The coextinction of parasites, commensals, and mutualists – a call for more natural history studies!

By definition, mutualists, commensals, and parasites (hereafter “affiliates” in this post) depend on their hosts for resources or services. Therefore, if a host species goes extinct, the affiliates associated with that host may go extinct, too. And in fact, coextinction events like this should be as common as – or even more common than – extinctions of hosts, because we know that every host species has many mutualists, commensals, and parasites. Just think about the mites living in your eyebrows, the bacteria living in your intestines, and that one time that you had lice in third grade. If humans disappeared, all of those affiliate species might also go extinct!

Of course, you might not believe in the intrinsic value of all species; you might be wondering why you should care if some tiny species that you’ve never heard of goes extinct. Extinctions of mutualistic species – such as the gut microbes that help you digest food and the pollinators that keep our agricultural systems running – have obvious implications for our economy and health. But parasites, too, play important roles in our lives. For instance, they regulate populations of wildlife host species, and they may prevent you from having allergic reactions to things that you shouldn’t be allergic to. And of course, species exist in intricate webs of interactions, and by accidentally (or purposely!) adding or removing species from ecosystems, we have often learned that one species can have huge impacts on ecological communities.

So, coextinctions of affiliates are important, and these coextinctions should be common. That means that we have documented tons of these coextinction events, right? Actually, we haven’t! There are very few examples of documented coextinctions (Dunn 2009, Colwell et al. 2012), and some of those are not entirely open and shut cases. But why?

Say that you document the extinction of a particular host species: Host A. Should every affiliate associated with Host A also go extinct? Because some affiliates likely use multiple host species, some of the affiliates of Host A probably survived on other host species. Also, even an affiliate that historically only used Host A might be able to continue existing if it can switch to a new host species. For instance, maybe Host B, a close relative of Host A, is a suitable alternative host.

Now imagine that you’re trying to document affiliate coextinctions as Host A disappears. What evidence might you use to figure out which affiliates have also disappeared? There might be published accounts of some of the affiliates of Host A, but there are very few host species (if any) for which every affiliate species has been documented. Therefore, the loss of one host species means that several unnamed and undescribed invertebrate species will be lost before ever being documented by humans. Even if you had a perfect list of every affiliate species, it might be really difficult to confirm whether each affiliate was now extinct. That’s because we rarely (if ever) have perfect lists of every host species used by a given affiliate species. So, if one affiliate species frequently uses three host species, but you think it is a specialist on Host A, you might think the affiliate has gone extinct, only to find it happily hanging out on Hosts B and C when you survey those species three decades later.

To summarize, we predict many coextinctions of affiliates to occur as hosts go extinct, but we have hardly documented any such coextinctions. It may be that that affiliate species are much less vulnerable than we expect due to the use of multiple host species or host species switching as a primary host goes extinct, and/or it may be that we are just very poorly equipped to observe and document these coexinctions. Clearly, if we’re going to get better estimates of affiliate coextinction rates, we need more data! Specifically, we need:

  • Better understanding of the natural histories of these systems. We need complete lists of affiliates for each host species, complete lists of host species for each affiliate species, preserved specimens of affiliates for genetic identification, and information on the strengths of the interactions between each affiliate and host species.
  • Better estimates of how frequently affiliates shift host species, and whether jumps to new host species are associated with declines in the availability of the current species. In other words, how often do we expect affiliates to sink with the ship versus swimming to safety? (For further reading about this, see Kiers et al. 2010.)

(It’s been too long since my last pirate worm cartoon….)


Some related reading:

Conservation – save the parasites along with the hosts?

Are pubic lice going extinct?


Colwell, R.K., R.R. Dunn, N.C. Harris, and D.J. Futuyme. 2012. Coextinction and Persistence of Dependent Species in a Changing World. Annual Review of Ecology Evolution and Systematics 43: 183-203.

Dunn, R.R., N.C. Harris, R.K. Colwell, L.P. Koh, and N.S. Sodhi. 2009. The sixth mass coextinction: are most endangered species parasites and mutualists? Proc. R. Soc. B 276: 3037–3045.

Kiers, E.T., T.M. Palmer, A.R. Ives, J.F. Bruno, and J.L. Bronstein. 2010. Mutualisms in a changing world: an evolutionary perspective. Ecology Letters 13(12): 1459-1474.

Inferring intra and interspecific parasite transmission from parasite population genetic structure

It’s often important to know how frequently parasites are transmitted among hosts of the same species (intraspecific transmission) or among hosts of different species (interspecific transmission). But observing parasite transmission events can be very difficult in wildlife populations, so we often have to use proxies instead of measuring transmission rates directly. For instance, we might use the frequency with which two bird species share a nesting site as a proxy for how frequently we think that transmission should happen between the two species.

