Patterns and predictability in the intra-annual organic carbon variability across the boreal and hemiboreal landscape

Losses of poly- and perfluoroalkyl substances to syringe filter materials

Syringe filters are used to separate solids from liquids prior to analysis of poly- and perfluoroalkyl substances (PFASs). This is a critical step in sample preparation, as losses of PFASs to the filter material can be significant and lead to underestimation. This study evaluated losses of 21 PFASs in three different matrices (methanol, MilliQ water, and water containing 10 mg L^-1 dissolved organic carbon (DOC)) to six different types of syringe filter (0.45 and 0.22 μm). Regarding sample matrix, the lowest average ∑₂₁PFAS losses were observed in methanol (13%), followed by DOC water (19%) and MilliQ water (26%). Regarding syringe filter material, the lowest average losses of ∑₂₁PFAS in DOC water and MilliQ water were observed for a recycled cellulose filter (average losses 16% and 21%, respectively), while a polypropylene filter had the lowest ∑₂₁PFAS losses in methanol (9%). A smaller polyethersulfone (PES) filter (0.22 μm, 17 mm Ø) showed significantly (p < 0.05) lower ∑₂₁PFAS losses in DOC water (on average 7.3%) than a larger PES filter (0.45 μm, 37 mm Ø) (23%). In DOC water, losses to the filter increased by 3.8%, 5.1%, and 8.4% per CF₂-moiety for C₃-C₁₁ perfluoroalkyl carboxylates (PFCAs), perfluoroalkyl sulfonates (PFSAs), and fluorotelomer sulfonic acids (FTSAs), respectively. Comparing different functional groups of PFASs, losses increased as follows: PFCAs < PFSAs < FTSAs < perfluorooctanesulfonamides (FOSAs). Thus, care is needed when including filtration in PFAS analysis, since losses can be significant (up to 100%) depending on the type of syringe filter, target PFAS, and matrix.

Boreal forest riparian zones regulate stream sulfate and dissolved organic carbon

In boreal forest catchments, solute transfer to streams is controlled by hydrological and biogeochemical processes occurring in the riparian zone (RZ). However, RZs are spatially heterogeneous and information about solute chemistry is typically limited. This is problematic when making inferences about stream chemistry. Hypothetically, the strength of links between riparian and stream chemistry is time-scale dependent. Using a ten-year (2003-2012) dataset from a northern Swedish catchment, we evaluated the suitability of RZ data to infer stream dynamics at different time scales. We focus on the role of the RZ versus upslope soils in controlling sulfate (SO4(2)(-)) and dissolved organic carbon (DOC). A priori, declines in acid deposition and redox-mediated SO4(2)(-) pulses control sulfur (S) fluxes and pool dynamics, which in turn affect dissolved organic carbon (DOC). We found that the catchment is currently a net source of S, presumably due to release of the S pool accumulated during the acidification period. In both, RZ and stream, SO4(2-) concentrations are declining over time, whereas DOC is increasing. No temporal trends in SO4(2-) and DOC were observed in upslope mineral soils. SO4(2-) explained the variation of DOC in stream and RZ, but not in upslope mineral soil. Moreover, as SO4(2-) decreased with time, temporal variability of DOC increased. These observations indicate that: (1) SO4(2-) is still an important driver of DOC trends in boreal catchments and (2) RZ processes control stream SO4(2-) and subsequently DOC independently of upslope soils. These phenomena are likely occurring in many regions recovering from acidification. Because water flows through a heterogeneous mosaic of RZs before entering the stream, upscaling information from limited RZ data to the catchment level is problematic at short-time scales. However, for long-term trends and annual dynamics, the same data can provide reasonable representations of riparian processes and support meaningful inferences about stream chemistry.