Bayesian multi-species modelling of non-negative continuous ecological data with a discrete mass at zero

Bayesian statistical decision theory;Bird populations--Mathematical models;Bird populations--Statistical methods;Predation (Biology)--Mathematical models;Predation (Biology)--Statistical methods;Songbirds

Swallow, Ben ; Buckland, S. T. (Stephen T.) ; Buckland, S. T. (Stephen T.)

University:University of St Andrews

Department:Mathematics & Statistics (School of)

关键词: Bayesian statistical decision theory;    Bird populations--Mathematical models;    Bird populations--Statistical methods;    Predation (Biology)--Mathematical models;    Predation (Biology)--Statistical methods;    Songbirds;   

Others  :  https://research-repository.st-andrews.ac.uk/bitstream/handle/10023/9626/BenSwallowPhDThesis.pdf?sequence=2&isAllowed=y

来源: DR-NTU

【 摘 要 】

Severe declines in the number of some songbirds over the last 40 yearshave caused heated debate amongst interested parties. Many factorshave been suggested as possible causes for these declines, includingan increase in the abundance and distribution of an avian predator,the Eurasian sparrowhawk Accipiter nisus. To test for evidence for apredator effect on the abundance of its prey, we analyse data on 10species visiting garden bird feeding stations monitored by the BritishTrust for Ornithology in relation to the abundance of sparrowhawks.We apply Bayesian hierarchical models to data relating to averagedmaximum weekly counts from a garden bird monitoring survey. Thesedata are essentially continuous, bounded below by zero, but for manyspecies show a marked spike at zero that many standard distributionswould not be able to account for. We use the Tweedie distributions,which for certain areas of parameter space relate to continuous nonnegativedistributions with a discrete probability mass at zero, andare hence able to deal with the shape of the empirical distributions ofthe data.The methods developed in this thesis begin by modelling single preyspecies independently with an avian predator as a covariate, usingMCMC methods to explore parameter and model spaces. This modelis then extended to a multiple-prey species model, testing for interactionsbetween species as well as synchrony in their response to environmentalfactors and unobserved variation.Finally we use a relatively new methodological framework, namelythe SPDE approach in the INLA framework, to fit a multi-speciesspatio-temporal model to the ecological data.The results from the analyses are consistent with the hypothesis thatsparrowhawks are suppressing the numbers of some species of birdsvisiting garden feeding stations. Only the species most susceptible tosparrowhawk predation seem to be affected.