The Bowman Lab

Shelly and I completed our last round of sampling at Tent Island yesterday. Normally we get a bit of wind protection from the island but not this time, it was possibly the coldest we’ve been in the field on this trip! Two weeks ago when we sampled ice cores from that spot we saw the first signs of the spring ice algal bloom; a slight brown coloration to the bottoms of the ice cores. By now the algal bloom is in full swing, with a thick mat of algae covering the bottom of each core. The growth is so thick that water flooding out of our drill holes is a soupy brown color, very much like the surface of a stagnant marsh pond (except this water isn’t stagnant of course; fresh, nutrient rich water is circulating in from the Ross Sea).

Everything takes longer on a cold, windy day and we had a couple of extra samples to collect. By the time the helicopter came to pick us up we were exhausted. After a good night’s sleep we’re ready for our final sampling effort – our fourth and hopefully final attempt on Taylor Glacier! Right now the weather looks promising, if it holds for a few more hours we’ll have it…

Whenever we bring samples back from the field we start melting them immediately so that we can filter them as quickly as possible. When I opened our sampling bins last night to start melting the samples I was greeted with a very strong odor that I should have anticipated. The samples, particularly the more algae rich samples from the ice core bottoms, reeked of sulfur. The smell was very much like what you might find at a tidal mud flat, but much stronger. The source of the smell is the same in both cases; a compound with the fancy name dimethylsulfanopropianate, or DMSP for short. DMSP is a small hydrocarbon molecule that also contains the element sulfur. When algae or phytoplankton are stressed (for example when trapped in newly forming ice, or on a rapidly drying mudflat) they release this compound. In sea ice DMSP is though to help the ice algae cope with the rapidly changing salinity caused by ice formation.

DMSP has received a lot of attention in recent years because it has some implications for regional and global climate. Many marine bacteria can use DMSP as a source of food, a process which converts DMSP to the volatile compound dimethylsulfide (DMS). DMS is probably the largest pool of sulfur in the atmosphere with a biological source (there are large industrial inputs, as well as natural inputs that don’t directly involve biology). In the atmosphere DMS converts to sulfuric acid, a very efficient nucleator of the droplets that form clouds (check out this YouTube video for a further explanation of how cloud nucleation works).

Clouds are basically water vapor, and water vapor is a very potent greenhouse gas. Scientists are still unsure of how exactly clouds impact climate, and how clouds might change as climate changes. One idea is that high altitude clouds have a different impact on climate than low altitude clouds. Both reflect and absorb solar energy and energy radiated from the Earth’s surface. Clouds must radiate the energy they absorb, how that energy is radiated is determined by the cloud’s temperature. One implication might be that cold high altitude clouds warm the atmosphere less than warm low altitude clouds. Consider a bloom of ice algae in relatively warm water underlying a cold atmosphere and producing copious quantities of a powerful nucleator. The resulting clouds will be very low in the atmosphere.

We don’t really know how communities of ice algae and ice microbes are changing with a changing climate, nor do we know how communities of phytoplankton and marine bacteria are changing. Without that knowledge we can’t predict how biology will influence cloud formation in the future. Taking it a step further we can see the full circle: climate changes biology, which changes clouds, which impacts climate, which further changes biology. It’s an interesting problem, there are three different interactions here that we know very little about!

The last few days have been fast and furious as we try to wrap up our sampling program. We fly back to New Zealand in exactly one week which means that time is quickly running out! Fortunately we’re getting close to complete. We lucked out on Monday with superb weather for our flight to Beaufort Island. A large network of leads often forms around Beaufort at the interface between the freely drifting pack ice and the ice loosely anchored to the island. The most recent satellite images that we could obtain, from late September, looked promising. Unfortunately when reached Beaufort we couldn’t find open water or young sea ice anywhere. The thinnest ice that we could find was over 50 cm thick, suggesting that the leads had frozen over soon after the satellite images were taken.

