20 October 2009

2009-10-20: Nitrogen sudoku


Not a very eventful day - I was asleep for most of it! This jet lag is quite frustrating. What little of the afternoon remained I spent at the Internet cafe, and then as soon as I got home to make dinner, the power went out. Once it returned, I spent the rest of the evening on my computer. My most important task was reassigning the plots in MZ12 to include a second drought treatment and subtract a nitrogen treatment.

Main text:

I awoke at 8:30, but the idea of getting up was abhorrent; I hadn't been able to fall asleep until 5 AM. "I really should... go to Makoka...", I said to myself, and fell asleep again.

I half-awoke several more times, but each time my body responded that it was still very tired and needed more rest. Finally, my slumber ran its natural course and I sat up, wide awake. I looked at my watch.

"AAAAHHHHHH! It's almost five o'clock in the afternoon!"

You can see why this is a short blog entry.

Feeling discouraged and embarrassed, I got dressed and washed my face. At least I could go to the Internet cafe in the remaining hour before it closed.

I can't explain why I am still mired in jet lag after being in Malawi for a week. Usually I have no trouble with jet lag. The first couple of days I thought I was fine. But after that I just wasn't able to sleep at night. Well, I guess I need to be really disciplined with myself, and make sure I am appropriately tired at the appropriate time. Tonight that is going to be a challenge.

I spent an hour at All Seasons Internet Cafe, then came home and started cooking some potatoes and carrots to put in soup. I had just added the carrots to the frying pan when - blackness!

Oh, bother. No electricity = no soup for dinner. At least the potatoes were cooked; I could eat the carrots raw. That, plus a cup of Milo (from water I'd boiled before the power went out) and a piece of bread, was dinner.

* * * * *

As I ate and worked on my computer by candlelight, I thought about how power outages were a part of life in Zomba, which reminded me that the large majority of Malawi's population (and a large fraction of the world's population) have no have electricity in their homes at all. It is easy to forget about this when living in the industrialised world; even brief power outages are usually rare. For most Americans, electricity is like air or gravity; it's just there. But for most of the world's population, it is not.

I decided I didn't want my kids to grow up taking electricity for granted. Maybe, I thought, once a year we can have an electricity-free day. No lights, no TV, no computers. No washer, dryer, or hot water heater. No fridge, no freezer, oven, toaster, or microwave. No charging cell phones or iPods. It really gives you a new perspective on the world!

I'm sure we'll be the envy of the neighbourhood. The newspaper headline will read: Fanatical mother forbids electricity, endangers health and welfare of family.

* * * * *

The power came back on before 8 PM. An electricity-free evening is much more tolerable than an electricity-free day.

"All right," I said to myself, "left to your own devices, you would stay up till dawn again. But no! Bedtime is 1 AM. Or else."

I spent most of the ensuing hours doing non-research-related writing, but saved an hour at the night's end to figure out the plot reassignments at MZ12. As I described previously, both Festus and Sileshi independently suggested that I expand the drought treatment to include two different regimes (early- and mid-season drought). To make space for this extra treatment, I would eliminate the 50% nitrogen treatment (focusing just on the two extremes of 0% and 100%).

But I hadn't yet figured out how that was actually going to work. Where should I build the six new rain shelters? I had a feeling that there was no perfect solution, given all the constraints. Fortunately, MZ12 has many more plots than I am actually using, because the experiment was originally designed to incoporate different pruning regimes and phosphorus application rates, neither of which is a variable in my experiment. Unfortunately, I have to consider the following problems:

  • I can't just reassign nitrogen treatments at will; there is undoubtedly a strong residual effect of cumulative nitrogen application.

  • Although I don't expect a residual effect of phosphorus, I would nonetheless like to make all my treatments identical with respect to their former phosphorus regime, just in case.

  • In my experience, the 50% N plots are probably fairly similar to the 100% N plots, since the fertiliser is measured with bottlecaps, and the bottlecaps that are supposed to be "half-full" are in practice more than half full. So I could potentially reassign "50%" plots to "100%."

  • Plots with rain shelters should not be located next to other plots (whether with rain shelters or not), because of the potential effects on precipitation and microclimate.

  • Some of the plots in Rep 3 are always stunted and low-yielding (and are missing many of their Gliricidia trees) for reasons that seem to have nothing to do with the actual treatments. I tried to avoid those plots in my initial plot assignments.

So, really, the whole problem becomes a giant brainteaser. Nitrogen sudoku.

I puzzled over it for awhile, and as I expected, it seemed impossible to satisfy all these constraints. The best solution I could come up with involved (in two cases) putting rain shelters next to other rain shelters. It also required inequality in the former phosphorus treatments, and a slight but noteworthy inequality in which 50% N plots were reassigned to 100% N. I think none of these problems are too serious as long as I accurately record and compensate for them, and the experimental design will still be OK overall.

This is one of the aspects of my project on which I don't expect to get much input from my advisors. I'm the only one who understands all of these nitty-gritty details at present, and no one else has time to get fully acquainted with them. I suppose that's how a lot of science is carried out, for better or worse: on the basis of single-handed compromises and assumptions by ignorant grad students....

All right, I should sign off, it's 1:30 AM. I'm a little past my designated bedtime, but not too bad. Maybe this will be the night I finally beat jet lag.

Chichewa word of the day: sivuto = no problem, don't mention it


No comments:

Post a Comment