Predicting the occurrence of Salmonella, Escherichia coli, and pathogenic E. coli in agricultural ponds in the Georgia Coastal Plain.
- 1Department of Food Science and Technology, University of Georgia, Athens, Georgia, USA.
- 2USDA-ARS, Environmental Microbial and Food Safety Laboratory, Beltsville, Maryland, USA.
- 3USDA-ARS, Southeast Watershed Research Laboratory, Tifton, Georgia, USA.
Abstract
Escherichia coli is one measure of fecal contamination in water, which is expected to correlate with the presence of fecal pathogens, such as pathogenic E. coli or Salmonella. Supplementing E. coli measurements with other water quality measurements, for example, temperature or pH, may better identify situations when foodborne pathogens are likely to be present in water. This work used generalized and multinomial regression and machine learning (K-nearest neighbors, random forest (RF), and support vector machines) to create explanatory models for levels of E. coli, pathogenic E. coli, and Salmonella presence in irrigation ponds. Model performance was mixed, with RF models performing best for E. coli levels (root mean square error [RMSE] = 0.45) for all ponds (n = 1266), pond 1 (RMSE = 0.44, n = 563), and pond 2 (RMSE = 0.71, n = 505), while generalized regression performed best at pond 3 (RMSE = 0.41, n = 198). Pathogenic E. coli presence for all ponds was best explained by RF (receiver operating characteristic area under the curve [ROC AUC] = 0.94, n = 717). Salmonella presence for ponds 1 and 2 (n = 284) was best explained by multinomial regression (ROC AUC = 0.83). These results indicate that explanatory models struggled to identify situations conducive to pathogen occurrence in this system and were not useful in identifying factors associated with pathogen presence.
Related
- Measuring atmospheric microplastic composition and occurrence from an uncrewed aerial system
- Microplastic contamination in commercially important marine fishes from the Sohar Coast, Sea of Oman: occurrence, characterization, and polymer composition
- On the Role of the Antarctic Slope Front on the Occurrence of the Weddell Sea Polynya under Climate Change
- Environmental Occurrence, Toxicity Concerns, Degradation Behavior, and Plant Resistance of Glufosinate-Ammonium
- Long-term associations between land-surface indicators and dust-storm occurrence along the margins of the Taklimakan Desert
BibTeX
@article{widmer2026predicting,
title = {{Predicting the occurrence of Salmonella, Escherichia coli, and pathogenic E. coli in agricultural ponds in the Georgia Coastal Plain.}},
author = {J Andrew Widmer and Matthew Stocker and Jaclyn E Smith and Oliva Pisani and Alisa Coffin and Manan Sharma and Yakov Pachepsky and Laurel L Dunn},
journal = {Journal of environmental quality},
year = {2026},
month = nov,
doi = {10.1002/jeq2.70317},
url = {https://europepmc.org/article/MED/42972214},
note = {CC BY},
}
RIS
TY - JOUR
TI - Predicting the occurrence of Salmonella, Escherichia coli, and pathogenic E. coli in agricultural ponds in the Georgia Coastal Plain.
AU - J Andrew Widmer
AU - Matthew Stocker
AU - Jaclyn E Smith
AU - Oliva Pisani
AU - Alisa Coffin
AU - Manan Sharma
AU - Yakov Pachepsky
AU - Laurel L Dunn
PY - 2026
DA - 2026/11/01
JO - Journal of environmental quality
DO - 10.1002/jeq2.70317
UR - https://europepmc.org/article/MED/42972214
AB - Escherichia coli is one measure of fecal contamination in water, which is expected to correlate with the presence of fecal pathogens, such as pathogenic E. coli or Salmonella. Supplementing E. coli measurements with other water quality measurements, for example, temperature or pH, may better identify situations when foodborne pathogens are likely to be present in water. This work used generalized and multinomial regression and machine learning (K-nearest neighbors, random forest (RF), and support vector machines) to create explanatory models for levels of E. coli, pathogenic E. coli, and Salmonella presence in irrigation ponds. Model performance was mixed, with RF models performing best for E. coli levels (root mean square error [RMSE] = 0.45) for all ponds (n = 1266), pond 1 (RMSE = 0.44, n = 563), and pond 2 (RMSE = 0.71, n = 505), while generalized regression performed best at pond 3 (RMSE = 0.41, n = 198). Pathogenic E. coli presence for all ponds was best explained by RF (receiver operating characteristic area under the curve [ROC AUC] = 0.94, n = 717). Salmonella presence for ponds 1 and 2 (n = 284) was best explained by multinomial regression (ROC AUC = 0.83). These results indicate that explanatory models struggled to identify situations conducive to pathogen occurrence in this system and were not useful in identifying factors associated with pathogen presence.
SN - 0047-2425
SN - 1537-2537
ER -