B1998
Title: Predicting biodiversity with the generalised functional response model
Authors: Shaykhah Aldossari - University of Glasgow (United Kingdom) [presenting]
Dirk Husmeier - Biomathematics and Statistics Scotland, Edinburgh (UK)
Jason Matthiopoulos - University of Glasgow (United Kingdom)
Abstract: Biodiversity is a measure of variability and is widely used to describe the variation in different fields. The Shannon entropy score is the most frequently used measure of biodiversity in ecology. It summarises the information about species abundance within a sample or a group. We use the Shannon entropy score to investigate three different things. First, we use the entropy score to assess the transferability of the generalized function response (GFR) model by measuring the information content in the dataset under study. Second, we observe the relationship between biodiversity and land cover types using the GFR model and various recent extensions. Finally, we investigate the legacy effect using the GFR models of land cover types on biodiversity. The large-scale North American Breeding Bird Survey (BBS) dataset was used for this purpose. We discuss how the information in the dataset affects the predictive ability of the model. Our finding is that our recent extensions of the GFR model double the biodiversity prediction accuracy compared to the standard generalised linear model (GLM). Moreover, biodiversity in the BBS dataset is found not to occur with time lags in response to land cover covariates using the GFR models.
