Bottom line up front: Snow covers ca. 30% of Earth’s surface; we are losing the last permanently dry snows in the Northern hemisphere: Are these vast (new) microbial habitats, thanks to climate change? We published a paper which shows bacteria can proliferate rapidly in a decaying Arctic snowpack.
You would be forgiven for thinking we’re all cryoconite – obsessed swivel-eyed loons here at AberCryoconite Towers. In reality, only some of us are. If only because we are also interested in other microbial habitats associated with glacial systems. In fact, if we can get our (numb, nitrile- gloved) hands on it, we’ll give anything icy a go. So, I’m going to summarize one of our recent papers, published by ISME Journal on the dynamics of bacteria in High Arctic supraglacial snowpacks and released to the press last week.
Like the cryoconite metagenome paper, this paper was born out of a collaboration with the Innsbruck Crew supported by the Society for General Microbiology’s fund for research visits in 2010. Chatting with two of Professor Birgit Sattler’s grad students, Kathi Hell and Jakub Zarsky, it appeared that some of the molecular and statistical approaches I had been developing could complement their experimental fieldwork earlier in the summer, working on Larsbreen on Svalbard.
Kathi and Jakub had visited at a one-week interval, sampling in an up-glacier transect of three stations at kilometre intervals, digging three snowpits at each station to collect snow, slush and ice cores. The second visit had seen a thin layer of aeolian dust deposited on the snow surface, so they sampled that too. While Jakub measured the activity rates of the bacterial community, Kathi filtered the samples for DNA analysis and chemistry.
So, Kathi bought DNA extracts over to the Aber lab in March 2011 to do T-RFLP and pyrosequencing. This is what we found out:
So what is the story?
1. Welcome to the layer cake?
Essentially, profiling the bacterial community revealed differences between different layers of the snow-covered glacier. We think the particular differences could only be explained by changes in the bacterial population in situ, or in other words after deposition as postulated by Xiang et al. (2009). In particular, we found slush (i.e. snow, as it melts and goes horrible, mucky, slushy crap) to harbour a distinct bacterial community. This confirms the melting snowpack as an active habitat for bacteria.
2. Betaproteobacteria: Slush puppies?
Our amplicon pyrosequencing allowed us to track the fates of different taxonomic groups in the different layers between the two sampling days. We found that the major bacterial group (class) in our samples, Betaproteobacteria, remained consistently abundant, but the organisms in the group shifted around. In particular, the genus Polaromonas proved to be a smooth player, able to duck and dive, wheel and deal in the rapidly changing environs of the slush layer.
If I had to anthropomorphize the view I have of Polaromonas following this study, they would be supraglacial Del Boy type characters, playing it nice and cool, while all the other bacteria remain oblivious, just like Trigger. Polaromonas, as a genus, crops up fairly regularly in cold environments. You would therefore expect it to be something of a “Ronseal” type bug: does what it says on the tin. But members of the genus are expert in taking on dodgy deals elsewhere too: Polaromonas napthalenivorans, for instance loves to split horrible organic pollutants straight down the middle, sixty-forty. Because of this (and other, unpublished) work we are now taking a very close look at Polaromonas.
3. Pyrosequencing- the Aristotlean connnection.
When I first examined our amplicon reads, a good fraction were “unclassified bacteria” and I didn’t know what (if anything) to make of the data. Upon re-analysis, we found that these reads were in OTUs comprising members of Chlamydomonadaceae. Specifically, the reads aligned to plastid rRNAs; thanks to our primers struggling to remain specific to bacteria we gained some bonus data. Something of an Eureka! moment ensued: the reads’ taxonomy and physical distribution were consistent with snow algae.
Classicists might complain that I’ve got the wrong Greek: after all Eureka is apparently what Archimedes shouted, not Aristotle. But that isn’t the connection I was alluding to. Aristotle, amongst other things, was the first to observe snow algal blooms. These occur as green algae, Chlamydomonas nivalis, having grown within the snowpack, produce characteristic carotenoid-rich structures which can colour the top 10-20 cm of a snowpack blood red. It’s an unlikely spectacular. The first time I saw such a bloom with my own eyes it was several weeks into a trip in the Arctic. I thought someone had laid into the stash of rhodamine to create a practical joke on an epic scale…
The significance for our study is two-fold.
Firstly, snow algal blooms are well-documented phenomena. Our pyro data reflected a nascent algal bloom. 97% id Operational Taxonomic Units, especially when automatically delineated by the hundred in high-throughput sequencing datasets can appear quite arbitrary if not totally “virtual”. As ecologists of the invisible it’s nice to get a highly visible confirmation sometimes. Cheers Aristotle!
Secondly, snow algal blooms demonstrate that snowpacks can be loci of considerable biological activity, resulting in spatially extensive phenomena which reveal themselves at brief timescales. Algal blooms are well known at sea too:
4. Nitrogen pollution of glaciers is NO joke.
Serious chemists won’t like that subheading. Nevertheless, anthropogenic nitrogen pollution from faraway lands acts to fertilize nutrient-poor environments in the Arctic. Arctic glaciers and the Greenland ice sheet are no exception. Microbes in glacial ecosystems can respond to assimilate ammonium from even a single deposition event. Using 454, fingerprinting and qPCR methods we found specific populations of Betaproteobacteria associated with ammonia oxidation in the snowpack, and in particular on the ice surface. We also appeared to find correlations between the bacterial community revealed by 454 and the decoupling of nutrient and non-nutrient anion dynamics in the snowpack. This would suggest microbes in the snowpack environment are tucking in to the nitrogen pollution. Our observations add to the growing body of literature on microbial interactions with the molested nitrogen cycle of glaciers to reveal a picture of an unholy trinity between microbial processes, anthropogenic nitrogen pollution and unstable climate.
What does it all mean then?
I struggled (visibly, even in the final edition) to write the paper’s discussion in a coherent fashion. Fortunately one lunchtime in the café I bumped into Aberystwyth’s Greenland guru and BBC Frozen Planet / Operation Iceberg “duderino” (his words) Dr. Alun Hubbard. He took a few minutes out of revising a manuscript to a headline journal to describe conditions high on the Greenland Ice Sheet in July 2012. Slush was forming, and piling up, saturating the firn and running off. In just 4 days, the percentage of the Greenland ice sheet experiencing surface melting skyrocketed from ca. 40% to a maximum of 97% on the 12th of July. At that timescale, the melt is too “flashy” to let the algae get out the starting blocks. But our paper suggested the bacteria might.
Could it be that the epic melt on Greenland in 2012 triggered a massive bacterial bloom?
Who knows. From our Svalbard study we could see a heady mix of a globally ubiquitous genus of bacteria proliferating at similar timescales in similar conditions, interacting with nitrogen pollution to create a glacier-scale bacterial bloom. Plans to test this hypothesis have amounted to little thus far: 2013 saw far less melting, and attempts to reconstruct events from 2012 by looking at the firn from 2012 must take care as a bloom may not be conserved in the stratigraphy (according to our study).
What is certain though, is that we are seeing less and less dry snow in the Northern hemisphere. And where there is liquid water, there is life.