This was very disappointing. With time running short we decided to head to nearby Lewis Bay on the north side of Ross Island where we sampled on Saturday. The ice and frost flowers in Lewis Bay are pretty old, but a lot younger than what we found off Beaufort. Enroute to Lewis Bay however we stumbled across a very large, recently frozen lead with ice only 13 cm thick. Perfect! This is thick enough to work on with little risk of breaking through but thin enough that the lead must have been open just days before. Every surface on the lead was coated with frost flowers. They looked a little droopy in the heat (-15 C and sunny!), but good enough for us…

It takes a while to find solid landing sites next to thin sea ice like this; the pilots need a solid 30 inches to land the helicopter. By the time we found a spot and got sorted out for sampling it was getting late in the day, we are grateful for the helicopter crew’s willingness to stay out a little longer to get the job done!

Back home there was little time to rest. We had to re-sterilize our equipment and get turned around quickly for a trip to Taylor Glacier the next day. The weather didn’t look too promising from the McMurdo helicopter pad but we decided go ahead and try for it. We made it across the Sound and all the way up the Taylor Valley before high winds forced us to turn around. It was heartbreaking to look down on our sampling site and not be able to reach it, but the pilots here are an experienced bunch and we have to trust their judgment. Conditions continued to deteriorate as we crossed McMurdo Sound back to the Station. We were within 10 minutes of the helo pad when visibility and a rapidly descending ceiling forced us to turn back around and return to the west side of the Sound. Having just been chased east by bad weather it was a little like being caught in a vise.

This sort of thing happens fairly frequently however, and as a result there’s a way out. A small fueling station and field camp is maintained at a site called Marble Point, at the foot of the Wilson-Piedemont Glacier. Three people staff the fueling station and they’re always ready for visitors. The crew of another helicopter, also trapped on the west side of McMurdo Sound, was already there. We sat down for a large, delicious hot lunch, cookies, and lots of coffee before deciding what to do.

Marble Point is very close to our old sampling site on the Wilson-Piedemont Glacier so we decided to take the opportunity to collect fresh snow from nearby. This is a sample we were planning to collect later in the week. It didn’t take too long (it was probably the easiest sample we’ve collected on the whole trip), so we enjoyed the scenery while we waited for the helicopter to take us back to Marble Point for more waiting. And wait we did. Just as we were reconciling ourselves to spending the night at Marble Point (which with their cozy bunkhouse is not a bad thing) a brief weather window opened around McMurdo. We hurriedly loaded the helicopter and went for it. It was an interesting flight. A full ground blizzard was blowing at the sea ice surface as we flew into McMurdo, it looked like a vast river of snow moving underneath. At altitude the helicopter was riding on waves of air, it felt rather like being on a small boat in ocean swells. The motion was never alarming, just a constant reminder of the energy on the other side of the thin plexiglas windows.

Shortly after we landed the weather window closed. Our flight to Taylor was rescheduled today and canceled. We are now 0 for 3 on Taylor Glacier. Tomorrow we will try to visit our Tent Island site which should be easier than reaching a glacier deep in a windy valley. That leaves Friday and Saturday for renewed attempts at Taylor. After that we need to close up shop and get ready to head home!

The weather finally cooperated with us and we made it out for our reconnaissance flight on Saturday. We even managed to bag a couple of samples along the way. After spending the last 8 weeks on the south side of Ross Island it was very interesting to see the north side. The mountains drop more dramatically into the sea on here, and the ice over the open waters of Ross Sea presents a much more fractured and dynamic surface than we normally see in McMurdo Sound. All together it makes for a much more dramatic backdrop while we work.

We worked our way around Ross Island in the counter-clockwise direction. As soon as we rounded Cope Crozier we started probing the sea ice, looking for ice thick enough to land next to ice thin enough for us to sample. Finding this combination isn’t easy. When we spotted a suitable ice flow that we thought was thick enough to land on the helicopter would set down gently with the pilot keeping the rotors spinning at full power. Working under the spinning blades someone would hop out and quickly drill the ice to see how thick it is.

There’s a pattern to the distribution of ice thickness in both land fast ice and pack ice (the ice not anchored to the land), but it takes a little while to start to see it in a new area. Successive storms, cold snaps, and the underlying pattern of prevailing winds and currents establish a checkerboard of thick and thin flows. After we checked a few flows and found them too thin to power down the aircraft we started finding flows that were thick enough. Once we were dialed in on the thicker floes we were able to piece together a safe path along them to the younger sea ice.

