Poly(lactic-co-glycolic acid) (PLGA) particle carriers of synthetic DNA have recently received increased attention for environmental applications due to their biodegradability, customizability, and nearly limitless number of uniquely identifiable "labels". In this paper, we present methodologies for the preparation of DNA-labeled particles, control of particle size, extraction of DNA-labels, and analysis via quantitative polymerase chain reaction (qPCR). Characterization and analysis of the DNA-labeled particles reveal spherical particles of diameters ranging from 60 to 1000 nm, with consistent zeta potentials around -45 mV, that are stable to aggregation, even in the presence of concentrated mono- and divalent cations. A highly correlated and consistent relationship between particle concentration and DNA-label count was observed, with a detection range spanning 7 orders of magnitude, from 0.01 to 10,000 mg/L (10-10⁷ particles/μL). The results of two environmental applications of the DNA-labeled particles are also presented, highlighting their feasibility for use in environmental studies. Whether exploring size-dependent transport phenomena or identifying potential pathogen transport pathways, the DNA-labeled particle approach presented here provides a powerful tool for the identification of overlapping particle signals at a range of concentrations.

Although multiple experimental studies have proven the use of free synthetic DNA as tracers in hydrological systems, their quantitative fate and transport, especially through the vadose zone, is still not well understood. Here we simulate the water flow and breakthrough of deuterium (D) and one free synthetic DNA tracer from a 10-day experiment conducted in a transient variably saturated 1m³ 10° sloped lysimeter using the HYDRUS-2D software package. Recovery and breakthrough flux of D (97.78%) and the DNA tracer (1.05%) were captured well with the advection-dispersion equation (R² = 0.949, NSE = 0.937) and the Schijven and Šimůnek two-site kinetic sorption model recommended for virus transport modeling (R² = 0.824, NSE = 0.823), respectively. The degradation of the DNA tracer was very slow (estimated to be 10% in 10 days), because the "loamy sand" porous media in our lysimeter was freshly crushed basaltic tephra (i.e., crushed rocks) and the microbes and DNase that could potentially degrade DNA in regular soils were rare in our "loamy sand". The timing of the concentration peaks and the HYDRUS-2D simulated temporal and spatial distribution of DNA in the lysimeter both revealed the role of the solid-water-air contact lines in mobilizing and carrying DNA tracer under the experimental variably saturated transient flow condition. The free DNA was nearly non-selectively transported through the porous media, and showed a slightly early breakthrough, possibly due to a slight effect of anion exclusion or size exclusion. Our results indicate that free DNA have the potential to trace vadose zone water flow and solute/contaminant transport, and to serve as surrogates to trace viral pathogen pollution in soil-water systems. To our knowledge, this study is the first to simulate transport mechanisms of free synthetic DNA tracers through real soil textured porous media under variably saturated transient flow condition.

Colloid transport through complex and dynamic (i.e. non-steady-state) hydrologic systems is rarely studied, owing to the difficulty of constraining initial and boundary conditions and quantifying colloid-porous media and colloid-colloid interactions in transient flow systems. Here we present a particle tracer experiment conducted on a sloped lysimeter receiving periodic rainfall events for 10 days. Four unique, DNA-labelled particle tracers were injected both in sequence and in parallel, together with a conservative tracer (deuterium), over the course of the first day and allowed to move through the system. Discharge-particle tracer concentration curves and the spatial distribution of particle tracer mass retained in the soil at the end of the experiment were found to be highly dependent on the timing of the tracer injection and the precipitation input and subsequent dynamic response of the water table. Overall, neglecting the total DLT recovery rate, the DLT particle tracer breakthrough trend (DNA-labelled particle tracer 4) was similar to deuterium and decreased over time with the exception of a few peaks later in the experiment. The individual particle tracer breakthrough curves suggest a complex system with different fast transport mechanisms (e.g. capillary barrier and size exclusion effect) and slow retention-release mechanisms (e.g. straining, physical-chemical adsorption), which resulted in particle tracers transferring faster than deuterium in the first 10 h of the experiment but being exceeded by deuterium soon after deuterium started to break through. The experiment not only highlights the interaction of repeated colloidal pollution events in hydrologic systems with different pre-event saturation conditions, but also the benefits of using multiple synchronous or sequential tracer applications to dissect explicit formulations of water flow and colloid transport processes in complex and dynamic hydrological systems. Such explicit process formulations could help improve understanding hydrologically-controlled transport through catchments and the quantitative prediction of these processes with water quality models.