The ghost of infections past, present, and future
Summary: A 39-year record of host-parasite interaction, recovered from sediment layers in a pond, is consistent with rapid coevolution.
Link: Host-parasite /`Red Queen/' dynamics archived in pond sediment
As I've discussed previously, archival samples often prove useful for answering questions that weren't being asked when the samples were collected. But what if nobody collected and preserved the samples you need for your research? Maybe you can find a "natural archive" that has what you need.
That's what Ellen Decaester and her coauthors did. They retrieved what appears to be a 39-year record of coevolution between the water flea Daphnia and its bacterial parasite Pasteuria ramosa from successive layers of sediment (24 cm total) on the bottom of a pond. To measure the coevolution of the ability of the bacterium to infect its host, and the ability of the fleas to resist infection, they exposed fleas from each layer to bacteria from the same layer (their "present") and from the next layer below (bacteria from their recent "past") and above ("future").
Bacteria from the past were less infectious, presumably because the flea population had evolved resistance to bacteria to which their recent ancestors had been exposed -- resistant fleas were more likely to have descendants. Bacteria from the "future" were also less infectious. Apparently the bacteria had lost some ability to attack fleas from their recent past. Perhaps there was a trade-off between retaining the ability to infect genotypes that were common in the previous flea generation versus being able to infect the present generation. Since they were able to recover live fleas and bacteria from layers thought to represent different time periods, it should be possible to work out the molecular details of this evolutionary arms race.
I say "thought to represent different time periods" because I wonder whether sediment layers preserve a perfect record of past genotypes. I'm sure the sediment layers were laid down in the order assumed, and presumably at a fairly consistent rate. A dormant flea egg from a layer laid down 20 years ago presumably represents a genotype present at that time. Dormant eggs aren't going to migrate from one layer to another. Neither would bacterial spores. Active bacteria might, however. This could just add random variation that wouldn't bias conclusions one way or another, but what if bacteria have a preferred direction (up, say, or towards particular flea genotypes)? "Parasite isolates were obtained by exposing a random set of Daphnia clones to sediment from each depth." Some method of ensuring that these bacterial isolates came from nonmotile spores rather than motile active bacteria would seem useful in future work using this approach. And I assume we will see such future work, based on this very innovative paper.