Using spatiotemporal ratio analyses to quantitatively estimate water quality recovery of the Rio Doce.

The Fundão Dam breached on 5 November 2015 (the "Event"), resulting in the release of tailings, water, scoured sediment and/or soil, and other debris to downstream watercourses. Statistical analyses using historical and recent water quality measurements were conducted to assess the extent to which water quality in the Rio Doce was recovering to baseline conditions. A review of station- and/or parameter-specific water quality time series in the Rio Doce revealed two challenges: pre-Event data imbalance and seasonality. Due to the combined effects of these two factors, data gathered from Rio Doce water quality stations before the Event likely underestimated concentration ranges and limited the usefulness of common recovery assessment techniques such as times series and water quality standard exceedance analyses. These challenges were addressed by calculating quarterly and watershed-specific river-to-tributary ratios. R code was used to produce spatiotemporal time series for 44 investigated parameters that were measured both before and after the Event. The water quality recovery durations shown by the parameter- and/or region-specific river-to-tributary ratio time series indicated that (a) turbidity provides the most conservative measure for water quality recovery; (b) chemical parameters associated with the tailings, like manganese and iron recovered faster than turbidity; and (c) other investigated parameters unrelated to the tailings showed either no discernable impact or rapid recovery after the Event. The resulting parameter- and/or region-specific river-to-tributary ratio time series provided reliable and quantifiable estimates of water quality recovery durations. The water quality in the region furthest from Fundão Dam, in Espírito Santo, recovered one year after the Event, while water quality in the closest region to Fundão Dam, upstream of Risoleta Neves (Candonga) Dam, recovered 4.2 years after the Event. Integr Environ Assess Manag 2024;20:74-86. © 2023 Newfields Companies, LLC. Integrated Environmental Assessment and Management published by Wiley Periodicals LLC on behalf of Society of Environmental Toxicology & Chemistry (SETAC).

Determining recovery of marine water quality of the Rio Doce using statistical and temporal comparisons with nearby river systems.

The Fundão dam breach affected the Rio Doce's estuarine and marine environments with water, tailings, scoured soil and/or sediments, and other debris. Time series and standard exceedance analyses are typically used to assess water quality recovery to baseline conditions after deteriorating water quality events. In the absence of historical measurements, impacts to water quality from the Fundão Event were compared with measurements of nearby rivers. Similar river systems with available water quality measurements were grouped into affected and unaffected estuarine and marine waters. Statistical and temporal comparisons of marine waters unaffected by the Rio Doce with those affected by the Rio Doce were evaluated for systematic differences. Multivariate statistical techniques were also used to assess water quality differences. Our results demonstrate that the Fundão dam breach had a short-term water quality impact on marine waters near the Rio Doce mouth. Principal component and comparative time series analyses clearly demonstrated this impact during the 2015/2016 and 2016/2017 wet seasons. The ephemeral effects of the breach, however, did not compromise marine water quality. Exceedances of CONAMA standards for metals remained either at zero or at very low levels during the affected period (<5.7%). Before the start of the next wet season in October 2017, water quality impacts from the Event were statistically indistinguishable from unaffected marine waters, indicating recovery to baseline conditions. Integr Environ Assess Manag 2024;20:99-116. © 2023 NewFields Companies, LLC. Integrated Environmental Assessment and Management published by Wiley Periodicals LLC on behalf of Society of Environmental Toxicology & Chemistry (SETAC).

Applying multivariate techniques to fingerprint water quality impact of the Fundão Dam breach within the Rio Doce basin.

The Fundão Dam breach on 5 November 2015 (the "Event") released tailings, water, soil and/or sediments, and other debris to downstream watercourses. This breach included both direct and indirect impacts from scouring of soils and sediments along and within the affected courses. Multivariate statistical techniques were used to determine the potential of fingerprinting the impact of the breach compared to pre-Event water quality conditions and unaffected watercourses. The selection of key parameters is an important first step for multivariate analyses. Analysis of too many parameters can mask important trends and relationships, while analysis of too few may miss significant water quality indicators. A two-phased selection process was used to identify key parameters that indicated impact from the Event: (a) unbiased, principal component analysis to extract chemically dominant profiles among all measured parameters and (b) comparison of metals' concentrations between unaffected soils and/or sediments and tailings samples. Radar charts of key parameters along with statistical comparisons to pre-Event and not-affected waterways were then aggregated over space and time to assess impact and potential recovery to pre-Event conditions. Nine parameters were identified that characterize tailings-related (direct) and background soil and/or sediment-related (indirect) impacts. Spatially and temporally aggregated radar charts and nonparametric Mann-Whitney U tests were used to assess the statistical significance of these impacts during each wet season since the breach. Indirect parameters, like aluminum and lead, returned to pre-Event levels in the first wet season after the Event. By the 2018/2019 wet season, most of the direct and indirect parameters had returned to pre-Event levels. Integr Environ Assess Manag 2024;20:133-147. © 2023 NewFields Companies, LLC. Integrated Environmental Assessment and Management published by Wiley Periodicals LLC on behalf of Society of Environmental Toxicology & Chemistry (SETAC).

