A Shared Component Hierarchical Model To Represent How Fish Assemblages Vary as a Function of River Temperatures and Flow Regimes. Application to Three Groupings of Juveniles in the Upper River Rhone during the 1980-2005 Period
Jeremy Piffady2, Eric Parent1
1MORSE, Statistical Research Unit, AgroParistech, Paris, France; 2HydroBiology, Cemagref, Lyon, France

A hierarchical model is developed to understand how the inter annual variations of fish assemblages answer to temperature and flow regimes. A latent variable of interest represents the common source of variation for the environmental and the biological data. This shared component links two sets of variables with different types: on one side, the latent variable can be understood as a “factor” from the continuous explanatory variables recording the environmental variations while, on the other side, it can be considered as a “logistic regressor” in a multinomial response model for counts of various fish species juvenile abundance collected under poorly controlled electro-fishing experiments. Datasets of water temperatures, flows and electric fishing samples have been collected on the upper River Rhone over the 1980-2005 period. The response variable consists in three groups of species determined according to their synchronic reaction to environmental variations. Inference relies on Monte Carlo Markov Chain techniques under the Bayesian paradigm.

Keywords: Hierarchical modelling; Bayesian statistics; Fish assemblages; Environment

Biography: I work as a Professor in applied statistics and probabilistic modeling for environmental engineering. My research group belongs to AgroParistech, an academic institution from the Ministry of Agriculture. The laboratory enrols PhD students and postgraduate engineers trained to work at the intersection between statistics, decision-making theory and environmental engineering. I co-authored three books (in French), one on Bayesian statistics for environmental engineering, the second on theoretical and algorithmic aspects of Bayesian theory and the third one depicts various cases with real data from various domains of applications. With my colleague Etienne Rivot, we are preparing a book “Bayesian models for Ecological Data”, with special attention to fish sciences. My broader interests include “Bayesian Statistics at work”, especially in case studies from various fields while working with my PhD students, under contract with industrial companies or public institutions