We didn’t find any truly young sea ice. The thinnest ice that we found was 28 cm, probably about two weeks old, though this ice was covered with old, moderately salty (60 ppt, roughly twice the salinity of the ocean) frost flowers. Although the ice was plenty thick to walk across we took the opportunity to test out techniques for working over thin ice. Hopefully we will need these techniques for our next foray. If the weather allows on Monday we will head back out to the north side of Ross Island, this time working to the northwest toward Beaufort Island. From our reconnaissance flight we could see thick pack ice all the way to Beaufort and a glimmer of sunlight off of the lead that seems to persist in a semicircle around it. With luck that lead will be full of very young ice and frost flowers!

The weather’s still playing tricks with us. We were supposed to fly to Taylor Glacier today, high in the McMurdo Dry Valleys, to sample more glacial ice. It looked hazy all morning but flights were still heading out. Our flight was scheduled for noon but deteriorating conditions in the Valleys forced us to wash at the last moment. Trying not to lose another sampling day we hurriedly made alternate plans to revisit the Wilson-Piedemont Glacier to collect snow (an effort that was scheduled for next week). WP is much lower than Taylor Glacier and since we’ve already surveyed a landing site there, much less risky to travel to under marginal conditions. Before we could take off for our alternate target however conditions deteriorated to the point that not even WP was a viable target.

We left our gear at the helicopter hanger in the hope of better conditions tomorrow and trudged back up the hill to the Crary Lab. Shelly and I are both pretty tired today; we worked around the clock to get our WP samples processed prior to collecting at Taylor Glacier. We finished filtering the last WP sample only moments before we were due at the helicopter pad. There’s one good thing about that; with no lab work to do we can enjoy a relaxing afternoon off!

If the weather allows us to fly tomorrow we will try to circumnavigate Ross Island looking for open water where we might find new ice forming. That’s a flight few people get to make and I’m really looking forward to it. If we find promising leads next to solid ice floes we will try to land the helicopter and collect frost flowers. More importantly the flight will serve as a reconnaissance for the culmination of our frost flower collection efforts on Monday, a trip to Beaufort Island (if the weather is good…). Beaufort Island is a distinctive mountain of an Island that juts out from the Ross Sea. It is so remote that it is almost never visited, despite the fact that it contains Adélie and Emperor penguin colonies and an abundance of other wildlife. We can’t set foot on the Island – it’s specially protected – but we can visit a distinctive large lead that forms to the south of it. This lead is present in every springtime satellite image that we can find and will be ideal for collecting frost flowers.

We got lots of frost flowers around Cape Royds earlier in the season (and will hopefully get more around Ross Island tomorrow), so why go through all the effort of collecting frost flowers way out in the Ross Sea? The bacteria that are found in frost flowers originate from seawater, and the microbial communities that inhabit coastal seawater and open ocean seawater are very different. In Barrow, Alaska and from within McMurdo Sound we’ve collected lots of frost flowers enriched with coastal marine bacteria. We’ve yet to look at the microbial community within frost flowers forming over the open ocean and this will be a rare opportunity to do so!

Monday was a spectacular day and we had no trouble getting out to collect samples on the Wilson-Piedemont Glacier. WP is a broad coastal glacier at the northwestern most extent of McMurdo Sound. Because the wind often blows from the southeast the WP is ideally located to collect small particles eroded from the surface of sea ice in the Sound. WP, like other glaciers, is almost entirely fresh. Researchers have shown that glaciers do contain small amounts of sea salts however, and there is further evidence that some of these salts may come from frost flowers. Glacial sea salts are just a little bit depleted in sulfate relative to the other ions, suggesting that the salt must have come from ice so cold that it contained no sodium sulfate (see “Pass the salt” for a further explanation). Frost flowers, saline snow, and other similar cold salty features at the sea ice surface seem like a logical source for this salt.

Our task at WP was to collect enough ice to determine whether bacteria and bacterial genes present within frost flowers might also be present within glacial ice. There aren’t a lot of bacteria within glacial ice – maybe 1,000 bacteria per milliliter compared to 100 times that many for sea ice – so we have to collect a lot of it. It took two helicopter flights to get us and all of our sampling containers to and from the glacier. We didn’t count the number of ice cores we pulled from the glacier surface but we later estimated that it took around 60 to fill 12 of our 20 gallon sample containers. Loaded each container weighs around 90 lbs, getting them back on the helicopter for the flight out was no easy chore!

