I am trying to figure out the best approach to determine the effect of fragmentation on the standing genetic variation in a population. We have the relatedness data of the individuals in the form of a family pedigree (20-30 years of data, ~500 individuals). This means we can estimate a relatedness matrix for all living individuals. Till now, the population has been allowed to freely mix, and did so often over long distances. However, changes in the environment is going to restrict these movements to a degree. The sub populations are going to be rather variable in size from a less than 10 to more than 150. We will get an estimate of the reduction in these movements. What I want to do is model/estimate what the effect is of these reduced movements on the standing genetic variation of the population as a whole. I am thinking of using some genetic drift principles to model these things. What would others suggest?
I am not sure you can say much about SGV with only pedigree data. And depending on the degree of a barrier, you might not see very much in such a short period of time in terms of how your relatedness matrix correlates to habitat fragmentation.
But modelling drift is a good start - you could do some coalescence modelling of populations of different sizes with different degrees of migration and see if you can differentiate it from a null model.
If you want to model the genetic diversity change over time you could use the Marlin simulation environment, which is actually a user-friendly interface for the quantiNEMO software package, but limited to neutral markers. I have used it recently for estimating the effect of wildlife passages in a fragmented landscape. You can actually input your existing genotypes into a virtual landscape (the landscape change you are expecting), set specific parameters for growth, migration rate and dispersal kernels etcetera, and run it.
http://www.bentleydrummer.nl/software/software/Marlin.html
If you want to simulate the future impact of fragmentation on genetic variance, I would suggest you to use an individual based model where the initial population would follow your relatedness data. Building your own model will allow you to incorporate details about the mating system and dispersal patterns of your population.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size