Survival of Escherichia coli in Irrigation Water Canals

Abstract

Escherichia coli are fecal bacteria that are commonly used as indicators of water quality in agricultural waters. E. coli bacteria can grow and survive in streambed sediments in lakes and rivers leading to downstream contamination, but their survival has yet to be studied in irrigation water canal systems. Previous studies in natural water systems in rivers and lakes have found thatE. coli survives longer in sediment than in the overlying water. This study aimed to assess the survival of E. coli in sediments typical of irrigation canal systems found in Arizona. Both a generic laboratory strain of E. coli and canal sediments containing E. coli were used in experiments conducted at two different temperatures (22°C and 32.2°C) using two different sediment types containing variable clay contents and amended with dried cattle manure as a nutrient additive. Water without sediments was used as a control to assess the survival of E. coli relative to survival in sediment. Introduced E. coli survival in sediments was of longer duration than water. In water, introduced E. coli decreased by 4.77 log10 in seven days with a k inactivation constant of -2.34. Introduced E. coli survival was higher in sandy loam sediments then loamy sand sediments with a log reduction of 2.70 over 20 days and k inactivation constant of -0.399 in sandy loam as compared to a log reduction of 3.57 and k inactivation constant of - 0.469 in loamy sand. Introduced E. coli was inactivated faster in loamy sand at 32.2°C than 22°C with a log reduction of 2.16 in ten days and a k inactivation constant of -0.619 as compared to a k inactivation constant of -0.469 at 22°C. Addition of manure did not seem to have a great effect on the survival of introduced of E. coli at 22°C in sand or sandy loam sediments; however, a major spike in growth from day 0-1 was observed. Indigenous E. coli at 22°C survived longer in loamy sand sediment than introduced E. coli with a log reduction of 0.58 and 1.10 (k inactivation constants of -0.112 and -0.146 for both indigenous experiments).