The dependence on helicopter flights and large sample volumes meant that we had very little time for exploration. This was too bad as the view from WP is truly spectacular. Our site was located next to the Bay of Sails, so named because large icebergs calving from the Ross Ice Shelf on the other side of Ross Island are frequently trapped in an eddy and driven aground in the bay. It looks like someone anchored a fleet of icebergs here, and the more distant ones really do look like sails. It isn’t difficult to imagine the frustrations of early explorers anxiously awaiting the arrival of a relief ship and trying to convince themselves of what was and wasn’t an iceberg.

Fortunately we will be visiting WP one more time, to collect fresh snow drifted near the edge of the glacier. In the meantime we’ll be working hard to get caught up on labwork. We have a flight to Taylor Glacier on Friday and over a thousand pounds of ice to melt and filter before then!

The helicopters started flying last Thursday and with their support we’re getting back on track with sampling. Friday we made it back out to Tent Island for our third sample there. It was great to make the trip in less than 10 minutes after all the hours we’ve spent driving that route! The only downside to this mode of transportation is the need for good weather. We had ambitious plans to fly to the Wilson-Piedemont glacier on the far side of McMurdo Sound Saturday for our first glacial ice sample, but fog grounded the helos for the day. Landing on glaciers is tricky business; the pilots are only willing to attempt it with good visibility. Tomorrow we’ll try again, hoping to net two large samples from a single trip. Time is running short – we leave November 2 – so we need to maximize every trip out!

The helicopter crews take Sundays off so we couldn’t fly yesterday. It was a warm (for Antarctica), sunny day so we decided to take a snowmobile trip to Turtle Rock, a little island tucked into a back pocket of McMurdo Sound. Turtle Rock is an interesting feature because it guards a bay full of sea ice which has not melted in many years. We call this ice “multiyear ice”. Until last year most of the ice in the south end of McMurdo Sound was multiyear ice, but a massive breakup of the ice last spring removed almost all of it. The ice that we have been working on so far is first year ice that only formed this winter.

We’ve done a little work on the microbiology of multiyear ice in the Arctic but a lot of questions remain. When sea ice forms in the fall it traps lots of bacteria from the seawater. These organisms aren’t adapted to life within the sea ice, but they persist there through the winter. When spring comes and the ice algae begin photosynthesizing these seawater bacteria are outcompeted by specialized sea ice bacteria that metabolize well under low temperatures and with a lot of organic carbon (produced by the ice algae). If sea ice survives that first summer to become multiyear ice however, what happens to those specialized sea ice bacteria? The ice algae use up the nitrogen, phosphorous, and other nutrients present in sea ice the first summer and can’t maintain the same level of photosynthesis in later years. The sea ice bacteria (which can perform better with little organic carbon) are long gone from the sea ice matrix. Does biological activity slowly grind to a halt, or is there a succession to a new microbial community adapted to low temperatures and low levels of nutrients?

Before we can answer these questions we need to acquire some cores of multiyear ice, something easier said than done. Reaching the multiyear ice behind Turtle Rock was easy enough but coring it proved quite difficult. First we had to reach the ice surface. It’s not uncommon to find a couple feet of hard, wind packed snow on top of the ice you want to core. In this case we had 7 feet of snow to tunnel through! Shelly and I cleared ourselves a pit large enough to work in and started coring. With limited space maneuvering the drilling equipment proved challenging, we were constantly climbing in and out of the pit to help lift the corer in and out of the hole. Multiyear ice can be 5 meters thick or more, after only two meters we had to quit for the day. With luck, time, and a couple of intrepid volunteers we might be able to finish that sampling effort before we leave. In the meantime we have more important objectives… collect our critical glacial ice samples and then start using the helicopters to access the ice edge for more frost flowers!