But of course, transmission doesn’t necessarily happen when two species contact each other. So how can we determine whether interspecific transmission is really happening? There’s more than one method, but today, I just want to talk about a cool method that I’ve seen in a bunch of recent papers: comparisons of parasite population genetic structure within and among host species. If parasite populations are highly genetically differentiated among host populations or among host species (or even among individual hosts!), then there is evidence for low parasite transmission and thus genetic mixing among host populations or among host species (or individual hosts). Conversely, if there is no genetic differentiation in parasite populations among host populations or host species, then there may be high parasite transmission among host populations or host species. Here are a whole bunch of examples of how this idea has been explored in the literature recently:

Ectoparasitic flies on bats (Olival et al. 2013):

Olival et al. (2013) sampled bat flies on three species of bats in the Pteropus genus at eight sites in Malaysia, Cambodia, and Vietnam. Almost all of the bat flies were from a single species: Cyclopodia horsfieldi. An analysis of the molecular variance in the sampled bat flies showed that very little of the variation was explained by geographic region or host species. This suggests high rates of interspecific transmission of this bat fly species among the three Pteropus bat species. Previously, interactions between the three bat species, including roost sharing, were thought to be uncommon. But because the bat flies pupate off the host in the roosts, Olival et al. (2013) suggest that perhaps interspecific transmission can happen when the different bat species share the same roost locations sequentially, rather than at the same time.

Even though there was low genetic structuring in the sampled Cyclopodia horsfieldi bat flies, for one bat host species (Pteropus hypomelanus), there was relatively low gene flow in the parasite population at some isolated island sites. It turns out that bat gene flow is also low at those smaller, more isolated island sites. But if that’s the case, then why don’t those parasites have distinctly different genetic lineages from other sites and host species?  Olival et al. (2013) suggest that one of the bat species, Pteropus vampyrus, visits those more isolated island populations of Pteropus hypomelanus during long-distance dispersal, and that those visits provide enough population mixing to prevent divergence in the parasite lineages among sites and host species.

Ectoparasitic mites on bats (van Schaik et al. 2014):

Let’s stick with bats, but shift our geographic focus to central Europe and our parasite focus to mites in the genus Spinturnix. S. myoti mites live on Myotis myotis bats and S. bechsteini mites live on Myotis bechsteinii bats. Both mites have similar life histories, and they are only transmitted during direct contact; they can only survive for a few hours off a host bat, unlike the bat flies discussed above. S. myoti mites had high genetic diversity and panmictic genetic structure, with no differentiation among bat populations. S. bechsteini mites had low genetic diversity and high differentiation among bat populations. van Schaik et al. (2014) suggest that the differences in the genetic structure of the two mite species can be explained by the differences in the social systems of the two bat species. Myotis myotis bats have larger colony sizes, more inter-colony visits during the maternal season, and closer intraspecific associations during the mating season, and all of these factors could lead to more intraspecific transmission of S. myoti mites, both within and among colonies. That is so cool! (By the way, check out this post for more information about the relationship between host contacts and parasite transmission.)

Ectoparasitic flies on birds (Levin and Parker 2013):

In the Galapagos, great frigatebirds (Fregata minor) are parasitized by Olfersia spinifera hippoboscid flies, and Nazca boobies (Sula granti) are parasitized by Olfersia aenescens hippoboscid flies. The great frigatebirds have distinct genetic population structure among islands, but their hippoboscid flies and a pathogen transmitted by the flies (Haemoproteus iwa) have no genetic differentiation among islands (Levin and Parker 2013). Also, of the few Olfersia spinifera hippoboscid flies sampled on a second frigate species (F. magnificens), all flies had the most common fly haplotype on great frigatebirds. Similarly, the Nazca boobies had distinct genetic lineages among sites, whereas the hippoboscid flies on boobies showed no genetic differentiation among sites or among multiple booby host species.

So, what’s going on? How could the parasites be so well-mixed among sites, while their bird hosts are not? Levin and Parker (2013) suggest two hypotheses: 1) maybe alternative host species that weren’t considered in this study are doing lots of island hoping and carrying flies around with them. Remember that Pteropus vampyrus bats may play that kind of role in the bat fly example above. 2) Host genetic structure is distinct among islands because the birds are philopatric; they like to mate at their natal breeding site. But juvenile birds may still visit other sites without mating, and thus without influencing bird population genetic structures, and those visits could spread the parasites among the islands, thus mixing the parasite lineages.

