We conducted our extraction on Wednesday hunched over under an overcast sky, catching papers, tubes, and foil as they blew away in the wind, and layering our gloves in a futile attempt to keep our hands warm. By contrast, our last day in the field gave us a fortunate change in the weather—the skies cleared, the sun broke through, and the wind dissipated. Thursday opened as half of our group set out early to begin a sequencing run at our field site in Kerlingarfjöll, and the rest of us arrived shortly after, the rest of our gear in tow. We arrived just in time to watch Maggie load our flow cell with our DNA from Wednesday’s field extraction, and we gazed on as the data begin to roll in and NanoOK started to return our analysis.

With our sequencing run chugging along, we set up our “field lab” with a beautiful view. The top of our ridge was in full sun, and the light illuminated the snow packs on the surrounding mountains against the dark rock. Part of our group hiked back down into the valley, toward the glacial streams and their surrounding banks, to collect more samples. As they set out, we began preparing another round of field extractions. We used a modified PowerSoil kit protocol to extract DNA from the soil sample we used the day before—the same soil whose DNA we had begun sequencing in the morning. We also extracted from a newly collected sample from a similar site. When the rest of our group returned, we concentrated a water sample from a glacial stream with our InnovaPrep concentrating pipette—in a matter of minutes, we had concentrated a few hundred milliliters into around 1 mL, some of which we were able to use in our extraction protocol alongside our soils. We worked through the protocol, swapping out instruments plugged into the generator to give each one enough power, and using field substitutions for some of the equipment we’d normally use in the lab. Once we had worked through our protocol, we prepped our DNA for quantification in order to see which samples, if any, were going to be viable for sequencing—and to check for contamination in our blank, which is an even bigger threat in the field.

I don’t think I’ve ever held my breath so much or hoped so frantically that a blank would turn out clean as I did then—when our control didn’t show evidence of contamination, we felt comfortable moving ahead with our second field sequencing run of the day. We prepared the DNA library, like we did in the morning, using a newly developed kit for field sequencing, and we set up our MinION (wrapped in some felt, to stay warm!) to begin our run. After only about 30 minutes, we were able to look at the first analysis returned by the NanOK RT pipeline and get information about the bacteria in our sample in real time. After spending the summer running practice extractions and struggling to properly load flow cells, we were finally able to see our entire process laid out before us, in the field—from raw sample to analyzed data, in one shot. It was exhilarating and gratifying to see all of our efforts come together and give us information about our sample, and also about what experiments are possible in the field.

As we drove back to the lodge, we tried one final test, just for fun—keeping one of our MinION runs sequencing on our laps as we drove over the bumpy dirt and gravel roads. We laughed about the MinION’s “road test,” and we were able to keep our sequencing run going during the trip back from our field site. While our sequencers ran through the evening, we packed our field equipment back into its suitcases. We set out the next morning for the long drive back to Reykjavik, and we crossed the Mid-Atlantic Ridge again, from the Eurasian plate to the North American, another step on our way home.

Wednesday dawned cloudy and, you guessed it, cold at Hveravellir. We huddled over our morning breakfast of toast with jam, cold cuts, and fermented shark (hákarl) while contemplating how this could possibly be August. Nevertheless, we packed up the trucks and headed off into the day. Half of our group journeyed up to Kerlingarfjoll to sample streams, ices and sediment for further analysis while the other half stayed behind for a practical test of our field extraction protocol. We set up our remote extraction on a yoga mat far from signs of habitation. Our only sources of power were the gas-powered portable generator and the car engine, which we hooked up to an electrical inverter.

Due to complications with the CP Select, we were not able to use the concentrator pipet for our extraction protocol. Instead, we extracted from soils and biofilms collected at Kerlingarfjoll the day before: one sample from the bank of a glacial outwash stream, one soil sample further away from the water source, and one soil from the trailhead at the top of the valley. We hoped to compare the yield from each of these samples to understand what would be most promising for our field sequencing and analysis run. To extract, we used the Qiagen PowerSoil kit combined with the Terralyzer to lyse cells instead of a traditional vortex.

