Data with a large number of zeroes is a particular problem for ecologists. As this data does not meet the assumptions for most regular statstical tests (such as a normal distribution), applying such tests would give non-sensical results. Proper analysis of such data, giving meaningful results, requires its own set of statistical tools, which are lacking from most commercial software packages. The R community, however, has developed relevant models and this book describes appropriate techniques, such as the Poisson general linear model (GLM), negative binomial GLM, Poisson or negative binomial generalized additive model (GAM), or GLMs with zero inflated distribution.
The chapters contain case studies of real-life datasets and show the reader how to go about analysing these, covering typical problematic biological datasets that have added complicating factors, such as 2-way nested data, spatial correlation, or temporal auto-correlation. The first chapter provides a basic introduction to Bayesian statistics and Markov Chain Monte Carlo (MCMC) techniques, as these are needed for most analyses.