Road salt applications mobilize trace elements from roadside soil to shallow groundwater.

In regions where deicers are applied to roadways, micronutrients and toxic trace elements may be mobilized from soil material into soil porewater. These elements may subsequently migrate with soil porewater to surface waters and groundwaters, potentially leaching the soil of micronutrients or introducing toxins to water resources. Our study thus aims to quantify the timing and extent of trace element releases from soil material to soil porewater and groundwater in response to deicing events. We sampled soil porewater near a road at a rural site for trace elements and compared the results to salt applications and soil porewater Na and Cl levels. We also assessed trace element, Na, and Cl concentrations in a karst spring at the rural site and a karst spring at an urban site to evaluate the role of land use in conveying these contaminants to groundwater. We found that certain trace elements (e.g., As, Ba, Fe, Sr) peaked concomitantly with Na and Cl in soil porewater at the rural site after road deicing events, suggesting their release due to excess salt inputs to the soil. We did not observe increases in trace element concentrations at the rural karst spring following individual road salt applications, likely due to low deicer inputs and trace element levels across its recharge basin. However, at the urban site, we observed that other assemblages of trace elements (e.g., As, Cu, Li) in the karst spring peaked with deicing-related Na and Cl pulses. We also found positive and significant correlations between salt applications to the recharge basin and exports of some trace elements (e.g., As, Cu, Li, Se) at the urban karst spring, indicating deicing events triggered trace element releases to groundwater. Overall, we detected road salt-driven trace element release from soil material to soil porewater and groundwater that was exacerbated by urbanization.

Cave sediment sequesters anthropogenic microparticles (including microplastics and modified cellulose) in subsurface environments.

Anthropogenic microparticles (of synthetic, semisynthetic, or modified natural compositions) are globally pervasive, yet little is known about their distribution and storage in the subsurface despite their potential threats to belowground environments. We therefore assessed their amounts and characteristics in water and sediment from a cave in the United States. During a flood, water and sediment samples were collected at 8 sites every ~25 m along the cave passageways. Both sample types were evaluated for anthropogenic microparticles, while water was assessed for geochemistry (e.g., inorganic species) and sediment was evaluated for particle sizes. Additional water samples were collected during low flow at the same sites for further geochemical analysis to determine water provenance. We found anthropogenic microparticles in all samples that were mainly fibers (91 %) and clear (59 %). Both suspected (identified visually) and confirmed (identified with Fourier transform infrared spectroscopy; FTIR) anthropogenic microparticle concentrations were positively correlated between the compartments (r ≥ 0.83, p ≤ 0.01), but quantities in sediment were ~100 times those in water. These findings indicate that sediment sequesters anthropogenic microparticle pollution in the cave. Microplastic concentrations were similar among all sediment samples, but only one water sample at the main entrance contained microplastics. Treated cellulosic microparticle abundances in both compartments generally increased along the cave stream's flowpath, which we suspect is due to both their flood and airborne deposition. Water geochemical and sediment particle size data collected at a branch indicated at least two distinct water sources to the cave. However, anthropogenic microparticle assemblages did not differ between these sites, implying minimal variation in sourcing across the recharge area. Our results reveal that anthropogenic microparticles intrude karst systems and are stored in sediment. Karstic sediment consequently represents a potential source of "legacy" pollution to the water resources and fragile habitats found in these globally distributed landscapes.

Floods enhance the abundance and diversity of anthropogenic microparticles (including microplastics and treated cellulose) transported through karst systems.

Microplastics (plastics <5 mm) are emerging contaminants that have been detected in virtually all environments. While microplastic research in terrestrial surface waters has been proliferating, microplastic contamination in subsurface environments remains understudied. Karst terrains may be particularly susceptible to microplastic pollution because the presence of large dissolution openings allows fast transport of water through these systems, facilitating the introduction of surface contaminants into subsurface habitats. Furthermore, few studies address the prevalence and movement of microparticles composed of semisynthetic and modified natural materials, despite their known ecotoxicity. Our study therefore aims to identify anthropogenic (i.e., synthetic, semisynthetic, and treated natural) microparticle extent, sourcing, and transport in subsurface karst environments. To do so, we examined a cave spring under variable flow conditions, finding that anthropogenic microparticles were present in all samples and were most frequently fibrous and clear. The mean anthropogenic microparticle concentration during baseflow was 9.2 counts/L but increased up to 81.3 counts/L during floods, indicating their enhanced mobilization when relatively dilute, acidic, and sediment-rich event water entered the cave. These results suggest that anthropogenic microparticles may originate from surface recharge or sediment resuspension within the cave. When we analyzed a subset of microparticles with Fourier transform infrared spectroscopy (FTIR), we found that cellulose of known (i.e., dyed) and suspected (i.e., clear) anthropogenic origin was the most abundant material type. We nevertheless confirmed the presence of microplastics in the cave stream under all flow conditions, with the most common polymer being polyethylene. Both the concentrations and relative fractions of microplastics were higher during floods compared to baseflow, indicating their increased transport during high flow events. We also observed that microplastic polymer types diversified as discharge increased. Our study gives new insight into how anthropogenic microparticle contamination is transported through karst landscapes that can help inform debris mitigation strategies to protect ecosystems and water resources.

