Hydrobiologia

, Volume 623, Issue 1, pp 153–171

The MEDiterranean Prediction And Classification System (MEDPACS): an implementation of the RIVPACS/AUSRIVAS predictive approach for assessing Mediterranean aquatic macroinvertebrate communities

  • José Manuel Poquet
  • Javier Alba-Tercedor
  • Tura Puntí
  • Maria del Mar Sánchez-Montoya
  • Santiago Robles
  • Maruxa Álvarez
  • Carmen Zamora-Muñoz
  • Carmen Elisa Sáinz-Cantero
  • Maria Rosario Vidal-Abarca
  • Maria Luisa Suárez
  • Manuel Toro
  • Ana Maria Pujante
  • Maria Rieradevall
  • Narcís Prat
Primary research paper

DOI: 10.1007/s10750-008-9655-y

Cite this article as:
Poquet, J.M., Alba-Tercedor, J., Puntí, T. et al. Hydrobiologia (2009) 623: 153. doi:10.1007/s10750-008-9655-y

Abstract

In Spain, a national project known as GUADALMED, focusing on Mediterranean streams, has been carried out from 1998 to 2005 to implement the European water framework directive (WFD) requirements. One of the main objectives of the second phase of the project (2002–2005) was to develop a predictive system for the Spanish Mediterranean aquatic macroinvertebrate communities. A combined-season (spring, summer, and autumn) predictive model was developed by using the latest improvements on the selection of best predictor variables. Overall model performance measures were used to select the best discriminant function (DF) models, and also to evaluate their biases and precision. The final predictive model was based on the best five DF models. Each one of these models involved eight environmental variables. Final observed (O), expected (E), and O/E values for the number of macroinvertebrate families (NFAM) and two biotic indices (IBMWP and IASPT) were calculated by averaging their values, previously weighted by the quality of each DF model. Regression analyses among the final O and E values for the calibration dataset showed a high proximity to the ideal theoretical model, where the final E values explained 73–84% of the variation present in the macroinvertebrate communities of the Spanish Mediterranean watercourses. The ANOVA performed among the reference (calibration and validation) and test datasets showed clear differences for the O/E values. Finally, the assessments carried out by the predictive model were sensitive to anthropogenic pressure present in the study area and allowed the definition of five ecological status classes according to the WFD requirements.

Keywords

Predictive modellingGUADALMED projectBioassessmentWater framework directiveEcological status

Copyright information

© Springer Science+Business Media B.V. 2008

Authors and Affiliations

  • José Manuel Poquet
    • 1
  • Javier Alba-Tercedor
    • 1
  • Tura Puntí
    • 2
  • Maria del Mar Sánchez-Montoya
    • 3
  • Santiago Robles
    • 4
  • Maruxa Álvarez
    • 5
  • Carmen Zamora-Muñoz
    • 1
  • Carmen Elisa Sáinz-Cantero
    • 1
  • Maria Rosario Vidal-Abarca
    • 3
  • Maria Luisa Suárez
    • 3
  • Manuel Toro
    • 6
  • Ana Maria Pujante
    • 7
  • Maria Rieradevall
    • 2
  • Narcís Prat
    • 2
  1. 1.Departamento de Biología Animal, Facultad de CienciasUniversidad de GranadaGranadaSpain
  2. 2.Departamento de EcologíaUniversidad de BarcelonaBarcelonaSpain
  3. 3.Departamento de Ecología e HidrologíaUniversidad de MurciaMurciaSpain
  4. 4.Cimera Estudios Aplicados SL, Parque Científico de MadridMadridSpain
  5. 5.Área de Ecología Universidad de VigoVigoSpain
  6. 6.División de Ecología de los Sistemas Acuáticos Continentales CEDEXMadridSpain
  7. 7.Red-Control SL, Parque Tecnológico de ValenciaPaternaSpain