Monitoring habitat types by the mixed multinomial logit model using panel data

Research output: Contribution to journalArticleAcademicpeer-review

1 Citation (Scopus)

Abstract

Habitats in the Wadden Sea, a world heritage area, are affected by land subsidence resulting from naturalgas extraction and by sea level rise. Here we describe a method to monitor changes in habitat types byproducing sequential maps based on point information followed by mapping using a multinomial logitregression model with abiotic variables of which maps are available as predictors.In a 70 ha study area a total of 904 vegetation samples has been collected in seven sampling roundswith an interval of 2–3 years. Half of the vegetation plots was permanent, violating the assumptionof independent data in multinomial logistic regression. This paper shows how this dependency can beaccounted for by adding a random effect to the multinomial logit (MLN) model, thus becoming a mixedmultinomial logit (MMNL) model. In principle all regression coefficients can be taken as random, butin this study only the intercepts are treated as location-specific random variables (random interceptsmodel). With six habitat types we have five intercepts, so that the number of extra model parametersbecomes 15, 5 variances and 10 covariances.The likelihood ratio test showed that the MMNL model fitted significantly better than the MNL modelwith the same fixed effects. McFadden-R2for the MMNL model was 0.467, versus 0.395 for the MNL model.The estimated coefficients of the MMNL and MNL model were comparable; those of altitude, the mostimportant predictor, differed most. The MMNL model accounts for pseudo-replication at the permanentplots, which explains the larger standard errors of the MMNL coefficients. The habitat type at a givenlocation-year combination was predicted by the habitat type with the largest predicted probability. Theseries of maps shows local trends in habitat types most likely driven by sea-level rise, soil subsidence,and a restoration project.We conclude that in environmental modeling of categorical variables using panel data, dependencyof repeated observations at permanent plots should be accounted for. This will affect the estimatedprobabilities of the categories, and even stronger the standard errors of the regression coefficients.
LanguageEnglish
Pages108-116
JournalEcological Indicators
Volume67
DOIs
Publication statusPublished - 2016

Fingerprint

panel data
logit analysis
habitat type
monitoring
habitats
subsidence
sea level
vegetation
North Sea
Multinomial logit model
Panel data
Habitat
Monitoring
environmental modeling
sampling
Logit model
Coefficients
logistics
soil
testing

Cite this

@article{c1a2c74c6f9644079a3951c0bfa5445e,
title = "Monitoring habitat types by the mixed multinomial logit model using panel data",
abstract = "Habitats in the Wadden Sea, a world heritage area, are affected by land subsidence resulting from naturalgas extraction and by sea level rise. Here we describe a method to monitor changes in habitat types byproducing sequential maps based on point information followed by mapping using a multinomial logitregression model with abiotic variables of which maps are available as predictors.In a 70 ha study area a total of 904 vegetation samples has been collected in seven sampling roundswith an interval of 2–3 years. Half of the vegetation plots was permanent, violating the assumptionof independent data in multinomial logistic regression. This paper shows how this dependency can beaccounted for by adding a random effect to the multinomial logit (MLN) model, thus becoming a mixedmultinomial logit (MMNL) model. In principle all regression coefficients can be taken as random, butin this study only the intercepts are treated as location-specific random variables (random interceptsmodel). With six habitat types we have five intercepts, so that the number of extra model parametersbecomes 15, 5 variances and 10 covariances.The likelihood ratio test showed that the MMNL model fitted significantly better than the MNL modelwith the same fixed effects. McFadden-R2for the MMNL model was 0.467, versus 0.395 for the MNL model.The estimated coefficients of the MMNL and MNL model were comparable; those of altitude, the mostimportant predictor, differed most. The MMNL model accounts for pseudo-replication at the permanentplots, which explains the larger standard errors of the MMNL coefficients. The habitat type at a givenlocation-year combination was predicted by the habitat type with the largest predicted probability. Theseries of maps shows local trends in habitat types most likely driven by sea-level rise, soil subsidence,and a restoration project.We conclude that in environmental modeling of categorical variables using panel data, dependencyof repeated observations at permanent plots should be accounted for. This will affect the estimatedprobabilities of the categories, and even stronger the standard errors of the regression coefficients.",
author = "D.J. Brus and P.A. Slim and G. Gort and A.H. Heidema and {van Dobben}, H.F.",
year = "2016",
doi = "10.1016/j.ecolind.2016.02.043",
language = "English",
volume = "67",
pages = "108--116",
journal = "Ecological Indicators",
issn = "1470-160X",
publisher = "Elsevier",

}

Monitoring habitat types by the mixed multinomial logit model using panel data. / Brus, D.J.; Slim, P.A.; Gort, G.; Heidema, A.H.; van Dobben, H.F.

