A comprehensive evaluation of predictive performance of 33 species distribution models at species and community levels

Ecological Monographs - Tập 89 Số 3 - 2019
Anna Norberg1, Nerea Abrego2,3, F. Guillaume Blanchet4, Frederick R. Adler5,6, Barbara J. Anderson7, Jani Anttila1, Miguel B. Araújo8,9,10, Tad Dallas1, David B. Dunson11, Jane Elith12, Scott D. Foster13, Richard Fox14, Janet Franklin15, William Godsoe16, Antoine Guisan17,18, Robert B. O’Hara19, Nicole Hill20, Robert D. Holt21, Francis K. C. Hui22, Magne Husby23,23, John Atle Kålås24, Aleksi Lehikoinen25, Miska Luoto26, Heidi K. Mod18, Graeme Newell27, Ian Renner28, Tomas Roslin2,29, Janne Soininen26, Wilfried Thuiller30, Jarno Vanhatalo1, David I. Warton31, Matt White27, Niklaus E. Zimmermann32, Dominique Gravel4, Otso Ovaskainen3,1
1Organismal and Evolutionary Biology Research Programme, University of Helsinki, P.O. Box 65, Helsinki, FI-00014 Finland
2Department of Agricultural Sciences, University of Helsinki, P.O. Box 27, Helsinki, FI-00014 Finland
3Department of Biology, Centre for Biodiversity Dynamics, Norwegian University of Science and Technology, Trondheim, N-7491 Norway
4Département de Biologie Université de Sherbrooke 2500 boulevard de l'Université Sherbrooke Quebec J1K 2R1 Canada
5Department of Mathematics, University of Utah, 155 South 1400 East, Salt Lake City, Utah, 84112 USA
6School of Biological Sciences, University of Utah, 257 South 1400 East, Salt Lake City, Utah 84112, USA
7Manaaki Whenua Landcare Research, Private Bag 1930, Dunedin, 1954 New Zealand
8Center for Macroecology, Evolution and Climate, Natural History Museum of Denmark, University of Copenhagen, Copenhagen, 2100 Denmark
9Departmento de Biogeografía y Cambio Global Museo Nacional de Ciencias Naturales Consejo Superior de Investigaciones Científicas (CSIC) Calle José Gutiérrez Abascal 2 Madrid 28006 Spain
10Rui Nabeiro Biodiversity Chair Universidade de Évora Largo dos Colegiais Evora 7000 Portugal
11Department of Statistical Science, Duke University, P.O. Box 90251, Durham, North Carolina, 27708 USA
12School of Biosciences, University of Melbourne, Parkville, Victoria 3010, Australia
13Commonwealth Scientific and Industrial Research Organisation (CSIRO), Hobart, Tasmania, Australia
14Butterfly Conservation, Manor Yard, East Lulworth, Wareham, BH20 5QP United Kingdom
15Department of Botany and Plant Sciences, University of California, Riverside, California 92521, USA
16Bio-Protection Research Centre, Lincoln University, P.O.Box 85084, Lincoln 7647, New Zealand
17Department of Ecology and Evolution (DEE), University of Lausanne, Biophore, Lausanne, CH-1015 Switzerland
18Institute of Earth Surface Dynamics (IDYST), University of Lausanne, Geopolis, Lausanne, CH-1015 Switzerland
19Department of Mathematical Sciences, Norwegian University of Science and Technology, Trondheim, N-7491, Norway
20Institute for Marine and Antarctic Studies, University of Tasmania, Private Bag 49, Hobart, Tasmania, 7001, Australia
21Department of Biology, The University of Florida, Gainesville, Florida, 32611 USA
22Mathematical Sciences Institute, The Australian National University, Acton, Australian Capital Territory, 2601 Australia
23BirdLife Norway Sandgata 30B Trondheim 7012 Norway
24Norwegian Institute for Nature Research, P.O. Box 5685 Torgarden, Trondheim NO-7485, Norway
25The Helsinki Lab of Ornithology, Finnish Museum of Natural History, University of Helsinki, P.O. Box 17, Helsinki, FI-00014 Finland
26Department of Geosciences and Geography, University of Helsinki, P.O. Box 64, Helsinki 00014, Finland
27Biodiversity Division, Department of Environment, Land, Water & Planning, Arthur Rylah Institute for Environmental Research, 123 Brown Street, Heidelberg, Victoria, 3084 Australia
28School of Mathematical and Physical Sciences, The University of Newcastle, University Drive, Callaghan, New South Wales, 2308 Australia
29Department of Ecology, Swedish University of Agricultural Sciences, Box 7044, Uppsala 750 07, Sweden
30CNRS LECA Laboratoire d’Écologie Alpine University Grenoble Alpes Grenoble F‐38000 France
31School of Mathematics and Statistics, Evolution & Ecology Research Centre, University of New South Wales, Sydney, New South Wales, 2052 Australia
32Dynamic Macroecology, Swiss Federal Research Institute WSL, Zuercherstrasse 111, Birmensdorf, CH-8903 Switzerland

