Some random thoughts on dissertation topics

It's near the end of the semester, and I've been cranking away at term papers. If I'm not focusing on that delightful subject, I'm usually muddling through wedding planning.

Well, I'm tired of that. It's time to talk dissertation topics, bitches. You knew this was coming. No going back now. (Alright so maybe there really is no built up demand to discuss dissertations, but I needed a way to introduce this blog)

My subject area will most likely be in the Great Plains; probably centered on Nebraska. There are lots of research questions lying out there in the grass. I'm most interested in 1) paleoenvironments; 2) silica biogeochemical cycles; 3) vegetation dynamics; 4) developing new methodologies to improve paleoenvironmental reconstructions using phytoliths (but proxies in general, I suppose). Let's break these down one by one.

1) Paleoenvironments. It's always good to know what was happening in the past. There are many methods out there to determine this. I'm interested in what the vegetation of a given area was doing. If the vegetation can be determined, it can be used as a paleoclimatic proxy (for example, if I find tundra vegetation, it probably means there was a tundra climate there). For the Great Plains, it would be fairly easy to do. Phytoliths are abundant in grasslands, and they're stored in loess sections. Loess sections build up through time, so the deeper a person digs in the loess, the older the phytoliths are. So it is possible to dig out 10,000 year old phytoliths and look at them. Performing a simple phytolith analysis would be the simplest route of the four options listed. I'd probably expand on this to include some of the methods I used as a Masters student. But there is a problem with this. Just using standard methods really isn't good enough for a dissertation. I'd need to have some really interesting results, and then expand on that.

2) Silica biogeochemical cycles. This topic has interested me of late. Many papers have shown the importance of Si to the global carbon biogeochemical cycle. This would be fairly straightforward -- just quantify the flux of Si from the soil to the terrestrial biomass, and back again. It's actually kind of boring when you think about it, but it could have huge ramifications. For example, suppose during the LGM that there was very little Si emplacement into soils due to mineral weathering. This would lead to a dearth of Si export to watersheds and on to the world oceans. In the oceans is where Si does its part to sequester carbon out of the air and deposit it into sediments. Would this mean that carbon would build up in the atmosphere? On the other hand, there would be considerably less global biomass during the LGM. This would mean less Si entrained in vegetation and consequently more Si just sitting around in the soil, where it would be prone to leaching. This leaching could export large amounts of Si to the world oceans, where it could act to sequester very large amounts of carbon. This would be a positive feedback loop - cold temperatures from the ice age would ultimately lead to an even greater reduction of atmospheric CO2. Paleofluxes of Si would be a great topic to tackle. It would be a bit difficult to do, but I don't think it's beyond the realm of possibility. I'd need some way to determine how much Si was around in the soils back then. During the LGM, Nebraska presumably would have looked quite a bit like central or northern Canada. I'd need to look at a modern analog for these soils -- in Canada. This would tell me what the Si flux in these soils is now. Once I know that, I could infer that Nebraska soils during the LGM must have been similar. From that, I could estimate how much Si is being produced from mineral weathering, how much is being taken up by the biomass, and how much is being exported to the watershed. I think it would be way too ambitious to do this for the entire globe, so I think I'd stick to the Great Plains, or a section of it. Another way might be to look at the Germaniun-Silica isotope ratio to determine the chemical weathering rate.

3) Vegetation dynamics. There's many routes I could follow on this one. I could stay modern, and just see how the Great Plains vegetation is responding to climate change. Or I could look at changes in vegetation through time. This ties in with #1.

4) New methodologies. This has the most appeal. There are still many uncertainties in phytolith analysis. Namely, how does Si dissolution affect phytolith preservation? In my Masters work, I saw progressively less phytoliths as the age increased. This makes sense, since one would expect the odds are against a given phytolith of being preserved as time wears on. Si is especially reactive, so it's really quite amazing to think that any of these things last more than a few years in the soil. The preservation of phytoliths in grasslands is a testament to the aridity of the region. Many studies have shown that upon deposition, a large percentage of phytoliths dissolve. That makes sense, since they're just sitting at the top of the soil getting rained on. As the phytoliths are buried, they stand a better chance of surviving. But even then, groundwater percolation will slowly and steadily dissolve the phytoliths. I just wonder if there is some way to estimate the amount of phytolith dissolution through time. Of course, it would only be a rough estimate.

Looking at this in a reverse way, it might be possible to determine phytolith production for a given time period. This summer I'll be heading off to Nebraska to determine just this. I'll take a look at the amount of phytoliths present during different climatic events. The question I'm asking is, are there more phytoliths during cold or warm periods? A person would expect to find more phytoliths during warm periods, simply because there would be more vegetation making phytoliths. But this may not be the case, as dissolution may be more of a factor. Thus, cold times may actually be over-represented int the phytolith record, simply because there is little dissolution. Once I know how much phytoliths are present for each climatic period, I'd need to find a way to estimate the vegetation density. If this is known, I could compare it to the actual amount of phytoliths present -- thus deducing the dissolution rate! I couldn't use the phytolith proportions to infer vegetation density, as that would be circular. Another way might be to use other methods, such as stable isotopes, macrofossils, pollen analysis, or even terrestrial diatoms.

Other problem areas in phytolith analysis include bioturbation and other forms of transport. This means simply that phytoliths may not stay put upon deposition. They can move around quite a bit. I really don't see any way around this, except if there were a way to identify the ages of single phytolith grains. This is way beyond our technology at this point. As with dissolution, it would be possible to estimate phytolith movement based upon the amount of contemporary bioturbation, and ground water percolation. One would need to revise these numbers based upon climate change in the past. For example, it was much colder 15,000 years ago. How will this affect phytolith transport?