In: Ecological Indicators, Vol. 67, 2016, p. 108-116.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Monitoring habitat types by the mixed multinomial logit model using panel data

AU - Brus, D.J.

AU - Slim, P.A.

AU - Gort, G.

AU - Heidema, A.H.

AU - van Dobben, H.F.

PY - 2016

Y1 - 2016

N2 - Habitats in the Wadden Sea, a world heritage area, are affected by land subsidence resulting from naturalgas extraction and by sea level rise. Here we describe a method to monitor changes in habitat types byproducing sequential maps based on point information followed by mapping using a multinomial logitregression model with abiotic variables of which maps are available as predictors.In a 70 ha study area a total of 904 vegetation samples has been collected in seven sampling roundswith an interval of 2–3 years. Half of the vegetation plots was permanent, violating the assumptionof independent data in multinomial logistic regression. This paper shows how this dependency can beaccounted for by adding a random effect to the multinomial logit (MLN) model, thus becoming a mixedmultinomial logit (MMNL) model. In principle all regression coefficients can be taken as random, butin this study only the intercepts are treated as location-specific random variables (random interceptsmodel). With six habitat types we have five intercepts, so that the number of extra model parametersbecomes 15, 5 variances and 10 covariances.The likelihood ratio test showed that the MMNL model fitted significantly better than the MNL modelwith the same fixed effects. McFadden-R2for the MMNL model was 0.467, versus 0.395 for the MNL model.The estimated coefficients of the MMNL and MNL model were comparable; those of altitude, the mostimportant predictor, differed most. The MMNL model accounts for pseudo-replication at the permanentplots, which explains the larger standard errors of the MMNL coefficients. The habitat type at a givenlocation-year combination was predicted by the habitat type with the largest predicted probability. Theseries of maps shows local trends in habitat types most likely driven by sea-level rise, soil subsidence,and a restoration project.We conclude that in environmental modeling of categorical variables using panel data, dependencyof repeated observations at permanent plots should be accounted for. This will affect the estimatedprobabilities of the categories, and even stronger the standard errors of the regression coefficients.

AB - Habitats in the Wadden Sea, a world heritage area, are affected by land subsidence resulting from naturalgas extraction and by sea level rise. Here we describe a method to monitor changes in habitat types byproducing sequential maps based on point information followed by mapping using a multinomial logitregression model with abiotic variables of which maps are available as predictors.In a 70 ha study area a total of 904 vegetation samples has been collected in seven sampling roundswith an interval of 2–3 years. Half of the vegetation plots was permanent, violating the assumptionof independent data in multinomial logistic regression. This paper shows how this dependency can beaccounted for by adding a random effect to the multinomial logit (MLN) model, thus becoming a mixedmultinomial logit (MMNL) model. In principle all regression coefficients can be taken as random, butin this study only the intercepts are treated as location-specific random variables (random interceptsmodel). With six habitat types we have five intercepts, so that the number of extra model parametersbecomes 15, 5 variances and 10 covariances.The likelihood ratio test showed that the MMNL model fitted significantly better than the MNL modelwith the same fixed effects. McFadden-R2for the MMNL model was 0.467, versus 0.395 for the MNL model.The estimated coefficients of the MMNL and MNL model were comparable; those of altitude, the mostimportant predictor, differed most. The MMNL model accounts for pseudo-replication at the permanentplots, which explains the larger standard errors of the MMNL coefficients. The habitat type at a givenlocation-year combination was predicted by the habitat type with the largest predicted probability. Theseries of maps shows local trends in habitat types most likely driven by sea-level rise, soil subsidence,and a restoration project.We conclude that in environmental modeling of categorical variables using panel data, dependencyof repeated observations at permanent plots should be accounted for. This will affect the estimatedprobabilities of the categories, and even stronger the standard errors of the regression coefficients.

U2 - 10.1016/j.ecolind.2016.02.043

DO - 10.1016/j.ecolind.2016.02.043

M3 - Article

VL - 67

SP - 108

EP - 116

JO - Ecological Indicators

T2 - Ecological Indicators

JF - Ecological Indicators

SN - 1470-160X

ER -