Ad infinitum

I have not touched my project since mid-July due to other pursuits, but I have been asked every day this week what I did with my summer. I will post, more or less, what I tell people who ask me that.


Biology professors at William and Mary have discovered that mercury, put into the South River in the Shenandoah Valley of Virginia by a coal-fired power plant, reduces the probability that hatched eggs survive to leave the nest (“fledge”). They noted that in at least one species of bird, there is a basis for genetic adaptation to mercury; that is, there is at least one gene that can help hatchlings fledge in the presence of mercury. The problem, however, is that when these tolerant birds are not exposed to mercury, their hatchlings also have trouble fledging. In short, “normal” birds are reproductively impaired in the presence of mercury, and “tolerant” birds can be reproductively impaired in the absence of mercury. Since birds, depending on their species and location, can be very mobile creatures, we are curious about how these genes can move across a heterogeneous landscape where some habitats are mercury-contaminated and some are not. A potential question is, under what conditions does the contaminated environment’s population become entirely tolerant? Another question is, how much can emigration from the contaminated site lower the overall fitness of uncontaminated sites, introducing what is called in evolutionary biology a “migration load”

To solve this question, I (with a lot of help from my advisor, Drew LaMar) wrote a mathematical model that allows us to simulate the above. I then coded this up in the statistics-processing language R. The code allows the user of the program to enter a list of population sizes, while indicating which of these are contaminated and which are not. The model keeps track of the size of these populations over time; each population has two stage classes, one of adults and one of newborns which, after a year, mature into adults. There are a fixed number of breeding territories in each population, and once the number of birds surpasses the number of breeding territories, a “floater” class appears of non-reproductive birds which will compete for territories in the future. In addition to tracking the number of birds, the model keeps track of allele frequencies (i.e. the percentage of tolerant and intolerant individuals) in each population, so that we can monitor the prevalence of tolerant/intolerant individuals. The final task of the summer was to write a script to handle different floater movement patterns — how far can a floater move in a single year, and where can they compete for territories?

The next task after finishing writing different floater movement equations is to apply different landscape configurations into the simulation; there could be several small patches that are accessible to all, or small patches placed in a line such that one can only migrate into a patch from adjacent patches. Obviously, the layout of these islands of birds has the potential to affect the spread of different alleles throughout the metapopulation, but will it? There are also several other assumptions about the effects of mercury contamination we can test — what if mercury exposure lowers reproductive output by 20%? 30%? 50%? What if it affects the two allele copies differently?

This summer was well spent — it allowed me to lay the groundwork for what will hopefully be an interesting thesis related to landscape genetics, population dynamics and ecological modeling. I intend to continue this work throughout the upcoming school year.