Using in situ measurements of optical brighteners for rapid reconnaissance of wastewater inputs to water resources.

Untreated wastewater entering the environment through leaking infrastructure and sewer overflows threatens both human and aquatic health. Water managers therefore need low cost, in situ methods to detect sewage contamination in real time to promptly employ mitigation strategies. However, wastewater has traditionally been identified in waterbodies using chemical and microbial tracers and indicators that can be non-unique to wastewater and often require complex and expensive analyses. Optical brighteners (synthetic brightening compounds present in laundry detergents and paper products) are emerging as ideal tracers of wastewater because of their quick and inexpensive field detection using handheld fluorometers. To test the efficacy of optical brighteners as standalone, in situ wastewater tracers, field readings of their fluorescence were compared with traditional wastewater analytes (e.g., B, F-, microbial indicators) at multiple points in time and space for a suburban watershed (Fishpot Creek, Saint Louis, Missouri, United States). We also used chemical tracers in three mixing models of endmembers to assess the wastewater fraction across the watershed. Compared to other analytes, optical brightener fluorescence measurements had the strongest correlation with wastewater infrastructure density (r = 0.71, p < 0.05), indicating their utility as tracers. All our endmember mixing models employing optical brightener readings predicted positive and significant correlations between the untreated wastewater fraction in streamflow and sewer pipe density at each site (r ≥ 0.77, p < 0.05). While using optical brightener readings for wastewater detection has some limitations (e.g., minor photodegradation), we found them to be more robust tracers than other analytes. Thus, optical brightener fluorescence measurements are an ideal initial screening tool for identifying wastewater contributions to the environment.

Integrating land cover, point source pollution, and watershed hydrologic processes data to understand the distribution of microplastics in riverbed sediments.

Microplastics are emerging contaminants ubiquitously distributed in the environment, with rivers acting as their main mode of transport in surface freshwater systems. However, the relative importance of hydrologic processes and source-related variables for benthic microplastic distribution in river sediments is not well understood. We therefore sampled and characterized microplastics in river sediments across the Meramec River watershed (eastern Missouri, United States) and applied a hydrologic modeling approach to estimate the relative importance of river discharge, river sediment load, land cover, and point source pollution sites to understand how these environmental factors affect microplastic distribution in benthic sediments. We found that the best model for the Meramec River watershed includes both source-related variables (land cover and point sources) but excludes both hydrologic transport-related variables (discharge and sediment load). Prior work has drawn similar and dissimilar conclusions regarding the importance of anthropogenic versus hydrologic variables in microplastic distribution, though we acknowledge that comparisons are limited by methodological differences. Nevertheless, our findings highlight the complexity of microplastic pollution in freshwater systems. While generating a universal predictive model might be challenging to achieve, our study demonstrates the potential of using a modeling approach to determine the controlling factors for benthic microplastic distribution in fluvial systems.

Local Monitoring Should Inform Local Solutions: Morphological Assemblages of Microplastics Are Similar within a Pathway, But Relative Total Concentrations Vary Regionally.

Pathways for microplastics to aquatic ecosystems include agricultural runoff, urban runoff, and treated or untreated wastewater. To better understand the importance of each pathway as a vector for microplastics into waterbodies and for mitigation, we sampled agricultural runoff, urban stormwater runoff, treated wastewater effluent, and the waterbodies downstream in four regions across North America: the Sacramento Delta, the Mississippi River, Lake Ontario, and Chesapeake Bay. The highest concentrations of microplastics in each pathway varied by region: agricultural runoff in the Sacramento Delta and Mississippi River, urban stormwater runoff in Lake Ontario, and treated wastewater effluent in Chesapeake Bay. Material types were diverse and not unique across pathways. However, a PERMANOVA found significant differences in morphological assemblages among pathways (p < 0.005), suggesting fibers as a signature of agricultural runoff and treated wastewater effluent and rubbery fragments as a signature of stormwater. Moreover, the relationship between watershed characteristics and particle concentrations varied across watersheds (e.g., with agricultural parameters only being important in the Sacramento Delta). Overall, our results suggest that local monitoring is essential to inform effective mitigation strategies and that assessing the assemblages of morphologies should be prioritized in monitoring programs to identify important pathways of contamination.

