CMStatistics 2016: Start Registration
View Submission - CMStatistics
B0776
Title: Joint models for repeated counts of predators and prey Authors:  Rafael de Andrade Moral - Maynooth University (Ireland) [presenting]
Clarice Garcia Borges Demetrio - ESALQ - USP (Brazil)
John Hinde - NUI Galway (Ireland)
Abstract: In ecological field surveys it is often of interest to estimate the abundance of species. However detection is imperfect and hence it is important to model these data taking into account the ecological processes and sampling methodologies. In this context, $N$-mixture models and extensions are particularly useful, as it is possible to estimate population size and detection probabilities under different ecological assumptions. In this framework, observed animal counts are assumed to be binomial distributed with the denominator being a latent random variable with a count distribution representing the true abundance in each site, and hence only a part of this abundance, subject to a probability of detection, is observed. We extend the $N$-mixture modelling framework to jointly model predator and prey species abundances and obtain the correlation between them in each surveyed site. We illustrate this approach using data from the North American Breeding Bird Survey on bald eagles and mallards, obtained in the 2015 campaign for the states of Oregon, Washington and British Columbia. In each state, several 50-stop routes were surveyed and the total counts for every ten stops are available for download from their website.