occAssess: An R package for assessing potential biases in species occurrence data

Abstract

Species occurrence records from a variety of sources are increasingly aggregated into heterogeneous databases and made available to ecologists for immediate analytical use. However, these data are typically biased, i.e. they are not a representative sample of the target population of interest, meaning that the information they provide may not be an accurate reflection of reality. It is therefore crucial that species occurrence data are properly scrutinised before they are used for research. In this article, we introduce occAssess, an R package that enables quick and easy screening of species occurrence data for potential biases. The package contains a number of discrete functions, each of which returns a measure of the potential for bias in one or more of the taxonomic, temporal, spatial and environmental dimensions. The outputs are provided visually (as ggplot2 objects) and do not include a formal recommendation as to whether data are of sufficient quality for any given inferential use. Instead, they should be used as ancillary information and viewed in the context of the question that is being asked, and the methods that are being used to answer it. We demonstrate the utility of occAssess by applying it to data on two key pollinator taxa in South America: leaf-nosed bats (Phyllostomidae) and hoverflies (Syrphidae). In this worked example, we briefly assess the degree to which various aspect of data coverage appear to have changed over time. We then discuss additional ways in which the package could be used, highlight its limitations, and point to where it could be improved in the future. Going forward, we hope that occAssess will help to improve the quality, and transparency, of assessments of species occurrence data as a necessary first step where they are being used for ecological research at large scales.

IUCNN – Deep learning approaches to approximate species' extinction risk

“…Accounting for geographic sampling bias in biodiversity science is an active field of research. The high flexibility of neural networks in IUCNN will make it easy for users to apply different approaches to factor geographic sampling biases into automated conservation assessments, for instance, by scrutinizing data structure (Boyd et al, 2021), spatial thinning (Aiello-Lammens et al, 2015) and modelling (Bruelheide et al, 2020;Varela et al, 2014).…”

“…In particu- that if records are split by time intervals, it may be relevant to explicitly consider temporal sampling bias (differences in sampling effort at different times; e.g. Boyd et al, 2021;Zizka, Rydén, et al, 2021).…”