Road salt retention and transport through vadose zone soils to shallow groundwater.

Increasing background salinity in watersheds has largely been attributed to road salt retention in groundwaters due to their long residence times. However, laboratory studies demonstrate that soils temporarily store salts, either in porewater or adsorbed onto particles. Field studies of road salt retention in soils are nevertheless rare, and mechanisms of salt transport across multiple hydrological reservoirs (e.g., from soil to groundwater) are unknown. Thus, we collected roadside soil porewater and karst spring water weekly for ~1.5 yr to determine salt transport through the vadose zone into the phreatic zone. We observed dual retention mechanisms of sodium (Na+) and chloride (Cl-) in soils due to slow porewater movement, causing ion movement through the soil as slow as 1.3 cm/day, and cation exchange processes, leading to initial Na+ retention followed by later release months after application. Cation exchange processes also caused base cation loss from exchange sites into mobile porewater. Rapid Na+ and Cl- delivery to groundwater occurred through karst conduits during the winter. However, elevated background levels of salt ions in groundwater during the non-salting months indicated accumulation in the catchment due to slower porewater flow in the soil and rock matrix and delayed Na+ release from soil exchange sites.

An R package for correcting continuous water quality monitoring data for drift.

Continuous water quality monitoring ins- truments are used to understand the chemical and physical behaviors of aquatic environments over time. However, the data generated from these instruments are susceptible to inaccuracies due to drift that can occur between site visits. While there are several software packages available to correct drift in water quality data, these packages are often proprietary, expensive, and/or do not offer the user control over the data corrections. This paper describes driftR, an R package that corrects drift in water quality data. driftR implements either one- or two-point variable data corrections based on the number of standards used to calibrate the sensor of interest, then linearly interpolates the correction over the period of interest. This program gives control to users to correct each parameter in a way that is ideal for their unique stu- dies and offers a free, reproducible method for drift correction.

Sampling, Sorting, and Characterizing Microplastics in Aquatic Environments with High Suspended Sediment Loads and Large Floating Debris.

The ubiquitous presence of plastic debris in the ocean is widely recognized by the public, scientific communities, and government agencies. However, only recently have microplastics in freshwater systems, such as rivers and lakes, been quantified. Microplastic sampling at the surface usually consists of deploying drift nets behind either a stationary or moving boat, which limits the sampling to environments with low levels of suspended sediments and floating or submerged debris. Previous studies that employed drift nets to collect microplastic debris typically used nets with ≥300 µm mesh size, allowing plastic debris (particles and fibers) below this size to pass through the net and elude quantification. The protocol detailed here enables: 1) sample collection in environments with high suspended loads and floating or submerged debris and 2) the capture and quantification of microplastic particles and fibers <300 µm. Water samples were collected using a peristaltic pump in low-density polyethylene (PE) containers to be stored before filtering and analysis in the lab. Filtration was done with a custom-made microplastic filtration device containing detachable union joints that housed nylon mesh sieves and mixed cellulose ester membrane filters. Mesh sieves and membrane filters were examined with a stereomicroscope to quantify and separate microplastic particulates and fibers. These materials were then examined using a micro-attenuated total reflectance Fourier transform infrared spectrometer (micro ATR-FTIR) to determine microplastic polymer type. Recovery was measured by spiking samples using blue PE particulates and green nylon fibers; percent recovery was determined to be 100% for particulates and 92% for fibers. This protocol will guide similar studies on microplastics in high velocity rivers with high concentrations of sediment. With simple modifications to the peristaltic pump and filtration device, users can collect and analyze various sample volumes and particulate sizes.

Transport of road salt contamination in karst aquifers and soils over multiple timescales.

