Canonical correlation between clogging agents and performance of drippers operating with aquaculture effluents

Abstract

ABSTRACT The canonical correlation is used to correlate groups of distinct variables. The objective of this study was to use canonical correlation analysis to evaluate the interaction between water quality attributes clogging-causing and the hydraulic performance of non-pressure compensating drippers operating with dilutions of aquaculture effluent in well water. The experiment was in a completely randomized design in a split-split-plot scheme with three replicates, with five dilutions in the plots (Dilution 1: 100% aquaculture effluent; Dilution 2: 75% aquaculture effluent + 25% well water; Dilution 3: 50% aquaculture effluent + 50% well water; Dilution 4: 25% aquaculture effluent + 75% well water; and Dilution 5: 100% well water), three emitters in the sub-plots (ST (1.6 L h-1), SL (1.6 L h-1) and NJ (1.7 L h-1)), and three operating times in the sub-subplots (0, 80 and 160 hours, in which 0 hours operating time means the start of evaluations). The statistical analyses performed were correlation matrix and canonical analysis. The canonical correlation analysis indicated that when the concentrations of attributes increase, the hydraulic performance of the drippers decreases. The interaction of this group of attributes indicates that the clogging of drippers operating with the five dilutions of aquaculture effluent in well water occurs by the combined action of physical, chemical, and biological agents. The ST dripper was the least sensitive to the group of quality attributes of the dilutions, whereas the most sensitive was the SL dripper.