Measuring discontinuities in the ter river

In this paper a method for measuring discontinuities in the spatial organization of a river produced by disturbances is presented. First of all, a way of doing comparable environmental gradients defined by different variables is proposed. To do so, all the variables have been transformed by means of the equation: Pi' = Pi ‐ Pmin/Pmax ‐ Pmin, as a previous measure of the parametric distances (Pij') between sites. When these parametric distances are plotted in relation to the geographical ones, the distortion between both river images appears as an expression of the changes occurring in each stretch, and the derivative dPij'/dx is a measure of its rate of change. To define a river gradient, we need to calculate the general trend of a variable, using direct gradient analysis. The fit of the results obtained from an undisturbed stretch to an empirical curve for all the river gives us a measure of the expected values for that river. The difference between expected and observed values can be a measure of the discontinuity. The same procedure has been used for an overall environmental and biological data. To this end, the Euclidean distance (ED) and a distance obtained from Factor score of a Principal Components analysis (PCA) have been calculated. Only the second method seems to be of general interest for studying discontinuities. Finally, the results of the first score loadings obtained from PCA of three groups of variables (physical and chemical, physiographical and morphological, and phytobenthic) have been used to do a comparison between the parametric images of the River Ter.