Feather lice on birds (Koop et al. 2014):

Let’s stick with birds in the Galapagos, but let’s change our focal host to hawks (Buteo galapagoensis) and our focal parasites to feather lice (Degeeriella regalis). Hawks are thought to cross open water far less often than the frigatebirds and boobies in the previous example. Unsurprisingly, Galapagos hawk populations have high genetic differentiation among islands, where the genetic differences among populations increase with the distance among islands (Koop et al. 2014). Hawk feather lice also show high genetic differentiation among islands, unlike in our previous parasite examples. This suggests that there is very little interpopulation dispersal of lice, and there isn’t an alternative host carrying lice to different islands, either. Furthermore, lice are mostly vertically transmitted from parent to offspring, rather than the host-roost-host or horizontal host-host transmission routes in the previous systems. As a result, there is also genetic differentiation of lice among individual hosts, so that each host acts like a parasite island! Neat!

Feather mites on birds (Dabert et al. 2015):

Birds again, but now let’s talk about feather mites on two species of skuas (arctic and long-tailed skuas) in Svalbard. The mites are thought to be transmitted only during direct host contact, either vertically from mother to offspring or horizontally among hosts. Even though the two skua species nest at the same sites during the breeding season, nests tend to be spaced far apart, so Dabert et al. (2015) predicted that the two skua species would have distinct mite species. Both skua species had mites in the Alloptes genus, which were morphologically very similar, but which were genetically distinct enough between the two host species to be classified as two different species. However, both skua species also had Zachvatkinia isolata mites, and those mites had a well-mixed population with no evidence for genetic differentiation among host species. How could that be? Well, the two skua species do contact each other, during very brief but common aerial fights. And it may be that Zachvatkinia isolata mites, which are more abundant on the host and specialize on a relatively external region of the feathers, are more likely to be transmitted during those brief aggressive encounters than the Alloptes mites that hang out in more protected parts of the plumage. UHM, AWESOME.

You might be wondering if similar studies have been done with host species that don’t fly, or with endoparasites instead of ectoparasites. There is some endoparasite work, like with schistosomes and whipworms, but I’m not going to cover it here. As for non-flying host species, check back next week for an example of how the insight gained from studies like this can be used in an applied way to manage parasite transmission.


(I was watching a lot of Fringe when I made this cartoon.)


Dabert, M, SJ Coulson, DJ Gwaizdowicz, B Moe, SA Hanssen, EM Biersma, HE Pilskog, and J Dabert. 2015. Differences in speciation progress in feather mites (Analgoidea) inhabiting the same host: the case of Zachvatkinia and Alloptes living on arctic and longtailed skuas. Exp Appl Acarol 65:163–179.

Olival, KJ, CW Dick, NB Simmons, JC Morales, DJ Melnick, and K. Dittmar. 2013. Lack of population genetic structure and host specificity in the bat fly, Cyclopodia horsfieldi, across species of Pteropus bats in Southeast Asia. Parasites & Vectors 6:231

Koop, JA, KE DeMatteo, PG Parker, and NK Whiteman. 2014. Birds are islands for parasites. Biology Letters 10: 20140255.

Levin, II, and PG Parker. 2013. Comparative host–parasite population genetic structures: obligate fly ectoparasites on Galapagos seabirds. Parasitology 140: 1061–1069.

van Schaik, J, G Kerth, N Bruyndonckx, and P Christe. 2014. The effect of host social system on parasite population genetic structure: comparative population genetics of two ectoparasitic mites and their bat hosts BMC Evolutionary Biology 14:18.

Host Control of Symbionts

Mutualisms are kind of a big deal.  In fact, our very existence depends on mutualisms.  From the pollinators that service our crops to the bacteria that help us digest our food, we just couldn’t survive without them.

Since mutualisms are so important, you might expect us to know a LOT about them.  But we don’t know as much as you might expect.  For instance, we don’t really understand how mutualisms are maintained in the long-term.  Theory predicts that at least some mutualists should cheat, because cheaters should have better fitness than perfect cooperators.  That is, mutualists should be like that used car dealer in Matilda that glues on the car bumpers.  We know that there are cheaters in many mutualisms – like in marine cleaning mutualisms – but something prevents all the mutualists from cheating.  Something prevents cooperation from evolving into parasitism.

That “something” may be partner control.  Partners can reward or punish their mutualists for good or bad behavior, thereby selecting for the mutualists that sell good cars don’t cheat.  But as I mentioned before, the benefits of having mutualists aren’t constant across all ecological conditions.  In fact, sometimes, mutualists can even act like parasites.  So, partner control may need to vary with ecological conditions.  Thus, enter this really cool branchiobdellidan-crayfish study by Skelton et al. (2014).


