Development of Empirical, Geographically Specific Water Quality Criteria: A Conditional Probability Analysis Approach

Abstract

The need for scientifically defensible water quality standards for nonpoint source pollution control continues to be a pressing environmental issue. The probability of impact at differing levels of nonpoint source pollution was determined using the biological response of instream organisms empirically obtained from a statistical survey. A conditional probability analysis was used to calculate a biological threshold of impact as a function of the likelihood of exceeding a given value of pollution metric for a specified geographic area. Uncertainty and natural variability were inherently incorporated into the analysis through the use of data from a probabilistic survey. Data from wadable streams in the mid‐Atlantic area of the U.S. were used to demonstrate the approach. Benthic macroinvertebrate community index values (EPT taxa richness) were used to identify impacted stream communities. Percent fines in substrate (silt/clay fraction, > 0.06 mm) were used as a surrogate indicator for sedimentation. Thresholds of impact due to sedimentation were identified by three different techniques, and were in the range of 12 to 15 percent fines. These values were consistent with existing literature from laboratory and field studies on the impact of sediments on aquatic life in freshwater streams. All results were different from values determined from current regulatory guidance. Finally, it was illustrated how these thresholds could be used to develop criterion for protection of aquatic life in streams.

Considerations for analyzing ecological community thresholds in response to anthropogenic environmental gradients

“…Threshold estimation depends on the selection of response variable, assumed shape of the response, and appropriateness of the corresponding statistical model, any of which can contribute to different interpretations regarding the location of a threshold or whether a threshold exists (e.g., Walsh et al 2005b, Moore and Palmer 2005, Dodds et al 2010. Compounding this problem is the wide array of statistical approaches available for identifying ecological thresholds (e.g., Toms and Lesperance 2003, Qian et al 2003, 2004, Paul and McDonald 2005, Brenden et al 2008, Andersen et al 2009, Sonderegger et al 2009, Baker and King 2010. Selection of an appropriate technique for a specific application from this increasingly lengthy list might not be obvious even to experienced analysts.…”

Associations between chlorophyll a and various microcystin-LR health advisory concentrations

“…Additional details of the specific implementation are available at https://github.com/USEPA/microcystinchla. A more detailed discussion of CPA is beyond the scope of this paper, but see Paul et al 15 and Hollister et al 16 for greater detail. All analyses were conducted using R version 3.2.2 and code and data from this analysis are freely available as an R package at https://github.com/USEPA/microcystinchla.…”

Métodos Estadísticos Para La Detección De Umbrales De Contaminación en Ecosistemas Acuáticos De Agua Dulce

“…El método de los intervalos de confianza (IC) que no se traslapan determina el umbral x_c, como el primer valor para el cual el IC de P(Q > q_c | X > r_c) no traslapa con el IC de P(Q > q_c) (2) . En casos donde el comportamiento no es estrictamente monótono el resultado no es razonable.…”

“…Un modelo que parte en dos el conjunto de los datos, dando lugar a dos parámetros que definirán el punto de cambio, a partir de un estimador de buen ajuste como la Devianza (2) . El comando nls() es la aplicación en R (5,8,10) que trabaja con este tipo de modelos.…”

“…El comando nls() es la aplicación en R (5,8,10) que trabaja con este tipo de modelos. En R, la diferencia principal entre los modelos lineales y modelos no lineales es tener que especificar la naturaleza exacta de la ecuación como parte de la fórmula del modelo no lineal (6,9,11) . En nuestro caso p_1,p_2,…,p_n son las probabilidades condicionales, x_1,x_2,…,x_n los valores de la variable de interés y f una función no lineal que se ajusta a los datos.…”