Evaluation of Statistical Treatments of Left-Censored Environmental Data Using Coincident Uncensored Data Sets. II. Group Comparisons.

The main classes of statistical treatments that have been used to determine if two groups of censored environmental data arise from the same distribution are substitution methods, maximum likelihood (MLE) techniques, and nonparametric methods. These treatments along with using all instrument-generated data (IN), even those less than the detection limit, were evaluated by examining 550 data sets in which the true values of the censored data were known, and therefore "true" probabilities could be calculated and used as a yardstick for comparison. It was found that technique "quality" was strongly dependent on the degree of censoring present in the groups. For low degrees of censoring (<25% in each group), the Generalized Wilcoxon (GW) technique and substitution of √2/2 times the detection limit gave overall the best results. For moderate degrees of censoring, MLE worked best, but only if the distribution could be estimated to be normal or log-normal prior to its application; otherwise, GW was a suitable alternative. For higher degrees of censoring (each group >40% censoring), no technique provided reliable estimates of the true probability. Group size did not appear to influence the quality of the result, and no technique appeared to become better or worse than other techniques relative to group size. Finally, IN appeared to do very well relative to the other techniques regardless of censoring or group size.

Evaluation of wastewater contaminant transport in surface waters using verified Lagrangian sampling.

Contaminants released from wastewater treatment plants can persist in surface waters for substantial distances. Much research has gone into evaluating the fate and transport of these contaminants, but this work has often assumed constant flow from wastewater treatment plants. However, effluent discharge commonly varies widely over a 24-hour period, and this variation controls contaminant loading and can profoundly influence interpretations of environmental data. We show that methodologies relying on the normalization of downstream data to conservative elements can give spurious results, and should not be used unless it can be verified that the same parcel of water was sampled. Lagrangian sampling, which in theory samples the same water parcel as it moves downstream (the Lagrangian parcel), links hydrologic and chemical transformation processes so that the in-stream fate of wastewater contaminants can be quantitatively evaluated. However, precise Lagrangian sampling is difficult, and small deviations - such as missing the Lagrangian parcel by less than 1h - can cause large differences in measured concentrations of all dissolved compounds at downstream sites, leading to erroneous conclusions regarding in-stream processes controlling the fate and transport of wastewater contaminants. Therefore, we have developed a method termed "verified Lagrangian" sampling, which can be used to determine if the Lagrangian parcel was actually sampled, and if it was not, a means for correcting the data to reflect the concentrations which would have been obtained had the Lagrangian parcel been sampled. To apply the method, it is necessary to have concentration data for a number of conservative constituents from the upstream, effluent, and downstream sites, along with upstream and effluent concentrations that are constant over the short-term (typically 2-4h). These corrections can subsequently be applied to all data, including non-conservative constituents. Finally, we show how data from other studies can be corrected.

In-stream attenuation of neuro-active pharmaceuticals and their metabolites.

In-stream attenuation was determined for 14 neuro-active pharmaceuticals and associated metabolites. Lagrangian sampling, which follows a parcel of water as it moves downstream, was used to link hydrological and chemical transformation processes. Wastewater loading of neuro-active compounds varied considerably over a span of several hours, and thus a sampling regime was used to verify that the Lagrangian parcel was being sampled and a mechanism was developed to correct measured concentrations if it was not. In-stream attenuation over the 5.4-km evaluated reach could be modeled as pseudo-first-order decay for 11 of the 14 evaluated neuro-active pharmaceutical compounds, illustrating the capacity of streams to reduce conveyance of neuro-active compounds downstream. Fluoxetine and N-desmethyl citalopram were the most rapidly attenuated compounds (t1/2 = 3.6 ± 0.3 h, 4.0 ± 0.2 h, respectively). Lamotrigine, 10,11,-dihydro-10,11,-dihydroxy-carbamazepine, and carbamazepine were the most persistent (t1/2 = 12 ± 2.0 h, 12 ± 2.6 h, 21 ± 4.5 h, respectively). Parent compounds (e.g., buproprion, carbamazepine, lamotrigine) generally were more persistent relative to their metabolites. Several compounds (citalopram, venlafaxine, O-desmethyl-venlafaxine) were not attenuated. It was postulated that the primary mechanism of removal for these compounds was interaction with bed sediments and stream biofilms, based on measured concentrations in stream biofilms and a column experiment using stream sediments.