The first thing we discovered is that it is possible to stretch latex gloves over warm wool ones to retain feeling in one’s hands. The second thing we discovered is that doing so gives one far less dexterity. Faced with the choice between dropping bits of Kerlingarfjoll all over Hveravellir and losing feeling in our fingers, we opted for the latter but climbed gratefully into the warm car whenever we needed engine power.
After about one-and-a-half hours, we looked triumphantly at the microcentrifuge tubes filled with 30 microliters of clear liquid. Had we managed to isolate the code for life itself? That question would require more work. We took a break to warm up, during which time I confirmed that none of my toes had frostbite and found a bag of grapes in the truck’s backseat, which gave us the fortification to continue. Back to the yoga mat we went, now pulling out the Qubit to ascertain our DNA concentrations. Then we had our answer: we had managed to get DNA from all of our samples, the most from the soil near the trailhead and the least from the soil on the bank of the stream. Moving forward, we hoped to sequence our results from sample x in the field, as well as completing a field extraction, sequencing, and NanoOK analysis all in the same day.

Dawn greeted us over Greenland the morning of August 6th. We landed on her smaller, warmer neighbor in Keflavík, Iceland. International tourists milled about in plush coats and new daypacks, but beyond the airport, our trip diverged from the norm. A local researcher outfitted us with 4X4 trucks, we stocked up at an Icelandic Bónus (“Are we sure this means peanut butter?”), and then we were bound for Úthlíd Cottages.

The ocean disappeared behind rolling green hills, and sturdy Icelandic ponies replaced seagulls. We crossed the mid-Atlantic ridge in Thingvellir National Park and spent the evening sampling the stream and tributary pools beneath Bruarfoss Waterfall. Field observation inevitably raises more questions than it answers. Why do the biofilms prefer standing pools to the swift stream? Why do the red and green biofilms grow in separate pools, and what processes produce their color? Do they differ in metabolism, or perhaps in symbionts? The horses pricked their ears at our excited chatter. We used these curious samples to rehearse our field protocols on the cottage’s back porch.

Our mobile field lab could be reborn daily from two suitcases. We concentrated our sample and extracted the community DNA by dinnertime. The sunlit evening was perfect for setting up our first MinION sequencing run, and the NanoOK analysis tool worked the night shift for us.

Monday at Úthlíd Cottages eased us into Icelandic fieldwork, but Tuesday we left behind the hay fields. We began to see why Iceland is considered a Martian analog; the landscape bears igneous rubble discarded by glacial migration. Hungry for something different, the eye settles on the arêtes, pointed as Giza, but formed by glaciers rather than human labor. Glaciers leaked hundred foot waterfalls down icy peaks, and beneath lay three black sheep, chewing steadily.

We stopped at Kerlingarfjöll to evaluate our main site. Kerlingarfjöll is a stunning study in contrast, when you can see it. Thick fog rolls off around eleven in the morning. Even hiking blind, the contrasts are palpable. We descended from a brisk ridge into a steaming valley. Green snowmelt and blue geothermal precipitate merged. When our hands grew cold, we unwisely warmed them over sulfuric fumes. The moisture turned to frost once we regained elevation. We sampled green biofilms, which were present on the glacial but not the geothermal banks of the stream.

Iceland grew bleaker as we neared Hveravellir, the furthest inland we had time to venture. Mud pots and fumaroles create an oasis; unlike Kerlingarfjöll, Hveravellir has small wading birds that seem to feed on aquatic insects. They forage with their feet in boiling water and their heads in freezing wind. Hveravellir has fumaroles of all ages, from strong new plumes to old seeps. Like volcanic islands, they seem to grow out of precipitated compounds and then erode. We had observed three unique habitats in a day, and we hoped to spend the rest of the week working to comprehend a sliver of their complexity.