Swift, Lynn C (2005) Calibrating SORTIE's recruitment subroutine for southeastern Québec : verifying the consistency of parameters. Masters thesis, Concordia University.
- Accepted Version
In recent years foresters have used inverse modelling as a tool to predict stand dynamics for use in research and management. Several models predicting recruitment density by species have been developed. There has, however, never been a serious attempt to see if the parameter values for a recruitment model are more or less constant from one site to another. I compared the performance of two dispersal functions (Weibull and lognormal) that can be used in the recruitment subroutine of SORTIE to determine which resulted in higher likelihoods, as well as whether there was a tendency toward species-specific parameter values among sites. Specifically, I calibrated the model for seven species in the deciduous forest of southern Québec at four sites located within 200km of each other. My results support the findings of Greene et al. , 2004, that the lognormal function is a better predictor of recruitment than the Weibull function. I also show that the previously suggested value for the parameter converting tree diameter into recruit production is hardly ideal for all species or for a single species across sites. Further, I show that while the estimated mean dispersal distances tended to be species-specific, they were not significantly so as they were swamped by inter-site differences within species. Indeed, it is not at all clear that inverse modelling permits us to characterize the species-specific dispersal parameters for any tree species.
|Divisions:||Concordia University > Faculty of Arts and Science > Biology|
|Item Type:||Thesis (Masters)|
|Authors:||Swift, Lynn C|
|Pagination:||viii, 71 leaves : ill. ; 29 cm.|
|Degree Name:||M. Sc.|
|Thesis Supervisor(s):||Greene, David|
|Deposited By:||Concordia University Libraries|
|Deposited On:||18 Aug 2011 18:25|
|Last Modified:||18 Aug 2011 19:29|
Repository Staff Only: item control page
Downloads per month over past year