Pirate worms will be groomed off and eaten on sight!

Skelton et al. (2014) wanted to know if crayfish were less resistant to branchiobdellidans – if they varied how much they controlled their worms – when the crayfish were bigger.  When crayfish are small, they molt frequently, so there isn’t much time for epibiotic material to accumulate on the crayfish.  Having cleaning worms probably isn’t very useful for small crayfish.  But when crayfish get big, they molt less frequently, so having worms around to keep them tidy might be very beneficial.  Therefore, Skelton et al. (2014) performed three experiments where they looked at how crayfish responded to worms when the crayfish were small, intermediate, or large in size.

NOW FOR THE COOLEST PART.  Crayfish ‘control’ their worms by grooming them off (and eating them).  They just reach up with their little walking legs and pluck the worms off.  To experimentally manipulate crayfish grooming, Skelton et al. (2014) just snipped off the dactyls on the walking legs so that the crayfish couldn’t pick off their worms.  That is, in each experiment, there were control crayfish with functional pinchers and other crayfish with their dactyls removed.

Small crayfish strongly resisted worms.  As soon as they had worms put on their backs, the small crayfish started trying to groom them off.  And by the end of the experiment, there were almost no worms remaining on small crayfish if the crayfish had their dactyls intact and could groom themselves.  Intermediate-sized crayfish reacted similarly, but by the end of the experiment, they still had 25% of their worms, even when they could groom themselves.  Conversely, big crayfish liked their worms – when they had their dactyls, they groomed off a few worms during the experiment, but they mostly allowed the worms to hang around.  So, large crayfish had very low resistance to worms, but smaller crayfish had very high resistance, and that’s probably because larger crayfish need the cleaning services, whereas smaller crayfish don’t need the cleaning services.

But why did intermediate-sized crayfish keep 25% of their worms, even when they had their dactyls intact?  The intermediate-sized crayfish had to keep 25% because they couldn’t groom off the worms: the worms hung out on the dorsal carapace in the one place that the crayfish couldn’t reach.  Tricky!  When the crayfish were small enough, there was no place for the worms to hide.  On large crayfish, worms would hang out in a variety of attachment sites, with few occurring on the dorsal carapace, because they could hang out anywhere without (much) risk of being groomed off.  But on intermediate-sized crayfish, they’d get groomed off if they didn’t avoid the areas that the pinchers could reach.  And what’s more, in the field, Skelton et al. (2014) also saw worms mostly hanging out on the dorsal carapace of intermediate-sized crayfish but attaching in a variety of places on large crayfish.  So, not only do crayfish change their control behavior with ontogeny, but worms also change their behavior with crayfish ontogeny.  Cooooooool.

I have a better cartoon of this, but I’m going to wait until next week.  Stay tuned!


Skelton, J., R.P. Creed, and B.L. Creed. 2014. Ontogenetic shift in host tolerance controls initiation of a cleaning symbiosis. Oikos.


Climate Change and Host-Parasite Temporal Mismatches

Climate change is (appropriately) a big topic in the ecological literature.  We have a pressing need to understand how climate change is going to affect individuals, populations, communities, and ecosystems.  And of course, we need to know how climate change is going to alter disease dynamics, so that we can be ready to control/manage important parasites and pathogens if we need to.

As I’ve discussed before, at one point, there was a lot of concern that the geographic ranges of diseases (like malaria) were going to increase dramatically with climate change.  In 2009, Lafferty published a paper where he pointed out that warmer isn’t always better for parasites.  He argued that we’re probably more likely to see range shifts than range expansions.  Similarly, a few months ago, I posted about a paper by Pickles et al. (2013) that supported Lafferty’s hypothesis.  Pickles et al. (2013) used models to show that in some parts of its range, the adult meningeal worm is likely to go extinct because its hosts will not be there, and in other places, the adult meningeal worm will be able to invade new areas along with its hosts.  In total, the parasite’s range is expected to shift, but not to increase.

The Pickles et al. (2013) paper was about how hosts and parasites might end up having a mismatch in space, so that parasites suddenly can’t find their hosts in areas that used to have hosts.  There has also been a lot of interest in how climate change might cause temporal mismatches among interacting species (e.g., predators and prey, plants and pollinators).  These mismatches result from climate-induced changes in the phenology of species.  And now we ask: will climate change cause temporal mismatches between hosts and parasites?

