Skip to content

Research at St Andrews

Model selection with overdispersed distance sampling data

Research output: Contribution to journalArticle

DOI

Standard

Model selection with overdispersed distance sampling data. / Howe, Eric J; Buckland, Stephen T; Després-Einspenner, Marie-Lyne; Kühl, Hjalmar S.

In: Methods in Ecology and Evolution, Vol. 10, No. 1, 01.2019, p. 38-47.

Research output: Contribution to journalArticle

Harvard

Howe, EJ, Buckland, ST, Després-Einspenner, M-L & Kühl, HS 2019, 'Model selection with overdispersed distance sampling data' Methods in Ecology and Evolution, vol. 10, no. 1, pp. 38-47. https://doi.org/10.1111/2041-210X.13082

APA

Howe, E. J., Buckland, S. T., Després-Einspenner, M-L., & Kühl, H. S. (2019). Model selection with overdispersed distance sampling data. Methods in Ecology and Evolution, 10(1), 38-47. https://doi.org/10.1111/2041-210X.13082

Vancouver

Howe EJ, Buckland ST, Després-Einspenner M-L, Kühl HS. Model selection with overdispersed distance sampling data. Methods in Ecology and Evolution. 2019 Jan;10(1):38-47. https://doi.org/10.1111/2041-210X.13082

Author

Howe, Eric J ; Buckland, Stephen T ; Després-Einspenner, Marie-Lyne ; Kühl, Hjalmar S. / Model selection with overdispersed distance sampling data. In: Methods in Ecology and Evolution. 2019 ; Vol. 10, No. 1. pp. 38-47.

Bibtex - Download

@article{f0f1853b7149429795cc3503c6c09b48,
title = "Model selection with overdispersed distance sampling data",
abstract = "1. Distance sampling (DS) is a widely used framework for estimating animal abundance. DS models assume that observations of distances to animals are independent. Non‐independent observations introduce overdispersion, causing model selection criteria such as AIC or AICc to favour overly complex models, with adverse effects on accuracy and precision.2. We describe, and evaluate via simulation and with real data, estimators of an overdispersion factor (ĉ), and associated adjusted model selection criteria (QAIC) for use with overdispersed DS data. In other contexts, a single value of ĉ is calculated from the “global” model, that is the most highly parameterised model in the candidate set, and used to calculate QAIC for all models in the set; the resulting QAIC values, and associated ΔQAIC values and QAIC weights, are comparable across the entire set. Candidate models of the DS detection function include models with different general forms (e.g. half‐normal, hazard rate, uniform), so it may not be possible to identify a single global model. We therefore propose a two‐step model selection procedure by which QAIC is used to select among models with the same general form, and then a goodness‐of‐fit statistic is used to select among models with different forms. A drawback of thi approach is that QAIC values are not comparable across all models in the candidate set.3. Relative to AIC, QAIC and the two‐step model selection procedure avoided overfitting and improved the accuracy and precision of densities estimated from simulated data. When applied to six real datasets, adjusted criteria and procedures selected either the same model as AIC or a model that yielded a more accurate density estimate in five cases, and a model that yielded a less accurate estimate in one case. 4. Many DS surveys yield overdispersed data, including cue counting surveys of songbirds and cetaceans, surveys of social species including primates, and camera‐trapping surveys. Methods that adjust for overdispersion during the model selection stage of DS analyses therefore address a conspicuous gap in the DS analytical framework as applied to species of conservation concern.",
keywords = "Animal abundance, Camera trapping, Cue counting, Distance sampling, Model selection, Overdispersion, QAIC",
author = "Howe, {Eric J} and Buckland, {Stephen T} and Marie-Lyne Despr{\'e}s-Einspenner and K{\"u}hl, {Hjalmar S.}",
note = "We thank the Robert Bosch Foundation, the Max Planck Society and the University of St Andrews for funding.",
year = "2019",
month = "1",
doi = "10.1111/2041-210X.13082",
language = "English",
volume = "10",
pages = "38--47",
journal = "Methods in Ecology and Evolution",
issn = "2041-210X",
publisher = "John Wiley & Sons, Ltd (10.1111)",
number = "1",

}

RIS (suitable for import to EndNote) - Download

TY - JOUR

T1 - Model selection with overdispersed distance sampling data

AU - Howe, Eric J

AU - Buckland, Stephen T

AU - Després-Einspenner, Marie-Lyne

AU - Kühl, Hjalmar S.