Important considerations in the derivation of background at sediment sites.

In the United States, there is an absence of federal guidance related to deriving and applying background concentrations at contaminated sediment sites. This absence has resulted in significant variability, uncertainty, and disagreement regarding how representative background concentrations of chemicals of concern should be derived for these sites. The present article discusses important considerations in the derivation of representative background concentrations to be used in the evaluation of contaminated sediment sites. Specifically, a thorough understanding of a site is critical to selecting the background reference areas from which representative background concentrations can be derived, representative background concentrations should account for contributions from those background chemical inputs (natural and anthropogenic sources) that will continue affecting the site even after remediation, perceived outliers should not be eliminated from the background data set just because they are the highest or lowest values, and geochemical evaluation of trace metals is a useful tool for deriving representative background concentrations. On a site-specific level, representative background concentrations are critical for putting site-related risk into context, developing a cost-effective and technically feasible remedial approach, understanding the potential for recontamination, and ensuring long-term remedy success. In a broader context, clear guidance from the United State Environmental Protection Agency (USEPA) for deriving and applying background concentrations for contaminated sediment sites would help promote national consistency in site assessment and remedy decision making. Integr Environ Assess Manag 2019;00:000-000. © 2019 The Authors. Integrated Environmental Assessment and Management Published by Wiley Periodicals, Inc. on behalf of Society of Environmental Toxicology & Chemistry (SETAC).

Chemical Preservation of Semi-volatile Polycyclic Aromatic Hydrocarbon Compounds at Ambient Temperature: A Sediment Sample Holding Time Study.

Site investigations require the collection and analysis of representative environmental samples to delineate impacts, risks, and remediation options. When environmental samples are collected, concentrations of semi-volatile polycyclic aromatic hydrocarbons (PAHs) begin to change due to several processes, such as evaporation, adsorption, precipitation, photo, and microbial degradation. Preservation techniques are used to minimize these changes between collection and analysis. The most common techniques are refrigeration, freezing, and acidification. In the mid 1970 s, regulatory agencies developed a holding time limit of 14 days for PAHs in soil/sediment samples stored at < 6 °C. The technical basis for this limit is not well defined yet failing to meet this limit may force resampling. This study examined the effectiveness of preservatives at maintaining PAH concentrations in sediment samples to 60 days. Sediment samples were collected at three sites that were impacted with petrogenic and pyrogenic PAHs. Chemically preserved (sodium azide, NaN3) and unpreserved samples were analyzed at defined time intervals from 0 to 60 days. Statistical analysis indicated acceptable preservation of PAHs in the sediment samples preserved with sodium azide for 60 days when maintained at either ambient laboratory temperature or 4 ± 2 °C, and for up to 21 days with no preservative when maintained at 4 ± 2 °C.

Experimental Evaluation of Preservation Techniques for Benzene, Toluene, Ethylbenzene, and Total Xylenes in Water Samples.

An experiment was designed to address the validity of the prescribed maximum allowable holding-time limit of 14 days when acidified at < 2 pH and maintained at 4°C to prevent significant loss of benzene, toluene, ethyl benzene, and xylenes (BTEX) in preserved water samples. Preservation methods prescribed by the United State Environmental Protection Agency were used as well as adaptions of that procedure to determine stability between 3 and 21 days. Water samples preserved at 4°C and pH of < 2 with hydrochloric acid did not result in unacceptable (> 15%) BTEX losses during the study as defined by procedures and statistical methods described by the American Society for Testing and Materials International. In addition, water samples preserved only with acid (pH < 2) at ambient temperatures (20-27°C) also provided acceptable results during the 21-day study. These results have demonstrated the acceptability of BTEX data derived from water samples exceeding the standard holding-time and/or temperature limits.

