Siliceous and Delicious

Monday

This morning’s schedule called for a return to the diatom diversity project, at long last. I had a hard time getting started, largely because I’m not sure what to do next. I’ve replicated the most basic of Rabosky’s results, and am on track to replicate the rest of what he’s done. At this stage, should I continue stepping through his code and replicating the results he obtained? Or should I already be working on my own take on the issue? The former is easier, and more comfortable—I know how to proceed along that line. But what do I start doing beyond his work? There was such a flood of ideas and approaches last summer and early fall, which was really the last time I fully and actively engaged with that aspect of the project, that I don’t feel that I have the overview. I suppose I am afraid of continuing my efforts to replicate Rabosky’s results lest I be doing that only as a way to avoid tackling the difficult question of what I am to do next. Which is certainly true. But there is also a strong argument to be made that I have made tremendous progress in understanding Rabosky’s code and have gotten very close to fully replicating his results—such that it would be a shame to stop now, where things are getting interesting. I will blaze ahead with his code, then. You never know what I’ll encounter along the way that might guide my own investigations from there.

Minor clerical observation: it seems that over the weeks before xmas I had the most success working on the R project from Darwin’s—and this morning I struggled to get into the groove precisely because of the sort of distractions that are absent there. Perhaps I should transplant myself to Darwin’s on those mornings or afternoons when I am scheduled to work on the diversity project to optimize my chances of success…

Achievement #1 of the day: put in an order for SEM supplies (stubs, sample containers, carbon tape), which Diane promised to submit “today”. Achievement #2: turned down a request to give a research talk to the Austrian Scientist and Scholars in North America Association. It still makes the hair on the back of my neck stand upright to tell someone to fuck off (in not so many words), but knowing that I’ve just saved myself countless hours of stress and frustration is really more deeply satisfying that I would have thought. We are all the masters of our own destiny. Or something.

The next conundrum to tackle is the silicon isotope project and Dave’s suggestion to start working with different material and with a different expert. I downloaded the paper by Takahashi he had suggested and took a look at it—it’s an impressive data set about the horizontal and vertical distribution of a couple of hundred different radiolarian species from transects of the Equatorial Pacific (west of Papua-New Guinea)… But I’m still not sure how to go ahead.

Spent some time on the microscope in the afternoon, picking through more of the Ohio material to see if I could find any radiolarians. Found what once again looked like exciting radiolarian specimens, but I will have to see what they look like under the microscope before I get too excited. Slightly improved my picking technique—it was getting very annoying picking up a radiolarian on the brush, moving the petri dish away, putting the SEM stub under the scope, finding the brush, and then transferring the specimen… I ended up using a multi-well and transferring the rads to water in one of the wells first, then to the stub in a second step when I had adjusted the stub so it would be in the same plane of focus as the bottom of the well. Kind of worked. Perhaps I can take a look at them under the SEM tomorrow morning!

Tuesday

Attempted to improve upon last week’s first attempt at staining diatom frustules with fluorescein. This time, dissolve much less fluorescein in methanol, but put the dry diatom sample directly in the methanol (rather than in water). The major challenge is what to do next. I placed an 11 µm filter on a vacuum plate and dropped some of the suspended sample onto it, then gently squirted DI water onto the filter while the pump was still running, in an attempt to clean excess fluorescein off the sample and filter. I then put the filter in a falcon tube with DI water and shook it to try to get the diatoms back off the filter paper and into suspension. Unfortunately I could not see any diatoms at all under the microscope when I made a smear slide of the resulting sample. Either I just didn’t take enough sample onto the slide (possible), or the diatoms are being washed off by the DI squirt bottle, or I simply didn’t put enough diatoms on the filter in the first place. More work is needed here, but it’s progress. I am starting to figure out how (and how not) to do this.

In the afternoon, I thought I’d take a look at the radiolarians I picked yesterday, so I went across the street and sputter coated the stub I had prepared. Alas, the EVO did not give me any joy and I could not get a focused image nor find anyone who would help me do so. A disappointing afternoon!

Wednesday

A day off to play in the snow—the geobiology group’s first ever ski trip, to Wachusset mountain. It felt incredibly good to come home completely exhausted and with glowing cheeks from a day of exertion in the cold, fresh air.

telemark

Thursday

Perhaps as a result of yesterday’s vigorous physical activity, a slow start to the day. Constructed the pick-o-matic, a 6-well cell culture cluster with a sawed-off SEM stub holder in one of the wells. I am hoping this will make my picking somewhat easier. We’ll see. I then talked to Ben Gill about how to switch between different acids. He explained that you could decant the HCl and then pour nitric acid on the sample, and that this would also help oxidize some of the organic matter. He mentioned that it was important, however, to rinse the sample in between (i.e. add DI water, let settle, and decant again) since HCl and nitric acid combine to form aqua regia, a substance so deeply nasty that it will dissolve even gold (!). Unfortunately the fume hood was being used by an undergraduate minion working on Phoebe’s samples, so I tried to do this decanting business on the lab bench. This turned out to be a bad idea as I soon felt the inside of my nose burning furiously, as well as feeling suddenly faint… I quickly aborted this mission and resigned myself to finishing the task once the scary fume hood is available for use again.

