differential sources of host species heterogeneity influence the transmission and control of multihost parasites. Ecol Lett, 2013. 16(8): p. 975-84.

Behavior and Parasite Transmission Risk

What determines parasite risk for a host? We often focus on factors such as the host’s immune system and the parasite’s ability to find or infect hosts. However, host behavior can determine a large proportion of infection risk. Take the 10-second rule for example. Some people like to say that food that has been on the ground for less than 10 seconds is safe to eat. That may or may not be true, but it is likely a riskier behavior (in terms of possible food contamination) than not eating food that has fallen to the ground. Engaging in such risky behaviors could thus make 10 second rule followers more likely to obtain parasites than people who do not eat food that has touched the ground. Importantly, this logic applies beyond humans and the 10 second rule. Within a species, one might predict that individuals that are more bold would be more likely to obtain parasites. One might predict this assuming that bolder individuals would be more likely to engage in risky behaviors, such as eating obviously bad meat. A recent study analyzing the contribution of host behavior to parasite transmission, for example, discovered that bold behavior (assessed through number of aggressive interactions and a few other measures) predicted Hanta virus infection in deer mice [1]. Specifically, bold deer mice were much more likely to show positive infection status for Hanta virus than shy deer mice. Cool, right? There are a few qualifications to point out here though. A significant percentage of shy mice were infected with Hanta, meaning there are other factors that need to be researched in the Hanta virus-deer mice system. This study did also not illuminate whether bold deer mice are bold because they were infected with Hanta (Hanta made them bold), or whether deer mice are more likely to obtain Hanta if they are bold (deer mice are bold prior to Hanta infection). Lastly, conclusions about the role of behavior in parasite transmission need to be relatively host and parasite specific. In the Hanta virus and deer mice system, for example, aggressive interactions lead to virus transmission because the virus is spread through bodily fluids. Aggressive interactions in which deer mice wound each other thus promote transmission of Hanta virus. However, for parasites that spread through other means, more sedentary or shy individuals could be more beneficial to parasite transmission.
Literature Cited

1.    Dizney, L. and M.D. Dearing, The role of behavioural heterogeneity on infection patterns: implications for pathogen transmission. Anim Behav, 2013. 86(5).