PointedSDMs: An R package to help facilitate the construction of integrated species distribution models

Abstract

Ecological data are being collected at a large scale from a multitude of different sources, each with their own sampling protocols and assumptions. As a result, the integration of disparate datasets is a rapidly growing area in quantitative ecology, and is subsequently becoming a major asset in understanding the shifts and trends in species' distributions. However, the tools and software available to construct statistical models to integrate these disparate datasets into a unified framework is lacking. This has made these methods inaccessible to general practitioners and has stagnated the growth of data integration in more applied settings. We therefore present PointedSDMs: an easy to use R package used to construct integrated species distribution models. It provides functions to easily format the data, fit the models in a computationally efficient way and presents the output in a format that is convenient for additional work. This paper illustrates the different uses and functions available in the package, which are designed to simplify the modelling of integrated models. A case study using the package is also presented: combining three datasets coming from different sampling protocols, all containing records of Setophaga caerulescens across Pennsylvania state.

intSDM: anRpackage for building a reproducible workflow for the field of integrated species distribution models

“…Recently the PointedSDMs R package (Mostert & O’Hara, 2023) has been proposed to estimate these complex models in an easy-to-use framework. PointedSDMs is built around the R-INLA package (Martins et al ., 2013), which uses the integrated nested Laplace approximation (INLA) methodology (Rue et al ., 2009) to approximate Bayesian latent Gaussian models.…”

“…Recently the PointedSDMs R package (Mostert & O'Hara, 2023) has been proposed to estimate these complex models in an easy-to-use framework.…”

“…Nevertheless, validation of results and model comparison is an important step in any scientific analysis. Therefore our workflow incorporates two types of cross-validation: a spatial-block crossvalidation (using functionality from the blockCV R package [Valavi et al, 2019]) as well as a leave-one-one cross-validation (see for example [Mostert & O'Hara, 2023]). Cross-validation for the workflow may be specified by using $crossValidation.…”

‘RISDM‘: species distribution modelling from multiple data sources in R

“…Alternative R‐packages to ‘RISDM' are: ‘spOccupancy' (Doser et al 2022), ‘PointedSDMs' (Mostert and O'Hara 2022), and ‘ibis.iSDM' (Jung 2023). ‘RISDM' differs from ‘spOccupancy' in the type of target data, with ‘spOccupancy' focussing entirely on occupancy data (a particular type of PA or AA data) and therefore is not appropriate for the types of analyses that ‘RISDM' accommodates (and vice versa).…”