CoChat
Sign Up
Cited by 464

The DOI matches its Crossref record; 3 of the 3 references it cites were found in a registry or an index.

MaxEnt’s parameter configuration and small samples: are we paying attention to recommendations? A systematic review

Morales, Narkis S., Fernández, Ignacio C., Baca-González, Victoria·2017·PeerJ
DOI 10.7717/peerj.3093PMC PMC5354112

Published version of a preprint

Open this paper in CoChat — free account

Abstract

Environmental niche modeling (ENM) is commonly used to develop probabilistic maps of species distribution. Among available ENM techniques, MaxEnt has become one of the most popular tools for modeling species distribution, with hundreds of peer-reviewed articles published each year. MaxEnt’s popularity is mainly due to the use of a graphical interface and automatic parameter configuration capabilities. However, recent studies have shown that using the default automatic configuration may not be always appropriate because it can produce non-optimal models; particularly when dealing with a small number of species presence points. Thus, the recommendation is to evaluate the best potential combination of parameters (feature classes and regularization multiplier) to select the most appropriate model. In this work we reviewed 244 articles published between 2013 and 2015 to assess whether researchers are following recommendations to avoid using the default parameter configuration when dealing with small sample sizes, or if they are using MaxEnt as a “black box tool.” Our results show that in only 16% of analyzed articles authors evaluated best feature classes, in 6.9% evaluated best regularization multipliers, and in a meager 3.7% evaluated simultaneously both parameters before producing the definitive distribution model. We analyzed 20 articles to quantify the potential differences in resulting outputs when using software default parameters instead of the alternative best model. Results from our analysis reveal important differences between the use of default parameters and the best model approach, especially in the total area identified as suitable for the assessed species and the specific areas that are identified as suitable by both modelling approaches. These results are worrying, because publications are potentially reporting over-complex or over-simplistic models that can undermine the applicability of their results. Of particular importance are studies used to inform policy making. Therefore, researchers, practitioners, reviewers and editors need to be very judicious when dealing with MaxEnt, particularly when the modelling process is based on small sample sizes.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.7717/peerj.3093

Cites

3 of 3 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Number of published articles (2004–2015) containing both “MaxEnt” and “species distribution” within the topic in the Web of Knowledge Databases (see ‘Methods’ section for databases details).
Figure 1 Number of published articles (2004–2015) containing both “MaxEnt” and “species distribution” within the topic in the Web of Knowledge Databases (see ‘Methods’ section for databases details).CC BY · PubMed Central
Figure 2 PRISMA flow diagram of the used search protocol following Moher et al. (2009).
Figure 2 PRISMA flow diagram of the used search protocol following Moher et al. (2009).CC BY · PubMed Central
Figure 3 Feature classes (A) and regularization multipliers (B) reported to be used for modelling in the analyzed articles. Columns show the percentage of articles using user-defined, software default, and articles not providing information. Numbers on top of columns represent the number of articles pertaining to each category per year. Columns on the right of each category show the percentage and number of articles for the 2013–2015 period.
Figure 3 Feature classes (A) and regularization multipliers (B) reported to be used for modelling in the analyzed articles. Columns show the percentage of articles using user-defined, software default, and articles not providing information. Numbers on top of columns represent the number of articles pertaining to each category per year. Columns on the right of each category show the percentage and number of articles for the 2013–2015 period.CC BY · PubMed Central
Figure 4 Replicability of the modelling process performed in analyzed articles. Columns show the percentage of articles providing information about GC, geographical coordinates, FC, feature classes, RM, regularization multiplier. Numbers above columns report the number of articles pertaining to each category. Only articles providing information regarding the three inputs (i.e., GC + F + RM column) are considered to provide enough information for replicating the modelling process.
Figure 4 Replicability of the modelling process performed in analyzed articles. Columns show the percentage of articles providing information about GC, geographical coordinates, FC, feature classes, RM, regularization multiplier. Numbers above columns report the number of articles pertaining to each category. Only articles providing information regarding the three inputs (i.e., GC + F + RM column) are considered to provide enough information for replicating the modelling process.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Research works printed but not linked: 47

