Coliform Bacteria in New Jersey Domestic Wells: Influence of Geology, Laboratory, and Method

Abstract

Following passage of the New Jersey Private Well Testing Act, 50,800 domestic wells were tested between 2002 and 2007 for the presence of total coliform (TC) bacteria. Wells containing TC bacteria were further tested for either fecal coliform or Escherichia coli (FC/E. coli) bacteria. Analysis of the data, generated by 39 laboratories, revealed that the rate of coliform detections in groundwater (GW) was influenced by the laboratory and the method used, and also by geology. Based on one sample per well, TC and FC/E. coli were detected in wells located in bedrock 3 and 3.7 times more frequently, respectively, than in wells located in the unconsolidated strata of the Coastal Plain. In bedrock, detection rates were higher in sedimentary rock than in igneous or metamorphic rock. Ice-age glaciers also influenced detection rates, most likely by removing material in some areas and depositing thick layers of unconsolidated material in other areas. In bedrock, coliform bacteria were detected more often in wells with a pH of 3 to 6 than in wells with a pH of 7 to 10 whereas the reverse was true in the Coastal Plain. TC and FC/E. coli bacteria were detected in 33 and 9.5%, respectively, of sedimentary rock wells with pH 3 to 6. Conversely, for Coastal Plain wells with pH 3 to 6, detection rates were 4.4% for TC and 0.6% for FC/E. coli.

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

“…This study confirms the findings of a previous analysis of a smaller population of wells that showed that wells located in bedrock have much higher coliform detection rates compared to wells located in the Coastal Plain (Atherholt et al ). Soils are characterized by numerous, directly connected macropores (Hoang et al ).…”

“…The method used had less influence on the FC or EC detection rates than for TC (Table S3). Because there are inter‐laboratory differences in detection rates (Atherholt et al ), the observed detection rates should be viewed only as approximations of the true rate.…”

“…Sampling often occurred at locations downstream of the pressure tank (e.g., a kitchen, bathroom, or outdoor faucet in homes without treatment). The effects of the pressure tank and/or household plumbing on the coliform detection rates are unknown but are thought to be minimal, considering the documented effect of other parameters such as geology (Atherholt et al ). Well location and analytical method information aside, it is not known how many errors exist in this data set.…”

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

“…We observed the highest and most significant likelihood of TC and FC/EC occurrences when pH was between 5.0 and 6.99. In contrast, when reviewed from a more regional perspective, Atherholt et al () documented higher rates of coliform detections in wells in the BK region when pH was less than 5 as compared to wells with higher pH values. This study population included nearly twice as many wells in the BK as the Atherholt et al study, and of the 10,047 wells with a pH less than 5 in this study, only 226 (2.2%) were located in the BK region.…”

“…The third phase evaluated the likelihood of coliform occurrences given the effect of precipitation and other factors for which data were available. The additional factors included season (Atherholt et al ), geological setting, laboratory method used (Atherholt et al ), pH (Atherholt et al ), and nitrate concentrations. The goal of the last phase was to provide a useful method to estimate the likelihood of encountering a coliform‐positive reading given certain readily available information in the absence of a coliform test.…”