Dominic Cyr
Hi everyone,
We are currently setting up simulation experiments that will be conducted with LANDIS-II in order to assess the effect of climate change on relatively large landscapes throughout Canada. We have 8 landscapes ranging from 2.5 Mha and 10 Mha in size, and we are working with a 250-m cell size.
We are in the process of choosing the method for setting species establishment probabilities (SEP).
We are starting to have a good general picture of the possible options, their advantages and caveats, but we still need to clarify some aspects before we make our final choice (and live with it).
I have two points for which feedback would be very appreciated. The first one is about the use of PICUS for setting SEPs, and the second one, about the scale-dependence of SEPs is the topic of a another message.
1- We are lucky enough to be part of a larger, integrated, multi-scale assessment of forest change through the next century. In that process, the stand scale patch model PICUS, which explicitly simulate physiological processes and is sensitive to climate, has been parameterized for all combinations of land types and species.
We are already using PICUS outputs to set growth parameters for our LANDIS-II simulations and we are now considering using it for setting establishment probabilities. However, we are still unsure about the most appropriate method to use to derivate SEP from PICUS outputs.
Has anyone used PICUS for that purpose? or has considered using it, and would be willing to share about info? Maybe something similar to previously applied methods with LINKAGES could be done (ex. Scheller et al. see citation below, or He et al. 1999)? Any thoughts about that?
"For each LINKAGES replication, if the stand exhibited positive biomass growth during the first 10 years, the replication represented a successful establishment and growth to age 10; 100 replications were used for each species. The SEC was calculated as the number of successful establishments (replications) divided by the total number of replications." (from Scheller et al. 2005)
That appears simple enough, but I wonder whether the stochastic components of PICUS will be strong enough to generate anything else that 0 and 1 values...
Thanks
Dominic
