Can niche-based distribution models outperform spatial interpolation?

By V. Bahn, Brian J. McGill. 2007

213 citations recorded by the source corpus.

Subfields: Ecological Modeling, Ecology, Nature and Landscape Conservation

View Paper: https://www.semanticscholar.org/paper/af76090d05900a44b6505b7ed5da2bd65addcd0f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Ecological Modeling

In Ecology

In Nature and Landscape Conservation

Citations in the corpus

Modelling exploration of the future of European beech (Fagus sylvatica L.) under climate change—Range, abundance, genetic diversity and adaptive response (2010, methodology): “Although extensively used and also criticized (Bahn and McGill, 2007), some analyses having thoroughly tested their predictive power have shown relatively good performance to predict the current distribution based on independent data (Araújo et al., 2005).”
Joint dynamic species distribution models: a tool for community ordination and spatio‐temporal monitoring (2016, background): “Spatial autocorrelation can bias estimates of habitat associations, inflate Type-I errors for statistical tests of habitat associations (Bahn & McGill, 2007) and decrease explanatory power for SDMs (Dormann, 2007).”
Evaluating presence‐only species distribution models with discrimination accuracy is uninformative for many applications (2019, background): “…correlations that a model might treat as biological truth, resulting in models that produce high discrimination accuracy even when occurrence data are random (Raes & ter Steege, 2007) or the predictors are biolog‐ ically meaningless (Bahn & McGill, 2007, 2013; Fourcade, Besnard, & Secondi, 2018).”
Cross-validation strategies for data with temporal, spatial, hierarchical, or phylogenetic structure (2017, background): “Resulting model predictions may perform fine in a situation where the correlation structure between non-causal and the “true” predictors (i.e. underlying structures) remains unchanged (Bahn and McGill 2007), but they could completely fail when predicting to novel situations.”
Predicting worldwide invasiveness for four major problematic decapods: an evaluation of using different calibration sets (2011, background): “While previous studies have found that simple spatial autocorrelation models can perform as well as NBM (Bahn and McGill 2007), this only indicates that we cannot distinguish between habitat suitability and spatial autocorrelation.”
Is there a correlation between abundance and environmental suitability derived from ecological niche modelling? A meta‐analysis (2017, background): “One way of addressing spatial variation of abundance within a species range is interpolating local abundance data over a geographic domain of interest (Bahn and McGill 2007) relying on the assumption that abundance is spatially autocorrelated (Brown 1995).”
Testing the predictive performance of distribution models (2013, unknown): “If true, it would be a gloomy assessment of our state of understanding and ability to predict the effects of global change at such a scale partly supported by Bahn and McGill (2007).”
Uses and misuses of bioclimatic envelope modeling. (2012, background): “Others have suggested that the assumption that climate determines species’ distributions is neither supported nor contradicted by data (e.g., Bahn and McGill 2007, Beale et al. 2008, Chapman 2010).”
Empirical Evidence for the Scale Dependence of Biotic Interactions (2015, methodology): “Geographic proximity We performed spatial interpolation to estimate the contribution of geographic proximity to other occupied sites to species occurrence probability, independent of environmental suitability (Bahn & McGill, 2007).”
Comparing species distribution models constructed with different subsets of environmental predictors (2015, background): “Distributions of rare species are often more accurately predicted by SDMs than common species (Bahn & McGill, 2007; Franklin et al., 2009), which is reflected by the generally high accuracy of the models.”
Climate change, genetic markers and species distribution modelling (2015, unknown): “As Bahn & McGill (2007) have shown, models of spatial variation in abundance of North American birds that contain only spatial autocorrelation can often outperform species distribution models that include standard climate variables and assume that samples are statistically independent of one…”
Niche and area of distribution modeling: a population ecology perspective (2010, background): “there are also dissenting views (Bahn and McGill 2007, Beale et al. 2008) and the fields of SDM and ENM still require conceptual clarification (Pearson and Dawson 2003, Araújo and Guisan 2006, Kearney 2006, Jiménez-Valverde et al. 2008, Elith and Graham 2009, Lobo et al. 2010).”
Predicting the future of species diversity: macroecological theory, climate change, and direct tests of alternative forecasting methods (2009, background): “Potentially useful inclusion of spatial effects in distribution models has been demonstrated in North American birds (Bahn and McGill 2007, Allouche et al. 2008).”
Realising the full potential of citizen science monitoring programs (2013, methodology): “…within these datasets is also 233 useful for improving the performance of spatial analysis methods such as species distribution modelling (e.g. 234 Araujo et al. 2005; Bahn and McGill 2007; Brotons et al. 2007; Luoto et al. 2007), and estimation of species 235 richness (Kery and Royle 2008).”
Species Distribution Modeling in the Tropics: Problems, Potentialities, and the Role of Biological Data for Effective Species Conservation (2009, background): “To merge several data sources into one homogenous dataset is an enormous challenge that usually dwarfs the time required to analyze these data [7].”
Spatial factor analysis: a new tool for estimating joint species distributions and correlations in species range (2015, background): “Bahn & McGill (2007) highlighted the good performance of spatial models like SFA when imputing between sampled locations, but we hypothesize that JSDMs that include measured covariates will often perform better than SFA when extrapolating outside the sampled domain.”