MaxEnt brings comparable results when the input data are being completed; Model parameterization of four species distribution models

Mohsen Ahmadi; Mahmoud‐Reza Hemami; Mohammad Kaboli; Farzin Shabani

doi:10.1002/ece3.9827

journal article Open Access Feb 01, 2023

MaxEnt brings comparable results when the input data are being completed; Model parameterization of four species distribution models

Mohsen Ahmadi

Mahmoud‐Reza Hemami

Mohammad Kaboli

Farzin Shabani

Ecology and Evolution Vol. 13 No. 2 · Wiley

View at Publisher Save 10.1002/ece3.9827

Abstract

AbstractSpecies distribution models (SDMs) are practical tools to assess the habitat suitability of species with numerous applications in environmental management and conservation planning. The manipulation of the input data to deal with their spatial bias is one of the advantageous methods to enhance the performance of SDMs. However, the development of a model parameterization approach covering different SDMs to achieve well‐performing models has rarely been implemented. We integrated input data manipulation and model tuning for four commonly‐used SDMs: generalized linear model (GLM), gradient boosted model (GBM), random forest (RF), and maximum entropy (MaxEnt), and compared their predictive performance to model geographically imbalanced‐biased data of a rare species complex of mountain vipers. Models were tuned up based on a range of model‐specific parameters considering two background selection methods: random and background weighting schemes. The performance of the fine‐tuned models was assessed based on recently identified localities of the species. The results indicated that although the fine‐tuned version of all models shows great performance in predicting training data (AUC > 0.9 and TSS > 0.5), they produce different results in classifying out‐of‐bag data. The GBM and RF with higher sensitivity of training data showed more different performances. The GLM, despite having high predictive performance for test data, showed lower specificity. It was only the MaxEnt model that showed high predictive performance and comparable results for identifying test data in both random and background weighting procedures. Our results highlight that while GBM and RF are prone to overfitting training data and GLM over‐predict nonsampled areas MaxEnt is capable of producing results that are both predictable (extrapolative) and complex (interpolative). We discuss the assumptions of each model and conclude that MaxEnt could be considered as a practical method to cope with imbalanced‐biased data in species distribution modeling approaches.

Topics

No keywords indexed for this article. Browse by subject →

References

82

[1]

10.1007/s10980-020-01033-5

[2]

10.1038/s41598-019-42792-9

[3]

10.1186/s12862-021-01863-0

[4]

10.1016/j.ympev.2018.06.019

[5]

10.1111/jbi.12703

[6]

10.1126/sciadv.aat4858

[7]

10.1111/j.1365-2699.2006.01584.x

[8]

10.1093/jhered/esy005

[9]

Spatial filtering to reduce sampling bias can improve the performance of ecological niche models

Robert A. Boria, Link E. Olson, Steven M. Goodman et al.

Ecological Modelling 10.1016/j.ecolmodel.2013.12.012

[10]

10.1007/s10530-022-02838-y

[11]

10.1111/jbi.12652

[12]

Random Forests

Leo Breiman

Machine Learning 10.1023/a:1010933404324

[13]

10.1016/j.biocon.2009.08.001

[14]

Candel A. (2016)

[15]

10.1111/ecog.05973

[16]

10.1111/geb.13383

[17]

10.1007/978-1-4419-9326-7_5

[18]

10.1890/0012-9658(2000)081[3178:cartap]2.0.co;2

[19]

10.1111/j.2007.0906-7590.05171.x

[20]

10.1111/ecog.03149

[21]

10.1002/ece3.3834

[22]

Novel methods improve prediction of species’ distributions from occurrence data

Jane Elith*, Catherine H. Graham*, Robert P. Anderson et al.

