Spatially explicit metapopulation model - Incidence Function Model
Section outline
-
The Incidence Function Model describes presence/absence of a species in the patches of a highly fragmented landscape at discrete time intervals (years) as the result of colonization and extinction processes. The IFM ignores local dynamics since they are faster than metapopulation dynamics in producing changes in the size of local populations (Hanski, 1994).
In the IFM, the process of occupancy of patch is described by a first-order Markov chain with two states, {O, i} (empty and occupied, respectively). The extinction probability of a population in a patch is constant in time and is assumed to decrease with increasing patch area, and the colonization probability is assumed to be a sigmoidal function increasing with connectivity. The IFM is the best known spatially explicit metapopulation model in literature.
This model has been applied to conservation problems and to area-wide pest-management.
This training will start with a short introduction to discrete time, finite space, homogeneous Markov chain, with the aim of understanding the basic mathematics of the IFM. Then, the IFM model will be discussed by deeply considering (a) the role of the parameters and how they affect metapopulation dynamics; (b) variations to the basic model (rescue effect, time-dependent colonization probabilities). The training will conclude with a section focusing on the use of the free software R to deal with simulation and parameter estimation. -
This section presents the basic Incidence Function model and its modifications (Hanski,1994) as well as the practical and theoretical differences about the models.
-
This section uses simulations to analyze the effect of moving from the “true” parameters to different parameter values, aiming at gaining further insight into the role of the IFM parameters. This will emphasize the role of parameters, and especially the effects of getting wrong parameters estimates.
