We have a new paper out in Polar Research (an open access journal, see link below), entitled Contrasts between the cryoconite and ice-marginal bacterial communities of Svalbard glaciers. This is the result of a collaboration between Aberystwyth University, University of Bristol and Innsbruck University in Austria.
Some of the earliest questions we had (first scribbled on my 2006 field season plan – a crumpled, coffee stained post-it note rediscovered in my office the other day following an avalanche of reprints) about the microbiology of cryoconite related to how the community was structured and assembled. Were the bacteria in cryoconite simply dumped on the ice surface along with dusts from neighbouring habitats, as earlier research suggested, or were cryoconite communities distinct in structure because of different selective pressures on the ice surface itself?
Some of our earlier work has shown that bacteria related to those from a diverse range of habitats were present and that the fungal communities in cryoconite samples were very different to those nearby, but as the idea that how the glacier surface gathers dust preconditions primary succession as the ice retreats gathers momentum, we thought it was worth revisiting our early question. Further impetus came from the confluence of attending an introductory R for dummies course held at Aberystwyth last summer, reading a paper on using rank abundance distributions to explore fungal community assembly, and having previously co-supervised a PhD student who routinely used Akaike Information Criterion for selecting models. These three strands came together in deciding to have a stab at modelling phylotype abundance distributions to some fingerprint data I had lying around.
So, after a lot of coffee and cursing of my rudimentary abilities with R, we found that cryoconites have their own, distinct functional microbial communities that are different from those of the surrounding moraines and tundra. This is the case even though much of the materials and cells found in cryoconite will have been blown onto the glacier from the surrounding areas. Models consistent associated with deterministic (niche based) rather than neutral processes best fit the rank abundance distributions of community structures revealed by the fingerprinting data. The optimal models in almost all cases were related to communities formed through succession. This lends some tentative support to the idea that succession kicks off before the ice is gone.
However, there are some problems with this interpretation. Co-occurence network analysis revealed only a minority – about 10% – of the phylotypes were consistently distributed across both cryoconite and other habitats. This would suggest that cryoconite bacteria have difficulty thriving after being inoculated into the glacier forefield. It would be interesting to follow up with some experiments to test this in more detail. But right now it’s nice to have crossed off something on the to-do list circa 2006!
Find the paper here: Polar Research 2013, 32, 19468, http://dx.doi.org/10.3402/polar.v32i0.19468 (open access)
Authors: Arwyn Edwards, Sara M. Rassner, Alexandre M. Anesio, Hilary J. Worgan, Tristram D.L. Irvine-Fynn, Hefin Wyn Williams, Birgit Sattler, Gareth Wyn Griffith