  1. Aguiar, 2015, Effect of chronological addition of records to species distribution maps: the case of Tonatia saurophila maresi (Chiroptera, Phyllostomidae) in South America, Austral Ecology, 40, 836, 10.1111/aec.12261 · DOI
  2. Alamgir, 2015, Modelling spatial distribution of critically endangered Asian elephant and Hoolock gibbon in Bangladesh forest ecosystems under a changing climate, Applied Geography, 60, 10, 10.1016/j.apgeog.2015.03.001 · DOI
  3. Alfaro-Saiz, 2015, Incorporating bioclimatic and biogeographic data in the construction of species distribution models in order to prioritize searches for new populations of threatened flora, Plant Biosystems, 149, 827, 10.1080/11263504.2014.976289 · DOI
  4. Amici, 2015, Habitat suitability and landscape structure: a maximum entropy approach in a mediterranean area, Landscape Research, 40, 208, 10.1080/01426397.2013.774329 · DOI
  5. Anderson, 2011, Species-specific tuning increases robustness to sampling bias in models of species distributions: an implementation with Maxent, Ecological Modelling, 222, 2796, 10.1016/j.ecolmodel.2011.04.011 · DOI
  6. Bernardes, 2013, Habitat characterization and potential distribution of Tylototriton vietnamensis in northern Vietnam, Journal of Natural History, 47, 1161, 10.1080/00222933.2012.743611 · DOI
  7. Bouchet, 2015, Drifting baited stereo-videography: a novel sampling tool for surveying pelagic wildlife in offshore marine reserves, Ecosphere, 6, 1, 10.1890/ES14-00174.1 · DOI
  8. Brambilla, 2013, Modelling distribution and potential overlap between Boreal Owl Aegolius funereus and Black Woodpecker Dryocopus martius: implications for management and monitoring plans, Bird Conservation International, 23, 502, 10.1017/S0959270913000117 · DOI
  9. Cao, 2013, Using Maxent to model the historic distributions of stonefly species in Illinois streams: the effects of regularization and threshold selections, Ecological Modelling, 259, 30, 10.1016/j.ecolmodel.2013.03.012 · DOI
  10. Carvalho, 2015, Predicting priority areas for conservation from historical climate modelling: stingless bees from Atlantic Forest hotspot as a case study, Journal of Insect Conservation, 19, 581, 10.1007/s10841-015-9780-7 · DOI
  11. Ceccarelli, 2015, Global climate change effects on Venezuela’s vulnerability to chagas disease is linked to the geographic distribution of five triatomine species, Journal of Medical Entomology, 52, 1333, 10.1093/jme/tjv119 · DOI
  12. Chetan, 2014, Delineating ecological boundaries of Hanuman Langur species complex in Peninsular India using MaxEnt modeling approach, PLOS ONE, 9, e87804, 10.1371/journal.pone.0087804 · DOI
  13. Chłond, 2015, Current and potential geographical distribution of Platymeris biguttatus (Linnaeus, 1767) with description of nymphs, Zoological Studies, 54, 9, 10.1186/s40555-014-0092-5 · DOI
  14. Chunco, 2013, Predicting environmental suitability for a rare and threatened species (Lao Newt, Laotriton laoensis) using validated species distribution models, PLOS ONE, 8, e59853, 10.1371/journal.pone.0059853 · DOI
  15. Conflitti, 2015, The speciation continuum: ecological and chromosomal divergence in the Simulium arcticum complex (Diptera: Simuliidae), Biological Journal of the Linnean Society, 115, 13, 10.1111/bij.12480 · DOI
  16. Crafton, 2015, Modeling invasion risk for coastal marine species utilizing environmental and transport vector data, Hydrobiologia, 746, 349, 10.1007/s10750-014-2027-x · DOI
  17. De Castro Pena, 2014, Assessing the conservation status of species with limited available data and disjunct distribution, Biological Conservation, 170, 130, 10.1016/j.biocon.2013.12.015 · DOI
  18. Elith, 2006, Novel methods improve prediction of species’ distributions from occurrence data, Ecography, 29, 129, 10.1111/j.2006.0906-7590.04596.x · DOI · Paper page
  19. Elith, 2009, Species distribution models: ecological explanation and prediction across space and time, Annual Review of Ecology, Evolution, and Systematics, 40, 677, 10.1146/annurev.ecolsys.110308.120159 · DOI · Paper page
  20. Fernández, 2016, A spatial multicriteria decision analysis for selecting priority sites for plant species restoration: a case study from the Chilean biodiversity hotspot, Restoration Ecology, 24, 599, 10.1111/rec.12354 · DOI
  21. Fitzpatrick, 2013, MaxEnt versus MaxLike: empirical comparisons with ant species distributions, Ecosphere, 4, 1, 10.1890/ES13-00066.1 · DOI
  22. Fois, 2015, A practical method to speed up the discovery of unknown populations using Species Distribution Models, Journal for Nature Conservation, 24, 42, 10.1016/j.jnc.2015.02.001 · DOI
  23. Gelviz-Gelvez, 2015, Ecological niche modeling under climate change to select shrubs for ecological restoration in Central Mexico, Ecological Engineering, 74, 302, 10.1016/j.ecoleng.2014.09.082 · DOI
  24. Guisan, 2000, Predictive habitat distribution models in ecology, Ecological Modelling, 135, 147, 10.1016/S0304-3800(00)00354-9 · DOI · Paper page