Paull and Johnson (2014) performed a mesocosm experiment to consider how one aspect of climate change – increased temperature – might affect disease dynamics in a trematode system.  They studied a particularly important trematode (Ribeiroia ondatrae) that causes mortality and severe limb malformations in some amphibian species.  The internet is rather devoid of good diagrams of the life cycle of R. ondatrae, but here’s a link to one such copyrighted diagram.  (If asked very nicely, I might cartoon one myself.)  Briefly, the adult trematode uses a heron as the definitive host.  The adult trematode releases eggs in the aquatic environment (e.g., pond), and those eggs hatch into free-living miracidia that swim around looking for a snail to infect.  After the parasite has sufficient time to develop and asexually reproduce in the snail, the snails release another free-living parasite stage called a cercaria.  (You guys should be familiar with cercariae at this point because they’re my go-to cartoon parasite.)  Cercariae swim around until they find a tadpole to infect, and then they encyst in the tadpole’s tissues.  If the tadpole lives long enough, it metamorphoses into an adult frog – possibility with limb malformations – and gets eaten by a heron.  Etc.

Paull and Johnson (2014) put snails in 36 mesocosms (miniature ponds) in August.  Half of the mesocosms had greenhouse lids that caused them to be heated to 3˚C warmer than the other mesocosms.  Then they added R. ondatrae eggs to the mesocosms so that the eggs could hatch into miracidia and infect the snails.  For the next year, they checked every few weeks to quantify 1) snail mortality, 2) what percentage of snails were infected, and 3) how many cercariae the infected snails were releasing per night.  Also, starting in April, they added one bullfrog tadpole to each mesocosm each month, let the tadpoles hang out for two days getting infected, and then dissected them to quantify infection.  Then in May, they added 20 chorus frog tadpoles to each mesocosm and left them until they started metamorphosing.  Then they quantified survival, limb malformations, and infection of the chorus frog tadpoles.  They also quantified a bunch of other stuff, but all that is less important to the story.

When the mesocosms weren’t heated, infected snails started releasing cercariae in late spring.  In the heated mesocosms, infected snails started releasing cercariae nine months earlier, in the fall!  The warmer temperatures vastly increased the developmental rates of the parasites.  Furthermore, more second-generation snails got infected in the heated treatment.  So, did this increase in parasite development rates and the percentage of infected second-generation snails result in more tadpole infection?  Actually, no!  First of all, more adult snails died in the heated treatment and there were fewer total second-generation snails in that treatment, too.  So, the total number of shedding snails was actually lower in the heated treatment.  Second, in the unheated treatment, snail infection peaked just when the tadpoles were available in the spring, whereas snail infection in the heated treatment rapidly declined in the spring.  The total number of cercariae was the same in both treatment groups, but due to the temporal mismatch and increased snail mortality, tadpole infection actually declined under the warmer temperature scenario.

Sometimes, I'm the only one who understands my cartoons....

Sometimes, I’m the only one who understands my cartoons…. Yes, those are soldier cercariae.

So, this was a very cool study.  Climate change experiments are really difficult to tackle because running long-term experiments with frequent sampling events is hard work!  And trying to manipulate multiple species in biologically relevant ways can be tricky.  In this case, the results of this experiment would likely be very sensitive to the timing of the input of R. ondatrae eggs.  Paull and Johnson (2014) did just one pulse of eggs in the fall.  If eggs also enter the system in the early spring, infection dynamics might be quite different under the +3˚C warming scenario.

Ok!  I’m ready to see more work like this.


Paull, S.H., and P.T.J. Johnson.  2014.  Experimental warming drives a seasonal shift in the timing of host-parasite dynamics with consequences for disease risk. Ecology Letters.

Host Susceptibility to Parasites: ‘Vicious Circles’

Today’s post is all about host susceptibility.  In mathematical models, we usually assume that every host is equally susceptible to infection (i.e., constant success rate, v).  That may be mostly true in laboratory settings, because laboratory hosts tend to 1) be maintained under optimal resource conditions and 2) have similar genotypes.  However, in the field, both host genetics and resource availability vary widely, and we expect hosts to vary in their ability to resist infection by parasites.  Specifically, we expect hosts in poor condition to be more susceptible to parasite infection.

As it turns out, variation in susceptibility among hosts is really important for two reasons:

  1. If there is variation in host susceptibility to infection among individuals (v), there must be variation in the per capita parasite transmission rates for each individual (β).  Without variation in β, you cannot have aggregation of parasites among hosts.  So, variation in host susceptibility is one potential cause of aggregation of parasites among hosts.
  2. Variation in host susceptibility to infection is one way to generate superspreader hosts.  In this case, highly susceptible hosts end up being super infected hosts.  (Seriously, I love that super host paper.)

