Cryptic connections and pathogen transmission

Happy Thanksgiving, Everyone! The origins of this holiday aside, I find it worthwhile to spend a day feasting and reflecting on all of the things that I’m thankful for. This year, one of those things is my new postdoc position studying white-nose syndrome (WNS). I’ve blogged about WNS before (e.g., here and here), but I’ve yet to blog about my favorite WNS paper, because it only just came out this week in Nature! I might be a bit biased in my evaluation, but it was certainly worth coming out of my blogging torpor to write about. Give it a read!

Let me tell you about a lovely dream that I share with many other disease ecologists: a new wildlife pathogen emerges; funding to study it becomes immediately available; we rush in and quickly figure out how the pathogen is transmitted by observing how hosts contact other hosts and/or pathogens in the environment; we thus quickly figure out how to interrupt pathogen transmission, our control efforts save an imperiled host species, and the crowd goes wild. Most of that scenario is still just wishful thinking, but today I’ll focus specifically on the difficulties associated with observing and quantifying the contacts that matter for pathogen transmission. There are two scenarios that can turn my lovely dream into a nightmare: the contacts I can observe are not important for transmission and/or the contacts that I cannot observe are important for transmission. Here are some examples: 

(1) The mycoplasma pathogen that causes house finch conjunctivitis seems like it should be transmitted from one bird eyeball to the next when birds physically contact each other. Direct contacts between birds aren’t necessarily easy to observe, but they can be quantified with proximity loggers and similar technology. But those obvious, quantifiable bird–bird contacts don’t really explain mycoplasma transmission dynamics. Instead, transmission seems to occur only when birds visit the same bird feeders subsequently – an infected bird visits, deposits some pathogen, and leaves, and then a susceptible bird visits later and gets exposed. These infected and susceptible birds are “connected” across time in a way that would be completely missed if we didn’t record videos of bird feeders or do feeder RFID experiments.

(2) Mountain brushtail possums spend their days in tree hollow dens and often share their dens with other individuals, especially their pair-bonded mates. Obvious contacts! But contact networks based on den-sharing contacts did a poor job of predicting E. coli strain sharing among possums. Spatial overlap in home ranges (and thus exposure to the same E. coli contaminated environments) wasn’t a great predictor of E. coli strain sharing either. Instead, brief (~4 min), nocturnal, cryptic contacts best explained E. coli transmission.

(3) And finally, we have the new white-nose syndrome example. It’s hard to imagine a more adorable and obvious contact than two bats snuggling for days at time while they hibernate. On average, each cave-hibernating bat in the Midwest is snuggled up with ~2% of the other bats in the cave during visual surveys. But if you cover individual bats in ultraviolet-fluorescent powder and leave them for a few months, you’ll come back to find that during their occasional bouts of arousal, they have actually contacted ~15% of the other bats and much of the cave environment, leaving little puffs of powder in their wakes. And it turns out that those cryptic contacts – the ones that were illuminated by powder trails but not by counting snuggling bats – do a much better job of predicting fungus transmission within and between bat species. For instance, northern long-eared bats were usually seen roosting alone, but the powder revealed a wealth of cryptic connections to individuals of the same and other species. Those cryptic connections likely explain why most northern long-eared bats are infected by the white-nose syndrome fungus by the end of the hibernation season. In contrast, tri-colored bats are rarely seen cuddling and were rarely contaminated by powder from other bats, confirming that they’re the loners of the cave world and explaining why so few tri-colored bats are infected by the end of the hibernation season. Really cool stuff!

These examples illustrate an important point that is easy to forget: if you have gone into the field and quantified a contact network for a host species, you have not necessarily also quantified a transmission network for that host species. To construct transmission networks, you need to know all contact types, and you need to actually quantify transmission.

References:

Adelman, J.S., S.C. Moyers, D.R. Farine, and D.M. Hawley. 2015. Feeder use predicts both acquisition and transmission of a contagious pathogen in a North American songbird. Proc Biol Sci. 282(1815): 20151429.

Blyton, M.D.J., S.C. Banks, R. Peakall, D.B. Lindenmayer, and D.M. Gordon. 2014. Not all types of host contacts are equal when it comes to E. coli transmission. Ecology Letters 17: 970–978

Hoyt, J. R., K. E. Langwig, J. P. White, H. M. Kaarakka, J. A. Redell, A. Kurta, J. E. DePue, et al. In press. Cryptic Connections Illuminate Pathogen Transmission within Community Networks. Nature.