Lagrangian mass-flow investigations of inorganic contaminants in wastewater-impacted streams.

Understanding the potential effects of increased reliance on wastewater treatment plant (WWTP) effluents to meet municipal, agricultural, and environmental flow requires an understanding of the complex chemical loading characteristics of the WWTPs and the assimilative capacity of receiving waters. Stream ecosystem effects are linked to proportions of WWTP effluent under low-flow conditions as well as the nature of the effluent chemical mixtures. This study quantifies the loading of 58 inorganic constituents (nutrients to rare earth elements) from WWTP discharges relative to upstream landscape-based sources. Stream assimilation capacity was evaluated by Lagrangian sampling, using flow velocities determined from tracer experiments to track the same parcel of water as it moved downstream. Boulder Creek, Colorado and Fourmile Creek, Iowa, representing two different geologic and hydrologic landscapes, were sampled under low-flow conditions in the summer and spring. One-half of the constituents had greater loads from the WWTP effluents than the upstream drainages, and once introduced into the streams, dilution was the predominant assimilation mechanism. Only ammonium and bismuth had significant decreases in mass load downstream from the WWTPs during all samplings. The link between hydrology and water chemistry inherent in Lagrangian sampling allows quantitative assessment of chemical fate across different landscapes.

Evaluation of statistical treatments of left-censored environmental data using coincident uncensored data sets: I. Summary statistics.

The main classes of statistical treatment of below-detection limit (left-censored) environmental data for the determination of basic statistics that have been used in the literature are substitution methods, maximum likelihood, regression on order statistics (ROS), and nonparametric techniques. These treatments, along with using all instrument-generated data (even those below detection), were evaluated by examining data sets in which the true values of the censored data were known. It was found that for data sets with less than 70% censored data, the best technique overall for determination of summary statistics was the nonparametric Kaplan-Meier technique. ROS and the two substitution methods of assigning one-half the detection limit value to censored data or assigning a random number between zero and the detection limitto censored data were adequate alternatives. The use of these two substitution methods, however, requires a thorough understanding of how the laboratory censored the data. The technique of employing all instrument-generated data--including numbers below the detection limit--was found to be less adequate than the above techniques. At high degrees of censoring (greater than 70% censored data), no technique provided good estimates of summary statistics. Maximum likelihood techniques were found to be far inferior to all other treatments except substituting zero or the detection limit value to censored data.

Accumulation of contaminants in fish from wastewater treatment wetlands.

Increasing demands on water resources in arid environments make reclamation and reuse of municipal wastewater an important component of the water budget. Treatment wetlands can be an integral part of the water-reuse cycle providing both water-quality enhancement and habitat functions. When used for habitat, the bioaccumulation potential of contaminants in the wastewater is a critical consideration. Water and fish samples collected from the Tres Rios Demonstration Constructed Wetlands near Phoenix, Arizona, which uses secondary-treated wastewater to maintain an aquatic ecosystem in a desert environment, were analyzed for hydrophobic organic compounds (HOC) and trace elements. Semipermeable membrane devices (SPMD) were deployed to investigate uptake of HOC. The wetlands effectively removed HOC, and concentrations of herbicides, pesticides, and organic wastewater contaminants decreased 40-99% between inlet and outlet. Analysis of Tilapia mossambica and Gambusia affinis indicated accumulation of HOC, including p,p'-DDE and trans-nonachlor. The SPMD accumulated the HOC detected in the fish tissue as well as additional compounds. Trace-element concentrations in whole-fish tissue were highly variable, but were similar between the two species. Concentrations of HOC and trace elements varied in different fish tissue compartments, and concentrations in Tilapia liver tissue were greater than those in the whole organism or filet tissue. Bioconcentration factors for the trace elements ranged from 5 to 58,000 and for the HOC ranged from 530 to 150,000.