So, what determines how susceptible an individual is?  The answer is: many things.  But those things mostly boil down to 1) genetics and 2) environment.  Genetics example:  some host genotypes may be more susceptible to a given parasite than others.  Environment example: hosts in low resource environments may be in poor condition, and may not have enough energy to devote to mounting an immune response to parasites.

And that brings me to a series of cool TREE papers about “vicious circles” of host susceptibility and infection.  The idea is that parasite infection is probably “both a cause and consequence of host condition.”1  That is, hosts might be in poor condition because they have many parasites, and/or hosts may have many parasites because they are in poor condition.  This sets up a potential negative feedback loop, where host condition continuously degrades while parasite burdens increase: the host is in poor condition, the host gets more parasites, the host’s condition further declines, the host gets more parasites, etc.

These “vicious circles” may be important to parasite transmission at both the individual and population levels.  At the individual level, ‘vicious circles’ may cause some individuals to have (increasingly) lower fitness and/or survival.  It may also be one way to generate super infected hosts.  At the population level, we might see population declines as individuals experience decreased fitness and/or survival.  Therefore, we might expect that the effects of parasites/pathogens on populations will be stronger when populations are already experiencing some environmental stress that causes the individuals to be in poor condition.

Do we always expect to see vicious circles of infection?  Nope!  For instance:

  1. If predators tend to pick off super infected hosts, then there may be no chance for the vicious circles to perpetuate.  Therefore, parasites/pathogens may have stronger effects on host populations where predators have been eliminated/reduced than in host populations regulated by healthy predator populations.  COOL.
  2. If the parasite is trophically transmitted, we might not expect hosts in poor condition to be more likely to get infected.  For instance, a definitive host might need to be in good condition to eat enough intermediate hosts to get infected.  If that’s the case, then we might expect a stabilizing feedback loop, instead of a negative feedback loop.

Saddest cartoon ever.


  1. Beldomenico, P. M., and M. Begon. 2010. Disease spread, susceptibility and infection intensity: vicious circles? Trends Ecology Evolution 25(1): 21–27.
  2. Loot, G., F. Thomas, and S. Blanchet. 2010. “Vicious circles” and disease spread: elements of discussion. Trends in ecology & evolution 25(3): 131.
  3. Beldomenico, P., and M. Begon. 2010. Response to comment by Loot et al. Trends in Ecology & Evolution 25(3): 32.

Bats and Emerging Infectious Diseases

Bats are amazing.  It’s easy to forget about bats because we don’t usually see them, but they’re out there, performing important ecosystem services that we often take for granted.  If you don’t believe me, you can read this review by Kunz et al. (2011), which outlines some important services that bats provide for humans:

  1. Pollination
  2. Seed dispersal
  3. Arthropod suppression
  4. Guano for fertilizer
  5. Tourism – caving, etc.
  6. “Witches and sorcerers used bats in ancient magic to induce desire and drive away sleep.”  (Seriously, without bats, there would be no ancient magic.)
  7. Bats are frickin’ cute (I added this.)

Bats are also reservoirs for many viruses that cause serious human illnesses.  These include viruses like SARS, Ebola, and some paramyxoviruses like the Hendra and Nipah viruses.  Because these viruses are such a big deal, there has been a lot of recent attention to bats and their potential as reservoirs for high-impact emerging zoonotic viruses.  Specifically, two major questions arise:

  1. Are bats hosts for more zoonotic viruses than other wildlife?
  2. If yes, what characteristics make bats such good reservoirs for these emerging zoonotic viruses?

In a recent meta-analysis of viruses of bats and rodents, Luis et al. (2013) found that on average, bat species host more zoonotic viruses than rodent species.  So, perhaps there is something special about bats that make them particularly good reservoirs!  Of course, comparing bats and rodents doesn’t fully answer Question 1, but it is a start to say that if we look at bats and another taxonomic group that shares many life history characteristics with bats, bat species host more viruses, on average. However, because there are more rodent species than bat species in the world, rodents host more total zoonotic viruses than bats.  Therefore, in terms of global human risk, bats don’t contribute more than rodents.