N1 - We thank the Robert Bosch Foundation, the Max Planck Society and the University of St Andrews for funding.

PY - 2019/1

Y1 - 2019/1

N2 - 1. Distance sampling (DS) is a widely used framework for estimating animal abundance. DS models assume that observations of distances to animals are independent. Non‐independent observations introduce overdispersion, causing model selection criteria such as AIC or AICc to favour overly complex models, with adverse effects on accuracy and precision.2. We describe, and evaluate via simulation and with real data, estimators of an overdispersion factor (ĉ), and associated adjusted model selection criteria (QAIC) for use with overdispersed DS data. In other contexts, a single value of ĉ is calculated from the “global” model, that is the most highly parameterised model in the candidate set, and used to calculate QAIC for all models in the set; the resulting QAIC values, and associated ΔQAIC values and QAIC weights, are comparable across the entire set. Candidate models of the DS detection function include models with different general forms (e.g. half‐normal, hazard rate, uniform), so it may not be possible to identify a single global model. We therefore propose a two‐step model selection procedure by which QAIC is used to select among models with the same general form, and then a goodness‐of‐fit statistic is used to select among models with different forms. A drawback of thi approach is that QAIC values are not comparable across all models in the candidate set.3. Relative to AIC, QAIC and the two‐step model selection procedure avoided overfitting and improved the accuracy and precision of densities estimated from simulated data. When applied to six real datasets, adjusted criteria and procedures selected either the same model as AIC or a model that yielded a more accurate density estimate in five cases, and a model that yielded a less accurate estimate in one case. 4. Many DS surveys yield overdispersed data, including cue counting surveys of songbirds and cetaceans, surveys of social species including primates, and camera‐trapping surveys. Methods that adjust for overdispersion during the model selection stage of DS analyses therefore address a conspicuous gap in the DS analytical framework as applied to species of conservation concern.

AB - 1. Distance sampling (DS) is a widely used framework for estimating animal abundance. DS models assume that observations of distances to animals are independent. Non‐independent observations introduce overdispersion, causing model selection criteria such as AIC or AICc to favour overly complex models, with adverse effects on accuracy and precision.2. We describe, and evaluate via simulation and with real data, estimators of an overdispersion factor (ĉ), and associated adjusted model selection criteria (QAIC) for use with overdispersed DS data. In other contexts, a single value of ĉ is calculated from the “global” model, that is the most highly parameterised model in the candidate set, and used to calculate QAIC for all models in the set; the resulting QAIC values, and associated ΔQAIC values and QAIC weights, are comparable across the entire set. Candidate models of the DS detection function include models with different general forms (e.g. half‐normal, hazard rate, uniform), so it may not be possible to identify a single global model. We therefore propose a two‐step model selection procedure by which QAIC is used to select among models with the same general form, and then a goodness‐of‐fit statistic is used to select among models with different forms. A drawback of thi approach is that QAIC values are not comparable across all models in the candidate set.3. Relative to AIC, QAIC and the two‐step model selection procedure avoided overfitting and improved the accuracy and precision of densities estimated from simulated data. When applied to six real datasets, adjusted criteria and procedures selected either the same model as AIC or a model that yielded a more accurate density estimate in five cases, and a model that yielded a less accurate estimate in one case. 4. Many DS surveys yield overdispersed data, including cue counting surveys of songbirds and cetaceans, surveys of social species including primates, and camera‐trapping surveys. Methods that adjust for overdispersion during the model selection stage of DS analyses therefore address a conspicuous gap in the DS analytical framework as applied to species of conservation concern.