Road deicing has caused widespread environmental Na+ and Cl- release for decades, yet the transport and retention of these contaminants in karst aquifers and soils are poorly understood. We examined the transport dynamics of Na+ and Cl- from road salt in shallow groundwater during flooding and over seasonal timescales by intensively monitoring an urban and a rural karst spring over approximately 2 years. Furthermore, we used a 20-year dataset for the rural spring to determine how salt retention affected long-term geochemical trends in the shallow groundwater. Salt transport was governed by hydrologic pathways through karst aquifers: during winter and early spring floods, flow through preferential pathways rapidly transported salty meltwater or stormwater over hours to days, while the remaining salt-contaminated water moved diffusely through the rock matrix on timescales of months to years. Flood hydrograph separations revealed that event water constituted 61.2% of stormflow on average at the urban spring, leading to more extreme variability in salt concentrations during flooding and throughout the year. This variability indicates that baseflow contributions to urban streams overlying karst aquifers with preferential flowpaths are likely less effective at buffering salt concentrations. In contrast, salt concentrations were less variable in the baseflow-dominated rural spring (28.7% event water). Furthermore, salt was episodically released from soils to shallow groundwater throughout the year during first flush events. A Cl- mass balance indicates that Cl- applied during previous winters persists within the springs' recharge basins for more than a year, raising baseline concentrations as road salt is introduced faster than it can be flushed from the basin. Inter-annual salt retention by soils or slow groundwater movement likely caused significant Cl- and specific conductivity (SpC) increases at the rural spring from 1996 to 2016. Accumulation of salt in shallow groundwater can elevate baseflow concentrations in surface waters, where it threatens aquatic organisms.

Characterizing nutrient distributions and fluxes in a eutrophic reservoir, Midwestern United States.

Harmful algal blooms are increasingly common in aquatic ecosystems and have been linked to runoff from agricultural land. This study investigated the internal nutrient (i.e., phosphorus (P) and nitrogen (N)) dynamics of a eutrophic reservoir in the Midwestern United States to constrain the potential for sedimentary nutrients to stimulate harmful algal blooms. The spatial distribution of nutrients in the water column (soluble reactive P (SRP), nitrate/nitrite-N (NOx-N), and ammonium-N (NH4+-N)) and sediments (total P, total carbon (C), total N, and organic matter (OM)) were quantified and mapped. Water column nutrients varied spatially and temporally, with generally higher concentrations near the dam wall during normal lake levels. The upper portion of the lake, near the inlet, was sampled during a flood event and had overall higher nutrient concentrations and lower chlorophyll levels compared to normal lake level samples. Mean sedimentary total P (936mg/kg) was ~30% higher in the reservoir than the surrounding upland soils, with the highest concentrations near the dam wall (1661mg/kg) and a significant positive correlation found between sedimentary total P, total C, and OM. Additionally, 15 intact sediment cores were manipulated ex situ to examine mechanisms of nutrient flux across the sediment-water interface (SWI) that may trigger algal blooms. Core treatment conditions included advection (i.e., simulating potential nutrient fluxes during wind events through sediment resuspension) and diffusion. Core experiments indicated both advective and diffusive conditions at the SWI may trigger the flux of nutrients important for algal growth from lake sediments, with diffusion contributing both N and P to the water column, while intense advection increased water column N, but decreased P. Release of P to the water column may be more diffusion-driven than advection-driven, whereas N release to the water column appears to be both diffusion- and advection-driven.

Multiple sources of boron in urban surface waters and groundwaters.

Previous studies attribute abnormal boron (B) levels in streams and groundwaters to wastewater and fertilizer inputs. This study shows that municipal drinking water used for lawn irrigation contributes substantial non-point loads of B and other chemicals (S-species, Li, and Cu) to surface waters and shallow groundwaters in the St. Louis, Missouri, area. Background levels and potential B sources were characterized by analysis of lawn and street runoff, streams, rivers, springs, local rainfall, wastewater influent and effluent, and fertilizers. Urban surface waters and groundwaters are highly enriched in B (to 250µg/L) compared to background levels found in rain and pristine, carbonate-hosted streams and springs (<25µg/L), but have similar concentrations (150 to 259µg/L) compared to municipal drinking waters derived from the Missouri River. Other data including B/SO4(2-)-S and B/Li ratios confirm major contributions from this source. Moreover, sequential samples of runoff collected during storms show that B concentrations decrease with increased discharge, proving that elevated B levels are not primarily derived from combined sewer overflows (CSOs) during flooding. Instead, non-point source B exhibits complex behavior depending on land use. In urban settings B is rapidly mobilized from lawns during "first flush" events, likely representing surficial salt residues from drinking water used to irrigate lawns, and is also associated with the baseflow fraction, likely derived from the shallow groundwater reservoir that over time accumulates B from drinking water that percolates into the subsurface. The opposite occurs in small rural watersheds, where B is leached from soils by recent rainfall and covaries with the event water fraction.