Tóm tắt

AbstractA large array of species distribution model (SDM) approaches has been developed for explaining and predicting the occurrences of individual species or species assemblages. Given the wealth of existing models, it is unclear which models perform best for interpolation or extrapolation of existing data sets, particularly when one is concerned with species assemblages. We compared the predictive performance of 33 variants of 15 widely applied and recently emerged SDMs in the context of multispecies data, including both joint SDMs that model multiple species together, and stacked SDMs that model each species individually combining the predictions afterward. We offer a comprehensive evaluation of these SDM approaches by examining their performance in predicting withheld empirical validation data of different sizes representing five different taxonomic groups, and for prediction tasks related to both interpolation and extrapolation. We measure predictive performance by 12 measures of accuracy, discrimination power, calibration, and precision of predictions, for the biological levels of species occurrence, species richness, and community composition. Our results show large variation among the models in their predictive performance, especially for communities comprising many species that are rare. The results do not reveal any major trade‐offs among measures of model performance; the same models performed generally well in terms of accuracy, discrimination, and calibration, and for the biological levels of individual species, species richness, and community composition. In contrast, the models that gave the most precise predictions were not well calibrated, suggesting that poorly performing models can make overconfident predictions. However, none of the models performed well for all prediction tasks. As a general strategy, we therefore propose that researchers fit a small set of models showing complementary performance, and then apply a cross‐validation procedure involving separate data to establish which of these models performs best for the goal of the study.

Từ khóa


Tài liệu tham khảo

10.1126/science.1215320

10.1111/j.1365-2664.2006.01214.x

10.1111/ele.12155

10.1111/j.1365-2699.2006.01584.x

10.1111/j.1466-8238.2007.00359.x

10.1111/j.1365-2486.2005.01000.x

10.1111/j.1600-0587.2011.06919.x

Asher J., 2001, The millennium atlas of butterflies in Britain and Ireland

10.2307/2404755

10.1111/j.1365-2664.2006.01164.x

10.1016/j.ecolmodel.2006.05.023

10.1111/j.1365-2745.2009.01494.x

10.1111/j.1600-0706.2012.00299.x

10.1111/j.1466-8238.2009.00490.x

10.1111/j.1600-0587.2009.05856.x

10.1111/j.1365-2699.2010.02341.x

10.1098/rstb.2011.0178

10.2307/3237218

10.1034/j.1600-0587.2001.240106.x

10.1371/journal.pone.0054179

10.1111/j.1461-0248.2012.01772.x

10.1111/2041-210X.12163

10.1007/s00265-010-1029-6

10.1007/s12080-015-0281-9

10.1111/j.1365-2699.2011.02517.x

Chen T. et al.2018.xgboost: extreme gradient boosting. R package version 0.71.2.https://CRAN.R-project.org/package=xgboost

10.1111/j.1461-0248.2004.00702.x

10.1890/13-1015.1

10.1002/ecm.1241

10.1111/gcb.12598

10.18637/jss.v023.i10

10.1111/geb.12357

10.1111/brv.12222

De'Ath G. T. M.Therneau B.Atkinson B.Ripley andJ.Oksanen.2014.Mvpart: multivariate partitioning. R package version 1.6‐2.https://CRAN.R-project.org/package=mvpart

10.1007/978-0-387-48536-2

10.1046/j.1466-822X.2003.00322.x

10.1111/j.1466-8238.2006.00279.x

10.1111/j.2007.0906-7590.05171.x

10.1111/j.1365-2699.2011.02659.x

10.1111/j.1472-4642.2011.00792.x

10.1016/j.ecolmodel.2010.11.030

10.1111/j.1600-0587.2008.05505.x

10.1146/annurev.ecolsys.110308.120159

10.1111/j.2006.0906-7590.04596.x

10.1111/j.1461-0248.2004.00603.x

10.1086/521315

10.1111/j.1365-2664.2006.01149.x

10.1098/rstb.2010.0079

10.1046/j.1523-1739.1995.09061466.x

10.1111/j.1541-0420.2009.01263.x

10.1111/j.1467-9876.2011.01030.x

10.2307/3237291

10.18637/jss.v033.i01

10.1016/j.ecolmodel.2015.06.002

10.1201/b16018

10.1016/j.tree.2010.03.002

10.1198/016214506000001437

10.1002/ece3.2657

10.1016/j.tree.2017.03.008

Golding N. andD. J.Harris.2015.BayesComm: Bayesian community ecology analysis. R package version 0.1‐2.https://cran.r-project.org/package=BayesComm

10.1111/j.1469-185X.2011.00187.x

10.1111/ecog.02445

10.1111/2041-210X.12252

10.1016/S0304-3800(02)00204-1

10.1111/j.1365-2664.2006.01164.x

10.1111/j.1365-2699.2011.02550.x

10.1111/j.1461-0248.2005.00792.x

10.1023/A:1009841519580

10.1890/06-0539

10.1890/06-1060.1

10.1111/j.1472-4642.2007.00342.x

10.1214/13-STS446

10.1111/j.1461-0248.2011.01651.x

10.1111/2041-210X.12332

10.7717/peerj.4278

10.1111/j.1466-8238.2007.00345.x

Hijmans R. J. S. J.Phillips J. R.Leathwick andJ.Elith.2017.Dismo: Species distribution modeling. R package version 1.1‐4.https://CRAN.R-project.org/package=dismo