The dive team showed us this video that they took recently. It illustrates a couple of really interesting aspects of the relationship between sea ice and seawater. The stalactite-like projection extending down from the bottom of the sea ice is called a brine tube. Brine tubes form when some physical process causes very salty water to drain from porous ice. Since this brine is below the freezing point for relatively fresh seawater when the two come in contact the seawater freezes, forming a tube around the brine. You can still see the brine draining from the bottom of the tube. Anecdotal evidence suggests that these brine tubes are more active on sunny days suggesting that the ice warms up enough on these days to change its porosity. The ice where this video was taken is over 4 feet thick, only the bottom few inches likely warmed enough to drain. A rule of thumb exists for when the ice becomes so porous that the pores actually connect (like a block of very holey Swish cheese), called the “rule of 5s”. For ice that contains .5 % salt, at -5 C, 5 % of the total volume of water contained in the ice is liquid. This is a sufficient volume for the pores to connect and the brine to potentially drain from the ice interior.

This flushing has important implications for sea ice biology and the larger ecosystem. The brine flows out of the sea ice because it is denser than the underlying seawater that replaces it in the sea ice matrix. By now the ice algae have been active for some time and have probably depleted many of the nutrients within the sea ice brine. The incoming seawater replenishes these and allows primary production to continue. Recent research shows that the sea ice algae are not passive players in this process; they surround themselves with a gelatinous mixture of sugars and proteins that alters the physical structure of the ice. Among other things this enhances the porosity of the ice, encouraging an exchange of water.

Brine drainage might even effect have an impact on global climate. In an earlier post I talked about how specific salts form during the sea ice formation process (see “Pass the Salt”). One of the first salts to form, at a relatively warm temperature, is a form of calcium carbonate known as calcite. This is the same mineral that gives coral reefs their structure. It has been suggested that the flushing of sea ice “pumps” calcite into the water column. This pump is in essence a transfer of carbon from the atmosphere (where it exists as carbon dioxide) to the water column.

There’s a lot that we don’t know about brine drainage from sea ice. Here we’re observing it during the spring, but what about in the fall when sea ice is forming? A tremendous amount of salt is rejected from forming sea ice but due to logistical challenges researchers are rarely around to observe it (there have been some notable exceptions). And what about the sea ice bacteria that we study? Are they rejected along with the salt, or do they have some trick to help them stay in the ice?

Here’s a short YouTube video of the penguins we saw yesterday.  The sound comes on about halfway through for Shelly’s highly effective penguin call.

For the last couple of weeks we’ve been battling worsening sea ice conditions in McMurdo Sound.  Cracks between the different pans of land fast ice keep growing wider, and though they freeze quickly it can take weeks for the ice to be thick enough to travel over by McMurdo Station rules.  One particularly troublesome crack extends west from the Erebus Glacier Tongue effectively blocking all travel north of the tongue.  All throughout September we were able to travel over the crack safely, but it has widened significantly over the last two weeks.  Since our Tent Island sampling site is on the other side of the crack this presents a bit of a problem for us.

Today Shelly and I went out with Jen and another field safety staff member named Susan.  We thought that be traveling far out to the west, toward the center of McMurdo Sound, we’d be able to do an end-run around the crack.  We put a lot of miles on the snowmobiles heading west without the crack showing any sign of tapering off!  Unable to run around the end we measured the thickness of new ice over the crack in a few spots to try and find a safe crossing.  This was tough work.  It’s a wide crack and the irregular edges of the ice here had collected large rock-hard snowdrifts that had to be dug down to the ice surface.  When snow drifts and solidifies over a hollow space (like a crevasse or open water) it’s called a snow bridge.  Snow bridges and can be remarkably strong.  We had no doubt that the snow bridges and ice underneath would have supported our equipment and more, but the science teams have been issued strict instructions not to deviate from station rules.  Oh well!

As a consolation prize we were visited by a group of Emperor penguins at the farthest point west that we reached.  In contrast to the group that we saw a week ago, which was hunkered down for a storm, this group was quite active.  They spotted us from a ways away and, being naturally curious, came right on over.  We took a mandatory penguin-watching break and sat down to see what they would do.  It was a funny exchange; they got remarkably close to us and then stood and stared (while we stared back).  After about 45 minutes, cold and mindful of the fact that we still had a lot of work left to do, we finally get up and started getting ready to go.  Satisfied, all but one of the penguins wandered off to find more interesting things on the ice.  The remaining bird followed us over to our snowmobiles and kept staring even after we started them.  I think he (she?) would have taken a ride had it been offered.

