Parasites are ubiquitous. I remember watching a video (years ago, while I was teaching at secondary school) about parasites that make humans their home. Lice, eyelash mites (yes, really!), various intestinal worms... I tell you, I had psychosomatic itching for days after seeing that! Then I got my hands on Carl Zimmer's wonderful book, Parasite Rex - as well as learning all sorts of stuff about parasites & how they live, I also had it brought home to me that parasites aren't just some sort of passive, undesirable house guest - in many cases they actively influence the host's behaviour in ways that enhance the parasites' ability to complete their life cycles.
I was alerted to a recent paper in this area by a blog post from another Kiwi blogger: his sub-header was 'zombie ants controlled by parasitic fungus for 48 million years', which reall y took my fancy (the link will take you to a story in the Guardian, of which more later in this post). The authors of this paper (Pontoppidan et al. 2010) point out that it's not just a case of the parasite affecting individual ants - they can structure the entire host population in terms of its distribution in time and space & thus influence their own distribuiton: the parasite's 'extended phenotype', if you will.
The authors kick off by listing some rather dramatic ways in which other host species are influenced by their parasites, such as behavoural changes that make them more susceptible to predation, thus enabling the parasite to move to its next host; or effectively drowning themselves, which lets the adult stage of the parasite reproduce. (Their full list's available in the PLoSOne paper.) All this raises interesting questions about just how this manipulation of host behaviour is achieved, & the effects of such parasitism on the species' population as a whole (it's obviously a Bad Thing for the indivdiuals concerned). Pontoppidan & her colleagues asked a further topic: the impact of infection on the host species' distribution in space & time. They chose to look at the fungal parasite Ophiocordyceps unilateralis , and a tropical species of carpenter ants (Camponotus leonardi.).
This is really cool stuff (in a gruesome sort of way). An ant picks up the sticky fungal spores by walking over them on the forest floor; fungal hyphae then penetrate the unfortunate animal's cuticle & extend throughout its body. It can be just a few days from infection until death. Once the ant's dead, the fungus grows a 'fruiting body' out the back of its host's head. This produces large spores, too big & heavy to spread on the wind. Instead they fall to the forest floor, produce & release secondary spores, a hapless ant comes along... and the cycle repeats itself. So far, so good (for the fungus), but the really interesting part is that the ants don't die just anywhere, nor do they simply turn up their toes & drop dead on the ground.
Ants biting the underside of leaves as a result of infection by O. unilateralis. The top panel shows the whole leaf with the dense surrounding vegetation in the background and the lower panel shows a close up view of dead ant attached to a leaf vein. The stroma of the fungus emerges from the back of the ant's head and the perithecia, from which spores are produced, grows from one side of this stroma, hence the species epithet. The photograph has been rotated 180 degrees to aid visualization.
Instead, before an ant actually dies it bites into the surface of whatever plant it's standing on at the time. Pontopiddan et al. identify this behaviour as the fungus's extended phenotype: it holds the ant's corpse in place on the plant for long enough that the fungus can secrete a 'glue' that will stick the body there more permanently, which in turn gives time for the fungus to develop its fruiting body (the 'stroma' & 'perithecia' in the images above). What's more, the team had heard accounts of 'graveyards' containing large numbers of dead carpenter ants (cue images of zombie ants staggering along to some formicine cemetery). So they decided to determine whether these graveyards really do exist and, if they do, how various biotic & abiotic factors influenced the distribution of dead ants.
To do this they spent more than 5 weeks & >500 person-hours in a Thai rainforest, looking for ants. (This wasn't quite needle-in-a-haystack territory as these ants can be >4mm long, but still...) In all this time they found 2243 dead ants in their study plots (the great majority of which were Camponotus leonardi), but only 2 live C.leonardi. But there were lots of living ants from other species, doing what ants do, in the study area - which suggested that leonardi was definitely the main host for Ophiochordyceps unilateralis. It was 3 weeks before they saw an active trail of leonardi, which descended one tree & travelled only 5m on the ground before heading up another trunk, followed by yet another descent before disappearing into the canopy again. That trail led to a single leonardi nest, high in the canopy (20-25m above ground), with a network of trails running along twigs & branches & extending up to 100m from the nest.
On the basis of these observations, the team hypothesised that ants of this particular species actively avoid descending to the forest floor unless it's the only way to reach a new resource. (You can see how natural selection might achieve this: a colony where too many ants go down to the ground on an everyday basis is likely to lose large numbers of foragers. So if there's a genetic underpinning for such behaviour, a queen passing on a 'go to ground' gene would end up losing lots of her daughters & thus her nest would be at a competitive disadvantage to other colonies.) It turns out that there is some evidence supporting this hypothesis: in an area of forest where the parasitic fungus isn't present, C.leonardi is commonly found at ground level.
When the research team went on to look at just where the dead ants were found, it appeared that the bodies weren't randomly distributed. Instead they were in large aggregations (the 'graveyards') of up to 26/m2, separated by corpse-free zones. The now-deceased had bitten onto the undersides of leaves, on average about 30cm above the ground - an example of how the fungus influences its host's behaviour. The distribution of dead ants appeared to be related to temperature & absolute humidity - things which could influence the survival of fungal spores & thus the chances of an individual ant picking up the infection.
Zombie jokes aside, this really is a fascinating example of the complexity of ecosystem interrelationships. And their longevity. It also turns out that this particular parasitic relationship may have been in place for a very long time indeed. The 'death bite' leaves a characteristic scar on a leaf, and in a separate paper David Hughes & colleagues describe finding just such a scar on a leaf dating back 48 million years
, from rocks in what is now Germany.
Hughes, DP, Wappler , T & Lanadeira, CC (2010) Ancient death-grip leaf scars reveal ant-fungal parasitism. Biology Letters. Published online before print August 18, 2010, doi: 10.1098/rsbl.2010.0521
Pontoppidan MB, Himaman W, Hywel-Jones NL, Boomsma JJ, & Hughes DP (2009). Graveyards on the move: the spatio-temporal distribution of dead ophiocordyceps-infected ants. PloS one, 4 (3) PMID: 19279680