Environmental filters reduce the effects of sampling bias and improve predictions of ecological niche models

Abstract

Ecological niche models represent key tools in biogeography but the effects of biased sampling hinder their use. Here, we address the utility of two forms of filtering the calibration data set (geographic and environmental) to reduce the effects of sampling bias. To do so we created a virtual species, projected its niche to the Iberian Peninsula and took samples from its binary geographic distribution using several biases. We then built models for various sample sizes after applying each of the filtering approaches. While geographic filtering did not improve discriminatory ability (and sometimes worsened it), environmental filtering consistently led to better models. Models made with few but climatically filtered points performed better than those made with many unfiltered (biased) points. Future research should address additional factors such as the complexity of the species’ niche, strength of filtering, and ability to predict suitability (rather than focus purely on discrimination).

Impacts of Species Misidentification on Species Distribution Modeling with Presence-Only Data

“…For example, modeling methods have been developed that use presenceonly data [11]; the impact of limited sample size in modeling have been studied [5,12]; and solutions have been proposed for the problem of incorrectly located species records [13], uneven sampling effort [14][15][16], spatial autocorrelation [17][18][19][20], and scales [21,22]. However, not all deficiencies in species data sets have been fully studied.…”

Climatic niche shift of aquatic plant invaders between native and invasive ranges: a test using 10 species across different biomes on a global scale

“…Duplicate occurrences of recorded data for species in 5.0-arc-minute grid pixels (10 km at the equator) were removed to avoid georeferencing errors (Beck et al, 2014). To decrease the negative effect of sampling bias on the performance of ENMs, 10 species with more than 50 unique records for both native and invasive ranges were selected; the entire globe was used as the extent of the input data (Varela et al, 2014) (Tab. 1).…”

SDMvspecies: a software for creating virtual species for species distribution modelling

“…This method was originally reported in Varela et al (2014). This method defines virtual species that are not resistant to extreme environment conditions, and virtual species limited by extreme environments.…”