Population distribution models: species distributions are better modeled using biologically relevant data partitions

Abstract

BackgroundPredicting the geographic distribution of widespread species through modeling is problematic for several reasons including high rates of omission errors. One potential source of error for modeling widespread species is that subspecies and/or races of species are frequently pooled for analyses, which may mask biologically relevant spatial variation within the distribution of a single widespread species. We contrast a presence-only maximum entropy model for the widely distributed oldfield mouse (Peromyscus polionotus) that includes all available presence locations for this species, with two composite maximum entropy models. The composite models either subdivided the total species distribution into four geographic quadrants or by fifteen subspecies to capture spatially relevant variation in P. polionotus distributions.ResultsDespite high Area Under the ROC Curve (AUC) values for all models, the composite species distribution model of P. polionotus generated from individual subspecies models represented the known distribution of the species much better than did the models produced by partitioning data into geographic quadrants or modeling the whole species as a single unit.ConclusionsBecause the AUC values failed to describe the differences in the predictability of the three modeling strategies, we suggest using omission curves in addition to AUC values to assess model performance. Dividing the data of a widespread species into biologically relevant partitions greatly increased the performance of our distribution model; therefore, this approach may prove to be quite practical and informative for a wide range of modeling applications.

A framework for species distribution modelling with improved pseudo-absence generation

“…5), with a strong under-estimation of high probability values. We argue that these results are due to the spatially clustered distribution of targeted group presences used as pseudoabsences, leading to spatially autocorrelated background samples resulting in inflated AUC values (Gonzàlez et al, 2011), and also to an over-estimated suitability for a large proportion of non-sampled areas (Figs. 6 and 7), as compared to the other methods.…”

“…With this regard, Hernández et al (2006) suggested that research in environmental niche modelling should focus on broad distributional subunits based on distinct genetic linages. For instance, Gonzàlez et al (2011) demonstrated that omission error is reduced when "biologically meaningful" data (in reference to genetically distinct populations of the same species) are modelled. Hence, in this study we consider genetically differenced groups of Quercus sp.…”

Mapping Species Distributions with MAXENT Using a Geographically Biased Sample of Presence Data: A Performance Assessment of Methods for Correcting Sampling Bias

“…This method has been used for species occurring over a large distribution range and extended environmental gradients [44], [51], [71]. We split our biased datasets in a northern and a southern stratum.…”

Avian SDMs: current state, challenges, and opportunities

“…This integration began with phylogenetic studies on niche evolution (summarized in Chan et al 2011, Alvarado-Serrano and Knowles 2014; see section above and Eyres et al 2017 for an overview in birds), but the use of finer-scale genetic information together with SDMs is a much more recent innovation (Fordham et al 2014). Using information generated from genetic markers systems such as microsatellites or single nucleotide polymorphisms can add relevant insights into population structure and refine SDMs below the species level (Gonzalez et al 2011). Due to local adaptation, species-environment relationships may change among genetic groups, and data pooling at the species level may consequently lead to an overestimation of potential distribution (Oney et al 2013).…”