The need of hypothesis-driven designs and conceptual models in impact assessment studies: An example from the free-living marine nematodes
Fonseca, Gustavo Fernandes Camargo [UNIFESP]
Gallucci, Fabiane [UNIFESP]
MetadataShow full item record
The literature on the use of marine nematodes as bioindicators of anthropogenic impact is extensive. Nevertheless, review studies have reported a high degree of variability among results and no consistent overall pattern has so far emerged. This lack of congruence might be partially because hypotheses formulation in environmental assessment studies has been largely inductive or abductive and not deductive or hypothesis-driven. In the present study, we emphasize the need of using hypothesis-driven designs and conceptual models in impact assessment studies. Hypotheses for individual and population level studies can be derived from the dynamic energy budget model (DEB). By means of differential equations, DEB model can infer whether a stressor promotes a shift in energy allocation along the life history of the individuals. For community/assemblage level studies, the predictions of the dynamic equilibrium model (DEM) for species richness is presented and extended for abundance, evenness, taxonomic distinctness, and changes in assemblage structure and sample dispersion. While it is predicted that species richness peaks at intermediate levels of disturbances and enrichment, evenness decreases with increasing disturbances and reducing enrichment/pollutant concentrations. Based on DEM,enrichment and pollutants may promote change in community structure by favoring the tolerant species, while physical disturbances may promote sample dispersion as a result of unselective mortality. Finally, we discuss the benefits of using niche-based models to select the indicator species, instead of using classical ordination methods, twinspan and similarity percentage analysis. The selection of indicator species have to be independent from the other species and must consider the set of environmental conditions. The use of conceptual models to select the best ecological indicators is highly recommended. It allows a logical way of testing for causalities and of scaling the different studies for comparisons. (C) 2016 Elsevier Ltd. All rights reserved.
- IMar - Artigos