10.1111/j.1600-0587.2009.05495.x

10.1111/oik.03726

Hui F. K. C.2017.Boral: Bayesian ordination and regression analysis. R package version 1.4.https://CRAN.R-project.org/package=boral

10.1890/12-1322.1

10.1007/978-1-4614-7138-7

10.1016/j.actao.2007.02.001

10.1016/j.baae.2008.10.006

10.1111/j.1365-2699.2011.02663.x

10.1890/04-0609

10.1111/2041-210X.12123

10.1890/0012-9658(2001)082[2560:CIBTSI]2.0.CO;2

10.1016/j.ecolmodel.2006.05.022

10.1016/0304-3800(95)00142-5

Liaw A., 2002, Classification and regression by random forest, R News, 2, 18

10.5253/arde.v103i1.a1

10.1111/j.0906-7590.2005.03957.x

10.1111/j.1523-1739.2003.00233.x

10.1111/j.1600-0587.2013.00127.x

10.1111/j.1365-2699.2006.01465.x

Maguire K. C., 2016, Controlled comparison of species‐ and community‐level models across novel climates and communities, Proceedings of the Royal Society B, 283, 1

10.1111/ddi.12776

10.1046/j.1365-2664.1999.00440.x

10.1111/j.1472-4642.2008.00491.x

10.1046/j.1365-2427.1997.00209.x

10.1111/j.1365-2745.2004.00838.x

10.1023/A:1011944805738

10.1111/j.1365-2699.2010.02405.x

10.1111/ecog.00845

Meyer D. E.Dimitriadou K.Hornik A.Weingessel andF.Leisch.2017.e1071: Misc Functions of the Department of Statistics Probability Theory Group (Formerly: E1071) TU Wien. R package version 1.6‐8.https://cran.r-project.org/package=e1071

10.1111/j.1365-2699.2007.01720.x

Milborrow S.2017.Earth: Multivariate adaptive regression splines. Derived from mda:mars by Trevor Hastie and Rob Tibshirani. Uses Alan Miller's Fortran utilities with Thomas Lumley's leaps wrapper. R package version 4.5.1.https://cran.r-project.org/package=earth

10.1177/0309133312442522

10.1177/0309133314521448

10.1111/ecog.01129

10.1016/S0304-3800(02)00197-7

10.1111/j.1365-2745.2008.01369.x

10.1371/journal.pbio.1001569

10.1111/2041-210X.12936

Niittynen P., 2017, The importance of snow in species distribution models of arctic vegetation, Ecography, 40, 1

Norberg A., 2019, aminorberg/SDM‐comparison: Norberg et al. (2019) (Version publication)

10.1111/j.1466-8238.2009.00451.x

10.1046/j.1365-2427.2002.00945.x

10.1111/2041-210X.12501

10.1111/ele.12757

10.1890/10-1251.1

10.1016/j.ecolmodel.2013.04.011

10.1093/acprof:oso/9780199577859.001.0001

10.1111/j.1365-2699.2006.01460.x

10.1111/j.1600-0587.2010.06386.x

10.1111/j.0906-7590.2007.05102.x

10.23943/princeton/9780691136868.001.0001

10.1111/2041-210X.12180

R Core Team, 2018, R: a language and environment for statistical computing

10.1111/j.1365-2699.2006.01466.x

10.1111/geb.12017

Ridgeway G.2017.Gbm: Generalized boosted regression models. R package version 2.1.3.https://cran.r-project.org/package=gbm

10.1111/ecog.02881

10.1111/j.1466-8238.2010.00581.x

10.1016/j.ecolmodel.2009.06.004

10.1111/j.1365-2699.2004.01076.x

10.1016/j.ecolmodel.2017.03.006

10.1016/j.fishres.2004.08.009

10.1111/j.1600-0587.2009.05883.x

10.1111/geb.12395

The MathWorks Inc, 2015, MATLAB and statistics toolbox release 2015a

10.1111/2041-210X.12203

10.1111/2041-210X.12359

10.1111/j.1365-2486.2004.00859.x

10.1111/j.1654-1103.2003.tb02199.x

10.1111/ele.12104

10.1111/2041-210X.12723

10.1111/ele.12760

10.1126/science.aad8466

10.2307/3236575

10.1086/652373

10.1007/978-0-387-21706-2

10.1111/j.2041-210X.2012.00190.x

10.1016/j.tree.2015.09.007

10.1098/rstb.2011.0056

10.1111/j.2041-210X.2011.00170.x

10.1111/2041-210X.13106

10.1890/070037

10.1111/j.1472-4642.2008.00482.x

10.1111/j.1469-185X.2012.00235.x

10.1111/j.1467-9868.2010.00749.x

10.1111/ecog.03571

10.1111/j.1600-0587.2010.06953.x

10.1111/gcb.13251

Thông tin
Thông tin xuất bản

Ecological Monographs

Tập 89 Số 3

Thông tin tác giả