About Darkling

I’ve called myself a population biologist, so what does that mean?
Several years back I was out tramping (hiking for Nth Americans) with a group of people. I knew a few out of the dozen or so people on the trip, which was a good mix of internationals and kiwis. At any point where I started to explain what I did, one of the people I did know would attempt to say that I was doing a math degree. This annoyed me (which is probably why she brought it up every time I started to elaborate), since I’ve always viewed myself as a biologist.

So what was I doing that got me accused of being a math PhD? I was carrying out an experimental study using feral house mice. I had established eight trapping grids on the Otago Peninsula, where I was running a mark recapture study. Every two months I went out to my study sites and live-trapped for either three or four days. Every mouse was individually marked allowing me to create encounter histories for each mouse. These encounter histories are simply a record of when each mouse was caught.

Using these records I could do several things. My study design is what is called the robust design in the capture-mark-recapture history. Using the encounter histories for each individual trapping session I could estimate mouse abundances on each of my grids. Because each mouse was individually marked I could work out what the proportion of mice recaptured on the second day was, and what proportion of mice were newly caught for the session. I could then do this for the third day and fourth day as well. Because not all mice were caught on the second and third day, and news ones were caught I could work out what the probability of recapturing a mouse was, and then based on the total number of mouse captures, I could then calculate the abundance of mice on the grid (although in practice I used a piece of software called MARK). In addition to mouse abundance these encounter histories could be used to calculate survival and recruitment between each of my trapping sessions.

So now I have three useful pieces of information; mouse survival, mouse recruitment and mouse abundances. The other part to this project was an experimental manipulation. The goal was to compare the predictions from a matrix model describing mouse population dynamics to what happened when the wild populations were perturbed.

Matrix models are useful tools for describing population dynamics and using them we can calculate various summary variables, which describe the state of the population. Sensitivity analysis is a tool for predicting how the dependent factor in the population model will change in response to changes in a model parameter. By dependent parameter, I mean a parameter that is calculated by the model such as growth rate or population size. For my study I reduced female mouse survival on some of my trapping grids and compared the observed population growth rates on those grids to the predicted population growth rates from my matrix models. I used estimates of mouse survival and recruitment from my control grids to construct my models.

This project took about two and a half years. My field sites were established in late 1999, and I removed the last trap from them in June 2002. While my dissertation contains one chapter describing how well the matrix model predicted the effects of my experimental manipulation, it has three chapters describing mouse population dynamics. One of those describes how mouse survival and recruitment rates responded to my experimental manipulation. I have two chapters looking at mouse population growth rates, one of those where the population growth rate is estimated from the mark-recapture rates, and another “decomposing” the variance in population growth rates to changes in various parameters (trapping grid, time, mouse abundances, site…).

So have I managed to answer the question as to what I do? I guess a shorter version is that I’m interested in looking at the processes affect population dynamics. What are the factors that lead to changes in population sizes, and how do demographic rates vary over time and factors affect them.