Scale-dependence in polychlorinated biphenyl (PCB) exposure effects on waterbird habitat occupancy.

Spatial scale is rarely considered in population-level assessments of contaminant impacts on wild animals; as a result misinterpretation of the relationship between contaminant exposure and population status may occur. We assessed the strength of expression of polychlorinated biphenyl (PCB) exposure effects at local vs. regional spatial scales on population status in five species of waterbirds, "bioaccumulators" often promoted as indicators of contaminant effects in aquatic ecosystems. Our focus was the upper Hudson River where PCBs occur at levels reported to have adverse impacts on wild birds. At the local scale, waterbird habitat occupancy was estimated from 220 repeat surveys made between 2001 and 2010 along the same survey route divided into 25 contiguous river segments with markedly different PCB concentrations. At the regional scale, waterbird habitat occupancy in relation to proximity to the upper Hudson River was estimated across 1248 Breeding Bird Atlas survey blocks while controlling for region-wide variation in habitat availability. At the local scale, many associations of habitat and sampling covariates with species detection probabilities were evident but none, including PCB concentration, with habitat occupancy, extinction or colonization of a given river segment. At the regional scale, survey effort and habitat factors not related to PCB exposure were the most important drivers of waterbird occurrence although two species were more likely to occur farther from the contaminated river segment. Spatial scale clearly mediates expression of contaminant impacts on wild bird populations; large-scale, expert-generated databases provide an underused opportunity for better delineating the spatial scales at which population impacts occur and risk assessments should be performed.

Assessing the footprint and volume of oil deposited in deep-sea sediments following the Deepwater Horizon oil spill.

The lateral and vertical extents of Macondo oil in deep-sea sediments resulting from the 2010 Deepwater Horizon oil spill were determined using chemical forensics and geostatistical kriging of data from 2397 sediment samples from 875 cores collected in 2010/2011 and 2014. The total mass of Macondo-derived hopane on the seafloor in 2010/2011 was conservatively estimated between 2.00 and 2.26metric tons, derived from 219,000 to 247,000barrels of oil; or 6.9 to 7.7% of the 3.19millionbarrels spilled. Macondo-derived hopane was deposited over 1030 to 1910km2 of the seafloor, mostly (>97%) in surface (0-1cm) and near-surface (1-3cm) sediments, which is consistent with short-term oil deposition. Although Macondo oil was still present in surface sediments in 2014, the total mass of Macondo-derived hopane was significantly lower (~80 to 90%) than in 2010/2011, affirming an acute impact from the spill and not long-term deposition from natural seeps.

Impacts of the Deepwater Horizon oil spill on the salt marsh vegetation of Louisiana.

The coastal wetland vegetation component of the Deepwater Horizon oil spill Natural Resource Damage Assessment documented significant injury to the plant production and health of Louisiana salt marshes exposed to oiling. Specifically, marsh sites experiencing trace or greater vertical oiling of plant tissues displayed reductions in cover and peak standing crop relative to reference (no oiling), particularly in the marsh edge zone, for the majority of this four year study. Similarly, elevated chlorosis of plant tissue, as estimated by a vegetation health index, was detected for marsh sites with trace or greater vertical oiling in the first two years of the study. Key environmental factors, such as hydrologic regime, elevation, and soil characteristics, were generally similar across plant oiling classes (including reference), indicating that the observed injury to plant production and health was the result of plant oiling and not potential differences in environmental setting. Although fewer significant impacts to plant production and health were detected in the latter years of the study, this is due in part to decreased sample size occurring as a result of erosion (shoreline retreat) and resultant loss of plots, and should not be misconstrued as indicating full recovery of the ecosystem.

Shoreline oiling from the Deepwater Horizon oil spill.

We build on previous work to construct a comprehensive database of shoreline oiling exposure from the Deepwater Horizon (DWH) spill by compiling field and remotely-sensed datasets to support oil exposure and injury quantification. We compiled a spatial database of shoreline segments with attributes summarizing habitat, oiling category and timeline. We present new simplified oil exposure classes for both beaches and coastal wetland habitats derived from this database integrating both intensity and persistence of oiling on the shoreline over time. We document oiling along 2113km out of 9545km of surveyed shoreline, an increase of 19% from previously published estimates and representing the largest marine oil spill in history by length of shoreline oiled. These data may be used to generate maps and calculate summary statistics to assist in quantifying and understanding the scope, extent, and spatial distribution of shoreline oil exposure as a result of the DWH incident.

