Bayesian meta-analysis to synthesize decay rate constant estimates for common fecal indicator bacteria.
Water Res
; 104: 262-271, 2016 Nov 01.
Article
em En
| MEDLINE
| ID: mdl-27543910
ABSTRACT
For decades, fecal indicator bacteria have been used as proxies to quantitatively estimate fecal loading into water bodies. Widely used cultured indicators (e.g. Escherichia coli and Enterococcus spp.) and more recently developed genetic markers are well studied, but their decay in the environment is still poorly understood. We used Hierarchical Bayesian Linear Modeling to conduct a series of meta-analyses using published decay rate constant estimates, to synthesize findings into pooled estimates and identify gaps in the data preventing reliable estimates. In addition to the meta-analysis assuming all estimates come from the same population, meta-regressions including covariates believed to contribute to decay were fit and used to provided synthesized estimates for specific combinations of significant variables. Additionally, statements regarding the significance of variables across studies were made using the 95% confidence interval for meta-regression coefficients. These models were used to construct a mean decay rate constant estimate as well as credible intervals for the mean and the distribution of all likely data points. While synthesized estimates for each targeted indicator bacteria were developed, the amount of data available varied widely for each target, as did the predictive power of the models as determined by testing with additional data not included in the modeling. Temperature was found to be significant for all selected indicators, while light was found to be significant only for culturable indicators. Results from the models must be interpreted with caution, as they are based only on the data available, which may not be representative of decay in other scenarios.
Palavras-chave
Texto completo:
1
Coleções:
01-internacional
Base de dados:
MEDLINE
Assunto principal:
Teorema de Bayes
/
Fezes
Tipo de estudo:
Prognostic_studies
/
Systematic_reviews
Idioma:
En
Ano de publicação:
2016
Tipo de documento:
Article