A1112
Title: Improving spatial maps with preferential sampling via hierarchical modeling
Authors: Giacomo Zoppi - University of Torino (Italy) [presenting]
Natalia Golini - University of Turin (Italy)
Rosaria Ignaccolo - University of Turin (Italy)
Anna Lo Presti - Università di Torino (Italy)
Michela Cameletti - Universita degli Studi di Bergamo (Italy)
Abstract: Mapping is essential to understanding the spatial pattern of species over a region. Recently, spatial modelling has been developed to explain the presence/absence or abundance of one or more species over a region, as well as through the use of environmental variables available at locations across the region. Nevertheless, the sampling location choice may not be completely random but guided by some kind of relationship with the variable of interest (preferential sampling). Misusing preferential sampling can introduce bias in the parameter estimates and in the predictions reported in the map. A hierarchical modelling approach is considered that takes into account preferential sampling for abundance data (e.g., counts, per cent cover, or biomass) and for the integration of abundance data with abundance-only data to improve prediction accuracy.
