MODELOS DE DISTRIBUCIÓN DE ÁRBOLES Y ARBUSTOS DE LA PENÍNSULA IBÉRICA Y BALEARES.
Fecha
2019-10-14
Título de la revista
ISSN de la revista
Título del volumen
Editor
Jaén: Universidad de Jaén
Resumen
[ES]Los modelos de distribución de especies (SDM), están aprovechando el auge actual
de los Sistemas de Información Geográfica (SIG), incorporándolos como una potente
herramienta que proporciona a los modelos una cantidad enorme de información ambiental
muy valiosa, para estudiar el nicho ecológico de las especies. En este trabajo, se ha
comparado dos bases de datos, CHELSA y Worldclim, con información bioclimática
consistente en 19 variables relacionadas con la temperatura y las precipitaciones. El marco
geográfico del estudio consiste en la Península Ibérica y Baleares, y el objeto 25 especies
arbóreas y arbustivas. Gbif proporciona los datos de localización de las especies. La
distribución potencial se modeló utilizando 7 algoritmos diferentes y una variante de uno de
ellos: BIOCLIM, DOMAIN, Mahalanobis, GLM con logit link, GLM con probit link,
RandomForests, Support Vector Machine (SVM) y un modelo combinado (CM). Aunque no se
encontraron diferencias significativas entre los resultados arrojados por CHELSA y Worldclim,
sí se encontraron diferencias notables entre modelos, especies y manejo de ambas bases de
datos.
Palabras clave: Modelos de distribución de especies, nicho ecológico, algoritmo, base de
datos, AUC, Península Ibérica.
[EN]Species distribution models (SDM), are taking advantage of the current rise of Geographic Information Systems (GIS), incorporating them as a powerful tool that provides models with a huge amount of environmental information very valuable, to study the ecological niche of the species. In this work, two databases, CHELSA and Worldclim, have been compared with bioclimatic information consisting of 19 variables related to temperature and rainfall. The geographical framework of the study consists of the Iberian Peninsula and the Balearic Islands, and the object 25 tree and shrub species. Gbif provides the location data of the species. The potential distribution was modeled using 7 different algorithms and a variant of one of them: BIOCLIM, DOMAIN, Mahalanobis, GLM with logit link, GLM with probit link, RandomForests, Support Vector Machine (SVM) and a combined model (CM). Although no significant differences were found between the results produced by CHELSA and Worldclim, there were notable differences between models, species and management of both databases. Keywords: Species distribution models, ecological niche, algorithm, database, AUC, Iberian Peninsula.
[EN]Species distribution models (SDM), are taking advantage of the current rise of Geographic Information Systems (GIS), incorporating them as a powerful tool that provides models with a huge amount of environmental information very valuable, to study the ecological niche of the species. In this work, two databases, CHELSA and Worldclim, have been compared with bioclimatic information consisting of 19 variables related to temperature and rainfall. The geographical framework of the study consists of the Iberian Peninsula and the Balearic Islands, and the object 25 tree and shrub species. Gbif provides the location data of the species. The potential distribution was modeled using 7 different algorithms and a variant of one of them: BIOCLIM, DOMAIN, Mahalanobis, GLM with logit link, GLM with probit link, RandomForests, Support Vector Machine (SVM) and a combined model (CM). Although no significant differences were found between the results produced by CHELSA and Worldclim, there were notable differences between models, species and management of both databases. Keywords: Species distribution models, ecological niche, algorithm, database, AUC, Iberian Peninsula.