Repeat Sampling and Coliform Bacteria Detection Rates in New Jersey Domestic Wells

Abstract

In compliance with the New Jersey Private Well Testing Act, 78,546 wells (93,787 samples, including samples from 13,290 wells that were analyzed more than once) were analyzed for total coliform (TC) bacteria by one or more of 39 laboratories over a 10‐year period. Samples containing TC bacteria were further analyzed for the presence of either fecal coliform or E. coli (FC/EC) bacteria. The large population of wells sampled multiple times permitted a systematic study of the effect of repeat sampling on coliform bacteria detection rates. The detection rate increased with the number of times wells were sampled. In bedrock, TC bacteria were detected in 21% of the population of wells analyzed only once, 33% in the population sampled twice, and 43% in the population sampled three times. It was estimated that TC bacteria would be detected in 90% of all wells if each well was analyzed 10 times. For FC/EC bacteria, it was estimated that 21 and 68 samples, respectively, would be required to reach the 50% and 90% population detection rates. In the Coastal Plain (CP), many more samples would be required to achieve the same estimated population detection rates. The population detection rate estimates were also dependent on the type of method used, the pH of the well water, and the geologic formation in which wells were located. A single sample was not sufficient to detect coliform bacteria when present in well water.

Seasonality of Coliform Bacteria Detection Rates in New Jersey Domestic Wells

“…We review the potential coliform sources and explain why some may not indicate the potential presence of pathogens or do so in an inconsistent fashion. We also verified that the observed seasonality was not a sampling artifact of previously examined influences on coliform detection rates in GW, such as the location of a well in certain types of geology (Atherholt et al ), the number of times a well is sampled (Atherholt et al ), or a seasonal difference in the number of sampled wells that were located in agriculture land.…”

“…The efficiency with which coliform bacteria can be transported from surface and subsurface sources to GW, and hence increase in their concentration in domestic wells, is determined by a number of factors including the amount of precipitation (Hynds et al 2012), the geological formation in which the well is located (Embrey and Runkle 2006;DeSimone 2009;Toccalino et al 2010;Atherholt et al 2013), well depth (Glanville et al 1997;Gonzales 2008;Allevi et al 2013), and deficiencies in well construction or integrity (Bacci and Chapman 2011;Swistock et al 2013;Hynds et al 2014). Once in the GW, the efficiency with which these bacteria are detected is influenced by the laboratory and type of detection method used (Francy et al 2000;Griffith et al 2006;Atherholt et al 2013), and also by the number of times a well is sampled (Atherholt et al 2015).…”

“…Two previous studies used the PWTA coliform data to examine the influence of several factors on GW coliform detection rates. These factors included the geologic setting of a well, the laboratory and method used to generate the data (Atherholt et al ), and the number of times a well is sampled (Atherholt et al ).…”

The Likelihood of Coliform Bacteria in NJ Domestic Wells Based on Precipitation and Other Factors

“…A detailed map showing the location, geologic setting, and yield of the aquifers in each region is presented by Herman et al (). Because coliform detection rates in the CP and BK regions are so different (Atherholt et al , ), regional data are analyzed independently and presented separately when applicable.…”

“…If TC bacteria are detected, that sample is further tested for the presence of either FC or E. coli bacteria. Strengths and limitations of the PWTA data have been highlighted in recent publications (Atherholt et al ; Atherholt et al ).…”