KW - Animal abundance

KW - Camera trapping

KW - Cue counting

KW - Distance sampling

KW - Model selection

KW - Overdispersion

KW - QAIC

U2 - 10.1111/2041-210X.13082

DO - 10.1111/2041-210X.13082

M3 - Article

VL - 10

SP - 38

EP - 47

JO - Methods in Ecology and Evolution

T2 - Methods in Ecology and Evolution

JF - Methods in Ecology and Evolution

SN - 2041-210X

IS - 1

ER -

Related by author

  1. Review of potential line-transect methodologies for estimating abundance of dolphin stocks in the eastern tropical Pacific

    Lennert-Cody, C. E., Buckland, S. T., Gerrodette, T., Webb, A., Barlow, J., Fretwell, P. T., Maunder, M. N., Kitakado, T., Moore, J. E., Scott, M. D. & Skaug, H. J., 25 Jan 2019, In : Journal of Cetacean Research and Management. 19, p. 9-21 13 p.

    Research output: Contribution to journalReview article

  2. Corrigendum: The number and distribution of polar bears in the western Barents Sea

    Aars, J., Marques, T. A., Lone, K., Andersen, M., Wiig, Ø., Fløystad, I. M. B., Hagen, S. B. & Buckland, S. T., 22 May 2018, In : Polar Research. 37, 1, 1457880.

    Research output: Contribution to journalComment/debate

  3. Attributing changes in the distribution of species abundance to weather variables using the example of British breeding birds

    Oedekoven, C. S., Elston, D. A., Harrison, P. J., Brewer, M. J., Buckland, S. T., Johnston, A., Foster, S. & Pearce-Higgins, J. W., Dec 2017, In : Methods in Ecology and Evolution. 8, 12, p. 1690-1702

    Research output: Contribution to journalArticle

  4. Point process models for spatio-temporal distance sampling data from a large-scale survey of blue whales

    Yuan, Y., Bachl, F. E., Lindgren, F., Borchers, D. L., Illian, J. B., Buckland, S. T., Rue, H. & Gerrodette, T., Dec 2017, In : Annals of Applied Statistics. 11, 4, p. 2270-2297

    Research output: Contribution to journalArticle

  5. Distance sampling with camera traps

    Howe, E. J., Buckland, S. T., Després-Einspenner, M-L. & Kühl, H., Nov 2017, In : Methods in Ecology and Evolution. 8, 11, p. 1558-1565

    Research output: Contribution to journalArticle

Related by journal

  1. Methods in Ecology and Evolution (Journal)

    Michael Blair Morrissey (Member of editorial board)
    1 Jan 20171 Jan 2020

    Activity: Publication peer-review and editorial work typesEditor of research journal

  2. Methods in Ecology and Evolution (Journal)

    Theoni Photopoulou (Editor)
    2017 → …

    Activity: Publication peer-review and editorial work typesEditor of research journal

  3. Methods in Ecology and Evolution (Journal)

    Oscar Eduardo Gaggiotti (Member of editorial board)
    1 Sep 2014 → …

    Activity: Publication peer-review and editorial work typesEditor of research journal

Related by journal

  1. A 2.6-gram sound and movement tag for studying the acoustic scene and kinematics of echolocating bats

    Stidsholt, L., Johnson, M., Beedholm, K., Jakobsen, L., Kugler, K., Brinkløv, S., Salles, A., Moss, C. F. & Madsen, P. T., Jan 2019, In : Methods in Ecology and Evolution. 10, 1, p. 48-58 11 p.

    Research output: Contribution to journalArticle

  2. State-switching continuous-time correlated random walks

    Michelot, T. & Blackwell, P. G., 14 Feb 2019, In : Methods in Ecology and Evolution. Early View

    Research output: Contribution to journalArticle

  3. The Langevin diffusion as a continuous-time model of animal movement and habitat selection

    Michelot, T., Gloaguen, P., Blackwell, P. G. & Etienne, M-P., 24 Aug 2019, In : Methods in Ecology and Evolution. Early View

    Research output: Contribution to journalArticle

  4. inlabru: an R package for Bayesian spatial modelling from ecological survey data

    Bachl, F. E., Lindgren, F., Borchers, D. L. & Illian, J. B., Jun 2019, In : Methods in Ecology and Evolution. 10, 6, p. 760-766 7 p.

    Research output: Contribution to journalArticle

ID: 255735220