Premium
Identification of a minimal adequate model to describe the biomass dynamics of river epilithon
Author(s) -
Boulêtreau Stéphanie,
Izagirre Oihana,
Garabétian Frédéric,
Sauvage Sabine,
Elosegi Arturo,
SánchezPérez JoséMiguel
Publication year - 2008
Publication title -
river research and applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.679
H-Index - 94
eISSN - 1535-1467
pISSN - 1535-1459
DOI - 10.1002/rra.1046
Subject(s) - akaike information criterion , biomass (ecology) , abiotic component , limnology , goodness of fit , environmental science , ecology , trophic level , model selection , hydrology (agriculture) , biology , mathematics , statistics , geology , geotechnical engineering
The present study sought to identify a minimal adequate model to describe the biomass dynamics of river epilithon, a functional indicator of river health. Identification of minimal adequate models is particularly necessary in river management, given the reduced number of variables authorities are willing to measure routinely. A model previously developed for epilithon dynamics in a pre‐alpine river was applied to epilithon biomasses recorded in contrasting hydrological, trophic and light conditions at various sites in the Agüera stream (Spain) over 3 years (11 case studies). A model selection tool, the Akaike Information Criterion (AIC), was used to determine the optimal combination of parameters. In nine of 11 case studies, the best model described epilithon biomass dynamics as the equilibrium between phototrophic growth and discharge‐dependent loss and ignored light, temperature and nutrient influences. The best adequate minimal model i.e. the model that is the best trade‐off between goodness‐of‐fit and model simplicity performed best, in years in which clearly contrasting short low and high water periods occurred. During years with less marked hydrodynamics, many other abiotic or biotic processes influenced epilithon biomass dynamics. In these cases, weaker goodness‐of‐fit had to be accepted to avoid excessively increasing model complexity. Copyright © 2007 John Wiley & Sons, Ltd.