This was the case after lunch, and I continued the task at hand under the hood. The hood (and the whole lab, for that matter) scares the shit out of me. There’s no knowing what horrendous chemicals have infused all of the surfaces, switches, and floors, what bone-dissolving substances have seeped into all of the tools and beakers and containers, nor what long-expired, biohazardous or sharp-edged object will fall out of the cabinet you open in search of something. I never, ever touch a thing in there without wearing gloves, and I never set foot in it without wearing shoes and a lab coat. And still, every time I’m in there, I start to burn and itch all over, I start to feel dizzy and nauseated, and generally get the heeby-jeebies. If only Francis’ lab could be built faster and I could start working there… this is too much like gardening on a superfund site.

Anyhow, it took me an incredibly long time to do it, but I washed the four samples I had gotten from Ben Gill that had been steadfastly refusing to dissolve in HCl. I rinsed each one in DI water countless times, filling up two gallon-jugs with waste HCl + water. After about an hour and a half, I began to wonder whether there was something wrong with the pH indicator strips I was using—they had been registering a pH of about 4 for several rinses, and the pH didn’t seem to be rising anymore. I tested the DI water, and according to the strips that was at pH 4 as well. Clearly the strips were not performing right, and I decided to call the samples successfully rinsed (and made a note to buy new pH strips). Added full strength nitric acid (15N), just enough to cover the samples, and called the labwork done for the day. There was no fizz, which I had expected given my experience with HCl, but there was some ominous smoky vapor coming off the sample, which I interpreted as indication there was some sort of reaction going on, hopefully leading to the dissolution of these pesky rocks. We’ll see. I certainly need to register this as a success in that I’ve learned something about how to switch between two acids today, but I have also learned that there are at least two downsides to this (besides the terrifying environment in which I need to do this): firstly, it takes a very long time, and second, it produces a lot of bottled waste (which I presume we eventually pay to be disposed, although the amount of waste that’s sitting in the “satellite waste accumulation area” is nothing short of horrifying).

Friday

Spent the morning checking to see if there weren’t, after all, some diatoms to be found in the product of my second fluorescein-dying attempt. I made another smear slide, but this time using a much larger drop of water than before (covering most of the slide), and found two or three diatom fragments. This was encouraging enough for me to try to make some proper slide mounts so I could see under the CLSM whether the staining had worked using so much less fluorescein. I made (for the first time) slide mounts, using Epotek EP-301 epoxy. This took a long time but worked fairly well, considering it was my first shot. The epoxy apparently hardens when exposed to heat, so in one of the slides I made it had hardened before I put the coverslip on. Another had far too many bubbles, but a third one was more or less OK. In any case, much like my first try, none of the slides contained abundant diatom material and at best fragments, but this was good enough to check to see whether the staining had worked.

The observations that there was (a) only broken diatom material, and (b) very little material indeed in the products of my staining process are odd and will need further attention. Are diatoms getting sucked and embedded into the 11-µm filter and not being released when the filter is placed in the vial? Are they being washed away during the DI water rinse? I doubt that there are too few diatoms to begin with, since I know the ‘one-pipette-tip-in-a-couple-of-mils-of-water’ yields dozens to hundreds of specimens per drop from my handful of SEM experiences.

In any case, I chose to break with the schedule and (after wolfing down an early lunch) used my scheduled lunch hour and email hour to look at the morning’s slides under the CLSM. In a nutshell, the fluorescein staining appeared to have worked fairly well, though (a) the fluorescence appeared to be weaker than before, and (b), it appeared to be patchy—not all parts of a diatom fragment seemed to be uniformly stained. All in all, this leads me to want to up the fluorescein dose quite a bit, though perhaps not quite to the levels I used the last time. If I can get stronger and more even fluorescence, without inundating the final slide with background noise, that would be ideal.

Some examples of the sorts of images I obtained today, mostly showing the patchy nature of the staining and the fragmentary material:

I also discovered a small radiolarian:

And, after having acquired my images, figured out how to make a video showing the Z-stacks I had taken in a 3-D-ish movie. These, I began to understand today, are not ‘3-D reconstructions’ in the sense of what the Amira software does, but rather just the stack of the 2-D images placed atop one another, using some defined opacity characteristics, and viewed from some particular angle (in this case a rotation through a whole 360˚ worth of angles). It was very quick to do this—my understanding is that a real 3-D reconstruction, which would clearly define zones of material in 3-D space (rather than a raster-cloud of variously-valued color points), takes much, much longer.

Rotating Diatom Rotating Radiolarian

An interesting observation from the radiolarian specimen, which could potentially limit the applicability of this technique quite dramatically, is that the fluorescein seems to have concentrated around the pores, and in some areas looks like it may actually be occluding or being lodged in the pores. This is basically giving precisely the opposite effect of what is desired—it is highlighting where material (silica) is not, rather than where it is located.

In summary, it’s still not perfect, but I am learning more about CLSM and sample preparation as I go along. Writing up the days’ work so far has taken up my scheduled time for the diatom diversity project, which I still haven’t revisited since the holidays. But then again, it’s only two weeks since I’ve been back at work, and I think I’ve made great progress since then.

For the remainder of the afternoon I procrastinated over replying to Dave and writing to Kozo Takahashi about my silicon isotope idea. For the first time since the ‘miraculous transformation’ of the new year and its positive attitude, I feel the spectre of unwillingess return. It’s because I’m somewhat stuck on what to do and what to say, exactly. I know what I want—I want to brazenly push ahead and feed radiolarians and diatoms from the Benguela into Rama’s mass spec setup to see what δSi values come out. And I know there are many caveats. But this is a tangible, doable sized project. And Dave’s conception of the all-encompassing modern-system study just doesn’t appeal to me. It has the flavor of the built-in failure of the overly ambitious projects I’ve attempted in the past.