Elucidating spatial patterns of E. coli in two irrigation ponds with empirical orthogonal functions

•Microbial water quality substantially varied across the studied irrigation ponds.•Persistent spatial patterns of E. coli concentrations were identified as EOFs.•Surrounding land use and flow conditions helped to interpret significant patterns.•EOFs of E. coli concentrations and water quality parameters were significantly correlated. Fecal contamination of water used for irrigation is assessed through the measurement of E. coli concentrations. The concentrations of this fecal indicator bacteria have been shown to vary highly in both space and time which may affect the results of microbial monitoring surveys. Determining if stable patterns of E. coli can exist in irrigation sources will improve our ability to design effective monitoring campaigns. The objective of this work was to research spatial and temporal variability of E. coli concentrations in two working irrigation ponds using an Empirical Orthogonal Function (EOF) analysis. E. coli, pH, specific conductance, dissolved oxygen, and temperature were measured biweekly at 23 and 34 locations at ponds P1 and P2, respectively, from 2016 to 2018. The first three spatial patterns (EOFs) of E. coli concentrations accounted for >80 % of the total variance in the original dataset each year at both ponds. In a given year, EOF1 explained from 34 to 54% and from 45% to 60 % of the total variance in ponds P1 and P2, respectively. Examination of EOF1 revealed distinct zones of relatively high and low E. coli concentrations related to known microorganism sources and transport conditions. The spatial patterns of E. coli showed a significant negative correlation with spatial patterns of temperature (rs = −0.706 and −0.321 for P1 and P2, respectively), pH (rs = 0.491 and −0.403), and DO (rs = −0.591 and −0.460) in both ponds. Overall, the EOF analysis provided substantial information about the temporally stable spatial patterns of E. coli by examining the deviations from the spatial averages across the ponds. The EOF analysis holds the promise to be a valuable tool for effective microbial water quality monitoring design.