  25. Halvorsen, 2015, Opportunities for improved distribution modelling practice via a strict maximum likelihood interpretation of MaxEnt, Ecography, 38, 172, 10.1111/ecog.00565 · DOI
  26. Hu, 2014, Unveiling the conservation biogeography of a data-deficient endangered bird species under climate change, PLOS ONE, 9, e84529, 10.1371/journal.pone.0084529 · DOI
  27. Kramer-Schadt, 2013, The importance of correcting for sampling bias in MaxEnt species distribution models, Diversity and Distributions, 19, 1366, 10.1111/ddi.12096 · DOI · Paper page
  28. Merow, 2013, A practical guide to MaxEnt for modeling species’ distributions: what it does, and why inputs and settings matter, Ecography, 36, 1058, 10.1111/j.1600-0587.2013.07872.x · DOI · Paper page
  29. Mestre, 2015, Combining distribution modelling and non-invasive genetics to improve range shift forecasting, Ecological Modelling, 297, 171, 10.1016/j.ecolmodel.2014.11.018 · DOI
  30. Meyer, 2014, Assessing the exposure of lion tamarins (Leontopithecus spp.) to future climate change, American Journal of Primatology, 76, 551, 10.1002/ajp.22247 · DOI
  31. Miller, 2010, Species distribution modeling, Geography Compass, 4, 490, 10.1111/j.1749-8198.2010.00351.x · DOI
  32. Mweya, 2013, Predicting distribution of Aedes aegypti and Culex pipiens complex, potential vectors of Rift Valley fever virus in relation to disease epidemics in East Africa, Infection Ecology & Epidemiology, 3, 21748, 10.3402/iee.v3i0.21748 · DOI
  33. Orr, 2014, Taxonomic utility of niche models in validating species concepts: a case study in Anthophora (Heliophila) (Hymenoptera: Apidae), Zootaxa, 3846, 10.11646/zootaxa.3846.3.5 · DOI
  34. Palmas-Pérez, 2013, Diversidad y distribución de Marcgraviaceae en México, Revista Mexicana de Biodiversidad, 84, 170, 10.7550/rmb.29666 · DOI
  35. Pedersen, 2014, Modelling spatial distribution of snails transmitting parasitic worms with importance to human and animal health and analysis of distributional changes in relation to climate, Geospatial Health, 8, 10.4081/gh.2014.23 · DOI
  36. Phillips, 2006, Maximum entropy modeling of species geographic distributions, Ecological Modelling, 190, 231, 10.1016/j.ecolmodel.2005.03.026 · DOI · Paper page
  37. Phillips, 2008, Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation, Ecography, 31, 161, 10.1111/j.0906-7590.2008.5203.x · DOI · Paper page
  38. Radosavljevic, 2014, Making better Maxent models of species distributions: complexity, overfitting and evaluation, Journal of Biogeography, 41, 629, 10.1111/jbi.12227 · DOI · Paper page
  39. Shcheglovitova, 2013, Estimating optimal complexity for ecological niche models: a jackknife approach for species with small sample sizes, Ecological Modelling, 269, 9, 10.1016/j.ecolmodel.2013.08.011 · DOI
  40. Simo, 2014, A predictive modeling approach to test distributional uniformity of Uruguayan harvestmen (Arachnida: Opiliones), Zoological Studies, 53, 50, 10.1186/s40555-014-0050-2 · DOI
  41. Stigall, 2014, Relating environmental change and species stability in Late Ordovician seas, GFF, 136, 249, 10.1080/11035897.2013.852619 · DOI
  42. Syfert, 2013, The effects of sampling bias and model complexity on the predictive performance of MaxEnt species distribution models, PLOS ONE, 8, e55158, 10.1371/journal.pone.0055158 · DOI
  43. Urbani, 2015, Maximum entropy modeling of geographic distributions of the flea beetle species endemic in Italy (Coleoptera: Chrysomelidae: Galerucinae: Alticini), Zoologischer Anzeiger, 258, 99, 10.1016/j.jcz.2015.08.002 · DOI
  44. Velásquez-Tibatá, 2016, Using measurement error models to account for georeferencing error in species distribution models, Ecography, 39, 305, 10.1111/ecog.01205 · DOI
  45. Vergara, 2015, More on the Mesopotamian-Yungas disjunction in subtropical and temperate Argentina: bioclimatic distribution models of the harvestman Discocyrtus dilatatus (Opiliones: Gonyleptidae), Zoologia (Curitiba), 32, 445, 10.1590/s1984-46702015000600003 · DOI
  46. Yackulic, 2013, Presence-only modelling using MAXENT: when can we trust the inferences?, Methods in Ecology and Evolution, 4, 236, 10.1111/2041-210x.12004 · DOI
  47. Yu, 2014, A model-based method to evaluate the ability of nature reserves to protect endangered tree species in the context of climate change, Forest Ecology and Management, 327, 48, 10.1016/j.foreco.2014.04.020 · DOI

Cited by

Sorted by citations

See all 464

Metadata

Authors
Morales, N. S., Fernández, I. C., & Baca-González, V.
Affiliations
  • Department of Biological Sciences, Faculty of Science and Engineering, Macquarie University, Sydney, New South Wales, Australia
  • Fundación Ecomabi, Santiago, Región Metropolitana, Chile
  • Facultad de Ciencias Biológicas, Universidad Complutense de Madrid, Madrid, Spain
Venue
PeerJ
Publisher
PeerJ
Published
2017-03-14
Volume
5
Pages
e3093
ISSN
2167-8359
DOI
10.7717/peerj.3093
PMC
PMC5354112
Cited by
464
References
3
Language
en
Licence
creativecommons.org/licenses/by/4.0/ (unspecified)

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review