Field assessment of the impacts of Deepwater Horizon oiling on coastal marsh vegetation of Mississippi and Alabama.

The Deepwater Horizon incident, which occurred in April 2010, resulted in significant oiling of coastal habitats throughout the northern Gulf of Mexico. Although the most substantial oiling of coastal salt marshes occurred in Louisiana, oiling of salt marshes in Mississippi and Alabama was documented as well. A field study conducted in Mississippi and Alabama salt marshes as a component of the Deepwater Horizon Natural Resource Damage Assessment determined that >10% vertical oiling of plant tissues reduced live vegetation cover and aboveground biomass (live standing crop) relative to reference sites in this region through fall 2012. This reduction of live vegetation cover and aboveground biomass appears to have largely resulted from diminished health and vigor of Juncus roemerianus, a key salt marsh species in Mississippi and Alabama. Fewer significant reductions in live vegetation cover and aboveground biomass were detected by the fall 2013 sampling, suggesting that vegetation in oiled salt marshes in this region may have begun to recover. This is corroborated by low levels of Deepwater Horizon oil contamination in these salt marsh soils. However, these findings should be interpreted in the context of the restricted sampling intensity of the present study. Environ Toxicol Chem 2016;35:2791-2797. © 2016 The Authors. Environmental Toxicology and Chemistry Published by Wiley Periodicals, Inc. on behalf of SETAC.

Impacts of the Deepwater Horizon Oil Spill on Salt Marsh Periwinkles (Littoraria irrorata).

Deepwater Horizon was the largest marine oil spill in U.S. waters, oiling large expanses of coastal wetland shorelines. We compared marsh periwinkle (Littoraria irrorata) density and shell length at salt marsh sites with heavy oiling to reference conditions ∼16 months after oiling. We also compared periwinkle density and size among oiled sites with and without shoreline cleanup treatments. Densities of periwinkles were reduced by 80-90% at the oiled marsh edge and by 50% in the oiled marsh interior (∼9 m inland) compared to reference, with greatest numerical losses of periwinkles in the marsh interior, where densities were naturally higher. Shoreline cleanup further reduced adult snail density as well as snail size. Based on the size of adult periwinkles observed coupled with age and growth information, population recovery is projected to take several years once oiling and habitat conditions in affected areas are suitable to support normal periwinkle life-history functions. Where heavily oiled marshes have experienced accelerated erosion as a result of the spill, these habitat impacts would represent additional losses of periwinkles. Losses of marsh periwinkles would likely affect other ecosystem processes and attributes, including organic matter and nutrient cycling, marsh-estuarine food chains, and multiple species that prey on periwinkles.

Monitoring and identification of spatiotemporal landscape changes in multiple remote sensing images by using a stratified conditional Latin hypercube sampling approach and geostatistical simulation.

This study develops a stratified conditional Latin hypercube sampling (scLHS) approach for multiple, remotely sensed, normalized difference vegetation index (NDVI) images. The objective is to sample, monitor, and delineate spatiotemporal landscape changes, including spatial heterogeneity and variability, in a given area. The scLHS approach, which is based on the variance quadtree technique (VQT) and the conditional Latin hypercube sampling (cLHS) method, selects samples in order to delineate landscape changes from multiple NDVI images. The images are then mapped for calibration and validation by using sequential Gaussian simulation (SGS) with the scLHS selected samples. Spatial statistical results indicate that in terms of their statistical distribution, spatial distribution, and spatial variation, the statistics and variograms of the scLHS samples resemble those of multiple NDVI images more closely than those of cLHS and VQT samples. Moreover, the accuracy of simulated NDVI images based on SGS with scLHS samples is significantly better than that of simulated NDVI images based on SGS with cLHS samples and VQT samples, respectively. However, the proposed approach efficiently monitors the spatial characteristics of landscape changes, including the statistics, spatial variability, and heterogeneity of NDVI images. In addition, SGS with the scLHS samples effectively reproduces spatial patterns and landscape changes in multiple NDVI images.