Spatial confounding in Bayesian species distribution modeling ...

• Main Authors: Mäkinen, Jussi, Numminen, Elina, Niittynen, Pekka, Luoto, Miska, Vanhatalo, Jarno
• Format: Dataset
• Language: English
• Published: Dryad 2022
• Subjects: estimation bias; Gaussian process; spatial confounfing; spatial random effect; Species distribution model; FOS Earth and related environmental sciences; Norway; Northern Norway
• Online Access: https://dx.doi.org/10.5061/dryad.hdr7sqvm5; https://datadryad.org/stash/dataset/doi:10.5061/dryad.hdr7sqvm5

Species distribution models (SDMs) are currently the main tools to derive species niche estimates and spatially explicit predictions for species geographical distribution. However, unobserved environmental conditions and ecological processes may confound the model estimates if they have a direct impact on the species and, at the same time, they are correlated with the observed environmental covariates. This, so-called spatial confounding, is a general property of spatial models but it has not been studied in the context of SDMs before. Here we examine how the estimation accuracy of SDMs depends on the type of spatial confounding. We construct two simulation studies where we alter spatial structures of the observed and unobserved covariates and the level of dependence between them. We fit generalized linear models with and without spatial random effects applying Bayesian inference and record the bias induced to model estimates by spatial confounding. After this, we examine spatial confounding also with real ... : The study analyzes simulated and empirical species occurrence data sets. The simulated data set was created by using Gaussian process regression to generate spatial covariates, compute a species presence probability with probit-transformed linear combination of the covariates, and sample species occurrences with the presence probabilities. The empirical data set was collected in-situ in Northern Norway. ...