So, given that bats may be somewhat unique in their ability to host zoonotic viruses, what causes them to be such good hosts?  Good question!  At this point, no one can really say, but it’s probably a combination of some of these unique bat characteristics:

  1. Bats have unique feeding ecology, where they tend to spit out their food. They suck on the fruit/flower of choice and swallow the nectar/juice, but then spit out the remaining material.  If another animal comes along and eats the pulp off the ground, it might ingest virus particles from the bat, and the virus will have the opportunity to jump/spillover into a novel host species.
  2. Bats and humans tend to overlap in habitat, which provides opportunities for bat viruses to spillover into human populations.  This is particularly likely in places where humans are altering landscapes so that livestock operations and bat habitat get mixed together.  For instances, in places where livestock pigs have access to fruit that bats have spit out.
  3. Bats can be gregarious, where they may roost in extremely high densities.  Furthermore, multiple bat species may share the same roost.  High densities of susceptible individuals provide a virus’ dream population.
  4. Some bats migrate, and their long-distance travel may help them to spread viruses.
  5. Some bats hibernate, and that reduced metabolic activity may be important for some viruses, like rabies.
  6. Because bats are evolutionarily ancient, their viruses may have highly conserved cell-receptor proteins that are good at invading the cells of many mammal species.

The take home message is that we need to study bats and emerging infectious diseases more.  We know very little about how and why and when viruses spillover from reservoir hosts to novel species, but in this era of global change, understanding those spillovers is becoming crucial for human health.  And as Luis et al. (2013) found, the more we study a given host species, the more viruses we find that infect that species.  So, if we want to know which viruses bats currently harbor in order to asses which viruses might be most likely to spillover into human populations, we should invest in more bat research!

That brown thing is a tree.


Kunz, T. H., E. Braun de Torrez, D. Bauer, T. Lobova, and T. H. Fleming. 2011. Ecosystem services provided by bats. Annals of the New York Academy of Sciences 1223: 1–38. (PDF link)

Luis, A. D., D. T. S. Hayman, T. J. O’Shea, P. M. Cryan, A. T. Gilbert, J. R. C. Pulliam, J. N. Mills, M. E. Timonin, C. K. R. Willis, A. a Cunningham, A. R. Fooks, C. E. Rupprecht, J. L. N. Wood, and C. T. Webb. 2013. A comparison of bats and rodents as reservoirs of zoonotic viruses: are bats special? Proceedings of the Royal Society B 280: 20122753. (PDF link)

Super Hosts and Parasite Transmission

In earlier posts, I told you that the “dilution effect” is a popular hypothesis in disease ecology that says that the risk of parasite infection decreases with host diversity.  I also pointed out that there is caveat: disease risk is only negatively correlated with biodiversity if the hosts in the communities with low diversity are particularly good hosts (e.g., they’re good at transmitting the parasites).

Let’ s call those particularly competent/important hosts “super hosts,” because that’s what Streicker et al. (2013) called them, and the cartoon potential is awesome.  What makes a host species a super host?  When parasites utilize multiple host species, why should one species be better than any other?  (Open access link!)

Using a combination of mathematical modeling and a field survey, Streicker et al. (2013) described three types of super hosts.  Super abundant host species are just that: they’re more abundant, on average, than the other host species in the community.  Super infected host species have higher prevalences of infection than other host species in the community.  And super shedder host species produce more infective stages than other host species in the community.  (I talked about individuals who are super shedders, like Typhoid Mary, in a different post.  Here, we’re talking about species, not individuals.)

The Super Host Species! From left to right: Super Abundant Host, Super Infected Host, and Super Shedder Host! (These look deceptive similar to the Three Muskrat-eers…)

Besides the super host cartoon potential, why should we care what mechanism leads to a host species being a super host species?  Streicker et al. (2013) did a great job of showing that the effectiveness of various management strategies (e.g., treating infected individuals, culling individuals) depends on the type of super host.  For instance, if the key species is a super abundant host species, then untargeted control is going to be pretty useless – if you just go out and kill a bunch of individuals, you probably won’t get the infected ones.  And if the key species is a super infected host species, it’s not worth spending as much time and money on diagnosing infected hosts – you should just treat/kill the ones you catch, because they’re probably infected.

There are, of course, many more cool details in the paper, and I highly recommend checking out the pretty math.

What kind of experiment would you like to see to go along with this paper?


Streicker, D.G., A. Fenton, and A.B. Pederson. 2013. Differential sources of host species heterogeneity influence the transmission and control of multihost parasites. Ecology Letters.

Parasite(ish)-Host Coevolution

Continuing on my not-parasite-but-kinda-similar trend, let’s talk about bacteria and phages.  This post was stimulated by a really cool talk by Britt Koskella, from the University of Exeter.  She has a wordpress site and has been tweeting about the EEID conference, so head that way for more cool stuff.  (Maybe she’ll also come correct places where I butcher her work.)

