Soil Solution Analysis With Untargeted GC–MS—A Case Study With Different Lysimeter Types

Abstract

Surface-sourced organic compounds in infiltrating waters and percolates are transformed during their belowground passage. Biotic and abiotic processes thereby lead to continuously changing chemical environments in subsurface compartments. The investigation of such transformations of organic compounds aims for tracing subsurface fluxes as well as biotic and abiotic activity. To collect samples of soil solution, different kinds of lysimeters are available, spanning simple free-draining devices that sample water based on gravimetric flow and tension lysimeters allowing for approximating natural hydraulic conditions. Protocols for untargeted analytical profiling of organic soil solution constituents are scarce. We report here a solid phase extraction followed by GC–MS analysis, utilizing two long-term sampling devices in the Hainich Critical Zone Exploratory in Thuringia, Germany. In addition, we introduce a new lysimeter constructed exclusively from inert materials that allows for obtaining samples with little background signals in GC–MS. Polyvinylchloride (PVC)-based lysimeters introduce substantial background signals from plasticizers. We show how signals from these contaminants can be lowered during data analysis using chemometric background removal. Applying multivariate statistics for data analysis, we demonstrate the ability for monitoring of several sugars, fatty acids and phenolic acids at the topsoil-subsoil boundary and even beyond, via an untargeted analytical approach. Statistical tools facilitated the detection of differences in chemical signatures at three different land use sites. Data mining methods for metabolomics led to the identification of 3-carboxyphenylalanin as marker for a pasture site. The combined approach is suitable for the collection and extraction of topsoil and subsoil solution for untargeted metabolomics under near-natural flow conditions.

Groundwater metabolome responds to recharge in fractured sedimentary strata

“…Previously, it has been queried to which extent the surface vegetation or (agricultural) land use in a recharge area impacts underlying groundwater systems (Küsel et al, 2016). Metabolome investigations of soil waters sampled in lysimeters revealed discrete profiles for the input metabolome in forest, pasture and agricultural areas (Ueberschaar et al, 2021). For the groundwater wells in Figure 1, we have demonstrated elsewhere that the metabolomes of the wells studied on a single sampling day were discernible in relation to the factors land-use type, aquifer system and sampling depth (Sanchez-Arcos et al, 2022).…”

“…Therefore, previous studies focused on selected marker molecules and compound class categorization. However, fingerprint analyzes by multivariate approaches such as principal component analysis also proved a powerful tool for the analysis of aggregated monitoring data (Benk et al, 2018, 2019; Schwab et al, 2017; Ueberschaar et al, 2021). These approaches allowed for categorisation of similarities across compartments that can be linked to subsurface connectivity (Küsel et al, 2016; R. Lehmann & Totsche, 2020) and the composition of groundwater microbiomes (Benk et al, 2019; Yan et al, 2020).…”

“…Untargeted metabolomics techniques are emerging tools in environmental monitoring of compound transportation and transformation (Crimmins & Holsen, 2019). They have also been successfully used to characterize DOM in ground- (Aukes & Schiff, 2021;Benk et al, 2019;Hofmann et al, 2020), soil- (Danczak et al, 2020;Lu et al, 2020;Seifert et al, 2016;Ueberschaar et al, 2021) and surface-waters (Aukes & Schiff, 2021;Danczak et al, 2020;Drake et al, 2020;Garayburu-Caruso et al, 2020;Hawkes et al, 2020;Lu et al, 2020;Lynch et al, 2019;Patriarca & Hawkes, 2021;Pemberton et al, 2020;Raeke et al, 2017;Seifert et al, 2016;Tanentzap et al, 2019;Wologo et al, 2021). As sub-surface metabolites are not well characterized, exact compound annotation is challenging.…”