Hierarchical models and functional traits
| Autor: | van Loon, E.E., Shamoun-Baranes, J., Sierdsema, H., Bouten, W., Cramer, W., Badeck, F., Krukenberg, B., Klotz, S., Kühn, I., Schweiger, O., Böhning-Gaese, K., Schaefer, H.-C., Kissling, D., Brandl, R., Brändle, M., Fricke, R., Leuschner, C., Buschmann, H., Köckermann, B., Rose, L. |
|---|---|
| Přispěvatelé: | Computational Geo-Ecology (IBED, FNWI) |
| Rok vydání: | 2006 |
| Zdroj: | International Conference on Macroecological Tools for Global Change Research, 134-134 STARTPAGE=134;ENDPAGE=134;TITLE=International Conference on Macroecological Tools for Global Change Research |
| Popis: | Hierarchical models for animal abundance prediction are conceptually elegant. They are generally more parsimonous than non-hierarchical models derived from the same data, give relatively robust predictions and automatically provide consistent output at multiple (spatio-temporal) scales. Another attractive feature of hierarchical models for abundance prediction is the possibility to define a given hierarchical level in terms of species groups or functional traits. We explore this by applying a three-layer hierarchical model to describe the abundance of breeding birds in the Netherlands. At the highest level relative abundance is predicted with one model for the entire Netherlands and all breeding birds, at a subsequent level regional models are made for more specific groups (grassland birds, birds of prey, water birds, insectivores and a rest-group), and at the lowest level abundance is predicted for individual species at a 1 km2 grid. Results from a high-level model constrain those at a lower level. Explanatory variables at the different model hierarchies are related to functional traits. At the highest level, openness of the landscape in combination with closeness to water bodies determine bird abundancies. These habitat features relate to the cost of food collection and levels of predation. At the level of bird groups, feeding strategy, major food source and nest location are mostly related to the explanatory variables. Finally, at the species level a large diversity of explanatory variables start to play a role, the relation between these and functional traits has not been fully explored but most explanatory variables are related to vegetation structure, primary production and antropogenic influences. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|