A bacteriophage is a virus that infects bacteria.  In Britt’s case, the bacteria of interest are parasites of the horse chestnut tree.  She studies the co-evolution of these three groups: the trees, the bacteria, and the phages.  Importantly, because these organisms/phages have very different life spans, we expect phages to evolve faster than bacteria, which should in turn evolve faster than trees.

Question 1:  Do we see local adaptation of phages to the bacteria of a given tree?  Answer:  Yep!  If you take leaves from multiple trees and culture the bacteria from those leaves, and then test all the phages on all of those bacteria cultures, you find that phages do best on bacteria from the tree that they were collected on.  Also, phages did better on bacteria from the interior of the leaves, which makes sense because the exterior is likely highly controlled by abiotic processes (e.g., UV radiation).

My cartoon of Britt’s graph. Phages are more successful at infecting bacteria from the tree they were collected on (sympatric tree) than other trees (allopatric trees).

Question 2:  Given that we see local adaptation of phages to bacteria, does that adaptation vary with time?  Answer:  Yep!  In this part, Britt calculated “local adaptation” as an index comparing phage success on bacteria from sympatric vs. allopatric trees.  She cultured bacteria from trees from each month in the season, and tried the phages from the last month on all of those cultures.  Does that make sense?  So, September phages on September bacteria, September phages on August bacteria, September phages on July bacteria, etc.  Here’s what she found: phages were most adapted to the bacteria from the previous month (=August), and then adaptation declined as you went further back in time.  She suggested that this is demonstrative of fluctuating selection, rather than an arms race between bacteria and phages.  That is, in an arms race, you should never see a decline in phage success as you go backwards in time.

Phages are most adapted to the bacteria of the prior month, and then adaptation declines as you continue backwards in time.

So, that was a lot about the phage evolution, but what about the bacteria?  Question 3: Do bacteria evolve resistance to phages?  Answer: Yep!  Since she had all of those monthly bacteria  and phage samples, she tested bacterial resistance to phages that were from the past time step, the present time step, and the future time step.  Bacteria were most resistant to past phages, indicating that bacteria evolve to resist their phage.  They were least resistant to phages from the future, which indicates that phages also evolve to better infect bacteria.  Neat!

(Edit: Check out Britt’s comment below about whether this pattern is the result of pairwise coevolution or species sorting.  More coolness to come!)


All of the parasite-host co-evolution stuff is super cool.  Britt also looks at co-evolution in other disease systems, and you can check out some of that work here.

Don’t you really, really want to do experiments where you’re looking “into the future?”  Futuristic snails must be awesome.  I’m sensing an upcoming cartoon…


Britt Koskella. 2013. Bacteria-phage interactions within a long-lived host.  EEID.

Koskella, B., Thompson, J.N., Preston, G.M. & Buckling, A. 2011. Local biotic environment shapes the spatial scale of bacteriophage adaptation to bacteria. The American Naturalist177(4):440-51.

More coming soon!

Wicked Cool Host-Commensal-Parasite System

I really like symbionts.  I really, really like interactions among symbionts, and I especially like it when commensals/mutualists eat parasites.  

So, it is with great pleasure that I introduce to you this urchin-crab-snail system.  The common pencil sea urchin (Eucidaris galapagensis) is a host for parasitic snails (Sabinella shaskyi and Pelseneeria spp) and commensal crabs (Mithrax nodosus).  And the crabs eat the snails!

Sonnenholzner et al. (2011) did some neat field and lab work to figure out how fishing for urchin predators affects parasitism of urchins by snails in this cool system.  Hilariously, they sum up their findings in the first line of the discussion by saying that they “found that the enemy (fisher) of the enemies (fish and lobster) of the enemy (crab) of the urchin’s enemy (snail) was the urchin’s friend.”  Swag.

Here’s the quick (and simplified!) version of their results, but I highly recommend checking out the paper!

I drew this fishing pole myself.

Do you know of any other host-commensal-parasite systems?  Bonus points if you guess my FAVORITE system of all!


Sonnenholzner, J.I., K.D. Lafferty, and L.B. Ladah. 2011. Food webs and fishing affect parasitism of the sea urchin Eucidaris galapagensis in the Galapagos. Ecology, 92(12): 2276-2284.

Why did the chicken cross the road? Because it was a PARASITE ZOMBIE.

This month, I’ve done a lot of posting about parasites turning hosts (like snails and insects) into “zombies.”  But I never addressed one very important question:  why did the chicken cross the road?  BECAUSE IT WAS A ZOMBIE.  (Thank you, Saturday Morning Breakfast Cereal, for making my life complete.)