Is my species distribution model fit for purpose? Matching data and models to applications

Gurutzeta Guillera-Arroita, José J. Lahoz-Monfort, Jane Elith, Ascelin Gordon, Heini Kujala, Pia E. Lentini, Michael A. Mccarthy, Reid Tingley, Brendan A. Wintle

Research output: Contribution to journalArticleResearchpeer-review

256 Citations (Scopus)

Abstract

Species distribution models (SDMs) are used to inform a range of ecological, biogeographical and conservation applications. However, users often underestimate the strong links between data type, model output and suitability for end-use. We synthesize current knowledge and provide a simple framework that summarizes how interactions between data type and the sampling process (i.e. imperfect detection and sampling bias) determine the quantity that is estimated by a SDM. We then draw upon the published literature and simulations to illustrate and evaluate the information needs of the most common ecological, biogeographical and conservation applications of SDM outputs. We find that, while predictions of models fitted to the most commonly available observational data (presence records) suffice for some applications, others require estimates of occurrence probabilities, which are unattainable without reliable absence records. Our literature review and simulations reveal that, while converting continuous SDM outputs into categories of assumed presence or absence is common practice, it is seldom clearly justified by the application's objective and it usually degrades inference. Matching SDMs to the needs of particular applications is critical to avoid poor scientific inference and management outcomes. This paper aims to help modellers and users assess whether their intended SDM outputs are indeed fit for purpose.

Original languageEnglish
Pages (from-to)276-292
Number of pages17
JournalGlobal Ecology and Biogeography
Volume24
Issue number3
DOIs
Publication statusPublished - 1 Jan 2015
Externally publishedYes

Keywords

  • Ecological niche model
  • Habitat model
  • Imperfect detection
  • Presence-absence
  • Presence-background
  • Presence-only
  • Prevalence
  • Sampling bias

Cite this

Guillera-Arroita, G., Lahoz-Monfort, J. J., Elith, J., Gordon, A., Kujala, H., Lentini, P. E., ... Wintle, B. A. (2015). Is my species distribution model fit for purpose? Matching data and models to applications. Global Ecology and Biogeography, 24(3), 276-292. https://doi.org/10.1111/geb.12268
Guillera-Arroita, Gurutzeta ; Lahoz-Monfort, José J. ; Elith, Jane ; Gordon, Ascelin ; Kujala, Heini ; Lentini, Pia E. ; Mccarthy, Michael A. ; Tingley, Reid ; Wintle, Brendan A. / Is my species distribution model fit for purpose? Matching data and models to applications. In: Global Ecology and Biogeography. 2015 ; Vol. 24, No. 3. pp. 276-292.
@article{99c04a59c00c4420891317a9d663c252,
title = "Is my species distribution model fit for purpose? Matching data and models to applications",
abstract = "Species distribution models (SDMs) are used to inform a range of ecological, biogeographical and conservation applications. However, users often underestimate the strong links between data type, model output and suitability for end-use. We synthesize current knowledge and provide a simple framework that summarizes how interactions between data type and the sampling process (i.e. imperfect detection and sampling bias) determine the quantity that is estimated by a SDM. We then draw upon the published literature and simulations to illustrate and evaluate the information needs of the most common ecological, biogeographical and conservation applications of SDM outputs. We find that, while predictions of models fitted to the most commonly available observational data (presence records) suffice for some applications, others require estimates of occurrence probabilities, which are unattainable without reliable absence records. Our literature review and simulations reveal that, while converting continuous SDM outputs into categories of assumed presence or absence is common practice, it is seldom clearly justified by the application's objective and it usually degrades inference. Matching SDMs to the needs of particular applications is critical to avoid poor scientific inference and management outcomes. This paper aims to help modellers and users assess whether their intended SDM outputs are indeed fit for purpose.",
keywords = "Ecological niche model, Habitat model, Imperfect detection, Presence-absence, Presence-background, Presence-only, Prevalence, Sampling bias",
author = "Gurutzeta Guillera-Arroita and Lahoz-Monfort, {Jos{\'e} J.} and Jane Elith and Ascelin Gordon and Heini Kujala and Lentini, {Pia E.} and Mccarthy, {Michael A.} and Reid Tingley and Wintle, {Brendan A.}",
year = "2015",
month = "1",
day = "1",
doi = "10.1111/geb.12268",
language = "English",
volume = "24",
pages = "276--292",
journal = "Global Ecology and Biogeography",
issn = "1466-822X",
publisher = "Wiley-Blackwell",
number = "3",