Ecography 10.1111/j.2006.0906-7590.04596.x

[23]

10.1111/j.1600-0587.2008.05505.x

[24]

The art of modelling range-shifting species

Jane Elith, Michael Kearney, Steven Phillips

Methods in Ecology and Evolution 10.1111/j.2041-210x.2010.00036.x

[25]

A working guide to boosted regression trees

J. Elith, J. R. Leathwick, T. Hastie

Journal of Animal Ecology 10.1111/j.1365-2656.2008.01390.x

[26]

10.1016/j.biocon.2015.08.027

[27]

A review of methods for the assessment of prediction errors in conservation presence/absence models

Alan H. FIELDING, JOHN F. BELL

Environmental Conservation 10.1017/s0376892997000088

[28]

10.1111/2041-210x.12242

[29]

10.1017/cbo9780511810602

[30]

10.1111/ecog.02909

[31]

10.1111/gcb.13992

[32]

10.1111/geb.12268

[33]

10.1111/j.1472-4642.2007.00342.x

[34]

Hardin J. W. (2007)

[35]

The Elements of Statistical Learning

Trevor Hastie, Robert Tibshirani, Jerome Friedman

Springer Series in Statistics 2009 10.1007/978-0-387-84858-7

[36]

Hemami M.‐R. "Using ecological models to explore niche partitioning within a guild of desert felids" Hystrix, the Italian Journal of Mammalogy (2018)

[37]

10.1890/11-0826.1

[38]

Very high resolution interpolated climate surfaces for global land areas

ROBERT J. HIJMANS, Susan E. Cameron, Juan L. Parra et al.

International Journal of Climatology 10.1002/joc.1276

[39]

Development of a two-band enhanced vegetation index without a blue band

Z Jiang, A Huete, K Didan et al.

Remote Sensing of Environment 10.1016/j.rse.2008.06.006

[40]

Insights into the area under the receiver operating characteristic curve (AUC) as a discrimination measure in species distribution modelling

Alberto Jiménez‐Valverde

Global Ecology and Biogeography 10.1111/j.1466-8238.2011.00683.x

[41]

ENMeval 2.0: Redesigned for customizable and reproducible modeling of species’ niches and distributions

Jamie M. Kass, Robert Muscarella, Peter J. Galante et al.

Methods in Ecology and Evolution 10.1111/2041-210x.13628

[42]

10.1016/j.envsoft.2017.11.009

[43]

10.1111/ddi.12096

[44]

Kuhn M.(2021).Caret: Classification and regression training. R Package Version 6.0‐90.https://CRAN.R‐project.org/package=caret

[45]

10.1111/ecog.01252

[46]

10.1111/j.0906-7590.2005.03957.x

[47]

AUC: a misleading measure of the performance of predictive distribution models

Jorge M. Lobo, Alberto Jiménez‐Valverde, Raimundo Real

Global Ecology and Biogeography 10.1111/j.1466-8238.2007.00358.x

[48]

10.1002/ece3.9624

[49]

What do we gain from simplicity versus complexity in species distribution models?

Cory Merow, Mathew J. Smith, Thomas C. Edwards et al.

Ecography 10.1111/ecog.00845

[50]

ENMeval: An R package for conducting spatially independent evaluations and estimating optimal model complexity for Maxent ecological niche models

Robert Muscarella, Peter J. Galante, Mariano Soley‐Guardia et al.

Methods in Ecology and Evolution 10.1111/2041-210x.12261

Showing 50 of 82 references

Metrics

152

Citations

82

References

Details

Published: Feb 01, 2023
Vol/Issue: 13(2)
License: View

Authors

M

Mohsen Ahmadi

Department of Natural Resources Isfahan University of Technology Isfahan Iran

M

Mahmoud‐Reza Hemami

Department of Natural Resources Isfahan University of Technology Isfahan Iran

M

Mohammad Kaboli

Department of Environmental Sciences, Faculty of Natural Resources University of Tehran Karaj Iran

F

Farzin Shabani

Department of Biological and Environmental Sciences College of Arts and Sciences, Qatar University Doha Qatar

Cite This Article

Mohsen Ahmadi, Mahmoud‐Reza Hemami, Mohammad Kaboli, et al. (2023). MaxEnt brings comparable results when the input data are being completed; Model parameterization of four species distribution models. Ecology and Evolution, 13(2). https://doi.org/10.1002/ece3.9827

MaxEnt brings comparable results when the input data are being completed; Model parameterization of four species distribution models

You May Also Like