Novel methods improve prediction of species distributions from occurrence data Export

@article{elith2006_species, abstract = {Prediction of species' distributions is central to diverse applications in ecology, evolution and conservation science. There is increasing electronic access to vast sets of occurrence records in museums and herbaria, yet little effective guidance on how best to use this information in the context of numerous approaches for modelling distributions. To meet this need, we compared 16 modelling methods over 226 species from 6 regions of the world, creating the most comprehensive set of model comparisons to date. We used presence-only data to fit models, and independent presence-absence data to evaluate the predictions. Along with well-established modelling methods such as generalised additive models and GARP and BIOCLIM, we explored methods that either have been developed recently or have rarely been applied to modelling species' distributions. These include machine-learning methods and community models, both of which have features that may make them particularly well suited to noisy or sparse information, as is typical of species' occurrence data. Presence-only data were effective for modelling species' distributions for many species and regions. The novel methods consistently outperformed more established methods. The results of our analysis are promising for the use of data from museums and herbaria, especially as methods suited to the noise inherent in such data improve.}, author = {Elith, Jane and Graham, Catherine H. and Anderson, Robert P. and Dud\'{\i}k, Miroslav and Ferrier, Simon and Guisan, Antoine and Hijmans, Robert J. and Huettmann, Falk and Leathwick, John R. and Lehmann, Anthony and Li, Jin and Lohmann, Lucia G. and Loiselle, Bette A. and Manion, Glenn and Moritz, Craig and Nakamura, Miguel and Nakazawa, Yoshinori and Jacob and Peterson, Townsend A. and Phillips, Steven J. and Richardson, Karen and Scachetti-Pereira, Ricardo and Schapire, Robert E. and Sober\'{o}n, Jorge and Williams, Stephen and Wisz, Mary S. and Zimmermann, Niklaus E.}, citeulike-article-id = {585800}, citeulike-linkout-0 = {http://www.blackwell-synergy.com/doi/abs/10.1111/j.2006.0906-7590.04596.x}, citeulike-linkout-1 = {http://dx.doi.org/10.1111/j.2006.0906-7590.04596.x}, citeulike-linkout-2 = {http://www.ingentaconnect.com/content/mksg/eco/2006/00000029/00000002/art00001}, citeulike-linkout-3 = {http://www3.interscience.wiley.com/cgi-bin/abstract/118729877/ABSTRACT}, citeulike-linkout-4 = {http://www.blackwell-synergy.com/doi/pdf/10.1111/j.2006.0906-7590.04596.x}, doi = {10.1111/j.2006.0906-7590.04596.x}, issn = {0906-7590}, journal = {Ecography}, keywords = {algorithms, analysis, biology, cart, classification, distribution\_model, geospatial, geostatistics, machine-learning, modeling, prediction-error, presence-absence-models, presence-only, presence-only-models, r\_stats, vegetation}, month = {April}, number = {2}, pages = {129--151}, posted-at = {2007-01-11 14:23:12}, priority = {2}, publisher = {Blackwell Publishing}, title = {Novel methods improve prediction of species distributions from occurrence data}, url = {http://dx.doi.org/10.1111/j.2006.0906-7590.04596.x}, volume = {29}, year = {2006} }

TY - JOUR ID - elith2006_species L3 - citeulike-article-id:585800 N2 - Prediction of species' distributions is central to diverse applications in ecology, evolution and conservation science. There is increasing electronic access to vast sets of occurrence records in museums and herbaria, yet little effective guidance on how best to use this information in the context of numerous approaches for modelling distributions. To meet this need, we compared 16 modelling methods over 226 species from 6 regions of the world, creating the most comprehensive set of model comparisons to date. We used presence-only data to fit models, and independent presence-absence data to evaluate the predictions. Along with well-established modelling methods such as generalised additive models and GARP and BIOCLIM, we explored methods that either have been developed recently or have rarely been applied to modelling species' distributions. These include machine-learning methods and community models, both of which have features that may make them particularly well suited to noisy or sparse information, as is typical of species' occurrence data. Presence-only data were effective for modelling species' distributions for many species and regions. The novel methods consistently outperformed more established methods. The results of our analysis are promising for the use of data from museums and herbaria, especially as methods suited to the noise inherent in such data improve. IS - 2 JF - Ecography SN - 0906-7590 EP - 151 TI - Novel methods improve prediction of species distributions from occurrence data PB - Blackwell Publishing VL - 29 SP - 129 KW - algorithms KW - analysis KW - biology KW - cart KW - classification KW - distribution_model KW - geospatial KW - geostatistics KW - machine-learning KW - modeling KW - prediction-error KW - presence-absence-models KW - presence-only KW - presence-only-models KW - r_stats KW - vegetation AU - Elith, Jane AU - Graham, Catherine AU - Anderson, Robert AU - Dudík, Miroslav AU - Ferrier, Simon AU - Guisan, Antoine AU - Hijmans, Robert AU - Huettmann, Falk AU - Leathwick, John AU - Lehmann, Anthony AU - Li, Jin AU - Lohmann, Lucia AU - Loiselle, Bette AU - Manion, Glenn AU - Moritz, Craig AU - Nakamura, Miguel AU - Nakazawa, Yoshinori AU - Jacob AU - Peterson, Townsend AU - Phillips, Steven AU - Richardson, Karen AU - Scachetti-Pereira, Ricardo AU - Schapire, Robert AU - Soberón, Jorge AU - Williams, Stephen AU - Wisz, Mary AU - Zimmermann, Niklaus PY - 2006/04// UR - http://dx.doi.org/10.1111/j.2006.0906-7590.04596.x ER -