Tomorrow Shelly and I will head back out to try and get some ice cores.  Since we can’t get to Tent Island we will have to try and find a site on this side of the crack that mimics our old site as close as possible.  That’s far less than ideal, any number of differences might exist between ice collected at the two sites, but that’s how field work goes sometimes!  It often seems that it goes that way most of the time…

I had a great conversation this morning with a group of oceanography students from Soap Lake High School in Washington (thanks to teacher Matt Brewer for facilitating). They’d done their homework and asked some really good questions. Among other things they asked me to draw some comparisons between Soap Lake (just down the road from their school) and the environments we are studying in Antarctica. The comparison is interesting enough that it’s worth expanding upon here.

Soap Lake isn’t a normal lake, it’s one of several saline lakes in the eastern part of Washington state. In fact Soap Lake falls into a special category of salt lakes; it’s an alkaline or “soda” lake. This means it has an unusually high pH brought on by the presence of certain minerals (primarily carbonate). Other famous soda lakes include Mono Lake in California.

As I’ve mentioned before in this blog frost flowers, and anything else at the sea ice surface, tend to be very salt. So how similar are these two salty environments? If we based our answer purely on the amount of salt present they are pretty similar. Frost flowers contain up to 15 % salt (ocean water is 3.5 %). The most saline water in Soap Lake, which forms a permanent layer at the bottom of the lake, is also close to 15 %. But quantity of salt isn’t everything. The composition of the salt probably has a bigger impact on these ecosystems.

Frost flowers are basically frozen seawater but they don’t contain the same normal ions, or charged elements, found in seawater; sodium, chloride, magnesium, sulfate, and carbonate among others. Mix a bunch of different ions together in water and evaporate the water, the ions will pair off to form the salt left in the bottom of the container. For example sodium might pair off with chloride to form halite, or table salt. A similar thing happens when water freezes. The ions will pair off at specific temperatures, the resulting salt will leave the water and settle to the bottom. In seawater calcium joins up with carbonate to form the salt calcite at only -2.3 C, not long after freezing started. As the temperature keeps going down other salts are lost. Sodium comes together with sulfate to form the salt mirabilite at -7.6 C. So before frost flowers even get a chance to form they are already deficient in calcium, sulfate, and carbonate (there is so much sodium present in seawater that little is lost during the production of mirabilite).

Soap Lake has an excess of carbonate, frost flowers are depleted in it. This is a critical difference. No one has yet come up with a way to measure the pH on the miniscule amounts of water in frost flowers, if we could we might find that these brines are more acidic than seawater (pH of 8.6) for the same reason that Soap Lake is more basic. This difference in pH could have a substantial impact on biology in the ice.

There are however, some Antarctic environments that could be very much like Soap Lake. In the nearby McMurdo Dry Valleys (which we are scheduled to visit toward the middle of the month) there are many saline lakes including Don Juan Pond, possibly the most saline lake in the world*. These lakes aren’t alkaline lakes. Instead of carbonate salts these lakes are dominated by the bizarre calcium chloride mineral antarcticite. They do share one important feature with Soap Lake however; high concentrations of sulfur in a layer of deep water that never comes to the surface. We don’t yet know a whole lot about the microbiology of these lakes, but these two things – a high concentration of sulfur and a permanent layer of deep water – have produced an interesting microbial ecosystem in Soap Lake. This ecosystem includes an incredibly dense community of bacteria capable of making their own organic carbon from carbon dioxide and found, as of yet, in no other lake on Earth (see article here). Of course studying Soap Lake and similar exotic bodies of water in our own backyards yields a wealth of information and techniques that make it possible to understand the ecology of lakes in the Dry Valleys and other places where access is limited.

*By the measure of water activity, the most biologically useful measure for how salty something is. Antarcticite is extremely hygroscopic, meaning it has a high affinity for water. As a result adding a few grams of antarcticite to a test tube containing cells and water would have a much bigger effect than adding a few grams of sodium chloride, as the cells would not be able to “compete” effectively for the water.