Obtaining Environmental Favourability Functions from Logistic Regression

By Raimundo Real, A. Márcia Barbosa, J. M. Vargas. Environmental and Ecological Statistics, 2006

313 citations recorded by the source corpus.

Subfields: Ecological Modeling, Ecology, Nature and Landscape Conservation

View Paper: https://www.semanticscholar.org/paper/85c8457b00625d05aecbc15af20c1478124d0ad5

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 invasion for a habitat generalist and a specialist plant species (2008, background): “It is also reasoned that when models perform poorly, the analyses were lacking significant variables to predict suitable habitats (Waters et al., 2004) or relied on inconclusive field data (Hernandez et al., 2006; Real et al., 2006).”
Past, present and future of wild ungulates in relation to changes in land use (2010, methodology): “For this purpose, a threshold—cut-off point used to reclassify model outputs as presence (over the threshold) or absence (under the threshold)—should be set at 0.5 for the favourability function (see Real et al. 2006).”
The effect of prevalence and its interaction with sample size on the reliability of species distribution models (2009, background): “…model functions has generally been neglected in the literature on species distribution modeling (but see Hirzel et al. 2001, Reese et al. 2005, Real et al. 2006, Meynard and Quinn 2007), virtual species are nowadays one of the best ways to make testable experiments in this field (see Austin…”
A statistical explanation of MaxEnt for ecologists (2011, background): “Some users may in fact see the species-specific scaling as an opportunity, since the literature on favourability functions (e.g., Real et al., 2006) claims that probability of presence is itself hard to work with.”
Effects of species’ ecology on the accuracy of distribution models (2007, methodology): “Both training and test data contained an equal number of presence and absence samples, because this optimises accuracy in logistic models and ensures comparability between species (McPherson et al. 2004, but see Real et al. 2006 for an alternative approach).”
Exploring the effects of quantity and location of pseudo-absences and sampling biases on the performance of distribution models with limited point occurrence data (2011, background): “…a true presence point, one must use an appropriate cut-off value to convert decimal fraction probabilities to a binary variable (Liu et al. 2005; Jiménez-Valverde & Lobo 2006; Jiménez-Valverde & Lobo 2007), or re-scale logistic probabilities by applying a favourability function (Real et al. 2006).”
Effects of species and habitat positional errors on the performance and interpretation of species distribution models (2009, methodology): “…mapped outputs from control and scenario models, we first standardized the probability scores to favourability to remove the effects of prevalence (Real et al., 2006) and then calculated the mean absolute difference in the predictions across all pixels as: 1 ½RjðCi SiÞj=p ; where Ci is the value…”
Climate change‐linked range expansion of Nathusius’ pipistrelle bat, Pipistrellus nathusii (Keyserling & Blasius, 1839) (2010, background): “This is caused by the artificial maintenance of positive records at a prevalence of 50% within the analysed data, whilst the true prevalence of the species may be significantly different from this (i.e. occurring in 5% of a given area) (Real et al., 2006).”
Assessing the effect of prevalence on the predictive performance of species distribution models using simulated data (2011, background): “Using real data to assess aspects of species distribution modelling poses several problems due to lack of knowledge about the mechanisms that drive real species distributions (Austin et al., 2006; Real et al., 2006; Meynard & Quinn, 2007).”
AUC: a misleading measure of the performance of predictive distribution models (2008, methodology): “It is necessary to rescale these probabilities if we want to use continuous and not binary predictions (Jimenez-Valverde & Lobo, 2006; Real et al . 2006 ).”
The ghost of unbalanced species distribution data in geographical model predictions (2006, background): “Interestingly, whereas P -values for different species are not comparable site to site because of the prevalence bias, this is not the case for F -values, which are directly equivalent (Real et al ., 2006).”
Delimiting the geographical background in species distribution modelling (2012, methodology): “To compare the results of each model obtained from species with different prevalences, the favourability function was applied to convert logistic probabilities (P) into favourability values (F) that are independent of sample prevalence (for further details about this function see Real et al., 2006).”
Conservation biogeography of ecologically interacting species: the case of the Iberian lynx and the European rabbit (2009, methodology): “…390–400, Journal compilation © 2008 Blackwell Publishing Ltd model for the lynx and another for the rabbit using the favourability function described by Real et al. (2006), which may be written as follows: where n1 and n0 are the numbers of presences and absences, respectively, e is the basis of…”