The importance of correcting for sampling bias in MaxEnt species distribution models

Abstract

Aim Advancement in ecological methods predicting species distributions is a crucial precondition for deriving sound management actions. Maximum entropy (MaxEnt) models are a popular tool to predict species distributions, as they are considered able to cope well with sparse, irregularly sampled data and minor location errors. Although a fundamental assumption of MaxEnt is that the entire area of interest has been systematically sampled, in practice, MaxEnt models are usually built from occurrence records that are spatially biased towards better‐surveyed areas. Two common, yet not compared, strategies to cope with uneven sampling effort are spatial filtering of occurrence data and background manipulation using environmental data with the same spatial bias as occurrence data. We tested these strategies using simulated data and a recently collated dataset on Malay civet Viverra tangalunga in Borneo. Location Borneo, Southeast Asia. Methods We collated 504 occurrence records of Malay civets from Borneo of which 291 records were from 2001 to 2011 and used them in the MaxEnt analysis (baseline scenario) together with 25 environmental input variables. We simulated datasets for two virtual species (similar to a range‐restricted highland and a lowland species) using the same number of records for model building. As occurrence records were biased towards north‐eastern Borneo, we investigated the efficacy of spatial filtering versus background manipulation to reduce overprediction or underprediction in specific areas. Results Spatial filtering minimized omission errors (false negatives) and commission errors (false positives). We recommend that when sample size is insufficient to allow spatial filtering, manipulation of the background dataset is preferable to not correcting for sampling bias, although predictions were comparatively weak and commission errors increased. Main Conclusions We conclude that a substantial improvement in the quality of model predictions can be achieved if uneven sampling effort is taken into account, thereby improving the efficacy of species conservation planning.

Spatial Scale Effects of Sampling on the Interpolation of Species Distribution Models in the Southwestern Amazon

“…That is, the higher the number of quadrants used in constructing the model, the smaller the AUCdifference was, thereby minimizing the effects of sampling bias. Townsend and Kramer-Schadt et al (2013) reported similar results.…”

A Climate-Oriented Approach to Support Decision-Making for Seed Provenance in Ecological Restoration

“…The geographic extent of the models was constrained to the focal and immediately neighboring biogeographic regions occupied by each species, and was the same for the five species. To account for the spatial bias in survey effort, the records density of all plant species belonging to the study species' botanical families (Cyperaceae, Fabaceae, Myrtaceae, and Proteaceae) was used as a sampling effort bias layer (Kramer-Schadt et al, 2013). For selection of the best predictors, variables with a Pearson's pairwise correlation coefficient >|0.80| were not included in the same model and were tested separately.…”

Seven years of NeoBiota – the times, were they a changin’?

“…Unfortunately, it does have many biases in occurrence records as well as taxa (Beck et al 2014, Meyer et al 2015, 2016. Similarly, MaxEnt (Elith et al 2011) is an extremely powerful tool for distributional analyses, if properly used (Merow et al 2013, Kramer-Schadt et al 2013). Combining not revised GBIF data with standard settings of MaxEnt, though, leading to poor ecological results, can be a reason for an immediate rejection.…”