}

Guillera-Arroita, G, Lahoz-Monfort, JJ, Elith, J, Gordon, A, Kujala, H, Lentini, PE, Mccarthy, MA, Tingley, R & Wintle, BA 2015, 'Is my species distribution model fit for purpose? Matching data and models to applications', Global Ecology and Biogeography, vol. 24, no. 3, pp. 276-292. https://doi.org/10.1111/geb.12268

Is my species distribution model fit for purpose? Matching data and models to applications. / Guillera-Arroita, Gurutzeta; Lahoz-Monfort, José J.; Elith, Jane; Gordon, Ascelin; Kujala, Heini; Lentini, Pia E.; Mccarthy, Michael A.; Tingley, Reid; Wintle, Brendan A.

In: Global Ecology and Biogeography, Vol. 24, No. 3, 01.01.2015, p. 276-292.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Is my species distribution model fit for purpose? Matching data and models to applications

AU - Guillera-Arroita, Gurutzeta

AU - Lahoz-Monfort, José J.

AU - Elith, Jane

AU - Gordon, Ascelin

AU - Kujala, Heini

AU - Lentini, Pia E.

AU - Mccarthy, Michael A.

AU - Tingley, Reid

AU - Wintle, Brendan A.

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Species distribution models (SDMs) are used to inform a range of ecological, biogeographical and conservation applications. However, users often underestimate the strong links between data type, model output and suitability for end-use. We synthesize current knowledge and provide a simple framework that summarizes how interactions between data type and the sampling process (i.e. imperfect detection and sampling bias) determine the quantity that is estimated by a SDM. We then draw upon the published literature and simulations to illustrate and evaluate the information needs of the most common ecological, biogeographical and conservation applications of SDM outputs. We find that, while predictions of models fitted to the most commonly available observational data (presence records) suffice for some applications, others require estimates of occurrence probabilities, which are unattainable without reliable absence records. Our literature review and simulations reveal that, while converting continuous SDM outputs into categories of assumed presence or absence is common practice, it is seldom clearly justified by the application's objective and it usually degrades inference. Matching SDMs to the needs of particular applications is critical to avoid poor scientific inference and management outcomes. This paper aims to help modellers and users assess whether their intended SDM outputs are indeed fit for purpose.

AB - Species distribution models (SDMs) are used to inform a range of ecological, biogeographical and conservation applications. However, users often underestimate the strong links between data type, model output and suitability for end-use. We synthesize current knowledge and provide a simple framework that summarizes how interactions between data type and the sampling process (i.e. imperfect detection and sampling bias) determine the quantity that is estimated by a SDM. We then draw upon the published literature and simulations to illustrate and evaluate the information needs of the most common ecological, biogeographical and conservation applications of SDM outputs. We find that, while predictions of models fitted to the most commonly available observational data (presence records) suffice for some applications, others require estimates of occurrence probabilities, which are unattainable without reliable absence records. Our literature review and simulations reveal that, while converting continuous SDM outputs into categories of assumed presence or absence is common practice, it is seldom clearly justified by the application's objective and it usually degrades inference. Matching SDMs to the needs of particular applications is critical to avoid poor scientific inference and management outcomes. This paper aims to help modellers and users assess whether their intended SDM outputs are indeed fit for purpose.

KW - Ecological niche model

KW - Habitat model

KW - Imperfect detection

KW - Presence-absence

KW - Presence-background

KW - Presence-only

KW - Prevalence

KW - Sampling bias

UR - http://www.scopus.com/inward/record.url?scp=84922503905&partnerID=8YFLogxK

U2 - 10.1111/geb.12268

DO - 10.1111/geb.12268

M3 - Article

AN - SCOPUS:84922503905

VL - 24

SP - 276

EP - 292

JO - Global Ecology and Biogeography

JF - Global Ecology and Biogeography

SN - 1466-822X

IS - 3

ER -

Guillera-Arroita G, Lahoz-Monfort JJ, Elith J, Gordon A, Kujala H, Lentini PE et al. Is my species distribution model fit for purpose? Matching data and models to applications. Global Ecology and Biogeography. 2015 Jan 1;24(3):276-292. https://doi.org/10.1111/geb.12268